Current Protocols in Nucleic Acid Chemistry

Current Protocols in Nucleic Acid Chemistry Contents Preface Foreword Chapter 1 Synthesis of Modified Nucleosides 1. 2. ...

Author: Martin Egli | Piet Herdewijn | Akira Matusda

56 downloads 1617 Views 31MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Current Protocols in Nucleic Acid Chemistry Contents Preface Foreword Chapter 1 Synthesis of Modified Nucleosides 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

Introduction Unit 1.1 Palladium-Mediated C5 Substitution of Pyrimidine Nucleosides Unit 1.2 Enzymatic Synthesis of M1G-Deoxyribose Unit 1.3 Synthesis of N2-Substituted Deoxyguanosine Nucleosides from 2-Fluoro-6-O(Trimethylsilylethyl)-2′-Deoxyinosine Unit 1.4 Unnatural Nucleosides with Unusual Base Pairing Properties Unit 1.5 Development of a Universal Nucleobase and Modified Nucleobases for Expanding the Genetic Code Unit 1.6 Syntheses of Specifically 15N-Labeled Adenosine and Guanosine Unit 1.7 Synthesis of Protected 2′-Deoxy-2′-fluoro-β-d-arabinonucleosides Unit 1.8 Synthesis, Characterization, and Application of Substituted Pyrazolopyrimidine Nucleosides Unit 1.9 Synthesis of 1,5-Anhydrohexitol Building Blocks for Oligonucleotide Synthesis Unit 1.10 Synthesis and Properties of 7-Substituted 7-Deazapurine (Pyrrolo[2,3-d]pyrimidine) 2′Deoxyribonucleosides Unit 1.11 Reduction of Ribonucleosides to 2′-Deoxyribonucleosides Unit 1.12 Synthesis of Fluorinated Nucleosides Unit 1.13 Synthesis of Ribonucleosides by Condensation Using Trimethylsilyl Triflate Unit 1.14 Synthesis of 2′-O-β-d-Ribofuranosylnucleosides Unit 1.15 Preparation of 2′-Deoxy-2′-Methylseleno-Modified Phosphoramidites and RNA Unit 1.16 Palladium-Catalyzed Cross-Coupling Reactions in C6 Modifications of Purine Nucleosides Unit 1.17 Nucleobase-Caged Phosphoramidites for Oligonucleotide Synthesis Unit 1.18 Synthesis of Altritol Nucleoside Phosphoramidites for Oligonucleotide Synthesis

Chapter 2 Protection of Nucleosides for Oligonucleotide Synthesis Introduction Unit 2.1 Nucleobase Protection of Deoxyribo- and Ribonucleosides Unit 2.2 Protection of 2′-Hydroxy Functions of Ribonucleosides Unit 2.3 Protection of 5′-Hydroxy Functions of Nucleosides Unit 2.4 A Base-Labile Protecting Group (Fluorenylmethoxycarbonyl) for the 5′-Hydroxy Function of Nucleosides 6. Unit 2.5 2′-Hydroxyl-Protecting Groups that are Either Photochemically Labile or Sensitive to Fluoride Ions 7. Unit 2.6 Deoxyribo- and Ribonucleoside H-Phosphonates 8. Unit 2.7 Deoxyribonucleoside Phosphoramidites 9. Unit 2.8 Regioselective 2′-Silylation of Purine Ribonucleosides for Phosphoramidite RNA Synthesis 10. Unit 2.9 Preparation of 2′-O-[(Triisopropylsilyl)oxy]methyl-protected Ribonucleosides 11. Unit 2.10 Preparation of 5′-Silyl-2′-Orthoester Ribonucleosides for Use in Oligoribonucleotide Synthesis 12. Unit 2.11 Enzymatic Regioselective Levulinylation of 2′-Deoxyribonucleosides and 2′-OMethylribonucleosides 1. 2. 3. 4. 5.

13. Unit 2.12 Nucleobase Protection with Allyloxycarbonyl 14. Unit 2.13 Universal 2-(4-Nitrophenyl)ethyl and 2-(4-Nitrophenyl)ethoxycarbonyl Protecting Groups

for Nucleosides and Nucleotides

Chapter 3 Synthesis of Unmodified Oligonucleotides 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

Introduction Unit 3.1 Solid-Phase Supports for Oligonucleotide Synthesis Unit 3.2 Attachment of Nucleosides to Solid-Phase Supports Unit 3.3 Synthetic Strategies and Parameters Involved in the Synthesis of Oligodeoxyribonucleotides According to the Phosphoramidite Method Unit 3.4 Synthesis of Oligodeoxyribo- and Oligoribonucleotides According to the H-Phosphonate Method Unit 3.5 Strategies for Oligoribonucleotide Synthesis According to the Phosphoramidite Method Unit 3.6 Oligoribonucleotides with 2′-O-(tert-Butyldimethylsilyl) Groups Unit 3.7 Synthesis of Oligoribonucleotides Using the 2-Nitrobenzyloxymethyl Group for 2′Hydroxyl Protection Unit 3.8 Chemical Synthesis of RNA Sequences with 2′-O-[(Triisopropylsilyl)oxy]methyl-protected Ribonucleoside Phosphoramidites Unit 3.9 3-(N-tert-Butylcarboxamido)-1-propyl and 4-Oxopentyl Groups for Phosphate/Thiophosphate Protection in Oligodeoxyribonucleotide Synthesis Unit 3.10 DNA Synthesis Without Base Protection Unit 3.11 The 4-Methylthio-1-Butyl Group for Phosphate/Thiophosphate Protection in Oligodeoxyribonucleotide Synthesis Unit 3.12 Nucleoside Phosphoramidites Containing Cleavable Linkers Unit 3.13 Microwave-Assisted Functionalization of Solid Supports for Rapid Loading of Nucleosides Unit 3.14 Solution-Phase Synthesis of Di- and Trinucleotides Using Polymer-Supported Reagents Unit 3.15 DNA Synthesis Without Base Protection Using the Phosphoramidite Approach Unit 3.16 A Universal and Recyclable Solid Support for Oligonucleotide Synthesis

Chapter 4 Synthesis of Modified Oligonucleotides and Conjugates 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

Introduction Unit 4.1 A Brief History, Status, and Perspective of Modified Oligonucleotides for Chemotherapeutic Applications Unit 4.2 Modification of the 5′ Terminus of Oligonucleotides for Attachment of Reporter and Conjugate Groups Unit 4.3 Direct Attachment of Conjugate Groups to the 5′ Terminus of Oligodeoxyribonucleotides Unit 4.4 Synthesis and Characterization of Chimeric 2-5A-DNA Oligonucleotides Unit 4.5 Attachment of Reporter and Conjugate Groups to the 3′ Termini of Oligonucleotides Unit 4.6 3′-Modified Oligonucleotides and their Conjugates Unit 4.7 Synthesis and Purification of Oligonucleotide N3′ P5′ Phosphoramidates and their Phosphodiester and Phosphorothioate Chimeras Unit 4.8 Incorporation of Halogenoalkyl, 2-Pyridyldithioalkyl, or Isothiocyanate Linkers into Ligands Unit 4.9 Modification of the 5′ Terminus of Oligodeoxyribonucleotides for Conjugation with Ligands Unit 4.10 Conjugation of 5′-Functionalized Oligodeoxyribonucleotides with Properly Functionalized Ligands Unit 4.11 Synthesis and Purification of Peptide Nucleic Acids Unit 4.12 Locked Nucleic Acids: Synthesis and Characterization of LNA-T Diol Unit 4.13 Cellular Delivery of Locked Nucleic Acids (LNAs) Unit 4.14 Solid-Phase Synthesis of Branched Oligonucleotides

16. Unit 4.15 Solid-Phase Synthesis of 2′-Deoxy-2′-fluoro- β-d-Oligoarabinonucleotides (2′F-ANA) and

Their Phosphorothioate Derivatives

17. Unit 4.16 Chemistry of CpG DNA 18. Unit 4.17 Synthesis of Phosphorothioate Oligonucleotides with Stereodefined Phosphorothioate

Linkages

19. Unit 4.18 Synthesis of Oligonucleotide Conjugates via Aqueous Diels-Alder Cycloaddition 20. Unit 4.19 5′-Iodination of Solid-Phase-Linked Oligodeoxyribonucleotides 21. Unit 4.20 Reversible Biotinylation of the 5′-Terminus of Oligodeoxyribonucleotides and its

Application in Affinity Purification

22. Unit 4.21 Uridine 2′-Carbamates: Facile Tools for Oligonucleotide 2′-Functionalization 23. Unit 4.22 Stepwise Solid-Phase Synthesis of Nucleopeptides 24. Unit 4.23 Synthesis of Oligoribonucleotides Containing N6-Alkyladenosine and 2-Methylthio-N6-

Alkyladenosine

25. Unit 4.24 Oligodeoxyribonucleotide Analogs Functionalized with Phosphonoacetate and

Thiophosphonoacetate Diesters

26. Unit 4.25 Base-Modified Oligodeoxyribonucleotides: Using Pyrrolo[2,3-d]pyrimidines to Replace

Purines

27. Unit 4.26 An Aminooxy-Functionalized Non-Nucleosidic Phosphoramidite for the Construction of

Multiantennary Oligonucleotide Glycoconjugates on a Solid Support 28. Unit 4.27 Large-Scale Preparation of Conjugated Oligonucleoside Phosphorothioates by the High-

Efficiency Liquid-Phase (HELP) Method

29. Unit 4.28 Disulfide Conjugation of Peptides to Oligonucleotides and Their Analogs 30. Unit 4.29 Methoxyoxalamido Chemistry in the Synthesis of Tethered Phosphoramidites and

Functionalized Oligonucleotides 31. Unit 4.30 Using Morpholinos to Control Gene Expression 32. Unit 4.31 Solid-Phase Oligonucleotide Labeling with DOTA

Chapter 5 Methods for Cross-Linking Nucleic Acids 1. 2. 3. 4. 5. 6. 7. 8.

Introduction Unit 5.1 Engineering Disulfide Cross-Links in RNA Using Thiol-Disulfide Interchange Chemistry Unit 5.2 Chemical and Enzymatic Methods for Preparing Circular Single-Stranded DNAs Unit 5.3 Engineering Specific Cross-Links in Nucleic Acids Using Glycol Linkers Unit 5.4 Engineering Disulfide Cross-Links in RNA Via Air Oxidation Unit 5.5 Use of Electrophilic Substitution to Form Site-Specific Cross-Links in DNA Unit 5.6 Synthesis of Endcap Dimethoxytrityl Phosphoramidites for Endcapped Oligonucleotides Unit 5.7 Engineering Terminal Disulfide Bonds Into DNA

Chapter 6 Chemical and Enzymatic Probes for Nucleic Acid Structure 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

Introduction Unit 6.1 Probing RNA Structure with Chemical Reagents and Enzymes Unit 6.2 Probing Nucleic Acid Structure with Shape-Selective Rhodium and Ruthenium Complexes Unit 6.3 Probing RNA Structure by Lead Cleavage Unit 6.4 Probing Nucleic Acid Structure with Nickel- and Cobalt-Based Reagents Unit 6.5 Probing RNA Structures with Hydroxyl Radicals Unit 6.6 Chemical Reagents for Investigating the Major Groove of DNA Unit 6.7 Probing DNA Structure with Hydroxyl Radicals Unit 6.8 Probing RNA Structure and Metal-Binding Sites Using Terbium(III) Footprinting Unit 6.9 Probing RNA Structure and Function by Nucleotide Analog Interference Mapping

Chapter 7 Biophysical Analysis of Nucleic Acids

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

Introduction Unit 7.1 Biophysical Analysis of Nucleic Acids Unit 7.2 NMR Determination of Oligonucleotide Structure Unit 7.3 Optical Methods Unit 7.4 Calorimetry of Nucleic Acids Unit 7.5 Molecular Modeling of Nucleic Acid Structure Unit 7.6 Methods to Crystallize RNA Unit 7.7 Recent Advances in RNA Structure Determination by NMR Unit 7.8 Molecular Modeling of Nucleic Acid Structure: Energy and Sampling Unit 7.9 Molecular Modeling of Nucleic Acid Structure: Electrostatics and Solvation Unit 7.10 Molecular Modeling of Nucleic Acid Structure: Setup and Analysis Unit 7.11 Characterization of DNA Structures by Circular Dichroism Unit 7.12 Biophysical Analysis of Triple-Helix Formation

Chapter 8 Nucleic Acid Binding Molecules 1. 2. 3. 4. 5. 6.

Introduction Unit 8.1 Determination of Binding Mode: Intercalation Unit 8.2 Determination of Binding Thermodynamics Unit 8.3 A Competition Dialysis Assay for the Study of Structure-Selective Ligand Binding to Nucleic Acids Unit 8.4 Chemistry of Minor Groove Binder–Oligonucleotide Conjugates Unit 8.5 A Fluorescent Intercalator Displacement Assay for Establishing DNA Binding Selectivity and Affinity

Chapter 9 Combinatorial Methods in Nucleic Acid Chemistry 1. 2. 3. 4. 5. 6. 7.

Introduction Unit 9.1 Theoretical Principles of In Vitro Selection Using Combinatorial Nucleic Acid Libraries Unit 9.2 Design, Synthesis, and Amplification of DNA Pools for In Vitro Selection Unit 9.3 In Vitro Selection of RNA Aptamers to a Protein Target by Filter Immobilization Unit 9.4 Selection for Catalytic Function with Nucleic Acids Unit 9.5 In Vitro Selection of RNA Aptamers to a Small Molecule Target Unit 9.6 In Vitro Selection Using Modified or Unnatural Nucleotides

Chapter 10 Purification and Analysis of Synthetic Nucleic Acids and Components Introduction Unit 10.1 Analysis of Oligonucleotides by Matrix-Assisted Laser Desorption/Ionization Time-ofFlight Mass Spectrometry 3. Unit 10.2 Analysis of Oligonucleotides by Electrospray Ionization Mass Spectrometry 4. Unit 10.3 Overview of Purification and Analysis of Synthetic Nucleic Acids 5. Unit 10.4 Polyacrylamide Gel Electrophoresis (PAGE) of Synthetic Nucleic Acids 6. Unit 10.5 Analysis and Purification of Synthetic Nucleic Acids Using HPLC 7. Unit 10.6 Base Composition Analysis of Nucleosides Using HPLC 8. Unit 10.7 Cartridge Methods for Oligonucleotide Purification 9. Unit 10.8 Analysis of Oxidized DNA Fragments by Gel Electrophoresis 10. Unit 10.9 Capillary Electrophoresis of DNA 11. Unit 10.10 Sequencing Oligonucleotides by Enrichment of Coupling Failures Using Matrix-Assisted Laser Desorption/Ionization Time-of-Flight Mass Spectrometry 12. Unit 10.11 Mass Determination of Phosphoramidites 1. 2.

Chapter 11 RNA Folding Pathways

Introduction Unit 11.1 RNA Folding Pathways Unit 11.2 RNA Secondary Structure Prediction Unit 11.3 Thermal Methods for the Analysis of RNA Folding Pathways Unit 11.4 Probing RNA Folding Pathways by RNA Fingerprinting Unit 11.5 Characterization of Tertiary Folding of RNA by Circular Dichroism and Urea Unit 11.6 Time-Resolved Hydroxyl Radical Footprinting of RNA with X-Rays Unit 11.7 Rapid Magnesium Chelation as a Method to Study Real-Time Tertiary Unfolding of RNA Unit 11.8 Use of Fluorescence Spectroscopy to Elucidate RNA Folding Pathways Unit 11.9 Use of Chemical Modification To Elucidate RNA Folding Pathways Unit 11.10 Probing RNA Structural Dynamics and Function by Fluorescence Resonance Energy Transfer (FRET) 12. Unit 11.11 Site-Specific Fluorescent Labeling of Large RNAs with Pyrene 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

Chapter 12 Nucleic Acid-Based Microarrays and Nanostructures 1. 2. 3. 4. 5. 6. 7. 8.

Introduction Unit 12.1 Key Experimental Approaches in DNA Nanotechnology Unit 12.2 Preparation of Gold Nanoparticle–DNA Conjugates Unit 12.3 Synthesis of 5′-O-Phosphoramidites with a Photolabile 3′-O-Protecting Group Unit 12.4 Derivatization of Glass and Polypropylene Surfaces Unit 12.5 DNA Microarray Preparation by Light-Controlled In Situ Synthesis Unit 12.6 Preparation of α-Oxo Semicarbazone Oligonucleotide Microarrays Unit 12.7 Synthesis of Covalent Oligonucleotide-Streptavidin Conjugates and Their Application in DNA-Directed Immobilization (DDI) of Proteins

Chapter 13 Nucleoside Phosphorylation and Related Modifications 1. 2. 3. 4. 5. 6. 7.

Introduction Unit 13.1 Overview of the Synthesis of Nucleoside Phosphates and Polyphosphates Unit 13.2 Chemoenzymatic Preparation of Nucleoside Triphosphates Unit 13.3 Synthesis and Polymerase Incorporation of 5′-Amino-2′,5′-Dideoxy-5′-N-Triphosphate Nucleotides Unit 13.4 Nucleoside-5′-Phosphoimidazolides: Reagents for Facile Synthesis of Dinucleoside Pyrophosphates Unit 13.5 Synthesis of Methylenebis(phosphonate) Analogs of Dinucleotide Pyrophosphates Unit 13.6 Chemical Phosphorylation of Deoxyribonucleosides and Thermolytic DNA Oligonucleotides

Chapter 14 Biologically Active Nucleosides 1. 2. 3. 4. 5. 6.

Introduction Unit 14.1 Synthesis of Acyclic Analogs of Adenosine Unit 14.2 Synthesis of Acyclic Nucleoside Phosphonates Unit 14.3 Synthesis of β-l-2′-Deoxythymidine (l-dT) Unit 14.4 Synthesis of Carbovir and Abacavir from a Carbocyclic Precursor Unit 14.5 Synthesis of 2′- and 3′-C-Methylribonucleosides

Chapter 15 Nucleoside Prodrugs and Delivery Strategies 1.

Introduction

2. 3. 4.

Unit 15.1 Synthesis of Amino Acid Phosphoramidate Monoesters via H-Phosphonate Intermediates Unit 15.2 Synthesis of Cidofovir and (S)-HPMPA Ether Lipid Prodrugs Unit 15.3 Chemistry of bisSATE Mononucleotide Prodrugs

Appendix 1 Standard Nomenclature, Data, and Abbreviations 1. 2. 3. 4. 5.

1A Selected Abbreviations Used in This Manual 1B Characteristics of Nucleic Acids 1C IUPAC-IUB Joint Commission on Biochemical Nomenclature Abbreviations and Symbols for the Description of Conformations of Polynucleotide Chains 1D Nucleoside and Nucleotide Nomenclature 1E A Convenient Stereochemical Notation for P-Chiral Nucleotide Analogs

Appendix 2 Laboratory Stock Solutions and Equipment 1.

2A Common Buffers and Stock Solutions

Appendix 3 Commonly Used Techniques 1. 2. 3. 4. 5.

3A References to Commonly Used Techniques 3B Denaturing Polyacrylamide Gel Electrophoresis 3C Introduction to the Synthesis and Purification of Oligonucleotides 3D Thin-Layer Chromatography 3E Column Chromatography

Appendix 4 Resources 1.

4A Useful Nucleic Acid Chemistry Web Sites

Appendix Suppliers 1.

Selected Suppliers of Reagents and Equipment

PREFACE hemistry has played a pivotal role in the development of molecular biology and biotechnology. The value of synthetic oligonucleotides became apparent when they provided the means for deciphering the genetic code. Modern cloning techniques and nucleic acid sequencing depend on the ability to obtain synthetic primers. In recent years intensive research has driven the technological development of high-throughput methods for nucleic acid analysis. Moreover, modified nucleic acids have been extensively studied and identified as highly specific therapeutic agents. At the same time, research efforts have provided new means to probe the structure of nucleic acids for a better understanding of all aspects of nucleic acid function and interactions in biology. Future applications of nucleic acids are likely to extend into material science, as nucleic acid sequences provide a code for information storage and manipulation as well as the means for self-assembly of complex devices and systems on the nanoscale.

C

The pathways by which the nucleoside building blocks of nucleic acids are biosynthesized and incorporated into nucleic acids are important targets for antiviral and chemotherapies. A large variety of modified nucleosides and nucleotides have been designed as tools for studying biochemical processes involving nucleotide binding proteins (as enzyme substrates or as cofactors), and many more have been evaluated as potential drugs. Detailed protocols for much of the early chemistry can be found in a useful series of books starting with Synthetic Procedures in Nucleic Acid Chemistry, Vols. 1 and 2 (Zorbach and Tipson, 1968) and continuing with four volumes, Nucleic Acid Chemistry: Improved and New Synthetic Procedures, Methods and Techniques, Parts I-IV (Townsend and Tipson, 1978, 1986, 1991). The current methods applied to the synthesis of modified nucleosides and nucleic acids for structure-function studies, potential therapeutic agents, and as tools for molecular biology, have spawned a unique set of chemistries that provide the fundamental basis on which this volume of Current Protocols has evolved. In Current Protocols in Nucleic Acid Chemistry, the practical aspects of innovative methods for the preparation of modified nucleosides and nucleic acids are strongly emphasized. Chapters 1 to 4 describe detailed methodology for the synthesis of both natural and modified oligonucleotides. Chapter 1 focuses on the synthesis of a variety of modified nucleosides with the emphasis on compounds that could be used as components of oligonucleotides. Various methods for N-protection of nucleobases and functionalization of 5 -, 3 -, and/or 2 -hydroxyl groups of deoxyribo- and ribonucleosides are presented in Chapter 2 along with the preparation of their H-phosphonate or phosphoramidite derivatives to enable oligonucleotide synthesis. In Chapter 3, the development of solid supports and strategies for solid-phase synthesis of native DNA or RNA oligonucleotides, through a collection of properly functionalized H-phosphonate or phosphoramidite monomers, is emphasized. Such rapid and efficient methods for automated synthesis of oligonucleotides has led to the preparation of modified oligonucleotides for therapeutic applications. Methods for altering native oligonucleotide properties through nucleobase, carbohydrate, and internucleotidic phosphodiester modifications or through conjugation with ligands or reporter groups are the focus of Chapter 4. Chapters 5, 6, and 11 provide methods for probing DNA and RNA structures, including cross-linking strategies to stabilize secondary structure, chemical and enzymatic probes of structure, and techniques for elucidating RNA folding. Two chapters focus on purification Current Protocols in Nucleic Acid Chemistry Contributed by Serge L. Beaucage, Donald E. Bergstrom, Piet Herdewijn, and Akira Matsuda Current Protocols in Nucleic Acid Chemistry (2005) iii-vii C 2005 by John Wiley & Sons, Inc. Copyright

iii Supplement 24

and analysis. Chapter 7 covers important biophysical and computational methods of nucleic acid analysis, including X-ray crystallography, NMR, and molecular modeling strategies. Mass spectrometric analysis is covered in Chapter 10 along with common purification techniques including HPLC and electrophoresis. Two chapters cover different aspects of the binding of nucleic acids to other molecules. Chapter 8 focuses on protocols for measuring binding mode and affinity of small molecules to DNA, a topic of importance because of the significant numbers of drugs that function by binding to nucleic acids. Chapter 9 includes methods for generating nucleic acid aptamers. Chapter 12 focuses on nucleic acid chemistry as applied to the preparation of oligonucleotide arrays and introduces topics related to the development of nucleic acid related nanotechnology. In Chapter 13, protocols are provided for the synthesis of various phosphorylated derivatives of nucleosides. A newly added Chapter 14 provides protocols for synthesis of modified nucleosides that play an important role in the fight against infectious diseases and cancers.

HOW TO USE THIS MANUAL Format and Organization This publication is available in looseleaf, CD-ROM, Intranet, and online formats. For looseleaf purchasers, binders are provided to accommodate the growth of the manual via the quarterly update service. The looseleaf format of the binder allows easy insertion of new pages, units, and chapters that are added. The index and table of contents are updated with each supplement. Purchasers of the CD-ROM and Intranet versions receive a completely new disc every quarter and should dispose of their outdated discs. The material covered in all formats is identical. Subjects in this manual are organized by chapters, and protocols are contained in units. Units generally describe a method and include one or more protocols with listings of materials, steps and annotations, recipes for unique reagents and solutions, and commentaries on the “hows” and “whys” of the method; there are also “overview” units containing theoretical discussions that lay the foundation for subsequent protocols. Page numbering in the looseleaf version reflects the modular arrangement by unit; for example, page 2.3.5 refers to Chapter 2 (Protection of Nucleosides for Oligonucleotide Synthesis), UNIT 2.3 (Protection of the 5 -Hydroxy Function of Nucleosides), page 5 of that particular unit. Many reagents and procedures are employed repeatedly throughout the manual. Instead of duplicating this information, cross-references among units are used extensively. Crossreferencing helps to ensure that lengthy and complex protocols are not overburdened with steps describing auxiliary procedures needed to prepare raw materials and analyze results. For some widely used techniques (such as gel electrophoresis), readers are referred to APPENDIX 3.

Preface

Introductory and Explanatory Information Because this publication is first and foremost a compilation of laboratory techniques in nucleic acid chemistry, we have not offered extensive instructive material. We have, however, included explanatory information where required to help readers gain an intuitive grasp of the procedures. Some chapters begin with special overview units that describe

iv Supplement 24

Current Protocols in Nucleic Acid Chemistry

the state of the art of the topic matter and provide a context for the procedures that follow. Chapter and unit introductions describe how the protocols that follow connect to one another, and annotations to the actual protocol steps describe what is happening as a procedure is carried out. Finally, the Commentary that closes each protocol unit describes background information regarding the historical and theoretical development of the method, as well as alternative approaches, critical parameters, troubleshooting guidelines, anticipated results, and time considerations. All units contain cited references and many indicate key references to inform users of particularly useful background reading, original descriptions, or applications of a technique.

Protocols Many units in the manual contain groups of protocols, each presented with a series of steps. The basic protocol is presented first in each unit and is generally the recommended or most universally applicable approach. Alternate protocols are given where different equipment or reagents can be employed to achieve similar ends, where the starting material requires a variation in approach, or where requirements for the end product differ from those in the basic protocol. Support protocols describe additional steps that are required to perform the basic or alternate protocols; these steps are separated from the core protocol because they might be applicable to other uses in the manual, or because they are performed in a time frame separate from the basic protocol steps. Reagents and Solutions Reagents required for a protocol are itemized in the materials list before the procedure begins. Many are common stock solutions, others are commonly used buffers or media, whereas others are solutions unique to a particular protocol. Recipes for the latter solutions are supplied in each unit under the heading Reagents and Solutions. It is important to note that the names of some of these special solutions might be similar from unit to unit (e.g., SDS sample buffer) while the recipes differ; thus, it is essential to ensure that reagents are prepared from the proper recipes. On the other hand, recipes for commonly used stock solutions and buffers are listed once in APPENDIX 2A. These universal recipes are cross-referenced parenthetically in the materials lists rather than repeated with every usage. Commercial Suppliers In some instances throughout the manual, we have recommended commercial suppliers of chemicals, biological materials, or equipment. This has been avoided wherever possible, because preference for a specific brand is subjective and is generally not based on extensive comparison testing. Our guidelines for recommending a supplier are that (1) the particular brand has actually been found to be of superior quality, or (2) the item is difficult to find in the marketplace. The purity of chemical reagents frequently varies with supplier. Generally reagent grade chemicals are preferred. Special care must be paid to procedures that require dry solvents. Different suppliers provide special anhydrous grade solvents which may vary in water content depending on the supplier. Addresses, phone numbers, facsimile numbers, and Web sites of all suppliers mentioned in this manual are provided in the SUPPLIERS APPENDIX.

Safety Considerations Anyone carrying out these protocols will encounter the following hazardous or potentially hazardous materials: (1) radioactive substances, (2) toxic chemicals and carcinogenic or


v Current Protocols in Nucleic Acid Chemistry

Supplement 24

teratogenic reagents, (3) pathogenic and infectious biological agents, and (4) recombinant DNA. Most governments regulate the use of these materials; it is essential that they be used in strict accordance with local and national regulations. Cautionary notes are included in many instances throughout the manual, but we emphasize that users must proceed with the prudence and precaution associated with good laboratory practice, and that all materials must be used in strict accordance with local and national regulations.

Reader Response The protocols included in this manual are used routinely by the authors who contribute them, and many are used in our own laboratories. To make them work for you, the authors have annotated critical steps and included critical parameters and troubleshooting guides in the commentaries to most units. However, the successful evolution of this manual depends upon readers’ observations and suggestions. Consequently, a selfmailing reader-response survey is included with each print supplement; we encourage readers to send in their comments. Electronic contact information is available at http://www3.interscience.wiley.com/c p/index.htm. ACKNOWLEDGMENTS This manual is the product of dedicated efforts by many of our scientific colleagues who are acknowledged in each unit, and by the hard work of the Current Protocols editorial staff at John Wiley & Sons, including Joe Ingram, Scott Holmes, Tom Cannon, Susan Lieberman, Allen Ranz, and Joseph White. The publisher’s commitment and continuing support for a nucleic acid chemistry manual have been essential for realizing and continuing this ambitious project. We are extremely grateful for the critical contributions made by Ann Boyle, who played a key role in bringing the initial project to completion, and to Beth Harkins, who continues to keep editors and contributors on track. Finally, we gratefully acknowledge the significant role of past editors Roger Jones and Gary Glick in developing Current Protocols in Nucleic Acid Chemistry during its first half-decade.

LITERATURE CITED Zorbach, W.W. and Tipson, R.S. (eds.) 1968. Synthetic Procedures in Nucleic Acid Chemistry, Vols. 1 and 2. Interscience Publishers, New York. Townsend, L.B. and Tipson, R.S. (eds.) 1978. Nucleic Acid Chemistry: Improved and New Synthetic Procedures, Methods and Techniques, Parts I and II. John Wiley & Sons, New York. Townsend, L.B. and Tipson, R.S. (eds.) 1986. Nucleic Acid Chemistry: Improved and New Synthetic Procedures, Methods and Techniques, Part III. John Wiley & Sons, New York. Townsend, L.B. and Tipson, R.S. (eds.) 1991. Nucleic Acid Chemistry: Improved and New Synthetic Procedures, Methods and Techniques, Part IV. John Wiley & Sons, New York.

RECOMMENDED BACKGROUND READING Blackburn, G.M. and Gait, M.G. 1996. Nucleic Acids in Chemistry Biology, 2nd ed., IRL Press, London General introduction to fundamental aspects of nucleic acid chemistry. Hecht, S.M. 1996. Bioorganic Chemistry: Nucleic Acids. Oxford University Press, New York. Barton, D., Nakanishi, K., and Meth-Cohn, O. 2000. Comprehensive Natural Products Chemistry, Vol. 7: DNA and Aspects of Molecular Biology (E.T. Kool, ed.), Elsevier Science, London. Overviews of many important topics related to nucleic acid chemistry, structure, and biochemistry. De Clercq, E. and Herdewijn, P. 2005. Strategies in the design of antiviral drugs. In Drug Discovery Handbook, pp. 1135-1190. Wiley Intersciences, Hoboken, N.J. Overview of the design and synthesis of nucleoside analog antiviral agents.

Preface

Nelson, D.L. and Cox, M.M. 2004. Lehninger Principles of Biochemistry, 4th ed. W.H. Freeman and Company, New York. Basic biochemistry text that includes fundamentals of nucleic acid biochemistry.

vi Supplement 24


Neidle, S. 1999. Oxford Handbook of Nucleic Acid Structure. Oxford University Press, Oxford. Neidle, S. 2002. Nucleic Acid Structure and Recognition, Oxford University Press, Oxford. The 1999 Oxford Handbook contains the most extensive and comprehensive coverage on nucleic acid structure. Gesteland, R.F., Cech, T.R., and Atkins, J.F. 1999. The RNA World, 2nd ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York. Reviews on many current topics in RNA biochemistry, chemistry and structure. Hartmann, R.K., Binderelf, A., Schön, A. and Westhof, E. 2005. Handbook of RNA Biochemistry. John Wiley & Sons, Hoboken, N.J. Includes current laboratory techniques and protocols. Zubrick, J.W. 2003. The Organic Chem Lab Survival Manual, A Student’s Guide to Techniques, 6th ed. John Wiley & Sons, Hoboken, N.J. Laboratory manual that describes very basic techniques for synthetic chemistry. Silverstein, R.M., Webster, F.X., and Kiemle, D. 2004. Spectrometric Identification of Organic Compounds, 7th ed. John Wiley & Sons, Hoboken, N.J. Explores the identification of organic compounds by mass spectrometry, infrared spectrometry, and nuclear magnetic resonance spectrometry. Lee, T.A. 1998. A Beginner’s Guide to Mass Spectral Interpretation. John Wiley & Sons, New York. Identification of organic compound by mass spectrometry, emphasizing recognition of typical fragmentation patterns for different types of organic compounds. Macomber, R.S. 1998. A Complete Introduction to Modern NMR Spectroscopy. John Wiley & Sons, New York. Clear, accessible coverage of modern NMR spectroscopy methods.

Serge L. Beaucage, Donald E. Bergstrom, Piet Herdewijn, and Akira Matsuda


vii Current Protocols in Nucleic Acid Chemistry

Supplement 24

Foreword Nucleic acid chemistry has developed at an accelerating pace in recent years. Today one can rapidly synthesize long DNA and RNA strands of any desired nucleotide sequence on a multigram to microgram scale. “DNA chips” containing thousands of different synthetic oligonucleotides, each at a unique, defined site on a 1 × 1–cm glass plate, can also be made. Elegant methods are available to prepare and incorporate into oligonucleotides a great variety of monomers containing tailored bases, sugars, and bridging units. Powerful analytical techniques, including high performance liquid chromatography, capillary gel electrophoresis, UV and laser spectroscopy, and mass spectrometry, enable rapid separation and characterization of polynucleotides. Advances in NMR spectroscopy and X-ray crystallography have broadened the horizons for structural elucidation, and sophisticated computational programs can now generate useful theoretical models to guide research. These and related developments provide current researchers with unprecedented problem-solving power and opportunities for discovery and invention. Yet the very breadth of the knowledge and the pace of innovation pose significant challenges for serious investigators seeking to capitalize on these advances. This manual, Current Protocols in Nucleic Acid Chemistry, brings together in one volume state-of-the-art protocols for the synthesis, isolation, characterization, modification, and design of a wide variety of compounds in the nucleic acid family. It will prove very useful for experts and practitioners working in the nucleic acid field. An attractive additional feature is a series of quarterly supplements that will describe new and improved methodologies as they emerge. These will be incorporated into the existing manual. Oligonucleotides and related derivatives are unique recognition molecules. They bind reversibly with high selectivity to polymers containing complementary nucleotide sequences, and key rules governing formation of the complexes are well established. Since very large numbers of different oligomers can be generated from different combinations of the four main building blocks, oligonucleotide chemistry offers unique opportunities for rationally designing a wide variety of self-assembling systems exhibiting predesignated properties. Increasingly, scientists in other disciplines are using synthetic oligonucleotides, modified oligonucleotides, or oligonucleotide conjugates as tools in their research. The impact of the chemistry has been greatest in molecular biology, where synthetic oligonucleotides now play an essential role as primers in sequencing and amplifying DNA, as models and probes for elucidating the structure and function of DNA and RNA, and as building units for the synthesis and modification of genes. Applications of oligonucleotide conjugates as nonradioactive probes in diagnostic medicine are expanding rapidly; a variety of modified oligonucleotides show promise as highly selective therapeutic agents; organic and inorganic chemists are finding oligonucleotides useful as tethers in positioning functional groups in complex assemblies; and an active new research area in materials science is directed toward creating novel materials from nanoparticle-oligonucleotide conjugates. Some visionaries even dream of employing oligonucleotides in building molecular-scale machines and computers. For the researchers from other disciplines who wish to utilize nucleic acid chemistry and for those working at the frontiers of several disciplines, this manual will be a special boon. The comprehensive collection of concise, explicit protocols will enable them to capitalize quickly on the opportunities opened by the advances in nucleic acid chemistry.

CHAPTER 1 Synthesis of Modified Nucleosides INTRODUCTION odified nucleosides have a wide variety of uses—often after incorporation into oligonucleotides—such as attachment to reporter groups and studies of basepairing altered by mutagens, by carcinogens, or by design (Beaucage and Iyer, 1993). This chapter will focus on sugar- and base-modified nucleosides. In keeping with Current Protocols format, the goal of this chapter is not to review the subject, but to provide a set of specific and detailed procedures leading to specific compounds. Within this framework, however, the procedures should enable researchers to undertake syntheses of related compounds that have been reported in the chemical literature.

M

The most widely used reaction in nucleoside chemistry is the sugar-base condensation reaction using the Vorbrüggen procedure. This reaction is used for synthesis of sugarmodified as well as base-modified nucleosides. It is a very elegant way to introduce heterocyclic bases at the anomeric center of sugars in a stereoselective manner (in case a 2 -O-acyl group is present for neighboring group participation). This reaction is described in UNIT 1.13. In a closely related unit, the synthesis of 2 -O-β-D-ribofuranosyl nucleosides are described in UNIT 1.14. The same principle is used here for the Lewis acid– catalyzed O-glycosylation reaction, in this case leading to disaccharide nucleosides. This procedure demonstrates the reaction conditions needed for the stereospecific formation of O-glycosidic bonds in nucleosides (with base moieties attached at the anomeric center). Disaccharide nucleosides occur widely in nature, especially as one of the modified nucleosides in tRNA. In some cases, modified nucleosides may themselves serve as key intermediates from which other modified nucleosides of interest can be prepared. For example, the palladiummediated C5 substitution of pyrimidines pioneered by Bergstrom has proved over the years to be a uniquely valuable route to these important compounds (Goodchild, 1990). An understanding of the synthetic procedure detailed in UNIT 1.1 will provide access to a large number of C5-modified pyrimidines. Similarly, UNIT 1.8 describes the synthesis and use of aminoalkyl-modified purine analogs. The palladium-catalyzed cross-coupling reaction for C-C bond formation is also used extensively in the purine field, in efforts to generate bioactive compounds. UNIT 1.16 describes the reaction of 6-chloropurine nucleosides with several organometallic reagents using palladium as catalyst, and these reactions are representative for the field. In addition, the 6-hydroxymethyl congener is used as example for further derivatization reactions, mainly for preparation of fluorinated nucleosides. The enzymatic coupling methods in UNITS 1.2 & 1.6 use enzymatic transglycosylation for synthesis of 2 -deoxyribonucleosides and ribonucleosides, respectively. Although nucleoside 2 -deoxyribosyltransferase is not commercially available at this time, the alternative enzymes (thymidine phosphorylase and purine nucleoside phosphorylase) are available. These routes then provide access to a wide variety of both 2 -deoxyribo- and ribonucleosides. UNIT 1.11 describes the widely used Robins procedure for chemically converting ribonucleosides to 2 -deoxyribonucleosides. Modified nucleosides containing reactive functionality, which have been denoted as “convertible” nucleosides when incorporated into oligonucleotides, are an increasingly

Current Protocols in Nucleic Acid Chemistry 1.0.1-1.0.3, September 2007 Published online September 2007 in Wiley Interscience (www.interscience.wiley.com). DOI: 10.1002/0471142700.nc0100s30 C 2007 John Wiley & Sons, Inc. Copyright

Synthesis of Modified Nucleosides

1.0.1 Supplement 30

important class of compounds, of which the 2-fluoro-2 -deoxyinosine derivative described in UNIT 1.3 is a timely example (Huang et al., 1999). UNIT 1.4 provides a discussion on the topic of unnatural nucleotides with unusual base-pairing properties and of universal nucleotides. UNIT 1.5 continues this topic by providing step-by-step procedures for synthesizing N- or C-nucleosides with a number of specific base analogs. Introduction of a fluorine atom in nucleosides has often led to compounds with a potent biological activity. This was demonstrated in the anti-HIV field, and UNIT 1.12 describes the example of 2 -fluoro-2 ,3 dideoxyadenosine. UNIT 1.7 describes the synthesis of 2 -deoxy-2 -fluoroarabinonucleosides. Oligonucleotides containing these modified bases form stable heteroduplexes with RNA that are substrates for RNase H, which is important to antisense applications. UNIT 1.9 addresses the use of nucleosides with a modified sugar to create hexitol nucleic acids (HNAs). This unit decribes the synthesis of 1,5-anhydrohexitol nucleoside monomers as well as their starting sugar compound. These can be used in automated oligonucleotide synthesis with standard phosphoramidite chemistry to synthesize HNAs, which hybridize to RNA and DNA in a sequence-specific manner and have applications in antisense and antiviral approaches. The altritol building blocks, described in UNIT 1.18, differ from the hexitol building blocks by the presence of an additional hydroxyl group on the sugar moiety. Therefore, altritol nucleic acids (ANAs) can be considered the RNA analogs of HNA. The hybridization properties of HNA and ANA are very similar. ANA can also be used for antisense and siRNA applications, and show excellent hybridization properties when used on arrays. UNIT 1.10 describes preparation of pyrrolo[2,3-d]pyrimidine analogs of 2 -deoxyadenosine, 2 -deoxyguanosine, and 2 -deoxyisoguanosine. The incorporation of these compounds into DNA oligomers is described in UNIT 4.25.

A problem encountered in nucleic acid crystallography is the phasing of data, and the use of selenium-labeled DNA and RNA has proven to be very helpful in this process. The preparation of 2 -methylseleno nucleosides and their corresponding phosphoramidites are described in UNIT 1.15. Procedures are also given for synthesis and purification of the modified oligonucleotides as well as their ligation using T4 RNA/DNA ligase. For molecular diagnostics and gene regulation, the development of caged oligonucleotides provides a useful application of nucleoside modification. The principle is that a biologically active oligonucleotide is made inactive by derivatizing some of its nucleobases with a photolabile group, which is the cage. The active oligomer can be released using laser (light) at the desired time and location. These tools can also be used for physicochemical studies (for example, to study oligonucleotide folding) and for biological purposes (for example, for the modulation of protein function using aptamers). UNIT 1.17 describes the synthesis of phosphoramidites derivatized with a 2-nitrophenylethyl or 2-nitrophenylpropyl cage; these provide building blocks for synthesis of caged oligonucleotides on an automated synthesizer. Several other chapters in this book describe specific types of nucleoside modifications or modifications for specific purposes. Chapter 2, for instance, describes nucleoside protection, which is essentially the transient modification of a nucleoside moeity to allow selective modification of other moeities during subsequent reactions. Chapter 13 describes nucleoside phosphorylation and related modifications, and Chapters 14 and 15 describe modifications designed for specific biological functionality and for delivery of nucleosides in a prodrug form.

Introduction

1.0.2 Supplement 30


LITERATURE CITED Beaucage, S.L. and Iyer, R.P. 1993. The synthesis of modified oligonucleotides by the phosphoramidite approach and their applications. Tetrahedron 49:6123-6194. Goodchild, J. 1990. Conjugates of olignucleotides and modified oligonucleotides: A review of their synthesis and properties. Bioconjugate Chem. 1:165-187. Huang, H., Chopra, R., Verdine, G.L., and Harrison, S.C. 1999. Structure of a covalently trapped catalytic complex of HIV-1 reverse transcriptase: Implications for drug resistance. Science 282:1669-1675.

Roger Jones and Piet Herdewijn

Synthesis of Modified Nucleosides

1.0.3 Current Protocols in Nucleic Acid Chemistry

Supplement 30

Palladium-Mediated C5 Substitution of Pyrimidine Nucleosides

UNIT 1.1

One of the most efficient ways to link a reporter group to oligonucleotides is through the incorporation of a modified nucleoside during automated oligonucleotide synthesis. Most techniques, which make use of synthetic oligonucleotides, function by hybridization to a complementary sequence. In order to avoid interference with hybridization, reporter groups should ideally be attached so that they do not interfere with hybridization or destabilize dsDNA. Two different types of tethers are described here—a rigid amidopropynyl linker and a flexible aminoethylthioether linker. The rigid amidopropynyl tether, linked through C5 of deoxyuridine, is sufficiently long and positioned such that a reporter group attached at the distal end lies outside the major groove of a DNA duplex. Basic Protocol 1 describes a detailed procedure for the synthesis of one example of deoxyuridine modified by an amidopropynyl-linked reporter group, 5-(3-nicotinamidopropyn-1-yl)-5′-O-(4,4′-dimethoxytrityl)-2′-deoxyuridine (Fig. 1.1.1). The procedure is general and may be applied to other amidopropynyl-linked functional groups. The nicotinoyl group was used only as an illustration of the strategy for incorporating a functional group on the amimopropynyl tether. For use in oligonucleotide synthesis, the C5-modified deoxyuridine is converted to a 3′-phosphoramite as described in UNIT 3.3. Basic Protocol 2 outlines the synthesis of 5-(3-acetamido-1-thiapropyl)-2′-deoxyuridine (Fig. 1.1.2). In contrast to the amidopropynyl tether, the more conformationally flexible thioether tether was designed to allow positioning of a molecular tool (e.g., chemical cleavage reagent or cross-linking reagent) on a complementary nucleic acid by hybridization of the modified oligonucleotide. The thiapropyl linker is capable of bridging the

O

O l HO

I

NH O

N

4,4'-dimethoxytrityl chloride

O

DMTrO

NH O

N

O

pyridine HO

HO 5-iodo-2'-deoxyuridine

5'-O-(4,4'-dimethoxytrityl)-5-iodo-2'-deoxyuridine

NH2

O

O

propargylamine

Cl N H Cl

N H

nicotinoyl chloride hydrochloride

Pd(PPh3)4 Cul

N

pyridine

N-(3-propyn-1-yl)nicotinamide

O

OCH3

O

N H

NH

N DMTrO

DMTr = H3CO

O

N

O

HO 4,4'-dimethoxytrityl

5-(3-nicotinamidopropyn-1-yl)-5'-O-(4,4'-dimethoxytrityl)-2'-deoxyuridine

Figure 1.1.1 Synthetic scheme for the preparation of 5-(3-nicotinamidopropyn-1-yl)-5′-O-(4,4′-dimethoxytrityl)-2′-deoxyuridine from 5-iodo-2′-deoxyuridine. The structure of 4,4′-dimethoxytrityl (DMTr) is shown in the lower left. Contributed by Mohammad Ahmadian, Douglas A. Klewer, and Donald E. Bergstrom Current Protocols in Nucleic Acid Chemistry (2000) 1.1.1-1.1.18 Copyright © 2000 by John Wiley & Sons, Inc.

Synthesis of Modified Nucleosides

1.1.1

span between two helices. For use in oligodeoxyribonucleotide synthesis, the N-acylated 5-(3-amino-1-thiapropyl)-2′-deoxyuridine is transformed to the 5′-dimethoxytrityl (DMTr) derivative as illustrated in Basic Protocol 1 for 5-iodo-2′-deoxyuridine, and is then converted to the 3′-phosphoramite as described in UNIT 3.3. Support Protocols 1 and 2 describe the preparation of two reagents needed for Basic Protocol 2—N,N′-bis(trifluoroacetyl)cystamine and N-acetoxysuccinimide, respectively. BASIC PROTOCOL 1

SYNTHESIS OF 5-(3-NICOTINAMIDOPROPYN-1-YL)-5′-O-(4,4′DIMETHOXYTRITYL)-2′-DEOXYURIDINE The sequence of reactions outlined here (see Fig. 1.1.1) illustrate conditions that are useful for the synthesis of a wide variety of reporter groups linked through C5 of deoxyuridine. The protocol includes three steps: synthesis of the N-acylated 3-aminopropyne (3-nicotinamidopropyne, in this example), reaction of 5-iodo-2′-deoxyuridine with 4,4′-dimethoxytrityl chloride, and palladium-catalyzed coupling of 3-nicotinamidopropyne with 5′-O-(4,4′-dimethoxytrityl)-5-iodo-2′-deoxyuridine. For the introduction of a different reporter group, 3-aminopropyne can be N-acylated by RCOCl (an acid chloride) or RC(O)OC(O)R (an anhydride) to obtain RC(O)NHCH2C≡CH, in which R is the desired reporter group. CAUTION: All reactions should be run in a suitable fume hood to avoid inhalation of toxic vapors. Materials Nicotinoyl chloride hydrochloride Pyridine, anhydrous Nitrogen (N2) stream Triethylamine, freshly distilled (dried and purified by distillation at atmospheric pressure over calcium hydride; boiling point = 89° to 90°C) Propargylamine, reagent grade (typically 99% pure) Dichloromethane, reagent grade 10% (w/v) hydrochloric acid in water

O

O

O NH

N

dR

(CF3CONHCH2CH2S)2

NH

2. NaCl

O

N

ClHg

1. Hg(OAc)2

F3C

Li2PdCl4

O

O N H

S

dR

2'-deoxyuridine

NH O

N dR

5-[3-(trifluoroacetamido)-1-thiapropyl]-2'-deoxyuridine

5-chloromercurio-2'-deoxyuridine

O DCC

O N-OH

NH4OH

CH3COH

+

acetic acid

O N-hydroxysuccinamide

O

O

N O O HO

O

dR = HO

H3C

O

O N H

O

S

H2N

NH

S

NH

N-succinimidylacetate N

O

dR

N

O

dR

deoxyribosyl

PalladiumMediated C5 Substitution of Pyrimidine Nucleosides

5-(3-acetamido-1-thiapropyl)-2'-deoxyuridine

5-(3-amino-1-thiapropyl)-2'-deoxyuridine

Figure 1.1.2 Synthetic scheme for the preparation of 5-(3-acetamido-1-thiapropyl)-2′-deoxyuridine from 2′-deoxyuridine. The structure of deoxyribosyl (dR) is shown in the lower left. DCC, dicyclohexylcarbodiimide.


Sodium sulfate, anhydrous Silica gel (230 to 400 mesh) Methanol, reagent grade 5-Iodo-2′-deoxyuridine 4,4′-Dimethoxytrityl chloride Diethyl ether, anhydrous N,N-Dimethylformamide, anhydrous Argon gas (optional) Tetrakis(triphenylphosphine)palladium, [(C6H5)3P]4Pd Copper(I) iodide 5% (w/v) Na2EDTA in water Ethyl acetate, reagent grade 25- and 50-mL round-bottom flasks Inert atmosphere/vacuum manifold (see Fig. 1.1.3) 500-µL and 1-mL syringes with stainless steel needles 125- and 250-mL Ehrlenmeyer flask 100-mL separatory funnel Filter funnel and Whatman no. 1 filter paper Chromatotron and radial chromatography plate coated with silica gel (2-mm thickness; Harrison Research) Rotary evaporator with vacuum pump and water aspirator Glass column (2-cm i.d. × ≥20-cm length) with stopcock Additional reagents and equipment for thin-layer chromatography (TLC; APPENDIX 4D) Synthesize N-(3-propyn-1-yl)nicotinamide 1. In a dry 25-mL round-bottom flask containing a 1⁄2-in. magnetic stir bar, add 1.068 g nicotinoyl chloride hydrochloride (6 mmol) to 10 mL of anhydrous pyridine under a nitrogen stream. It is important that the flask be dry because nicotinoyl chloride reacts with water to give nicotinic acid. Glassware can be effectively dried by heating in a drying oven at 120°C for 2 hr. A small magnetic stir bar is generally dried at the same time as the flask and added to the flask prior to addition of the reagents. A general setup for running small-scale reactions under a dry nitrogen atmosphere is shown in Figure 1.1.3. The apparatus is configured for air-sensitive palladium-catalyzed reactions. For most other reactions, it is not necessary to bubble nitrogen through the reaction mixture at inlet (b).

2. Add 500 µL triethylamine (700 mg, 7 mmol) with a 1-mL syringe and stainless steel needle, and stir the mixture on a magnetic stirrer at room temperature until the triethylamine is completely in solution. CAUTION: Wear reagent-impermeable protective gloves. Triethylamine and propargylamine are corrosive.

3. Add 250 µL propargylamine (365 mg, 6.6 mmol) dropwise to the reaction mixture with a 500-µL syringe and stainless steel needle, and continue stirring under nitrogen at room temperature for 4 hr. 4. Transfer contents of the flask to a 125-mL Ehrlenmeyer flask containing 40 mL water. Stir the mixture briefly, transfer to a 100-mL separatory funnel, and extract three times with 40 mL reagent-grade dichloromethane.



Figure 1.1.3 Inert atmosphere/vacuum manifold setup for running reactions in a dry, oxygen-free atmosphere. As shown, the inert gas can be introduced into the reaction flask through stopcock (a) and a hypodermic needle inserted at (b; 14/20 standard taper joint). The slow stream of inert gas then passes through stopcock (d) and out through the gas bubbler. Care must be taken to avoid completely closing the system while the insert gas is being introduced under pressure through the manifold. The setup requires a source of dry nitrogen. For very oxygen-sensitive reactions, the solution is purged by bubbling the inert gas (nitrogen or argon) directly through the solution by lowering the stainless steal hypodermic needle into the solution. The needle is then pulled up above the level of the solution and the flask and condenser evacuated through stopcock (a) with stopcock (d) positioned to allow only the inert gas to pass through to the gas bubbler.

5. Wash the combined organic extracts twice with 20 mL of 10% hydrochloric acid and then once with 10 mL water. 6. Transfer the combined dichloromethane solution to a 250-mL Erlenmeyer flask and add 0.5 g anhydrous sodium sulfate. Swirl the solution for a few minutes and then allow to stand for 30 min. 7. Remove the drying agent by gravity filtration through a filter funnel fitted with Whatman no. 1 filter paper. PalladiumMediated C5 Substitution of Pyrimidine Nucleosides

8. Wash the solid with 10 mL dichloromethane and remove solvent under reduced pressure using a rotary evaporator and water aspirator at room temperature to obtain the crude product.


With a water aspirator and a water bath temperature of 25°C, the dichloromethane can generally be completely removed within 30 min. Longer periods of time may be required for complete removal of other higher-boiling-temperature solvents, such as methanol (step 9).

9. Purify the crude product by radial chromatography according to manufacturer’s instructions using a chromatotron plate with 2-mm silica gel thickness. Elute with 96:4 (v/v) dichloromethane/methanol and collect the effluent in ∼10-mL fractions. The silica gel plates may be purchased from the manufacturer or prepared according to the manufacturer’s instructions. Alternatively, the crude product can be purified by column chromatography on silica gel (230 to 400 mesh; 12 × 2 cm) and eluted with the same solvent to give ∼10-mL fractions.

10. Analyze fractions by thin-layer chromatrography (TLC) on silica gel. Develop TLC plates with 96:4 (v/v) dichloromethane/methanol. 11. Combine all fractions that contain the desired product (Rf = 0.26). Evaporate the solvent under reduced pressure (see step 8) to obtain N-(3-propyn-1-yl)nicotinamide (784 mg, 78%) as a white solid. The compound is stable at room temperature and can be stored in a capped amber glass vial that has been purged with nitrogen. The authors do not generally store the product for >1 month, but it may be stable for a longer period of time. Unless otherwise specified, all intermediates synthesized as part of this protocol are stored under these conditions.

12. Analyze the product by mass spectrometry (MS) and by proton and carbon nuclear magnetic resonance (NMR) spectroscopy. N-(3-Propyn-1-yl)nicotinamide has the following spectroscopic characteristics: MS-EI 160 (M+), 106, 78 MS-CI 161 (M + H)+ H NMR 250 MHz (CHCl3-d1) d 9.0 (d, J = 1.6 Hz, H-2 aromatic, 1H), 8.74 (m, H-6, 1H), 8.16 (m, H-5, 1H), 7.41 (m, H-4, 1H), 6.79 (br s, H-N, 1H), 4.28 (dd, J1 = 5.2 Hz, J2 = 2.5 Hz, N-CH2-, 2H), 2.32 (t, J = 2.5 Hz, CCH, 1H)

1

C NMR 62.9 MHz (CHCl3-d1) δ 29.90, 72.32, 79.0, 123.62, 135.26, 147.94, 152.61

13

Analysis calculated for C9H8N2O: C, 67.5; H, 5.0; N, 17.5; observed: C, 67.24; H, 4.82; N, 17.65. All values are given as percentages. The same procedure may be used for the synthesis of other amide derivatives of propargylamine from carboxylic acid chlorides or anhydrides. The Rf and the spectral characteristics will differ depending on the nature of the acyl group.

Synthesize 5′-O-(4,4′-dimethoxytrityl)-5-iodo-2′-deoxyuridine 13. In a 50-mL round-bottom flask containing a 3⁄4-in. egg-shaped magnetic stir bar, dissolve 354 mg of 5-iodo-2′-deoxyuridine (1 mmol) in 10 mL anhydrous pyridine. CAUTION: The reaction must be performed in a well-vented fume hood. This reaction is sensitive to water, and anhydrous solvent(s) must be used under inert atmosphere. Anhydrous pyridine obtained in Sure/Seal bottles (e.g., Aldrich) is suitable for use in this reaction without further drying. Otherwise, the pyridine should be dried over solid KOH and distilled over Linde type 5Å molecular sieves and solid KOH. Pyridine has a fairly high boiling point (115°C).

14. Evaporate approximately half the solvent using a rotary evaporator connected to a vacuum pump.



It is advisable to use a vacuum pump rather than a water aspirator in order to rapidly evaporate the pyridine. This procedure removes water from the reaction mixture that may have been associated with the nucleoside by way of a pyridine-water azeotrope. With a good vacuum (5 years.

8. Analyze the product by IR, UV, and 1H NMR spectroscopy. 5-Chloromercurio-2′-deoxyuridine has the following spectroscopic characteristics:



H NMR (1.0 M KCN/D2O) d 7.70 (s, 1H), 6.35 (t, 1H, J = 6.5 Hz), 4.47 (m, 1H), 3.99 (m, 1H), 3.83 (m, 2H), 2.36 (2H, dd, J = 6 Hz) 1

IR (KBr) 3365, 1714, 1642, 1440, 1275, 1089, 1040 cm −1 UV (pH 1.0) lmax 266 (ε = 10,440),lmin 239 (ε = 4,180); (pH 9.0)lmax 266 (e = 10,120), lmin 242 (e = 5,440); (pH 12.3) lmax 267 (e = 8,870),lmin 253 (e = 7,070) Analysis calculated for C9H11N2O5HgCl: C 23.34; H 2.39; N 6.05; observed: C 23.54; H 2.32; N 5.89.

Synthesize 5-[3-(trifluoroacetamido)-1-thiapropyl]-2′-deoxyuridine 9. Grind 5-chloromercurio-2′-deoxyuridine to a fine powder using a mortar and pestle. CAUTION: In the process of grinding, the powder tends to accumulate static electricity and may be difficult to contain in the mortar and pestle. The mortar should be placed on a surface that can be easily cleaned. It is preferable to carry out this step in a hood to avoid breathing the powder. Grind ≥10% more material than required for the subsequent step to make up for loss that occurs during this process.

10. Place 0.926 g finely ground 5-chloromercurio-2′-deoxyuridine (2.0 mmol) and 1.720 g N,N′-bis(trifluoroacetyl)cystamine (5 mmol) in a 100-mL round-bottom flask. 11. Add 40 mL of 0.1 M Li2PdCl4 solution to the flask and stir on a magnetic stir plate for 16 hr at ambient temperature. The mixture turns orange to yellow shortly after the reagents are combined, and usually yields a clear orange-yellow solution within a few hours.

12. Rapidly bubble hydrogen sulfide through the solution for 30 sec. Filter the mixture through a filter funnel by gravity filtration through Whatman no. 1 filter paper, and wash the solid with 20 mL reagent-grade methanol. CAUTION: Hydrogen sulfide gas is highly toxic. All operations should be conducted in a well-ventilated fume hood. Hydrogen sulfide may be obtained in 1⁄2-lb (227-g) lecture bottles.

13. Using a rotary evaporator with a water aspirator, evaporate the solvent from the filtrate under reduced pressure to give an oil. Because the methanol solution still contains hydrogen sulfide gas, this evaporation should be done using a rotary evaporator located inside a fume hood. Water aspirator pressure is normally sufficient to remove the methanol within 30 min at room temperature.

14. Purify the crude product on a 12 × 2–cm, 230- to 400-mesh silica gel column, eluting with a linear chloroform/methanol gradient ranging from 10% to 18% methanol. 15. Combine fractions that contain material with Rf = 0.30 (CH3OH-CHCl3 1:9 v/v) or Rf = 0.71 (CH3OH-CHCl3 1:3 v/v), and evaporate the solvent on a rotary evaporator using a water aspirator to obtain 5-[3-(trifluoroacetamido)-1-thiapropyl]-2′-deoxyuridine (0.41 g; 51% yield). 5-[3-(Trifluoroacetamido)-1-thiapropyl]-2′-deoxyuridine is neither water nor air sensitive. It can be stored without decomposition in amber bottles at room temperature for many years.

16. Analyze the product by MS and by IR, 1H, and 13C NMR spectroscopy. PalladiumMediated C5 Substitution of Pyrimidine Nucleosides

5-[3-(trifluoroacetamido)-1-thiapropyl]-2′-deoxyuridine has the following spectroscopic characteristics: MS-FAB m/z calculated for C13H16F3N3O6S: 399.071; observed: 400.079 (M + H)+


H NMR 300 MHz (CH3OH-d4) d 8.32 (s, 1H, H-6), 6.27 (t, 1H, J = 6.6 Hz, H-1′), 4.43 (m, 1H, H-3′), 3.95 (m, 1H, H-4′), 3.79 (m, 2H, H-5′), 3.47 (t, 2H, J = 6.0 Hz, H-3′′), 2.87 (m, 2H, H-2′), 2.32 (t, 2H, J = 6.0 Hz, H-2′′) 1

13

C NMR 125 MHz (CH3OH-d4) 164.5 (C4), 159.0 (q, J = 286.6 Hz, C5′′), 106.8 (C5), 89.0 (C1′), 86.8 (C4′), 72.0 (C3′), 62.7 (C5′), 41.4 (C2′), 39.7 (C3′′), 33.5 (C2′′) IR (KBr): 3550-2900 (br, O-H), 3423, 3443 (N-H), 1723, 1692, 1651 (C=O), 1660, 1553, 1461 (C=C), 1179, 1271 cm−1 Analysis calculated for C13H16F3N3O6S: C, 39.08; H, 4.01; N, 10.52; S, 8.03; observed: C, 39.35; H, 3.63; N, 10.38; S, 8.22.

Remove trifluoroacetyl protecting group 17. In a 100-mL round-bottom flask containing a 1-in. magnetic stir bar, dissolve 615 mg of 5-[3-(trifluoroacetamido)-1-thiapropyl]-2′-deoxyuridine (1.5 mmol) in 10 mL methanol. 18. Add 30 mL concentrated ammonium hydroxide, cap the reaction container, and stir at room temperature for 16 hr. 19. Remove excess ammonia and methanol under reduced pressure using a rotary evaporator with a water aspirator. Freeze the remainder of the solution with a mixture of dry ice and acetone. The methanol and ammonia can generally be removed on a rotary evaporator in 30 min or less at room temperature.

20. Remove water by lyophilization and dissolve the remaining solid in 3 mL anhydrous ethanol. Synthesize 5-(3-acetamido-1-thiapropyl)-2′-deoxyuridine 21. Prepare a solution of 315 mg N-acetoxysuccinimide (2 mmol) and 350 µL freshly distilled triethylamine (2.5 mmol) in 2 mL anhydrous tetrahydrofuran. Add to ethanolic solution and stir at room temperature for 4 hr. Other active esters may be used in place of N-acetoxysuccinimide to place a different functional group on the tether. The resulting chromatographic characteristics and spectroscopic properties will change accordingly.

22. Remove the solvent under reduced pressure using a rotary evaporator and a water aspirator to obtain the crude product. 23. Purify the crude product by column chromatography using a 12 × 2–cm silica gel column, and eluting with a gradient of 100% ethyl acetate to 80:20 (v/v) ethyl acetate/methanol. 24. Collect effluent in ∼10-mL fractions and analyze by TLC on silica gel, developing the plates with 85:15 (v/v) ethyl acetate/methanol. 25. Combine all fractions that contain the desired product (Rf = 0.32). Evaporate the solvent under reduced pressure using a rotary evaporator and a water aspirator to obtain 5-[3-acetamido-1-thiapropyl]-2′-deoxyuridine (462 mg; 87% yield) as a white solid. The product may be stored indefinitely under nitrogen in an amber bottle.

26. Analyze the product by MS and by UV, 1H, and 13C NMR spectroscopy. 5-(3-Acetamido-1-thiapropyl)-2′-deoxyuridine has the following spectroscopic characteristics:



MS-FAB m/z calculated for C13H19N3O6S: 345.1073; observed: 346.1072 (M + H)+ H NMR 300 MHz (CH3OH-d4) d 8.33 (s, H-6, 1H), 6.25 (t, J = 7 Hz, H-1′), 4.41 (m, H-3′, 1H), 3.93 (m, H-4′, 1H), 3.78 (m, H-5′, 2H), 3.31 (m, SCH2-, 2H), 2.78 (t, J = 6 Hz, -CH2N-, 2H), 2.29 (m, H-2′, 2H), 1.95 (s, acetyl group’s CH3, 3H) 1

13

C NMR 125 MHz (DMSO-d6) 169.3, 161.65, 150.0, 142.55, 106.8, 87.54, 84.58, 70.23, 61.1, 39.94, 37.88, 32.1, 22.6 UV (methanol) lmax: 282.4, 202.0 nm Analysis calculated for C13H19N3O6S: C, 45.2; H, 5.5; N, 12.2; S, 9.3; observed: C, 44.88; H, 5.37; N, 12.27; S, 9.32.

SUPPORT PROTOCOL 1

SYNTHESIS OF N,N′-BIS(TRIFLUOROACETYL)CYSTAMINE The trifluoroacetyl group has found wide application as a base-sensitive protecting group. It can be removed by ammonia under the same conditions used to deprotect oligodeoxyribonucleotides following synthesis by the phosphoramidite method. Amines are most commonly converted to trifluoracetyl derivatives by treatment with trifluoroacetic anhydride and a tertiary amine. CAUTION: Both triethylamine and trifluoroacetic anhydride are corrosive. Wear gloves and work only in a suitable hood. Materials Chloroform, reagent grade Cystamine dihydrochloride Triethylamine, freshly distilled (dried and purified by distillation at atmospheric pressure over calcium hydride; boiling point = 89° to 90°C) Trifluoroacetic anhydride 10% (w/v) NaHCO3 2 N HCl Sodium sulfate, anhydrous Methanol, reagent grade Ethyl acetate, reagent grade Hexane, reagent grade 1-liter round-bottom flask Drying tube containing Drierite 5-mL syringe 1-liter separatory funnel Rotary evaporator with water aspirator Vacuum oven at 35°C Buchner funnel and Whatman no. 1 filter paper Synthesize N,N′-bis(trifluoroacetyl)cystamine 1. Filter reagent-grade chloroform through a short column of basic alumina and add 500 mL to a 1-liter round-bottom flask containing a magnetic stir bar and capped by a drying tube containing Drierite. Reagent-grade chloroform typically contains ethanol to inhibit decomposition, which produces HCl. The basic alumina removes the ethanol.


2. Add 9.0 g cystamine dihydrochloride (40 mmol) and 20.2 g triethylamine (27.8 mL, 0.2 mol) to the filtered chloroform.


3. Cool the flask in a cold water bath (0° to 5°C) while slowly adding 1.5 mL trifluoroacetic anhydride (18.5 g, 88 mmol) with a 5-mL syringe. Trifluoroacetic anhydride, which hydrolyzes to trifluoroacetic acid when exposed to water, should be stored in a hood and protected from moisture.

4. Stir the reaction mixture at room temperature overnight. 5. Transfer the reaction mixture to a 1-liter separatory funnel and wash sequentially with 250 mL water, 10% NaHCO3, 2 N HCl, and water again. 6. Dry the chloroform solution over anhydrous sodium sulfate for 1 hr (see Basic Protocol 1, step 6). 7. Add 30 mL reagent-grade methanol, filter to remove the solid sodium sulfate (see Basic Protocol 1, step 7), and evaporate the solvent on a rotary evaporator using a water aspirator to yield a light yellow paste. 8. Add 100 mL of 2 N HCl to give a slurry of light yellow crystals, and stir for 20 min at room temperature. 9. Filter the slurry and wash the solid product with 100 mL of 2 N HCl and then with 100 mL water. 10. Dry the solid product in a vacuum oven at 35°C overnight. The product is of suitable purity for use in palladium coupling reactions. To obtain analytically pure product, recrystallize as described below. The product may be stored indefinitely in a closed bottle at room temperature.

Recrystallize product 11. Dissolve product in a minimum of hot methanol. 12. Slowly add ethyl acetate to the warm methanol solution and allow to cool. On cooling to room temperature, a colorless crystalline product separates (melting point = 111°C).

13. Collect crystals by vacuum filtration through a Buchner funnel using Whatman no. 1 filter paper. 14. Collect a second crop by adding hexane to the warmed methanol/ethyl acetate solution, and cooling and collecting crystals again. The total yield of crystalline product is 10.8 g (84%). The product may be stored indefinitely in a closed bottle at room temperature.

Analyze product 15. Analyze the product by 1H and 13C NMR spectroscopy. H NMR (250 MHz, acetone-d6) d 2.70 (t, CH2 4H, J = 10.4 Hz), 2.68 (q, CH2 4H)

1

C NMR (62.9 MHz, acetone-d6) d 37.2 (CH2, J = 10.4 Hz), 39.6 (CH2), 117.0 (CF3, J = 287 Hz), 157.8 (C=O).

13

SYNTHESIS OF N-ACETOXYSUCCINIMIDE N-Hydroxysuccinamide esters are generally prepared from carboxylic acids by reaction with dicyclohexylcarbodiimide and N-hydroxysuccinimide. The procedure outlined below can be applied to the synthesis of other N-hydroxysuccinimide esters. Active esters of some compound classes (e.g., amino acids) are commercially available.

SUPPORT PROTOCOL 2 Synthesis of Modified Nucleosides


Materials N-Hydroxysuccinamide Tetrahydrofuran, anhydrous Nitrogen (N2) gas Glacial acetic acid Dicyclohexylcarbodiimide (DCC) Silica gel (optional; 230 to 400 mesh) Ethyl acetate, reagent grade Methanol, reagent grade Vacuum manifold apparatus (Fig. 1.1.3) modified with a 10-mL conical flask and a 500-µL syringe Filter funnel and Whatman no. 1 filter paper Glass chromatography column (optional; 2-cm i.d. × 10-cm length) with stopcock 1. In a 10-mL conical flask containing a small magnetic stir bar, prepare a solution of 345.3 mg N-hydroxysuccinamide (3 mmol) in 1 mL anhydrous tetrahydrofuran under an inert atmosphere (e.g., nitrogen). The setup shown in Figure 1.1.3 may be used with the inert gas inlet at (b) replaced by a 500-mL syringe containing glacial acetic acid (step 2). The apparatus is maintained under a positive pressure of nitrogen by the inlet line at (c).

2. Add 174 µL glacial acetic acid (3 mmol). 3. Dissolve 620 mg dicyclohexylcarbodiimide (3 mmol) in 1 mL tetrahydrofuran and add to the reaction mixture. Stir the resulting solution at room temperature overnight (12 hr). 4. Remove the white precipitate (dicyclohexylurea) by filtration through Whatman no. 1 filter paper in a filter funnel. 5. Optional: Purify filtrate on a 10 × 2–cm silica gel column, eluting with 9:1 (v/v) ethyl acetate/methanol. Collect appropriate fractions (Rf = 0.45) and evaporate on a rotary evaporator using a water aspirator to give a white powder melting at 120°C. Although N-acetoxysuccinimide can be used without purification, it is generally preferable to purify if the product is to be stored. The product may be stored indefinitely under dry nitrogen in an amber bottle.

6. Analyze by MS and by proton NMR spectroscopy. The purified product has the following spectroscopic characteristics: Rf = 0.45 (CH2Cl2, silica) 1

H NMR (500 MHz, CHCl3-d1) d 2.84 (s, CH2 4H), 2.34 (s, CH3 3H)

MS-CI calculated for C6H7NO4: 157; observed m/z: 158 (M + H)+.

REAGENTS AND SOLUTIONS Use deionized, distilled water in all recipes and protocol steps. For common stock solutions, see APPENDIX 2A; for suppliers, see SUPPLIERS APPENDIX. PalladiumMediated C5 Substitution of Pyrimidine Nucleosides

0.1 M Li2PdCl4 solution 1.77 g PdCl2 (0.01 mol) 0.84 g LiCl (0.02 mol) 70 mL anhydrous methanol Stir at room temperature for 24 hr


Adjust volume to 100 mL with methanol The solution is generally not stored for more than a few weeks in a stoppered flask at room temperature.

COMMENTARY Background Information The C5 position of pyrimidine nucleosides is nearly ideal as a site for tethering molecular reporter groups and other molecular devices to oligodeoxyribonucleotides, as groups of different sizes may be attached without adversely affecting DNA duplex formation. In recent years, a variety of specialized probe moieties such as biotin (Langer et al., 1981; Shimkus et al., 1985; Cook et al., 1988), fluorophores (Prober et al., 1987; Tesler et al., 1989; Hagmar et al., 1995), paramagnetic probes (Spaltenstein et al., 1988, 1989; Kirchner et al., 1990), pendant catalytic moieties (Dreyer and Dervan, 1985; Bashkin et al., 1994; Kwiatkowski et al., 1994; Bergstrom and Chen, 1996; Shah et al., 1996), and cross-linkers (Gibson and Benkovic, 1987; Tabone et al., 1994; Chaudhuri and Kool, 1995; Meyer and Hanna, 1996) have been coupled to deoxyuridine and then incorporated into nucleic acids. The use of C5 linkers to functionalize nucleic acids has been reviewed (Goodchild, 1990). Although many kinds of linkers have been used to attach reporter groups to C5, alkynyl groups appear to be preferable because they enhance duplex stability (Sagi et al., 1993; Ahmadian et al., 1998). Synthesis of 5-carboxamidopropynyl-2′deoxyuridine derivatives was initiated by 4,4′dimethoxytrityl protection of the readily available 5-iodo-2′-deoxyuridine following a standard procedure (Jones, 1984). Tritylation of 5′-hydroxyl of 5-iodo-2′-deoxyuridine prior to the coupling reaction eliminated the need for toluyl protection and deprotection of the nucleoside hydroxyl groups (Robins and Barr, 1981, 1983). Palladium-mediated coupling reactions were carried out in dimethylformamide following a procedure similar to that reported by Hobbs (1989). The steps outlined in the preparation of 5-(3-nicotinamidopropyn-1-yl)-5′-O-(4,4′-dimethoxytrityl)-2′-deoxyuridine provide a blueprint and strategy for the incorporation of many different kinds of reporter groups. As long as the desired reporter group has a reactive acyl functional group (carboxylic acid anhydride, chloride, or active ester), it is likely that it can be coupled to propargylamine and sub-

sequently linked to deoxyuridine by the organopalladium coupling reaction. The thioether-linked deoxyuridine derivative 5-[3-(trifluoroacetamido)-1-thiapropyl]2′-deoxyuridine was synthesized by the procedure reported by Bergstrom et al. (Bergstrom et al., 1991; Ahmadian et al., 1998) via a palladium-mediated reaction of the disulfide N,N′bis(trifluoroacetyl)cystamine with 5-chloromercurio-2′-deoxyuridine. Trifluoroacetyl functions as a base-sensitive protecting group that can be easily removed with aqueous ammonia. Because the trifluoroacetamido group is a poor ligand for palladium, it does not interfere with the palladium-mediated coupling reaction. Groups that are substantially more electron rich (such as acetamido) interfere with the coupling reaction. For this reason, the coupling reaction must be carried out prior to attaching electron-rich ligands to the cystamine amino group. 5-[3-(Trifluoroacetamido)-1-thiapropyl]-2′deoxyuridine may be incorporated into oligonucleotides as its 5′-O-dimethoxytrityl-3′phosphoramidite derivative (procedure not given). The trifluoroacetyl protecting group is subsequently cleaved during the final ammonia deprotection of the oligonucleotide. This leaves the amino group available for conjugation to reporter groups or molecular tools at the oligonucleotide stage. Alternatively, 5-[3-(trifluoroacetamido)-1-thiapropyl]-2′-deoxyuridine may be deprotected and conjugated to the desired reporter groups or molecular tool prior to transformation to the 5′-O-dimethoxytrityl3′-phosphoramidite derivative and incorporation into the oligonucleotide. Synthesis of 5-[3-(trifluoroacetamido)-1thiapropyl]-2′-deoxyuridine requires 5-chloromercurio-2′-deoxyuridine. This intermediate is obtained in high yield by direct electrophilic mercuration of 2′-deoxyuridine with mercuric acetate in aqueous solution (Bergstrom and Ruth, 1977). An important difference between the synthesis of the alkynyl-linked dU and the thioether-linked dU is that the former requires use of a protected dU and must be done with complete exclusion of oxygen, while the latter is not air sensitive and does not require protecting groups.



Table 1.1.1

Estimated Times for Completion of Syntheses

Protocol Basic Protocol 1

Time (hr) 18 20 15

Basic Protocol 2

Support Protocol 1 Support Protocol 2

16 21 32 36 14

Synthesis N-(3-Propyn-1-yl)nicotinamide 5′-O-(4,4′-Dimethoxytrityl)-5-iodo-2′-deoxyuridine 5-(3-Nicotinamidopropyn-1-yl)-5′-O-(4,4′-dimethoxytrityl)2′-deoxyuridine 5-Chloromercurio-2′-deoxyuridine 5-[3-(Trifluoroacetamido)-1-thiapropyl]-2′-deoxyuridine 5-(3-Acetamido-1-thiapropyl)-2′-deoxyuridine N,N′-Bis(trifluoroacetyl)cystamine N-Acetoxysuccinimide

There are alternative types of linkers as well as alternative methods for preparing alkynyl and thioalkyl linkers. In addition to the alkynyl and thioether linkages described here, alkyl and alkenyl linkers may also be obtained through organopalladium coupling methodology (Bergstrom and Ogawa, 1978). Examples of reporter groups linked through alkyl and alkenyl linkers are included in Literature Cited. The advantage of the alkynyl linker is the ease of preparation of the C5-substituted deoxyuridine that contains the 5′ protecting group (dimethoxytrityl) needed for subsequent oligonucleotide synthesis via the phosphoramidite methodology. In addition to the nicotinamidopropyne coupling reaction, the preparation of a series of other carboxamidopropynyl derivatives has been described (Ahmadian et al., 1998). Since organopalladium reactions can tolerate a wide variety of functional groups (e.g., hydroxyl, amido, carboxamide, ester, cyano, nitro, keto), there may be relatively few limitations on the nature of the group that can be introduced at C5 as a component of carboxamidopropyne. Deoxyuridine C5 substitution is preferred over deoxycytidine N4 or deoxyadenosine N6 substitution because the latter two modifications destabilize doublestranded DNA.

Critical Parameters


The most critical parameter in each reaction is the purity of the reactants and the reagents. Even fresh commercial reagents should be checked by TLC and 1H NMR for purity and identity. 4,4′-Dimethoxytrityl chloride is sensitive to moisture and will deteriorate over time. The best results are obtained with a freshly opened bottle of the reagent. Alternatively, the authors store and transfer 4,4′-dimethoxytrityl chloride in a Vacuum Atmospheres dry box under a dry nitrogen atmosphere (any dry box

should be suitable). The condensation reaction between 5′-O-(4,4′-dimethoxytrityl)-5-iodo2′-deoxyuridine and the alkyne requires airsensitive tetrakis(triphenylphosphine)palladium. This reagent should also be stored and transferred under nitrogen or argon. Again, it is preferable to use freshly opened reagents, as they are supplied in sealed glass ampules that may be difficult to keep free of oxygen once opened. When planning the construction of a C5modified nucleoside for introduction into oligonucleotides, the intermediates must not contain functional groups that are likely to interfere either with phosphoramidite preparation (e.g., hydroxyl) or oligonucleotide synthesis.

Anticipated Results If the procedures are followed as described in this unit, the yields of isolated product should be comparable to that reported here. The palladium-catalyzed coupling reactions are especially sensitive to reagents and conditions. If an alkyne other than the one described in the protocol is used for the procedure, it may be necessary to try palladium catalysts other than tetrakis(triphenylphosophine)palladium. If the products are to be used to construct phosphoramidites for oligonucleotide synthesis, 50 mg of the C5-substituted 4,4′-dimethoxytrityl (DMTr) derivative is generally sufficient to obtain enough product to accomplish one coupling reaction on a 1-µmol synthesis scale.

Time Considerations The time required to complete each procedure is summarized in Table 1.1.1. The estimated times do not include the time necessary to purify solvents. In some cases, these may be purchased and used without further purification.


Literature Cited Ahmadian, M., Zhang, P., and Bergstrom, D.E. 1998. A comparative study of the thermal stability of oligodeoxyribonucleotides containing 5substituted-2′-deoxyuridines. Nucl. Acids Res. 26:3127-3135. Bashkin, J.K., Sondhi, S.M., Sampath, U., d’Avignon, D.A., and Modak, A.S. 1994. Synthesis and connectivity assignment (by 2D-NMR) of a nucleoside-dipeptide: 5-[3-[[2-[[2-[[[2-Amino]1-oxo-3-[1H-imidazol-4-yl]propyl]amino]-1oxo-3-[1H-imidazol-4-yl]propyl]amino]ethyl] amino]-3-oxopropyl]-2′-deoxyuridine. New J. Chem. 18:305-318. Bergstrom, D.E. and Chen, J. 1996. Sequence-specific oligodeoxyribonucletide cleavage by a major-groove-positioned metal-binding ligand tethered to C-5 of deoxyuridine. Bioorg. Med. Chem. Lett. 6:2211-2214. Bergstrom, D.E. and Ogawa, M.K. 1978. C-5 substituted pyrimidine nucleosides. 2. Synthesis via olefin coupling to organopalladium intermediates derived from uridine and 2′-deoxyuridine. J. Am. Chem. Soc. 100:8106-8112. Bergstrom, D.E. and Ruth, J.L. 1977. Preparation of C-5 mercurated pyrimidine nucleosides. J. Carbohydrates Nucleotides Nucleosides 42:257269. Bergstrom, D.E., Beal, P., Jenson, J., and Lin, X. 1991. Palladium-mediated synthesis of C-5 pyrimidine nucleoside thioethers from disulfides and mercurinucleosides. J. Org. Chem. 56:5598-5602. Chaudhuri, N.C. and Kool, E.T. 1995. Very high affinity DNA recognition by bicyclic and crosslinked oligonucleotides. J. Am. Chem. Soc. 117:10434-10442. Cook, A.F., Vuocolo, E., and Brakel, C.L. 1988. Synthesis and hybridization of a series of biotinylated oligonucleotides. Nucl. Acids Res. 16:4077-4095. Dreyer, G.B. and Dervan, P.B. 1985. Sequence-specific cleavage of single-stranded DNA: Oligodeoxynucleotide-EDTA.Fe(II). Proc. Natl. Acad. Sci. U.S.A. 82:968-972. Gibson, K.J. and Benkovic, S.J. 1987. Synthesis and application of derivatizable oligonucleotides. Nucl. Acids Res. 15:6455-6467. Goodchild, J. 1990. Conjugates of oligonucleotides and modified oligonucleotides: A review of their synthesis and properties. Bioconjugate Chem. 1:165-187. Hagmar, P., Bailey, M., Tong, G., Haralambidis, J., Sawyer, W.H., and Davidson, B.E. 1995. Synthesis and characterization of fluorescent oligonucleotides. Effect of internal labelling on protein recognition. Biochim. Biophys. Acta 1244:259-268. Hobbs, F.W. Jr. 1989. Palladium-catalyzed synthesis of alkynylamino nucleosides. A universal linker for nucleic acids. J. Org. Chem. 54:3420-3422. Jones, R.A. 1984. Preparation of protected deoxyribonucleosides. In Oligonucleotide Synthesis: A Practical Approach (M.J. Gait, ed.) pp. 27-28. IRL Press, Washington, D.C.

Kirchner, J.J., Hustedt, E.J., Robinson, B.H., and Hopkins, P.B. 1990. DNA dynamics from a spin probe: Dependence of probe motion on tether length. Tetrahedron Lett. 31:593-596. Kwiatkowski, M., Samiotaki, M., Lamminmaki, U., Mukkala, V.-M., and Landegren, U. 1994. Solidphase synthesis of chelate-labelled oligonucleoties: Application in triple-color ligase-mediated gene analysis. Nucl. Acids Res. 22:2604-2611. Langer, P.R., Waldrop, A.A., and Ward, D.C. 1981. Enzymatic synthesis of biotin-labeled polynucleotides: Novel nucleic acid affinity probes. Proc. Natl. Acad. Sci. U.S.A. 78:66336637. Meyer, K.L. and Hanna, M.M. 1996. Synthesis and characterization of a new 5-thiol-protected deoxyuridine for site-specific modification of DNA. Bioconjugate Chem. 7:401-412. Prober, J.M., Trainor, G.L., Dam, R.J., Hobbs, F.W., Robertson, C.W., Zagursky, R.J., Cocuzza, A.J., Jensen, M.A., and Baumeister, K. 1987. A system for rapid DNA sequencing with fluorescent chain-terminating dideoxynucleotides. Science 238:336-341. Robins, M.J. and Barr, P.J. 1981. Nucleic acid related compounds. 31. Smooth and efficient palladium-copper catalyzed coupling of terminal alkynes with 5-iodouracil nucleosides. Tetrahedron Lett. 22:421-424. Robins, M.J. and Barr, P.J. 1983. Nucleic acid related compounds. 39. Efficient conversation of 5-iodo to 5-alkynyl and derived 5-substituted uracil bases and nucleosides. J. Org. Chem. 48:1854-1862. Sagi, J., Szemzo, A., Ebinger, K., Szabolcs, A., Sagi, G., Ruff, E., and Otvos, L. 1993. Basemodified oligodeoxynucleotides. I. Effect of 5-alkyl, 5-(1-alkenyl) and 5-(1-alkynyl) substitution of the pyrimidines on duplex stability a n d h y d r o p h o b i c i t y. Tetrahedron Lett. 34:2191-2194. Shah, K., Neenhold, H., Wang, Z., and Rana, T.M. 1996. Incorporation of an artificial protease and nuclease at the HIV-1 Tat binding site of transactivation responsive RNA. Bioconjugate Chem. 7:283-289. Shimkus, M., Levy, J., and Herman, T. 1985. A chemically cleavable biotinylated nucleotide: Usefulness in the recovery of protein-DNA complexes from avidin affinity columns. Proc. Natl. Acad. Sci. U.S.A. 82:2593-2597. Spaltenstein, A., Robinson, B.H., and Hopkins, P.B. 1988. A rigid and nonperturbing probe for duplex DNA motion. J. Am. Chem. Soc. 110:12991301. Spaltenstein, A., Robinson, B.H., and Hopkins, P.B. 1989. Sequence- and structure-dependent DNA base dynamics: Synthesis, structure, and dynamics of site and sequence specifically spin-labeled DNA. Biochemistry 28:9484-9495.



Tabone, J.C., Stamm, M.R., Gamper, H.B., and Meyer, R.B. Jr. 1994. Factors influencing the extent and regiospecificity of cross-link formation between single-stranded DNA and reactive complementary oligodeoxynucleotides. Biochemistry 33:375-383. Tesler, J., Cruickshank, K.A., Morrison, L.E., and Netzel, T. 1989. Synthesis and characterization of DNA oligomers and duplex containing covalently attached molecular labels: Comparison of biotin, fluorescin, and pyrene labels by thermodynamic and optical spectroscopic measurements. J. Am. Chem. Soc. 111:6966-6976.

Contributed by Mohammad Ahmadian Cerus Corp. Concord, California Douglas A. Klewer Texas A&M University College Station, Texas Donald E. Bergstrom Purdue University West Lafayette, Indiana



Enzymatic Synthesis of M1G-Deoxyribose

UNIT 1.2

M1G-deoxyribose (M1G-dR, 1,N2-pyrimido[1,2-α]purin-10(3H)-one, or pyrimidopurinone; see structure in Fig. 1.2.2, below) is an endogenous exocyclic DNA adduct formed by the reaction of the dicarbonyl compound malondialdehyde (MDA) with a deoxyguanosine residue in DNA. M1G-dR is an intermediate in the synthesis of a class of modified oligodeoxyribonucleotides that are used to study the mutagenicity and repair of M1G. This unit presents two methods for synthesizing M1G-dR using enzymatic coupling. The Basic Protocol describes a procedure for coupling the nucleobase to deoxyribose, in a reaction mediated by the enzyme nucleoside 2′-deoxyribosyltransferase, followed by preparation of the modified base (see Fig. 1.2.1A). Preparation of the enzyme is described in the Support Protocol. The Alternate Protocol uses two commercially available enzymes, purine nucleoside phosphorylase and thymidine phosphorylase (see Fig. 1.2.1B). Although the enzyme preparation step is avoided, additional purification steps are required that increase the time needed to complete the synthesis and decrease the yield (see Commentary). NOTE: Use deionized, distilled water in all recipes and protocol steps.

BASIC PROTOCOL

ENZYMATIC COUPLING USING NUCLEOSIDE 2′-DEOXYRIBOSYLTRANSFERASE Nucleoside 2′-deoxyribosyltransferase (trans-N-deoxyribosylase or nucleoside:purine(pyrimidine) deoxyribosyltransferase; E.C. 2.4.2.6) catalyzes the transfer of the deoxyribosyl moiety from a deoxyribonucleoside to any other nucleoside base (see Fig. 1.2.1A). This enzyme is found exclusively in Lactobacilli and related microorganisms that require deoxynucleosides for growth (Carson and Wasson, 1988). The enzyme’s broad specificity makes it a useful tool for synthesizing modified deoxyribonucleotides. This protocol describes the use of this enzyme (see Support Protocol for preparation) to transfer a deoxyriboside from 2′-deoxycytidine to M1G and produce M1G-dR. The two synthesis steps can be carried out in a single flask, which decreases the time needed to purify M1G-dR and significantly increases the yield of the reaction. A

C

HO

O

HO

transferase

HO

O

M1G

HO

M1G

C

B T

HO

O HO

HO

TPase

O HO

HPO42-

T

HO

PNPase

O

M1G

OPO32M1G

HPO42-

HO

Figure 1.2.1 Enzymatic coupling reactions. (A) Reaction catalyzed by nucleoside 2′-deoxyribosyltransferase (Basic Protocol). (B) Reaction catalysed by thymidine phosphorylase and purine nucleoside phosphorylase (Alternate Protocol). Contributed by Nathalie C. Schnetz-Boutaud, Marie-Christine Chapeau, and Lawrence J. Marnett Current Protocols in Nucleic Acid Chemistry (2000) 1.2.1-1.2.8 Copyright © 2000 by John Wiley & Sons, Inc.


1.2.1

Materials Guanine hydrochloride (Sigma) 1 N HCl Tetraethoxypropane (Aldrich) Methanol (MeOH; Aldrich) Potassium carbonate (Aldrich) Nanopure water (water purified using Nanopure system from Barnstead/Thermolyne) MES (2-[N-morpholino]ethanesulfonic acid; Sigma) 2′-Deoxycytidine (dC; Sigma) 1 N NaOH Nucleoside 2′-deoxyribosyltransferase (transferase; see Support Protocol) Dichloromethane (CH2Cl2; Fisher) 250-mL round-bottom flask Oil bath, 70°C Magnetic stir plate and stir bar Ice bath pH indicator strips Büchner funnel Whatman No. 1 filter paper Shaking incubator, 37°C Silica-gel thin-layer chromatography (TLC) plates Lyophilizer Silica gel (60 to 100 mesh; Fisher) 8 × 50–cm chromatography column Prepare modified base 1. In a 250-mL round-bottom flask, dissolve 1 g (5.3 mmol) guanine hydrochloride in 100 mL of 1 N HCl that has been heated to 70°C using an oil bath. Stir on a magnetic stir plate until dissolved. The dissolution can take from 30 min to 1 hr.

2. Mix 1.3 mL (5.86 mmol) tetraethoxypropane with 1.25 mL methanol. Add this dropwise to the solution from step 1. Slow addition of tetraethoxypropane favors the formation of the modified base and avoids the polymerization of MDA.

3. Let the reaction mixture stir for 30 min on a magnetic stir plate, then cool to 0°C in an ice bath. Cooling moderates the neutralization reaction to follow.

4. Neutralize by slowly adding potassium carbonate to pH 6, verifying the pH using pH indicator strips. Take special care when approaching the desired pH. 5. Remove unreacted guanine by filtering through Whatman no. 1 filter paper on a Büchner funnel under vacuum. Wash the precipitate twice with 20 mL of Nanopure water. The filtrate contains the modified base. The yield of the reaction is estimated at 30%. Enzymatic Synthesis of M1G-Deoxyribose

6. Add MES to a final concentration of 0.5 M. MES is used to buffer the enzymatic reaction.


7. Add 1.2 g (5.3 mmol) 2′-deoxycytidine. 8. Equilibrate the solution to pH 6.0 with 1 N HCl or 1 N NaOH.

Perform enzymatic coupling 9. Add an appropriate amount of nucleoside 2′-deoxyribosyltransferase solution and incubate overnight at 37°C with shaking. Each enzymatic preparation has its own concentration and activity, which must be tested empirically by assaying the preparation’s ability to convert a small quantity of M1G to M1GdR to estimate the amount to use for the full-scale reaction.

10. Check the progress of the reaction by TLC on silica-gel plates using 9:1 (v/v) CH2Cl2/MeOH as the mobile phase. Add 50 mL methanol to 50 mL of reaction mix and spot on TLC plate using a capillary tube. If the reaction is not complete, add more enzyme and dC, but not more than one-fifth the amount used at the start of the reaction, and incubate an additional 6 to 12 hours.

11. Once the reaction is complete, lyophilize the reaction mixture to dryness. 12. Purify the crude product on a silica-gel column using 9:1 (v/v) CH2Cl2/MeOH for equilibration and elution. Elute by gravity and collect 50-mL fractions. The lyophilized reaction mixture may be added directly to the top of the packed column.

13. Confirm the purity of the product by 1H NMR. Store the product under nitrogen at –20°C. Under these conditions, it is stable for several years. A typical spectrum is presented in Figure 1.2.2. The estimated yield is 10% to 15%. O N N

HO

O

8 7

N N

N

6

HO

Figure 1.2.2

1

H NMR spectrum of M 1-GdR in D2O.



SUPPORT PROTOCOL

PREPARATION OF NUCLEOSIDE 2′-DEOXYRIBOSYLTRANSFERASE Nucleoside 2′-deoxyribosyltransferase was first isolated by McNutt (1952). Partial purification of the enzyme from Lactobacillus leichmannii has been described by Beck and Levin (1963), and its complete purification and crystallization by Uerkvitz (1971). This protocol is based upon the latter method. The partial purification described here is sufficient for obtaining enzyme to be used in the Basic Protocol. Materials Lactobacillus broth AOAC (see recipe) Lactobacillus helveticus culture 0.15 M NaCl (4°C) 50 mM potassium phosphate buffers, pH 6.0 and 6.9 (see APPENDIX 2A; dilute with Nanopure water to desired molarity) 50 mM potassium phosphate, pH 5.1, containing 10 g/L NaCl 250-mL Erlenmeyer flask Centrifuge and rotors (e.g., Sorvall GS-3 and SS-34) Microtip sonicator (Virsonic 100) BCA Protein Assay (Pierce; optional) or equivalent Purify the enzyme 1. Using aseptic technique, place 25 mL of Lactobacillus broth in an Erlenmeyer flask. Add a loopful of Lactobacillus helveticus commercial stock and incubate overnight at 37°C without shaking. 2. Transfer the 25-mL culture into 1000 mL of fresh Lactobacillus broth. Grow 18 hr at 37°C without shaking. 3. Cool to 4°C and divide into centrifuge tubes. Centrifuge the tubes 10 min at 7000 × g (6500 rpm in a GS-3 rotor), 4°C. 4. Resuspend pellets in 100 mL cold 0.15 M NaCl and centrifuge again as in step 3. Repeat. 5. Suspend cell pellets in a total of 20 mL of 50 mM potassium phosphate, pH 6.0. 6. Sonicate 10 times for 1 min each time with a Virsonic 100 microtip sonicator at a setting of 4 to 5. 7. Centrifuge 30 min at 28,000 × g (15,000 rpm in an SS-34 rotor), 4°C. Save the supernatant. 8. Wash the pellets twice by resuspending in a minimal volume of 0.15 M NaCl and centrifuging as in step 7. Save the supernatants. 9. Combine the supernatants from the previous two steps, and dialyze overnight against 4 L of 50 mM potassium phosphate, pH 5.1, containing 10 g/L NaCl. 10. Heat 10 min at 55°C, then immediately place on ice. The heating denatures all heat-sensitive proteins that are present in the extract, thereby enriching for transferase, which is not heat sensitive.

11. Centrifuge again as in step 7. Enzymatic Synthesis of M1G-Deoxyribose

12. Collect the supernatant and dialyze overnight in 50 mM potassium phosphate, pH 6.9.


To quantify the amount of total protein purified, the BCA Protein Assay (Pierce) can be used. For the purpose of this experiment, this level of purification is sufficient. Store the protein in aliquots at −20°C or proceed to coupling experiments (see Basic Protocol).

Assay the enzyme 13. Prepare five 5-mL aliquots of filtrate containing M1G (see Basic Protocol, step 8). 14. Add to the filtrate 50, 100, 150, 200, and 250 µL of supernatant containing the enzyme (from step 12). 15. Incubate overnight at 37°C with shaking. 16. Spot 10 µL of each reaction mixture three times on a TLC plate using 9:1 (v/v) CH2Cl2/MeOH as the mobile phase. Determine presence of M1G-dR. ENZYMATIC COUPLING USING PURINE NUCLEOSIDE PHOSPHORYLASE AND THYMIDINE PHOSPHORYLASE

ALTERNATE PROTOCOL

Purine nucleoside phosphorylase (PNPase; E.C. 2.4.2.1) catalyzes the displacement of phosphate from deoxyribose-1-phosphate on purines and purine analogs. The stereochemistry of purine attachment produces the naturally occurring β-isomers. Although ribose-1-phosphate is commercially available, it may be more conveniently and cost-effectively generated in situ by thymidine phosphorylase (TPase; E.C. 2.4.2.4)–catalyzed phosphorolysis of thymidine (see Fig. 1.2.1). This Alternate Protocol can be used to avoid the transferase preparation required for the Basic Protocol. First the modified base is synthesized and purified, then the enzymatic coupling is initiated, and finally the product is separated on a medium-performance liquid chromatography (MPLC) column. Due to the instability of M 1G, only 2% to 5% of product is recovered, regardless of the yields achieved in the transribosylation step. Additional Materials (also see Basic Protocol) Thymidine (e.g., Sigma) Purine nucleoside phosphorylase (PNPase; Sigma) Thymidine phosphorylase (TPase; Sigma) 20 mM potassium phosphate, pH 7.3 (APPENDIX 2A) MPLC buffer: 20% methanol in water UV lamp (254 and 365 nm) mPLC column (Baker C18-40 µm, 30 × 500 mm) Prepare modified base 1. Prepare modified base (see Basic Protocol, steps 1 to 5). 2. Concentrate the filtrate under vacuum. 3. Prepare a slurry of the residue with ∼10 g silica gel. 4. Purify by chromatography on a silica-gel column using 9:1 (v/v) CH2Cl2/MeOH for equilibration and elution. 5. Collect and combine the fractions containing M1G as determined by UV fluorescence. M1G base is fluorescent under the long-wavelength (365 nm) of the UV lamp.



6. Evaporate under vacuum and store under nitrogen at –20°C. 7. Verify the purity of M1G base by 1H NMR. The estimated yield is 2% to 5%.

Perform enzymatic coupling 8. Dissolve 37 mg (0.2 mmol) M1G base and 73 mg (0.6 mmol) thymidine in 200 mL of 20 mM potassium phosphate buffer. Adjust the pH to 7.3 with 1 N HCl or 1 N NaOH. 9. Add 20 U TPase and 30 U PNPase. 10. Incubate the solution 18 hr at 37° to 39°C with shaking. Verify the completion of the enzymatic coupling by thin-layer chromatography (see Basic Protocol, step 10). If the reaction is not complete, continue the purification without further incubation.

11. Concentrate the reaction under vacuum. This and all subsequent vacuum steps may be performed using a side-arm flask with water aspiration.

12. Prepare a slurry of the residue with ∼5 g silica gel. 13. Purify by chromatography on a silica-gel column using 9:1 (v/v) CH2Cl2/MeOH for equilibration and elution. Elute by gravity and collect 50-mL fractions. 14. Collect the fractions containing M1G-dR as determined by yellow fluorescence at 365 nm. M1G and M1G-dR have different elution times.

15. Evaporate to dryness under vacuum. Redissolve two aliquots in 2 mL MPLC buffer each. 16. Purify by MPLC on a Baker C18 40 µm, 30 × 500–mm column, eluting with 20% methanol in water (isocratic) at a flow rate of 3 mL/min. This second column is necessary to separate thymidine from M1G-dR.

17. Collect the fluorescent fractions and evaporate to dryness under vacuum. 18. Verify the purity of the product by 1H NMR. A typical spectrum is presented in Figure 1.2.2 (see Basic Protocol). The estimated yield is 50% to 60%.

REAGENTS AND SOLUTIONS Use Nanopure water (water purified using Nanopure system from Barnstead/Thermolyne) where indicated, and deionized, distilled water in all other recipes and protocol steps. For common stock solutions, see APPENDIX 2A; for suppliers, see SUPPLIERS APPENDIX.


Lactobacillus broth AOAC Mix 38 g of Lactobacillus broth AOAC (Difco) into 1000 mL Nanopure water. Heat to boiling for 2 min. Autoclave 30 min and allow to cool to room temperature. Prepare fresh for each run.


COMMENTARY Background Information Adducts formed between electrophiles and nucleic acid bases are believed to play a key role in chemically induced mutations and cancer (Singer and Grunnenberger, 1983). Chemical synthesis of deoxynucleoside adducts provides not only authentic standards for comparison to biologically derived materials but also reagents for the synthesis of adducted nucleotides (Basu and Essigmann, 1988). The preparation of certain classes of deoxynucleoside adducts is problematic because of the instability of the intermediates under the conditions used for synthetic manipulations (e.g., the acid lability of purine deoxyribosides). Synthetic approaches to preparing sensitive deoxynucleosides include coupling of adducted bases to activated deoxyribose derivatives (Srivasta et al., 1988) and attachment of deoxyribose moieties to modified bases (Garner and Ramakanth, 1988). However, there are inherent difficulties in controlling both the regiochemistry (e.g., to achieve attack at the N7 versus N9 atom of a purine) and the stereoselectivity (e.g., an SN2 attack on a sugar isomer is needed to generate the desired linkage). Chemical synthesis increases the difficulty of obtaining the correct regioselectivity, whereas enzymatic coupling on nonmodified bases generates only one regioselective isomer. Enzymatic coupling of purine or pyrimidine analogs to deoxyribose has been used to synthesize a number of compounds, including isotopically substituted deoxynucleosides, antitumor agents, and biological active molecules (Holy and Votruba, 1987; Krenitsky et al., 1981, 1986; Muller et al., 1996). When PNPase and TPase are used on M1G (see Alternate Protocol), N9 linkage is shown to be preferred over N7. With the use of transferase (see Basic Protocol), complete regioselectivity (for the N9 isomer only) is obtained.

the Alternate Protocol, which uses a combination of two commercially available enzymes, is more demanding in terms of time and purification. M1G base must be purified; moreover, when the crude mixture is reacted with the combination of PNPase and TPase, transribosylation is inefficient. Also, due to the close polarity of thymidine and M1G-dR (in a variety of solvents system used), a combination reversed-phase and straight-phase column must be employed to obtain pure material. The transferase used in the Basic Protocol and prepared in the Support Protocol has a higher selectivity and produces only the N9 isomer. The two-enzyme method of the Alternate Protocol produces a mixture of N7 and N9, increasing the difficulty of purification. Once a stock solution of transferase is prepared, the Basic Protocol is much easier to perform, less time-consuming, and gives better yields.

Anticipated Results The two protocols described in this unit allow preparation of the desired modified nucleoside in good yields. The Basic Protocol should give yields between 10% and 15% and the Alternate Protocol gives between 2% and 5%. Figure 1.2.2 shows the 1H NMR spectrum of M1G-dR.

Time Considerations Preparation of transferase (see Support Protocol) may take 3 to 4 days. The one-step condensation procedure using transferase (see Basic Protocol), from the synthesis of the modified base to the purification of the nucleoside, should take 24 hr with lyophylization. The alternative method using commercial phosphorylases (see Alternate Protocol) should take 2 or 3 days.

Literature Cited Critical Parameters Since the pyrimidopurinone M1G-dR synthesized in these procedures is base labile, the pH must stay below 7.5 for all protocols. M1G base is even less stable than M1G-dR. The Basic Protocol does not require purification of the modified base; the transferase can be added to the crude mixture. The desired nucleoside can be purified by simple column chromatography. This strategy not only shortens the time required for the synthesis, but also significantly improves the yield. In comparison

Basu, A.K. and Essigmann, J.M. 1988. Site-specifically modified oligonucleotides as probes for the structural and biological effects of DNA-damaging agents. Chem. Res. Toxicol. 1:1-18. Beck, W.S. and Levin, M. 1963. Purification, kinetics, and repression control of bacterial trans-Ndeoxyribosylase. J. Biol. Chem. 238:702. Carson, D.A. and Wasson, D.B. 1988. Synthesis of 2′,3′-dideoxynucleosides by enzymatic transglycosylation. Biochem. Biophys. Res. Comm. 155:829-834. Garner, P. and Ramakanth, S. 1988. A regiocontrolled synthesis of N7- and N9-guanine nucleosides. J. Org. Chem. 53:1294-1298.



Holy, A. and Votruba, I. 1987. Facile preparation of purine and pyrimidine 2-deoxy-β-D-ribonucleosides by biotransformation on encapsulated cells. Seventh symposium on the Chemistry of Nucleic Acid Components August 30–September 5, 1987. Nucleic Acids Symp. Ser. 18:69-72. Krenitsky, T.A., Kozallka, G.W., and Tuttle, J.V. 1981. Purine nucleoside synthesis, an efficient method employing nucleoside phosphorylases. Biochemistry 20:3615-3621. Krenitsky, T.A., Rideout, J.L., Chao, E.Y., Koszalka, G.W., Gurney, F., Crouch, R.C., Cohn, N.K., Wolberg, G., and Vinegar, R. 1986. Imidazo[4,5c]pyridines (3-deazapurines) and their nucleosides as immunosuppresive and antiinflammatory agents. J. Med. Chem. 29:138-143. McNutt, W.S. 1952. The enzymically catalysed transfer of the deoxyribosyl group from one purine or pyrimidine to another. Biochem. J. 50:384. Muller, M., Hutchinson, L.K., and Guengerich, F.P. 1996. Addition of deoxyribose to guanine and modified DNA bases by Lactobacillus helveticus trans-N-deoxyribosylase. Chem. Res. Toxicol. 9:1140-1144.

Singer, B. and Grunnenberger, D. 1983. Molecular Biology of Mutagens and Carcinogens. Plenum, New York. Srivasta, P.C., Robins, R.K., and Meyer, R.B. 1988. Synthesis and properties of purine nucleosides and nucleotides. In Chemistry of Nucleosides and Nucleotides (L.B. Townsend, ed.) pp. 113281. Plenum, New York. Uerkvitz, W. 1971. Purification of nucleoside 2-deoxyribosyltransferase from Lactobacillus helveticus. Eur. J. Biochem. 23:387-395.

Key Reference Uerkvitz, 1971. See above. Decribes the purification and crystallization of the transferase.

Contributed by Nathalie C. Schnetz-Boutaud, Marie-Christine Chapeau, and Lawrence J. Marnett Vanderbilt University Nashville, Tennessee



Synthesis of N2-Substituted Deoxyguanosine Nucleosides from 2-Fluoro-6-O(Trimethylsilylethyl)-2′-Deoxyinosine

UNIT 1.3

This unit describes the synthesis of 2-fluoro-6-O-(trimethylsilylethyl)-2′-deoxyinosine and gives examples of its use for the preparation of N2-substituted deoxyguanosine nucleosides. Such nucleoside derivatives are used for a variety of purposes including chemotherapy, enzyme mechanism studies, nuclear magnetic resonance (NMR) studies (when isotopically labeled), and as synthetic standards for identification of adducts formed by the reaction of DNA with xenobiotics. In addition, the O6-protected 2-fluoro2′-deoxyinosine compounds can be converted to phosphoramidites and used in the synthesis of oligonucleotides, thus allowing substitution reactions to be carried out after oligonucleotide assembly. 2-Halopurine derivatives have been used for many years for the preparation of N2-substituted guanosine derivatives, with the 2-fluoro substituent being the most easily displaced by nucleophiles (Montgomery and Hewson, 1960; Gerster and Robins, 1965, 1966). The 2-fluoro group is introduced by aqueous diazotization of guanosine in the presence of potassium fluoride or fluoroboric acid. However, these conditions are too harsh for 2′-deoxyguanosine and lead to depurination; hence, different synthetic methodology is needed for the deoxynucleoside. The fluorine atom can be introduced successfully by diazotization under anhydrous conditions with t-butyl nitrite as the diazotizing agent and HF in pyridine as the fluoride source (Robins and Uznanski, 1981; Lee et al., 1990; Harris et al., 1991). Success in the fluoridation step requires protection of the C6 oxygen group, which is done by Mitsunobu alkylation (Mitsunobu, 1981) with trimethylsilylethanol or other alcohols. The O6-protecting group also facilitates displacement of the halogen by nucleophiles. Basic Protocol 1 in this unit describes the synthesis of 2-fluoro-6-O-(trimethylsilylethyl)2′-deoxyinosine and comprises three separate procedures: (1) protection of the 2-NH2, 3′-OH, and 5′-OH groups of 2′-deoxyguanosine to make a triacetyl derivative, (2) protection of the O6 group by Mitsunobu alkylation with trimethylsilylethanol, and (3) introduction of the 2-fluoro group. Alternate Protocol 1 describes the preparation and use of 3′,5′-O-diacetyl-2′-deoxyguanosine and its use in the Mitsunobu reaction described in Basic Protocol 1. Alternate Protocol 1 gives lower yields than Basic Protocol 1, but is quicker and is better for the preparation of 6-O-(p-nitrophenethyl)-2′-deoxyguanosine, another commonly used O6-protected derivative. Basic Protocol 2 describes a general procedure for the synthesis of N2-substituted 2′-deoxyguanosines. Two specific examples are then given in Alternate Protocols 2 and 3, which give detailed directions for synthesis using an unhindered diamine to give a derivative with an alkylamine sidechain, and for using an amino alcohol to yield an N2 hydroxyalkenyl derivative. A Support Protocol outlines the procedure for carrying out reactions in an inert atmosphere. CAUTION: Several of the steps in these protocols involve the use of toxic, corrosive, and flammable chemicals. It is highly recommended that all operations be carried out in a fume hood. Good laboratory safety practices should be observed at all times, including the use of safety goggles, a laboratory coat, and disposable gloves. It is recommended that this synthesis be done only by personnel experienced in the handling of reactive and toxic chemicals. Contributed by Thomas M. Harris and Constance M. Harris Current Protocols in Nucleic Acid Chemistry (2000) 1.3.1-1.3.19 Copyright © 2000 by John Wiley & Sons, Inc.


1.3.1

NOTE: A variety of methods is used in these procedures to remove volatile solvents and reagents. The choice of method depends upon the boiling points of the volatiles, the stability of the products (e.g., sometimes heat cannot be used to hasten evaporation), and the volume to be removed. For additional details, see Critical Parameters and Troubleshooting. BASIC PROTOCOL 1

SYNTHESIS OF 2-FLUORO-6-O-(TRIMETHYLSILYLETHYL)-2′DEOXYINOSINE USING 2-N-3′,5′-O-TRIACETYL-2′-DEOXYGUANOSINE This protocol describes the synthesis of the 2-fluoro derivative of 6-O-(trimethylsilylethyl)2′-deoxyinosine (Fig. 1.3.1). It is divided into three basic procedures. (1) Protection of the 2-NH2, 3′-OH, and 5′-OH groups of 2′-deoxyguanosine (S.1) is performed by acetylation (S.2). (2) Protection of the O6 group is carried out via Mitsunobu alkylation with trimethylsilylethanol, diethyl azodicarboxylate, and triphenylphosphine. Sodium methoxide and methanol are then added to the reaction to remove the acetyl protecting groups, yielding the 6-O-trimethylsilylethyl (TMSE) derivative (S.3). (3) The 2-NH2 group is converted to the 2-fluoro substituent by performing nonaqueous diazotization and fluoridation at low temperature with t-butyl nitrite and HF/pyridine (Robins and Uznanski, 1981). An Alternate Protocol utilizing 3′,5′-O-diacetyl-2′-deoxyguanosine is also described (see Alternate Protocol 1). NOTE: Anhydrous solvents are required for several steps in these procedures. They can be purchased (e.g., from Aldrich in Sure/Seal bottles) or dried by distillation from appropriate desiccants and stored under nitrogen. Materials 2′-Deoxyguanosine (dG) monohydrate Pyridine, anhydrous (Aldrich; packed under nitrogen in a Sure/Seal bottle) Acetic anhydride, freshly distilled Triethylamine (d 0.726), distilled from calcium hydride 4-Dimethylaminopyridine (DMAP)

N

HO

O

N

NH N

NH2

a

N

AcO

O AcO

HO 1

Me3Si

O

O N

NH N

NHAc

O N

b,c

N

N

HO

N

O

NH2

HO 2

3

d

Me3Si

O N N

HO

O

N N

F

HO 4

Synthesis of N2-Substituted Deoxyguanosine Nucleosides

Figure 1.3.1 Synthesis of 2-fluoro-6-O-(trimethylsilylethyl)-2′-deoxyinosine (S.4) using 2-N-3′,5′O-triacetyl-2′-deoxyguanosine (S.2; see Basic Protocol 1). Reagents: (a) acetic anhydride, pyridine, 4-dimethylaminopyridine (steps 1 to 13): (b) triphenylphosphine, diethyl azodicarboxylate, 2-trimethylsilylethanol (steps 14 to 19); (c) sodium methoxide, methanol (steps 20 to 33); (d) HF/pyridine, t-butyl nitrite (steps 34 to 46).


Dry nitrogen (N2) or argon (Ar) Methanol, anhydrous Methylene chloride (CH2Cl2), anhydrous Anisaldehyde/sulfuric acid spray (see recipe) Acetonitrile Dioxane, anhydrous, distilled from sodium metal before use Triphenylphosphine 2-Trimethylsilylethanol Diethyl azodicarboxylate (DEAD; from a fresh, unopened bottle) 0.35 M sodium methoxide in methanol (see recipe) Aqueous acetic acid: 6.2 mL glacial acetic acid in 30 mL water Sodium sulfate (Na2SO4), anhydrous 63- to 200-mesh silica gel Sand Dry ice/acetonitrile cooling bath (−35° to −40°C) 70% HF/pyridine solution (Aldrich) t-Butyl nitrite Potassium carbonate (K2CO3) Ethyl acetate 2-liter round-bottom flask Rotary evaporator equipped with a condenser cooled with chilled water or a dry ice condenser Reflux condenser with 24/40 joint and gas inlet adapter Temperature-controlled oil bath (up to ∼115°C) 0.25-mm silica gel 60F-254 glass thin-layer chromatography (TLC) plates UV light source Vacuum system (oil pump) capable of creating G = T > C). By comparison, 3-nitropyrrole (S.29) pairs with little discrimination between each of the four natural bases (Bergstrom et al., 1995, 1997); however, it is significantly destabilizing relative to a natural base. This has been used to advantage in at least two separate applications. 3-Nitropyrrole has been used to increase the selectivity of hybridization-based detection of single-nucleotide polymorphisms (Guo et al., 1997), and as a tool to elevate the fidelity of thermostable Thermus thermophilus (Tth) DNA ligase for the ligation of oligonucleotide primers (Luo et al., 1996). The introduction of a second aromatic ring (e.g., 5-nitroindole; S.30) led to substantial improvement in duplex stability without too great of a loss in pairing nondiscrimination (Loakes and Brown, 1994). Neither 3-nitropyrrole nor 5-nitroindole deoxyribonucleosides are effective substrates or template components for DNA polymerase (Loakes et al., 1995). They preferentially direct the incorporation of the less polar natural bases A and T (Hoops et al., 1997). More recently it has been shown that both nondiscriminatory base pairing and high stability can be achieved with nucleosides based on the quinolone heterocycle (S.31; Berger et al., 2000b). The emphasis of the latter study has been to develop analogs that are effective substrates for DNA polymerases.

3. The corresponding deoxyribonucleoside triphosphate should be suitable as a substrate for DNA polymerase as assessed by steadystate kinetic experiments for single-nucleotide primer extension. 4. The base should be a substrate for DNA ligase as primer component as assessed by ligase chain reaction. This list is by no means complete, but represents those specifications that would be of the greatest use for the manipulation of DNA in conventional molecular biology techniques. The factors that contribute to base pair stability as assessed by thermal denaturation studies are not directly related to template preference by DNA polymerase. This has been illustrated by comparison of a set of azole carboxamide nucleosides for which the order of thermodynamic stabilities differs substantially from their template coding properties (Hoops et al., 1997).

Nonpolar Nucleobase Analogs The development of universal nucleic acid bases has progressed along two parallel lines. The first class of analogs includes molecules that cannot associate through hydrogen bonding, but because of size, shape, and hydrophobicity prefer to occupy the interior of a duplex (Fig. 1.4.8). The first analog of this class, phenyl deoxyribonucleoside (S.28), was reported in 1984 (Millican et al., 1984). It is highly destabilizing and pairs with significant

Evolution of the non-hydrogen bonding, hydrophobic universal base

NO2

NO2

N dR

dR 28

N dR

29

O 30

31

N dR

Hydrogen bonding capable universal base candidates

O N

Unnatural Nucleosides with Unusual Base Pairing Properties

N

H

NH2

N N dR

N

N 26

N dR

27

Deoxyinosine

Figure 1.4.8

Universal base candidates.

1.4.6 Supplement 5


Polar Hydrogen Bonding Nucleobase Analogs The first studies of nucleoside analogs specifically designed to base pair with more than one of the four primary DNA bases appeared over a decade ago (Millican et al., 1984; Eritja et al., 1986; Seela and Kaiser, 1986; Habener et al., 1988; Lin and Brown, 1989; Brown and Lin, 1991b). The most extensively studied example is 2′-deoxyinosine (S.26; Fig. 1.4.8), which has been in use as a putative universal nucleoside in oligonucleotide probes and primers since 1985 (Ohtsuka et al., 1985). Structural studies on deoxyinosine-modified oligonucleotides show that dI can base pair to dC, dA (Corfield et al., 1987; Uesugi et al., 1987), T (Cruse et al., 1989; Carbonnaux et al., 1990), and dG (Oda et al., 1991). However, it is not a true universal nucleoside because the base pairs dI-dX (X = dA, dC, dG, T) differ in stability by as much as 2 to 3 kcal/mol (Martin and Castro, 1985; Kawase et al., 1989). More importantly, primers constructed with multiple sites of deoxyinosine substitution frequently give undecipherable results in sequencing experiments. More recently Seela and Debelak (2000) have developed a nucleoside analog, N8-(2′-deoxyribofuranoside) of 8-aza-7-deazaadenine (S.27; Fig. 1.4.8), which pairs with all four natural bases with significantly less discrimination than inosine, but with relatively high affinity.

TRIPLEX CONSTITUENTS Duplex formation occurs through WatsonCrick pairing of purine and pyrimidine bases, which involves hydrogen bonding of NH3 and O4 of thymine with N1 and NH6 of adenine, and O2, N3, and NH4 of cytosine with NH2, NH1, and O6 of guanine. This leaves two sites on each of the purine bases (N7 and NH6 of adenine and N7 and O6 of guanine) free for hydrogen bonding. A third oligonucleotide strand can associate with a duplex through hydrogen bond formation to these sites within the major groove to form a triplex. Triplex formation with natural nucleotides generally assumes one of two themes: parallel association between a homopyrimidine strand and a homopurine-homopyrimidine duplex following the base association schemes C+•GC (S.32; Fig. 1.4.9) and T•AT (S.33), or antiparallel association involving the triplets G•GC (S.34), A•AT (S.35), and T•AT. In each of these cases, base-base recognition from the third strand involves recognition of N7 and NH6 (adenine) or O6 (guanine). The reason that these base configurations are more stable than other pos-

sible arrangements of bases in a triplet configuration stems from geometry preferences, hydrogen bonding, and base stacking factors. Neither cytosine nor thymine can occupy the center strand, since both have only a single site (O6 of thymine and NH4 of cytosine) available for hydrogen bonding to a base in a third strand located in the major groove. On the other hand, each of the base-triplets illustrated in the figure has two hydrogen bonds between the central purine and the third-strand base. As a result, triplex formation is generally limited to polypurine strings within one stand of the duplex. This significantly limits the number of potential targets in an organism. T•TA, C•CG, C•TA, T•CG, A•TA, A•CG, G•TA, and G•CG, all of which could have at best one hydrogen bond between the third strand base and the central pyrimidine, are not stable. Furthermore, since C must be protonated at N3 in order to hydrogen bond to N7 of G, triplex formation through this motif is favored only at low pH. For these reasons, significant effort has been expended to (1) design third-strand unnatural bases that can bind opposite the pyrimidine bases, and (2) develop structural variations of cytosine that are protonated at neutral pH. Nucleobase design for triplex formation has been extensively reviewed (Ganesh et al., 1996; Doronina and Behr, 1997; Gowers and Fox, 1999). Rather than reiterate the extensive studies that have been done on unnatural nucleosides as third-strand components, this review will only provide a few examples to illustrate the different types of approaches that attempt to solve the problem. The triplets composed of C+•GC and T•AT are isomorphous. On the other hand, the G•GC, A•AT, and T•AT are not, which leads to significant dependence of triplex stability on the duplex sequence as well as on the relative number of GC and AT base pairs. Since the third-strand association of C by G requires that C be protonated, contiguous protonated C’s may lead to some destabilization through charge-charge repulsion. Consequently, nucleic acid chemists have sought nucleobase analogs that are protonated at this site at physiological pH. The approaches include modification of the cytosine to increase the pKa of protonated N3, and replacement of cytosine with unnatural bases that have hydrogen bond donor-donor configuration required to pair with the acceptor G. Figure 1.4.10 (S.36 to S.38) illustrates a few of the reported analogs designed to mimic protonated C.



Supplement 5

H N H

H N dR N

H

N

O

H

O

H

dR N

H N H

H O

H

N

N

dR N

O

O

N 33

N O

N H N

H N

H

N dR

N

H

H

O

N N A•AT

dR

H N

N dR

N

H N H

N

G•GC

N

H N

N dR

N dR

N

H

N

N H T•AT

N H N

H N

H

32

O

N

N O

H

N

N dR

O

O dR N

O

N

N C+•GC

N dR

N

34

N

N

35

dR

Figure 1.4.9

Base association on triplex formation.

the base is a pyrazopyrimidine (P1) with both hydrogen-bond donating groups contained in the pyrimidine ring (Koh and Dervan, 1992). The third triple, S.38, contains a protonated 2-aminopyridine (designated as P in the figure; Cassidy et al., 1997). This is a more accurate structural mimic of protonated C than M or P1. The greater basicity of the pyridine (pKa ~ 6) means that the equilibrium will be shifted more towards the protonated form at physiological pH than in the case of C (pKa = 4.35).

The three structures shown in Figure 1.4.10 reflect three very different design strategies for mimicking protonated C. The first of these, S.36, contains the protonated-C mimic N6methyl-8-oxo-deoxyadenosine (designated as M in the figure; Krawczyk et al., 1992). The 8-oxo group shifts the equilibrium about the glycosidic bond toward a conformation (shown in the figure) that positions the hydrogen-bond donor sites on M for interaction with N7 and O6 of G. The second triple, S.37, also contains a neutral protonated-C mimic, but in this case

N N

H N H

Me N H

dR N

N

O

H

N

H

N N

N dR

M•GC

H

N

dR

P1•GC

H N H O

H N


P•GC

Figure 1.4.10 and P1.

N dR

N

N

N dR

O

H N N

N dR

N

H N H 37

N dR O

H N

O

H N

N H 36

H N

H

N O

H

H N H

H N

N

N N dR

O

N

O

N dR

H N H

38

Protonated cytosine analogs for triplex association. See text for definitions of M, P,

1.4.8 Supplement 5


The expansion of base pair recognition beyond C+•GC, T•AT, G•GC, A•AT, and T•AT is not the only issue. The Tm values for thirdstrand dissociation are typically far lower than for duplex DNA of equivalent length. Consequently, further modifications have been explored to enhance triplex stability. Commonly, this has been achieved by attaching intercalators to one end of the third strand. Significant effort has even been expended to develop triplex-specific intercalators. An alternative is to develop base analogs that provide additional stabilization elements. One example, recently reported by Fox and co-workers, is the use of the nucleoside analog 5-(3-aminopropargyl)2′-deoxyuridine as a component of the third strand in place of thymidine (Bijapur et al., 1999). The increase in triplex stability, presumably due to association of the protonated amino group with the phosphodiester anion, was substantial. A challenge for nucleic acid chemists has been to design and develop modified nucleobases that can recognize CG and TA base pairs through association from the pyrimidine side of the major groove. Two examples of modified nucleosides designed to bind the CG base pair are illustrated in Figure 1.4.11. In S.39, only one hydrogen bond to the cytosine is present (Prévot-Halter and Leumann, 1999). In contrast, S.40 was designed to extend across the base pair and hydrogen bond to G as well as C (Huang et al., 1996). Rothman and Richards have proposed a number of structures from modeling experiments for TA base pair recognition (Rothman and Richards, 1996). The proposed structures, which contain alkyne/alkene spacers to a five-membered ring heterocycle, are predicted to form hydrogen bonds to O4 of T and N6 of A. They should be interesting candidates for future investigation.

MODIFYING NATURAL BASES TO TUNE PAIRING AFFINITY There are applications that would benefit from either decreasing base pair stability or increasing base pair stability while retaining base pair specificity. For natural sequences, CG base pairs contribute more to duplex stability than AT base pairs. This creates a problem in DNA-array-based strategies where it is advantageous for all sequences of the same length to have melting temperatures within a narrow temperature range. One would expect that to accomplish this without complicating sequence-specific effects would be difficult. A number of examples have been published that report progress in this direction. Nguyen et al. (1998) have reported that N4-ethylcytosine destabilizes a CG base pair to the extent that it resembles an AT base pair in stability. Alternatively, it is possible to increase the stability of AT base pairs by appending certain substituents at C5 of T in place of methyl. This is exemplified by the nucleoside analog 5-propynyl-2′deoxyuridine, which has found application in antisense oligonucleotides (Wagner et al., 1993). One would think that it would be possible to increase the affinity of association between A and T by adding an amino group to C2 of A to give 2,6-diaminopurine (2,6-DAP), which should hydrogen bond to T with three hydrogen bonds. However, in practice, this is not the case. The effect of the 2,6-DAP-T base pair on duplex DNA stability is dependent on sequence, and in some instances is not as stabilizing as an AT base pair. However, Matray et al. (2000) have discovered that 2,6-DAP is consistently stabilizing when it is incorporated into oligonucleotides containing the N3′→P5′ phosphoramidate linkage. This effect may be related to the adoption of A-type helices by this backbone modification. This example illustrates the difficulty of designing unnatural bases for manipulating DNA properties. Too many parame-

N dR N N O

H

N

T•CG

Figure 1.4.11

N dR

O H H

4H

39

N

N

N N dR

N H

N N

H

NH O

H

N

H H N

40

N dR

O

H

O

N

N

dR N

N

N H

N N

H

O

dR

Nucleobase analogs for triplex recognition of CG base pairs.



Supplement 5

ters contribute to base pair stability to enable unambiguous prediction of the effects on duplex properties of even minor structural changes in the heterocycle. Yet another way that modified bases may be used to tune hybridization was reported by Kutyavin et al. (1996). 2-Thiothymine base pairs with adenine, but not with 2,6-DAP. This allowed Kutyavin and co-workers to develop a strategy for invasion of double-stranded DNA with formation of a stable three-arm junction. The complementary invading oligonucleotides contain 2-thiothymine and 2,6-DAP and do not form a stable duplex with each other, but they do hybridize effectively with the complementary natural sequences.

t-Bu

O

t-Bu O


O

N

t-Bu

O

N

N

H

H

41 H

N R

N

O

N H N

H N

O

H N

N R

N H

G•G mismatch recognition N

BASE PAIR RECOGNITION The unifying characteristic of the compounds discussed in the previous sections is that they are designed to function as nucleic acid components in place of natural nucleotides. However, the scope and potential uses for new molecules designed for nucleobase recognition extend well beyond those described above for unnatural nucleosides. Work on the recognition of base sequences through azole oligomers that bind to the nucleobases through the minor groove has led to the development of highly specific inhibitors of duplex DNA. Design of analogs that read single base pairs provides the opportunity for development of highly sensitive mismatch detection, which could in turn lead to tools for discovery of single-nucleotide polymorphisms. One example is the bisnaphthyridine intercalator (S.42) that binds specifically to GG base pairs (Fig. 1.4.12; Nakatani et al., 2001). It should be possible to design other analogs that selectively bind other mismatches. A second example illustrated in the figure is the hexylureido isoindolin-1-one derivative (S.41), which can associate with both bases of a CG base pair through hydrogen bonding in the major groove. The molecular details of recognition of multiple base pairs in a sequence through specific association of the bases with DNA-binding molecules such as netropsin are well recognized. Extensive research by Dervan and coworkers has cumulated in the design of polyamides capable of binding long duplex segments with high specificity through hydrogen bonding interactions in the minor groove (Dervan and Burli, 1999; Gottesfeld et al., 2000).

C•G base pair recognition

t-Bu

O N

N

H

H

N

N

N H

H

O N N R

O

N R

O H H N

N

N H

N

H

N

N N N

42

R

Figure 1.4.12 Molecules that recognize and bind combinations of natural bases.

CONCLUSION The objective of this review is to highlight a rapidly developing area of nucleic acid chemistry: unnatural base design. The design of new unnatural base pairs, universal bases, triplex components, and other bases with unusual base pairing specificities will continue to provide an attractive arena for molecular designers. It is hoped that the unit will provide some guidance and inspiration for synthetic chemists seeking problems in nucleic acid chemistry. Many potential uses for unnatural bases have been identified, and in most instances totally successful solutions have not yet been established.

LITERATURE CITED Berger, M., Ogawa, A.K., McMinn, D.L., Wu, Y., Schultz, P.G., and Romesberg, F.E. 2000a. Stable and selective hybridization of oligonucleotides with unnatural hydrophobic bases. Angew. Chem. Int. Ed. Engl. 39:2940-2942. Berger, M., Wu, Y., Ogawa, A.K., McMinn, D.L., Schultz, P.G., and Romesberg, F.E. 2000b. Universal bases for hybridization, replication and chain termination. Nucl. Acids Res. 28:29112914.

1.4.10 Supplement 5


Bergstrom, D.E., Zhang, P., Toma, P.H., Andrews, C.A., and Nichols, R. 1995. Synthesis, structure, and deoxyribonucleic acid sequencing with a universal nucleoside: 1-(2′-Deoxy-β-D-ribofuranosyl)-3-nitropyrrole. J. Am. Chem. Soc. 117:1201-1209.

Eritja, R., Horowitz, D.M., Walker, P.A., ZiehlerMartin, J.P., Boosalis, M.S., Goodman, M.F., Itakura, M., and Kaplan, B.E. 1986. Synthesis and properties of oligonucleotides containing 2′-deoxynebularine and 2′-deoxyxanthosine. Nucl. Acids Res. 14:8135-8153.

Bergstrom, D.E., Zhang, P., and Johnson, W.T. 1996. Design and synthesis of heterocyclic carboxamides as natural nucleic acid mimics. Nucleosides Nucleotides 15:59-68.

Ganesh, K.N., Kumar, V.A., and Barawkar, D.A. 1996. Synthetic control of DNA triplex structure through chemical modifications. In Supramolecular Control of Structure and Bonding (A.D. Hamilton, ed.) pp. 263-327. John Wiley & Sons, New York.

Bergstrom, D.E., Zhang, P., and Johnson, W.T. 1997. Comparison of the base pairing properties of a series of nitroazole nucleobase analogs in the oligod eoxyribonucleotide sequence 5′d(CGCXAATTYGCG)-3′. Nucl. Acids Res. 25:1935-1942. Bijapur, J., Keppler, M.D., Bergqvist, S., Brown, T., and Fox, K.R. 1999. 5-(1-Propargylamino)-2′deoxyuridine (Up): A novel thymidine analogue for generating DNA triplexes with increased stability. Nucl. Acids Res. 27:1802-1809. Brown, D.M. and Lin, P.K.T. 1991a. The structure and application of oligodeoxyribonucleotides containing modified, degenerate bases. Nucl. Acids Symp. Ser. 24:209-212. Brown, D.M. and Lin, P.K.T. 1991b. Synthesis and duplex stability of oligonucleotides containing adenine-guanine analogues. Carbohydr. Res. 216:129-139. Carbonnaux, C., Fazakerley, G.V., and Sowers, L.C. 1990. An NMR structural study of deaminated base pairs in DNA. Nucl. Acids Res. 18:40754081. Cassidy, S.A., Slickers, P., Trent, J.O., Capaldi, D.C., Roselt, P.D., Reese, C.B., Neidle, S., and Fox, K.R. 1997. Recognition of GC base pairs by triplex-forming oligonucleotides containing nucleosides derived from 2-aminopyridine. Nucl. Acids Res. 25:4891-4898. Corfield, P.W.R., Hunter, W.N., Brown, T., Robinson, P., and Kennard, O. 1987. Inosine-adenine base pairs in a B-DNA duplex. Nucl. Acids Res. 15:7935-7949. Cruse, W.B.T., Aymani, J., Kennard, O., Brown, T., Jack, A.G.C., and Leonard, G.A. 1989. Refined crystal structures of an octanucleotide duplex with I.T. mismatch base pairs. Nucl. Acids Res. 17:55-72. Day, J.P., Bergstrom, D., Hammer, R.P., and Barany, F. 1999a. Nucleotide analogs facilitate base conversion with 3′-mismatch primers. Nucl. Acids Res. 27:1810-1818. Day, J.P., Hammer, R.P., Bergstrom, D., and Barany, F. 1999b. Nucleotide analogs and new buffers improve a generalized method to enrich for low abundance mutations. Nucl. Acids Res. 27:18191827. Dervan, P.B. and Burli, R.W. 1999. Sequence-specific DNA recognition by polyamides. Curr. Opin. Chem. Biol. 3:688-693. Doronina, S.O. and Behr, J.-P. 1997. Towards a general triple helix mediated DNA recognition scheme. Chem. Soc. Rev. 26:63-71.

Gottesfeld, J.M., Turner, J.M., and Dervan, P.B. 2000. Chemical approaches to control of gene expression. Gene Expr. 9:77-91. Gowers, D.M. and Fox, K.R. 1999. Towards mixed sequence recognition by triple helix formation. Nucl. Acids Res. 27:1569-1577. Guckian, K.M. and Kool, E.T. 1997. Highly precise shape mimicry by a difluorotoluene deoxynucleoside, a replication-competent substitute for thymidine. Angew. Chem. Int. Ed. Engl. 36:2825-2828. Guckian, K.M., Morales, J.C., and Kool, E.T. 1998. Structure and base pairing properties of a replicable nonpolar isostere for deoxyadenosine. J. Org. Chem. 63:9652-9656. Guo, Z., Liu, Q., and Smith, L.M. 1997. Enhanced discrimination of single nucleotide polymorphisms by artificial mismatch hybridization. Nat. Biotechnol. 15:331-335. Habener, J.F., Vo, C.D., Le, D.B., Gryan, G.P., Ercolani, L., and Wang, A.H.J. 1988. 5-Fluorodeoxyuridine as an alternative to the synthesis of mixed hybridization probes for the detection of specific gene sequences. Proc. Natl. Acad. Sci. U.S.A. 85:1735-1739. Hill, F., Loakes, D., and Brown, D.M. 1998. Polymerase recognition of synthetic oligodeoxyribonucleotides incorporating degenerate pyrimidine and purine bases. Proc. Natl. Acad. Sci. U.S.A. 95:4258-4263. Hoops, G.C., Zhang, P., Johnson, W.T., Paul, N., Bergstrom, D.E., and Davisson, V.J. 1997. Template directed incorporation of nucleotide mixtures using azole-nucleobase analogs. Nucl. Acids Res. 25:4866-4871. Horlacher, J., Hottiger, M., Podust, V.N., Hubscher, U., and Benner, S.A. 1995. Recognition by viral and cellular DNA polymerases of nucleosides bearing bases with nonstandard hydrogen bonding patterns. Proc. Natl. Acad. Sci. U.S.A. 92:6329-6333. Huang, C.-Y., Bi, G., and Miller, P.S. 1996. Triplex formation by oligonucleotides containing novel deoxycytidine derivatives. Nucl. Acids Res. 24:2606-2613. Inoue, H., Imura, A., and Ohtsuka, E. 1985. Synthesis and hybridization of dodecadeoxyribonucleotides containing a fluorescent pyridopyrimidine deoxynucleoside. Nucl. Acids Res. 13:7119-7128.



Supplement 5

Johnson, W.T., Zhang, P., and Bergstrom, D.E. 1997. The synthesis and stability of oligodeoxyribonucleotides containing the deoxyadenosine mimic 1-(2′-deoxy-β- D-ribofuranosyl)imidazole-4carboxamide. Nucl. Acids Res. 25:559-567.

Matray, T., Gamsey, S., Pongracz, K., and Gryaznov, S. 2000. A remarkable stabilization of complexes formed by 2,6-diaminopurine oligonucleotide N3′→P5′ phosphoramidates. Nucleosides Nucleotides Nucleic Acids 19:1553-1567.

Kawase, Y., Iwai, S., and Ohtsuka, E. 1989. Synthesis and thermal stability of dodecadeoxyribonucleotides containing deoxyinosine pairing with four major bases. Chem. Phamacol. Bull. 37:599-601.

McMinn, D.L., Ogawa, A.K., Wu, Y., Liu, J., Schultz, P.G., and Romesberg, F.E. 1999. Efforts towards expansion of the genetic alphabet: DNA polymerase recognition of a highly stable, self pairing hydrophobic base. J. Am. Chem. Soc. 121:11585-11586.

Klewer, D., Zhang, P., Bergstrom, D.E., Davisson, V.J., and Liwang, A.C. 2001. Conformations of nucleoside analog 1-(2′-deoxy-β-D-ribofuranosyl)-1,2,4-triazole-3-carboxamide in DNA duplexes with different sequence contexts. Biochemistry 40:1518-1527. Koh, J.S. and Dervan, P.B. 1992. Design of a nonnatural deoxyribonucleoside for recognition of GC base pairs by oligonucleotide-directed triplehelix formation. J. Am. Chem. Soc. 114:14701478. Kool, E.T. 1998. Replication of non-hydrogen bonded bases by DNA polymerases: A mechanism for steric matching. Biopolymers 48:3-17. Krawczyk, S.H., Milligan, J.F., Wadwani, S., Moulds, C., Froehler, B.C., and Matteucci, M.D. 1992. Oligonucleotide-mediated triple helix formation using an N3-protonated deoxycytidine analog exhibiting pH-independent binding within the physiological range. Proc. Natl. Acad. Sci. U.S.A. 89:3761-3764. Kutyavin, I.V., Lukhtanov, E.A., Gorn, V.V., Meyers, R.B. Jr., and Gamper, H.B. Jr. 1996. Oligonucleotides containing 2-aminoadenine and 2-thiothymine act as selectively binding complementary agents. Biochemistry 35:11170-11176. Lin, T.K.T. and Brown, D.M. 1989. Synthesis and duplex stability of oligonucleotides containing cytosine-thymine analogues. Nucl. Acids Res. 17:10373-10383. Loakes, D. and Brown, D.M. 1994. 5-Nitroindole as an universal base analogue. Nucl. Acids Res. 22:4039-4043. Loakes, D., Brown, D.M., Linde, S., and Hill, F. 1995. 3-Nitropyrrole and 5-nitroindole as universal bases in primers for DNA sequencing and PCR. Nucl. Acids Res. 23:2361-2366. Luo, J., Bergstrom, D.E., and Barany, F. 1996. Improving the fidelity of Thermus thermophilus DNA ligase. Nucl. Acids Res. 24:3071-3078. Luyten, I. and Herdewijn, P. 1998. Hybridization properties of base-modified oligonucleotides within the double and triple helix motif. Eur. J. Med. Chem. 33:515-576. Martin, F.H. and Castro, M.M. 1985. Base pairing involving deoxyinosine: Implications for probe design. Nucl. Acids Res. 13:8927-8938. Matray, T.J. and Kool, E.T. 1999. A specific partner for abasic damage in DNA. Nature 399:704-708.

Meggers, E., Holland, P.L., Tolman, W.B., Romesberg, F.E., and Schultz, P.G. 2000. A novel copper-mediated DNA base pair. J. Am. Chem. Soc. 122:10714-10715. Millican, T.A., Mock, G.A., Chauncey, M.A., Patel, T.P., Eaton, M.A.W., Gunning, J., Cutbush, S.D., Neidle, S., and Mann, J. 1984. Synthesis and biophysical studies of short oligodeoxynucleotides with novel modifications: A possible approach to the problem of mixed base oligodeoxynucleotide synthesis. Nucl. Acids Res. 12:7435-7453. Morales, J.C. and Kool, E.T. 1999. Minor groove interactions between polymerase and DNA: More essential to replication than Watson-Crick hydrogen bonds? J. Am. Chem. Soc. 121:23232324. Moran, S., Ren, R.X.-F., and Kool, E.T. 1997a. A thymidine triphosphate shape analog lacking Watson-Crick pairing ability is replicated with high sequence selectivity. Proc. Natl. Acad. Sci. U.S.A. 94:10506-10511. Moran, S., Ren, R.X.-F., Rumney, S.I., and Kool, E.T. 1997b. Difluorotoluene, a nonpolar isostere for thymine, codes specifically and efficiently for adenine in DNA replication. J. Am. Chem. Soc. 119:2056-2057. Nakatani, K., Sando, S., and Saito, I. 2001. Scanning of guanine-guanine mismatches in DNA by synthetic ligands using surface plasmon resonance. Nat. Biotechnol. 19:51-55. Negishi, K., Williams, D.M., Inoue, Y., Moriyama, K., Brown, D.M., and Hayatsu, H. 1997. The mechanism of mutation induction by a hydrogen bond ambivalent, bicyclic N-4-oxy-2′-deoxycytidine in Escherichia coli. Nucl. Acids Res. 25:1548-1552. Nguyen, N.K., Bonfils, E., Auffray, P., Costaglioli, P., Schmitt, P., Asseline, U., Durand, M., Maurizot, J.C., Dupret, D., and Thuong, N.T. 1998. The stability of duplexes involving AT and/or G(4Et)C base pairs is not dependent on their AT/G(4Et)C ratio content. Implication for DNA sequencing by hybridization. Nucl. Acids Res. 26:4249-4258. Oda, Y., Uesugi, S., Ikehara, M., Kawase, Y., and Ohtsuka, E. 1991. NMR studies for identification of dI:dG mismatch base-pairing structure in DNA. Nucl. Acids Res. 19:5263-5267.


1.4.12 Supplement 5


Ogawa, A.K., Wu, Y., McMinn, D.L., Liu, J., Schultz, P.G., and Romesberg, F.E. 2000. Efforts toward the expansion of the genetic alphabet: Information storage and replication with unnatural hydrophobic base pairs. J. Am. Chem. Soc. 122:3274-3287. Ohtsuka, E., Matsuki, S., Ikehara, M., Takahashi, Y., and Matsubara, K. 1985. An alternative approach to deoxyoligonucleotides as hybridization probes by insertion of deoxyinosine at ambiguous codon positions. J. Biol. Chem. 260:26052608. Piccirilli, J.A., Krauch, T., Moroney, S.E., and Benner, S.A. 1990. Enzymatic incorporation of a new base pair into DNA and RNA extends the genetic alphabet. Nature 343:33-37. Pochet, S. and Marliére, P. 1996. Construction of a self-complementary nucleoside from deoxyguanosine. C. R. Acad. Sci. (Paris) 319:1-7. Prévot-Halter, I. and Leumann, C.J. 1999. Selective recognition of a C-G base pair in the parallel DNA triple-helical binding motif. Bioorg. Med. Chem. Lett. 9:2657-2660. Rothman, J.H. and Richards, W.G. 1996. Novel Hoogsteen-like bases for configurational recognition of the T-A base pair by DNA triplex formation. Biopolymers 39:795-812. Saenger, W. 1984. Principles of Nucleic Acid Structure. Springer-Verlag, New York. Schweitzer, B.A. and Kool, E. 1995. Hydrophobic, non-hydrogen-bonding bases and base pairs DNA. J. Am. Chem. Soc. 117:1864-1872. Seela, F. and Debelak, H. 2000. The N8-(2′-deoxyribofuranoside) of 8-aza-7-deazaadenine: A universal nucleoside forming specific hydrogen bonds with the four canonical DNA constituents. Nucl. Acids Res. 28:3224-3232. Seela, F. and Kaiser, K. 1986. Phosphoramidites of base-modified 2′-deoxyinosine isosteres and solid-phase synthesis of d(GCI*CGC) oligomers containing an ambiguous base. Nucl. Acids Res. 14:1825-1844. Switzer, C.Y., Moroney, S.E., and Benner, S.A. 1993. Enzymatic recognition of the base pair between isocytidine and isoguanosine. Biochemistry 32:10489-10496.

Tanaka, K. and Shionoya, M. 1999. Synthesis of a novel nucleoside for alternative DNA base pairing through metal complexation. J. Org. Chem. 64:5002-5003. Ueno, Y., Mikawa, M., and Matsuda, A. 1998. Synthesis and properties of oligodeoxynucleotides containing 5-[N-[2[N,N-bis(2-aminoethyl)amino]ethyl]carbamoyl]-2′-deoxyuridine and 5[N-[3-[N,N-Bis(3-aminopropyl)amino]propyl] carbamoyl]-2′- deoxyuridine. Bioconjugate Chem. 9:33-39. Uesugi, S., Oda, Y., Ikehara, M., Kawase, Y., and Ohtsuka, E. 1987. Identification of I-A mismatch base-pairing structure in DNA. J. Biol. Chem. 262:6965-6968. Voegel, J.J. and Benner, S.A. 1994. Nonstandard hydrogen bonding in duplex oligonucleotides. The base pair between an acceptor-donor-donor pyrimidine analog and donor-acceptor-acceptor analog. J. Am. Chem. Soc. 116:6929-6930. Wagner, R.W., Matteucci, M.D., Lewis, J.G., Gutierrez, A.J., Moulds, C., and Froehler, B.C. 1993. Antisense gene inhibition by oligonucleotides containing C-5 propyne pyrimidines. Science 260:1510-1513. Wu, Y., Ogawa, A.K., Berger, M., McMinn, D.L., Schultz, P.G., and Romesberg, F.E. 2000. Efforts toward expansion of the genetic alphabet: Optimization of interbase hydrophobic interactions. J. Am. Chem. Soc. 122:7621-7632. Yu, H., Eritja, R., Bloom, L.B., and Goodman, M.F. 1993. Ionization of bromouracil and fluorouracil stimulates base mispairing frequencies with guanine. J. Biol. Chem. 268:15935-15943. Zhang, P., Johnson, W.T., Klewer, D., Paul, N., Hoops, G., Davisson, V.J., and Bergstrom, D.E. 1998. Exploratory studies on azole carboxamides as nucleobase analogs: Thermal denaturation studies on oligodeoxyribonucleotide duplexes containing pyrrole-3-carboxamide. Nucl. Acids Res. 26:2208-2215.

Contributed by Donald E. Bergstrom Purdue University West Lafayette, Indiana



Supplement 5

Development of a Universal Nucleobase and Modified Nucleobases for Expanding the Genetic Code

UNIT 1.5

This unit presents protocols for the synthesis and characterization of nucleosides with unnatural bases in order to develop bases for the expansion of the genetic alphabet or for nonselective pairing opposite natural bases. The faithful pairing of nucleobases through complementary hydrogen-bond (H-bond) donors and acceptors forms the foundation of the genetic code. However, there is no reason to assume that the requirements for duplex stability and replication must limit the genetic alphabet to only two base pairs, or, for that matter, hydrogen-bonded base pairs. Expansion of this alphabet to contain a third base pair would allow for the encoding of additional information and would make possible a variety of in vitro experiments using nucleic acids with unnatural building blocks. Previous efforts to generate orthogonal base pairs have relied on H-bonding patterns that are not found with the canonical Watson-Crick pairs. However, in all cases, the unnatural bases were not kinetically orthogonal, and instead competitively paired with natural bases during polymerase-catalyzed DNA synthesis (Horlacher et al., 1995; Lutz et al., 1996, 1998a,b). Tautomeric isomerism, which would alter H-bond donor and acceptor patterns, likely contributes to this kinetic infidelity (Roberts et al., 1997a,b; Robinson et al., 1998; Beaussire and Pochet, 1999). An alternative strategy is centered around developing unnatural bases that form pairs based not on hydrogen bonds, but rather on interbase hydrophobic interactions. Such hydrophobic bases should not pair stably opposite natural bases due to the forced desolvation of the purines or pyrimidines. Additionally, the use of nucleobase analogs, which are not restricted to the shape of H-bonding topologies of natural bases, allows for the use of a wider range of analogs. This unit describes the design, synthesis, and characterization of unnatural base pairs involving 1-β-D-2-deoxyribosyl-N- and -C-nucleosides. To be reasonable candidates for the modification of the genetic code, unnatural nucleosides must meet certain criteria. First, the unnatural bases must pair stably and selectively in duplex DNA. Second, the unnatural bases must be good substrates for DNA polymerases, being replicated with good efficiency and fidelity. Determination of these thermodynamic and kinetic parameters of the unnatural nucleosides is accomplished by incorporation into oligonucleotides and subsequent evaluation as described herein. A general procedure for the synthesis of 1-β-D-2-deoxyribosyl-N-nucleosides containing pyrimidine-like unnatural hydrophobic bases using a modified Silyl-Hilbert-Johnson reaction (Hilbert and Johnson, 1930), namely Vorbrüggen glycosylation (Niedballa and Vorbrüggen, 1970), is first described (see Basic Protocol 1). This methodology is very useful for the synthesis of pyrimidine-like hydrophobic base pairs, because under the necessary reaction conditions the polar (but otherwise often rather insoluble) pyrimidine bases are readily converted by silylation into lipophilic silyl compounds, which are then soluble in organic solvents. Thus, the homogeneous coupling with sugar moieties is permitted. Next, a general procedure is included for the synthesis of 1-β-D-2-deoxyribosyl-N-nucleosides containing purine-like bases using the metal salt procedure (see Basic Protocol 2), which entails the condensation of nucleobase sodium salts with a sugar halide (Kazimierczuk et al., 1984). The sodium salts of acidic heterocyclic systems such as imidazole, purine, triazole, and pyrazole are prepared in situ with NaH or analogous bases. Contributed by Floyd E. Romesberg, Chengzhi Yu, Shigeo Matsuda, and Allison A. Henry Current Protocols in Nucleic Acid Chemistry (2002) 1.5.1-1.5.36 Copyright © 2002 by John Wiley & Sons, Inc.


1.5.1 Supplement 10

The synthesis of 3,5-dimethylphenyl-C-nucleoside by (1) condensation of in situ–generated Grignard reagent from 1-bromo-3,5-dimethylbenzene with 1-α-chloro-3,5-di-Otoluoyl-2-deoxyribofuranose and (2) methoxide-mediated deprotection of the bistoluoyl groups is also described (see Basic Protocol 3). In addition, the unit includes a procedure for the synthesis of 1,4-dimethylnaphthaleneC-nucleoside (see Basic Protocol 4). The sugar precursor containing the aldehyde group is used; this is synthesized from 2-deoxyribose in seven steps (Eaton and Millican, 1988) and does not require the generation of any stereocenters beyond C1′. The condensation of the aryllithium reagent of 2-bromo-1,4-dimethylnaphthalene with the aldehyde, followed by in situ cyclization and deprotection of hydroxyl groups, affords the desired nucleoside. An Alternate Protocol describes the synthesis of 3-methyl-2-naphthalene-C-nucleoside. A different sugar precursor containing an aldehyde group at C3′ is generated from Felkin-Anh addition of an allylzinc nucleophile to isopropylidene-protected glyceraldehydes with high diastereoselectivity (Solomon and Hopkins, 1993). Mesylation, followed by treatment with excess trifluoroacetic acid, yields a diastereomeric mixture of separable C-nucleosides. The purification of DNA containing these unnatural bases is then presented (see Basic Protocol 5). Purified DNA products are used to characterize the unnatural base pairs as candidate pairs for expanding the genetic code. Finally, the thermodynamic (see Basic Protocol 6) and kinetic characterization (see Basic Protocol 7) of the unnatural bases are described. NOTE: All glassware used for reactions should be evacuated, flame-dried, and flushed with argon before use. All operations should be carried out in a well-ventilated fume hood. All reactions involving moisture-sensitive reagents should be performed under argon atmosphere (see Reagents and Solutions). NOTE: It is recommended that all products in this unit be stored in a desiccator at ≤4°C. BASIC PROTOCOL 1

GENERAL GUIDELINES FOR SYNTHESIS OF PYRIMIDINE-LIKE 1-β-D-2-DEOXYRIBOSYL-N-NUCLEOSIDES This protocol outlines a general procedure for the synthesis of unnatural hydrophobic pyrimidine-like N-nucleosides (Figs. 1.5.1 and 1.5.2; McMinn et al., 1999). Specific protocols are given for the synthesis of 7-propynylisocarbostyril (PICS) triphosphate (S.7), as well as the phosphoramidite of PICS (S.6). Synthesis of the other unnatural nucleosides shown in Figure 1.5.2 can be accomplished using similar synthetic transformations, starting from the appropriate pyrimidine. The introduction of propynyl groups, the deprotection of toluoyl groups, and the formation of the phosphoramidite and triphosphate described here are generally applied to all synthesis protocols in this unit.

Development of a Universal Nucleobase and Unnatural Nucleobases

Materials Argon (see recipe) Isocarbostyril (S.1; Aldrich) or pyrimidine of choice Acetonitrile, anhydrous N,O-Bis(trimethylsilyl)acetamide (Aldrich) Bis-toluoyl-protected chloroglycoside: 1-α-chloro-3,5-di-O-toluoyl-2-deoxyribofuranose (S.2; Berry & Associates; Takeshita et al., 1987)

1.5.2 Supplement 10


TolO O N H 1

O

TolO

a

Cl

+

5'

OTol

HO

d N

O

N

O

O

O

4

TolO

2'

OTol 3

TolO

O 1'

3'

2

b, c

N O

4'

5

OH e, f g

DMTrO N

O

PiPiPiO

O

N

O

O O P

6

N(i -Pr) 2

O

OH

CN

7

Figure 1.5.1 General procedure for synthesis of unnatural hydrophobic pyrimidine-like N-nucleosides (see Basic Protocol 1). Reagents: (A) Bis-acetamide, SnCl4, CH3CN, 0°C to room temperature (steps 1 to 22); (B) ICl, CH2Cl2, reflux to room temperature (steps 23 to 30); (C) propyne, (Ph3P)2PdCl2, CuI, TEA, −78°C to room temperature (steps 31 to 43); (D) 0.5 M sodium methoxide, methanol, room temperature (steps 44 to 49); (E) DMTrCl, pyridine, room temperature (steps 50 to 62); (f) 2-cyanoethyl-diisopropylchlorophosphoramidite, DIPEA, CH2Cl2, 0°C (steps 63 to 71); (g) tetrabutylammonium pyrophosphate, POCl3, tributylamine, Proton-Sponge, trimethyl phosphate, 0°C (steps 72 to 80).

N

O

N

O

N

O

N

O

ICS

MICS

5MICS

Pyridone

(isocarbostyril)

(3-methylisocarbostyril)

(5-methylisocarbostyril)

(3,5-dimethyl-2-pyridone)

N

O

N

O

N

O

PIM

PPyridone

PICS

(7-propynyl-3-methylisocarbostyril)

(7-propynyl-3-methyl-2( 1H )-pyridone)

(7-propynylisocarbostyril)

Figure 1.5.2 Unnatural hydrophobic N-nucleosides (see Basic Protocol 1).



Supplement 10

SnCl4, anhydrous, freshly distilled in vacuo Ethyl acetate Hexanes Ammonium molybdate solution (see recipe) Saturated sodium bicarbonate (NaHCO3) solution Saturated sodium chloride (NaCl) solution Sodium sulfate (Na2SO4), anhydrous Silica gel, 200 to 400 mesh, 60 Å Dichloromethane (CH2Cl2), freshly distilled from calcium hydride Ethyl ether, anhydrous Iodine monochloride (ICl; Aldrich; packaged under nitrogen in Sure/Seal vials) Saturated sodium thiosulfate (Na2S2O3) solution Magnesium sulfate (MgSO4), anhydrous Triethylamine (TEA), freshly distilled from calcium hydride Dichlorobis(triphenylphosphine) palladium(II) [(Ph3P)2PdCl2] (Aldrich) Copper(I) iodide (CuI) Dry ice/ethyl ether bath (–100°C) and dry ice/acetone bath Propyne Sodium methoxide (Aldrich) Methanol, anhydrous, distilled from magnesium turnings and stored over 3A molecular sieves Ammonium chloride (NH4Cl) Pyridine, anhydrous, distilled and stored on NaOH protected from light 4,4′-Dimethoxytrityl chloride (DMTr-Cl; Aldrich) Diisopropylethylamine (DIPEA), freshly distilled from calcium hydride 2-Cyanoethyl diisopropylchlorophosphoramidite (Aldrich) Trimethyl phosphate Proton-Sponge (Aldrich) Phosphorous oxychloride (POCl3), freshly distilled Tributylamine Tetrabutylammonium pyrophosphate (TBAP), stored over Drierite 1 M triethylammonium bicarbonate (TEAB), pH 7


10-, 25-, and 100-mL two-neck round bottom flasks with 14/20 joints, oven-dried Rubber septa Vacuum system (oil pump) with manifold and cold trap Silica gel 60 F254 alumina-backed thin-layer chromatography (TLC) plates (Fisher) Heat gun UV light source 100-, 250-, and 500-mL separatory funnels 300-mL round bottom flask Cotton Glass funnel Rotary evaporator (Büchi) equipped with a dry ice condenser and a vacuum system Heavy-walled glass columns (1.5-cm i.d.; 10-, 15-, and 20-cm length) with glass adapters attached to compressed air or nitrogen source (see flash chromatography steps in APPENDIX 3E) Sea sand Reflux condenser with 14/20 joint 23- and 18-G needles 100-mL three-neck flask with 14/20 joints Propyne inlet adapter (Aldrich, cat. no. Z41.577-4) Teflon caps for 14/20 joints

1.5.4 Supplement 10


Speedvac evaporator (Savant) High-performance liquid chromatography (HPLC) system Glass or disposable plastic syringes 10-mL test tubes Oil bath with temperature controller Additional reagents and equipment for thin-layer chromatography (TLC; APPENDIX 3D) and flash chromatography (APPENDIX 3E) NOTE: Glass or disposable plastic syringes are generally used for addition of liquid reagents. The 10-mL test tubes are used for collection of eluting fractions in column chromatography. Oil baths with temperature controllers are used for heating reactions as necessary. Perform condensation of isocarbostyril (S.1) with bis-toluoyl-protected chloroglycoside (S.2) 1. Place a magnetic stir bar in an oven-dried 100-mL two-neck flask with two rubber septa, and place the flask on top of a magnetic stir plate. 2. Evacuate reaction flask on vacuum line, then flush with argon. Repeat this procedure three times and attach the flask to an argon line on the manifold. 3. Quickly remove the rubber septum on one of the side inlets, add 0.50 g (3.4 mmol) isocarbostyril (S.1) to the flask under argon, then immediately reinsert the septum. 4. Transfer 10 mL acetonitrile to the flask under argon. 5. Add, in dropwise fashion, 0.85 mL (3.4 mmol) N,O-bis(trimethylsilyl)acetamide and stir vigorously for 40 min at room temperature. During this time the suspension is cleared.

6. Add an additional 12 mL of acetonitrile to the clear reaction mixture, followed by 1.10 g (2.80 mmol) bis-toluoyl-protected chloroglycoside (S.2). Cool the reaction mixture in an ice-water bath. 7. Slowly add 0.18 mL (2 mmol) SnCl4 and continue to stir under argon at 0°C. Preparation of S.3 uses 1.0 eq S.1, 0.8 eq S.2, 1.0 eq N,O-bis(trimethylsilyl)acetamide, 0.6 eq SnCl4, and 129.4 eq acetonitrile. Increasing the quantity of SnCl4 from catalytic to substoichiometric amounts may improve yield.

8. Monitor the progress of the reaction by analytical TLC (APPENDIX 3D) as follows: a. Occasionally withdraw a small sample (1 to 5 µL) using a capillary tube and spot on silica gel 60 F254 alumina-backed plates. b. Develop the plate (APPENDIX 3D) with 4:1 (v/v) hexanes/ethyl acetate. c. Visualize by dipping plate into ammonium molybdate solution and heating with a heat gun. Examine with UV light source. For S.3, Rf = 0.4 (see APPENDIX 3D). To analyze the course of reactions by TLC, the anisaldehyde/sulfuric acid staining system (6 g of p-anisaldehyde, 50 mL of absolute ethanol and 2.5 mL of concentrated H2SO4) is useful for nucleosides. For nucleosides containing a DMTr group, 3% TFA/CH2Cl2 is also effective. Ammonium molybdate is used to assay the presence of carbohydrate derivatives and is sufficient for visualization of S.3. Spots generally appear in various shades of purple and blue.



Supplement 10

Work up and purify S.3 9. When TLC analysis indicates the reaction is complete, transfer the mixture to a 500-mL separatory funnel and add 250 mL ethyl acetate. 10. Extract twice, each time with 50 mL saturated NaHCO3 solution, then extract once with 100 mL saturated NaCl solution. 11. Insert cotton into a glass funnel and add a 2-cm-thick layer of anhydrous Na2SO4 over the cotton. Filter the organic layer by gravity into a 300-mL round-bottom flask. 12. Remove solvents under reduced pressure with a rotatory evaporator equipped with a dry ice condenser. It is recommended that the bath temperature not exceed 45°C.

13. Pack a 1.5 × 20–cm heavy-walled column with 60 g of 200 to 400 mesh silica gel in hexanes for flash column chromatography (APPENDIX 3E). Gas pressure (air or nitrogen) is applied via a glass adapter to the top of the column through the gas inlet to increase the flow rate to about 5 drops/sec.

14. Dissolve the crude product in 1 mL of dichloromethane. 15. Transfer the solution using a 3-mL glass pipet onto the silica gel and let the solution sink into the column bed. 16. Rinse the flask with two 1-mL aliquots of dichloromethane, add each rinse to the surface of silica gel, and allow the rinse to enter the column. 17. Carefully add a 0.5-cm layer of sea sand to the top of the silica gel bed. This step is optional. Keeping the surface of the silica gel flat is important for good separation.

18. Elute with a gradient of 8:1 to 4:1 (v/v) hexanes/ethyl acetate. Use a glass cylinder to freshly mix the solvents.

19. Collect 10-mL fractions and analyze by TLC (see step 8). To achieve the best separation, TLC using 5:1 hexanes/ethyl acetate should give an Rf value of ∼0.35 for the possible product.

20. Transfer fractions containing pure product to a 300-mL round-bottom flask and evaporate to dryness with a rotary evaporator. 21. Add 3 mL of anhydrous ethyl ether to triturate the product. Remove the solvent and evacuate the product until the weight of the flask is not changed. 22. Confirm the desired product by 1H NMR, NOESY, and COSY analysis (see, e.g., Lambert et al., 1998). Flash chromatography gives a pair of diastereoisomers as the α and β anomers. The desired product is the faster-migrating β anomer, which is a white foam. The assignment of β-stereochemistry at C1′ for each nucleoside is based on NOESY data, in which H1′ shows cross-peaks with both α-H4′ and α-H2′ (see Fig. 1.5.1). The optimal yield of 1′-β-3′,5′-O-toluoyl-2′-deoxyribosylisocarbostyril (S.3) is 48%.


Iodinate S.3 by Friedel-Crafts iodination 23. Assemble a 25-mL oven-dried two-neck flask with a 14/20 reflux condenser in the center neck, a rubber septum on the side neck, and a magnetic stirring bar. Maintain a smooth flow of argon using a 23-G needle inserted through another septum on top of the reflux condenser.

1.5.6 Supplement 10


24. Mix 0.40 g (0.80 mmol) S.3 (from step 21) with 5 mL freshly distilled dichloromethane (CH2Cl2) and transfer solution to the flask under argon atmosphere. 25. Add, in a dropwise fashion, 0.96 mL of 1 M ICl in CH2Cl2 (0.96 mmol ICl). 26. Reflux the reaction mixture 1 min at 45°C, then cool to room temperature. Perform TLC analysis on the mixture as in step 8. CAUTION: Iodine monochloride is highly toxic and volatile and should be handled with care in a well-ventilated hood.

Work up the iodinated product 27. When TLC analysis indicates that all of the starting material has disappeared during the cooling period, add 10 mL saturated NaHCO3 followed by dropwise addition of saturated Na2S2O3 until the reaction solution is clear. 28. Transfer the mixture to a 100-mL separatory funnel. Extract the aqueous layer three times, each time with 20 mL dichloromethane. 29. Combine the organic extracts and dry over anhydrous MgSO4. 30. Repeat steps 11 and 12 and attach a vacuum line to dry the residue. The resulting crude iodinated product (not shown in the figures) is pure enough for use in the next step. It should be kept in darkness and used immediately.

Introduce propynyl group by Sonogashira coupling 31. Equip a 100-mL, 14/20 three-neck flask with rubber septa on the two side necks, a 14/20 dry-ice condenser-trap in the center neck, and a magnetic stir bar. Place a gas inlet adapter on top of the condenser and attach it to an argon line on the manifold. 32. Evacuate the apparatus and then purge it with argon three times to remove air. 33. Transfer ∼0.80 mmol of the iodinated compound and 15 mL triethylamine (TEA) to the flask and quickly add 0.01 g (0.02 mmol) (Ph3P)2PdCl2 and 0.01 g (0.06 mmol) CuI under argon. (Ph3P)2PdCl2 and CuI are highly sensitive to oxygen and light, respectively. They should be handled very carefully and quickly. High quality of these reagents is essential to the success of Sonogashira coupling reactions.

34. Cool the reaction vessel in dry ice/diethyl ether bath (−100°C). 35. Fill the condenser-trap with pulverized dry ice. 36. Quickly remove the rubber septum on one of the side inlets and insert a propyne inlet adapter. 37. Purge the flask with propyne for 10 min or until final volume is 50 mL. Preparation of S.4 uses 1.0 eq S.3, 1.2 eq ICl, 116.2 eq CH2Cl2, 0.07 eq CuI, 0.02 eq (Ph3P)2PdCl2, 134.5 eq TEA, and >120 eq propyne.

38. Turn off the propyne cylinder. Quickly remove the rubber septa and the propyne inlet, and seal inlets tightly with Teflon caps. 39. Remove dry-ice condenser and seal the inlet tightly with a Teflon cap. Continue to stir at room temperature for 5 hr. Synthesis of Modified Nucleosides


Supplement 10

Work up and purify S.4 40. Cool the reaction mixture in dry ice/acetone bath for 5 min, then remove the cold bath. 41. Remove the caps carefully to vent the remaining propyne at ambient temperature. Evaporate the solvent with a rotary evaporatory. 42. Purify the crude product by flash chromatography (steps 13 to 19) using a 1.5 × 15–cm silica gel column packed with 30 g silica gel. Elute with a gradient of 5:1 to 3:1 (v/v) hexanes/ethyl acetate. Perform TLC on the eluate (see step 8) using 4:1 hexanes/ethyl acetate. For S.4, Rf = ∼0.35.

43. Combine appropriate fractions based on TLC analysis and evaporate to dryness. Flash column chromatography yields 0.11 g (64%) of S.4 as a white foam.

Deprotect 3′- and 5′-OH groups 44. Equip a 10-mL two-neck flask with two rubber septa and a stir bar. Keep the vessel under argon using an 18-G needle threaded through the septum. 45. Add 0.10 g (0.20 mmol) S.4 (from step 43) and 4 mL anhydrous methanol sequentially to the flask. 46. Add dropwise 1.2 mL of 0.5 M sodium methoxide in methanol (0.60 mmol sodium methoxide) to the reaction mixture and continue to stir under argon for 20 min. Preparation of S.5 uses 1.0 eq S.4, 3 eq sodium methoxide, and 617 eq methanol.

Work up and purify S.5 47. Add ∼30 mg NH4Cl to quench the reaction and keep stirring for 10 min. 48. Evaporate solvents and purify the product by flash chromatography (steps 13 to 19) using a 1.5 × 10–cm silica gel column packed with 20 g silica gel. Elute with a gradient of 2:100 to 5:100 (v/v) methanol/CH2Cl2. Perform TLC on the eluate (see step 8) using 4:100 methanol/CH2Cl2. For S.5, Rf = ∼0.35.

49. Combine appropriate fractions based on TLC analysis and evaporate to dryness. The reaction is approximately quantitative and the product, 1-β-2′-deoxyribosyl-7propynylisocarbostyril (S.5), is obtained as a white solid.

Tritylate 5′-OH by Williamson etherification 50. Equip an oven-dried two-neck 10-mL flask with a stir bar and two rubber septa. Keep the vessel under argon using an 18-G needle threaded through the septum. 51. Add 0.046 g (0.154 mmol) S.5 (from step 49) and 4 mL anhydrous pyridine to the flask. 52. Using an 18-G needle, insert a vacuum line through one of the septa to remove most of the pyridine by vacuum evaporation. 53. Purge the reaction flask with argon and remove the vacuum line immediately. Development of a Universal Nucleobase and Unnatural Nucleobases

54. Transfer 0.7 mL anhydrous pyridine to the flask to dissolve S.5.

1.5.8 Supplement 10


55. Prepare a solution of 0.08 g (0.23 mmol) 4,4′-dimethoxytrityl chloride in 0.3 mL pyridine and add in a dropwise fashion to the reaction mixture over a period of 20 min. Tritylation uses 1.0 eq S.5, 1.5 eq DMTr-Cl, and 82.3 eq pyridine.

56. Continue to stir at room temperature for an additional 20 min. Work up and purify tritylated nucleoside 57. Combine the reaction mixture with 20 mL of ethyl acetate, 5 mL saturated NaHCO3, and 5 mL saturated NaCl to partition. 58. Transfer the mixture to a 250-mL separatory funnel. Rinse the reaction vessel twice, each time with 5 mL of ethyl acetate, and transfer to the funnel. 59. Separate the organic layer and extract the aqueous layer twice, each time with 10 mL of ethyl acetate. 60. Filter and dry the organic layer as in steps 11 and 12. 61. Purify the crude product by flash chromatography (steps 13 to 19) using a 1.5 × 10–cm silica gel column packed with 20 g silica gel. Elute with a gradient of 5:1 to 3:1 (v/v) hexanes/ethyl acetate. Perform TLC on the eluate (see step 8) using 3.5:1 hexanes/ethyl acetate. For the DMTr-protected product of the above reaction, Rf = ∼0.35.

62. Combine appropriate fractions based on TLC analysis and evaporate to dryness. Chromatographic purification affords the DMTr-protected product (0.073 g, 79%) as a white foam.

Phosphitylate to give phosphoramidite S.6 63. Equip a 10-mL two-neck flask with two rubber septa and a stir bar. Keep the vessel under argon using an 18-G needle threaded through the septum. 64. Prepare a solution of 0.073 g (0.121 mmol) tritylated product (from step 62) in 1.3 mL of freshly distilled dichloromethane and transfer to the flask. 65. Add 0.084 mL (0.48 mmol) DIPEA to the solution and cool in ice-water bath. 66. Add, in a dropwise fashion, 0.041 mL (0.16 mmol) 2-cyanoethyl diisopropylchlorophosphoramidite and allow the reaction mixture to reach ambient temperature over 15 min. Relative to the starting nucleoside (S.5 in step 51), preparation of S.6 uses 3.2 eq DIPEA, 1.1 eq 2-cyanoethyldiisopropylchlorophosphoramidite, and 135.2 eq dichloromethane.

Work up and purify S.6 67. Dilute the reaction mixture with 50 mL dichloromethane. Transfer the reaction mixture to a 100-mL separatory funnel. 68. Wash with 5 mL saturated NaHCO3 and 5 mL saturated NaCl. 69. Filter and dry as in steps 11 and 12. 70. Perform flash chromatography (steps 13 to 19). Elute with a gradient of 6:1 to 4:1 (v/v) hexanes/ethyl acetate. Perform TLC (see step 8) using 5:1 hexanes/ethyl acetate. For S.6, Rf = ∼0.35.



Supplement 10

71. Combine appropriate fractions based on TLC analysis and evaporate to dryness using a rotary evaporator. The phosphoramidite product (S.6) is a white foam (63 mg), prepared with 65% yield, and is stored in a desiccator at −20°C. Phosphoramidite compounds are used for automated DNA synthesis; isolation and desalting of oligonucleotides is carried out by PAGE and electrophoretic dialysis (see Basic Protocol 5). Oligonucleotides are used for duplex oligonucleotide denaturation temperature measurements and kinetic studies of DNA extension (see Basic Protocols 6 and 7).

Prepare triphosphate S.7 72. Equip a 10-mL two-neck flask with two rubber septa and a stir bar. Keep the vessel under argon using an 18-G needle threaded through the septum. 73. Quickly remove the septum on one of the side inlets, add 0.008 g (0.027 mmol) S.5 (from step 49) under argon, and insert the rubber septum back to the flask. 74. Coevaporate moisture and methanol twice from 0.1 mL anhydrous pyridine. 75. Add 0.134 mL (0.23 mmol) trimethyl phosphate and 9 mg (0.04 mmol) ProtonSponge, dissolve completely, and bring to 0°C. 76. Add 0.003 mL (0.032 mmol) POCl3 and continue to stir at 0°C for 2 hr. 77. Add 0.042 mL (0.03 mmol) tributylamine and 0.023 g (0.040 mmol) TBAP and stir for 1 min. Work up and purify S.7 78. Add 2.7 mL of 1 M TEAB, pH 7. 79. Stir the reaction mixture for 10 min and concentrate to dryness in a Speedvac evaporator. 80. Dissolve the crude product in 200 µL of 1:1 (v/v) DMSO/isopropanol and purify by HPLC (UNIT 10.5) using the following conditions: Column: Rainin C18-Dynamax 60 Å column (7.8 × 300 mm) Buffer A: 0.1 M TEAB, pH 7 Buffer B: Acetonitrile Gradient: 98% to 90% buffer A over 5 min, 90% to 75% buffer A over 25 min, 75% to 0% buffer A over 5 min, return to 98% A over 5 min. Flow rate: 10 mL/min. Approximately 1 mg of the product triphosphate (S.7) is yielded with a retention time of 27.5 min, where the gradient is ∼65% buffer A. The triphosphate compound could be used directly for kinetic studies of DNA incorporation. BASIC PROTOCOL 2


GENERAL GUIDELINES FOR SYNTHESIS OF PURINE-LIKE 1-β-D-2-DEOXYRIBOSYL-N-NUCLEOSIDES This protocol outlines a general procedure for the synthesis of unnatural hydrophobic purine-like N-nucleosides. The coupling strategy between 5H-pyrrolo[2,3-b]pyrazine (PP) and S.2 (Fig. 1.5.1) to generate the nucleoside 1′-β-3′,5′-O-toluoyl-2′-deoxyribosylN-5H-pyrrolo[2,3-b]pyrazine (S.8; Fig. 1.5.3) can be generally applied for the synthesis of other purine-like nucleosides such as those depicted in Figure 1.5.4 by starting with the appropriate purine analog. This protocol describes the synthesis of the 3′,5′-toluoylprotected nucleoside (S.8). Synthesis of the free nucleoside, phosphoramidite, and triphosphate are performed as in Basic Protocol 1.

1.5.10 Supplement 10


N

TolO

N

TolO N

O

+

N H

N

O

Cl OTol

PP

N

OTol

2

8

Figure 1.5.3 General procedure for synthesis of unnatural hydrophobic purine-like N-nucleosides

N N

N

N

N

N

N

7AI

M7AI

ImPy

(7-azaindole)

(6-methyl-7-azaindole)

(imidazole pyridine)

N N

N

N

N

P7AI

PPP

(3-propynyl-7-azaindole)

(3-propynyl-4,7-diazaindole)

Figure 1.5.4 Unnatural hydrophobic purine-like N-nucleosides (see Basic Protocol 2).

Materials Argon (see recipe) 5H-Pyrrolo[2,3-b]pyrazine or purine analog of choice Acetonitrile, anhydrous 60% sodium hydride (NaH; see recipe) Bis-toluoyl-protected chloroglycoside: 1-α-chloro-3,5-di-O-toluoyl-2-deoxyribofuranose (S.2; Berry & Associates; Takeshita et al., 1987) Ethyl acetate Ethyl ether Saturated sodium bicarbonate (NaHCO3) Sodium sulfate (Na2SO4), anhydrous Silica gel, 200-400 mesh, 60 Å Hexanes 10-mL two-neck round-bottom flask with 14/20 joints, oven-dried Rubber septa 18-G needle Cotton Glass funnel 1.5 × 30–cm heavy-walled glass column with glass-adapters attached to compressed air or nitrogen source (see flash chromatography steps in APPENDIX 3E)



Supplement 10

Sea sand Silica gel 60 F254 alumina-backed thin-layer chromatography (TLC) plates (Fischer) Heat gun UV light source Rotary evaporator (Büchi) equipped with a dry ice condenser and a vacuum system Additional reagents and equipment for purification and analysis (see Basic Protocol 1). 1. Equip a 10-mL two-neck flask with two rubber septa and a stir bar. Keep the vessel under argon using an 18-G syringe needle threaded through the septum. 2. Add 100 mg (0.84 mmol) 5H-pyrrolo[2,3-b]pyrazine and 4.5 mL anhydrous acetonitrile and cool the solution 5 min in an ice-water bath. 3. Add 36 mg of 60% NaH (0.9 mmol) in three aliquots to the reaction mixture. Remove the cold bath and stir at ambient temperature for 10 min. 4. Add 326.8 mg bis-toluoyl-protected chloroglycoside (S.2) in three aliquots to the flask and stir at ambient temperature for 10 min. Preparation of S.8 uses 1.0 eq 5H-pyrrolo[2,3-b]pyrazine, 1.1 eq NaH, 1.2 eq S.2, and 523 eq CH3CN.

5. Add 25 mL of ethyl acetate and 20 mL of saturated aqueous NaHCO3 to partition the reaction mixture. 6. Extract the aqueous layer twice, each time with 20 mL ethyl acetate. 7. Dry and filter the organic layer (see Basic Protocol 1, steps 11 and 12). 8. Perform flash chromatography (see Basic Protocol 1, steps 13 to 19) using a 1.5 × 10–cm column packed with 20 g silica gel. Elute with a gradient of 6:1 to 3:1 (v/v) hexanes/ethyl acetate. Analyze fractions by TLC (see Basic Protocol 1, step 8) using 4:1 hexanes/ethyl acetate. For S.8, Rf = ∼0.35.

9. Combine the appropriate fractions and evaporate to dryness using a rotary evaporator. The bis-protected product (S.8) is prepared as a white foam (161 mg) at 41% yield. BASIC PROTOCOL 3


SYNTHESIS OF 3,5-DIMETHYLPHENYL-C-NUCLEOSIDE This protocol describes the synthesis of the 3,5-dimethylphenyl-C-nucleoside, involving the condensation of the Grignard reagent derived from 1-bromo-3,5-dimethylbenzene with the bis-toluoyl-protected chloroglycoside (S.2; Fig. 1.5.1; see Fig. 1.5.5 for the reaction). Deprotection yields the free nucleoside (S.11), followed by generation of the phosphoramidite (S.12) and triphosphate (S.13) derivatives. Two other C-nucleosides are prepared using a similar strategy (TM and 2Nap in Fig. 1.5.6A). Deprotection and formation of the phosphoramidite and triphosphate all follow the general procedures as described in Basic Protocol 1. Materials Argon (see recipe) Magnesium metal turnings Tetrahydrofuran (THF), anhydrous 1-Bromo-3,5-dimethylbenezene (S.9; Aldrich)



Bis-toluoyl-protected chloroglycoside: 1-α-chloro-3,5-di-O-toluoyl-2-deoxyribofuranose (S.2; Takeshita et al., 1987) Ethyl acetate Saturated ammonium chloride (NH4Cl) solution Sodium sulfate, anhydrous Silica gel, 200-400 mesh, 60 Å Hexanes Ethyl ether, anhydrous Methanol, anhydrous Sodium methoxide (Aldrich or synthesized from sodium metal and methanol) Dichloromethane (CH2Cl2), freshly distilled from calcium hydride Pyridine, anhydrous Triethylamine (TEA), freshly distilled from calcium hydride 4,4′-Dimethoxytrityl chloride (DMTr-Cl; Aldrich) Saturated sodium bicarbonate (NaHCO3) solution 4-Dimethylaminopyridine (DMAP; Aldrich)

Me

Me

TolO O Cl OTol

Br

2

9 a

Me

Me

Me b

TolO

HO O

O OTol

OH

10 c, d

Me

Me

Me

PiPiPiO

O

O

P N(i -Pr)2 O

11 e

Me

DMTr O

O

Me

CN 12

OH 13

Figure 1.5.5 Synthesis of 3,5-dimethylphenyl-C-nucleoside and its phosphoramidite and triphosphate derivatives (see Basic Protocol 3). Reagents: (A) Mg0, then S.9, THF (steps 1 to 9); (B) sodium methoxide, methanol (steps 10 to 15); (C) DMTrCl, TEA, pyridine, CH2Cl2 (steps 16 to 24); (D) 2-cyanoethyl-diisopropylchlorophosphoramidite, triethylamine, CH2Cl2 (steps 25 to 31); (E) ProtonSponge, trimethyl phosphate, POCl3, tributylamine, tributylammonium pyrophosphate, DMF (steps 32 to 38).



Supplement 10

2-Cyanoethyl diisopropylchlorophosphoramidite (Aldrich) Trimethyl phosphate (Aldrich) Proton-Sponge (Aldrich) Phosphorus oxychloride (POCl3) Tributylamine n-Tetrabutylammonium pyrophosphate (TBAP; Sigma) Dimethylformamide (DMF) Triethylammonium bicarbonate (TEAB; Fluka) Dimethylsulfoxide (DMSO) Isopropanol 25-mL and 100-mL two-neck round-bottom flasks with 14/20 joints, oven-dried Reflux condenser with 14/20 joint Rubber septa 18- and 23-G needles 125-mL separatory funnel Cotton Glass funnel Rotary evaporator (Büchi) equipped with a dry ice condenser and a vacuum system Heavy-walled glass columns (1.5-cm i.d.; 10- and 15-cm length) with glass adapters attached to compressed air or nitrogen source (see flash chromatography steps in APPENDIX 3E) Sea sand 300-mL round-bottom flask Filter paper Lyophilizer (e.g., Labconco freeze-dry system) Additional reagents and equipment for purification and analysis (see Basic Protocol 1) Perform condensation of 1-bromo-3,5-dimethylbenzene (S.9) with bis-toluoyl-protected chloroglycoside (S.2) 1. Assemble a 25-mL oven-dried two-neck round-bottom flask with a 14/20 reflux condenser in the center neck, a rubber septum on the side neck, and a magnetic stirring bar. Maintain a smooth flow of argon using a 23-G needle inserted through another septum on top of the reflux condenser. 2. Quickly add 52 mg (2.139 mmol) magnesium metal and 2 mL THF to the flask under argon atmosphere. 3. Add, in a dropwise fashion, 293 µL (2.156 mmol) 1-bromo-3,5-dimethylbenzene (S.9) and heat the resulting suspension to 50°C for 1 hr using a temperature-controlled oil bath. 4. Prepare a solution of 126 mg (0.324 mmol) bis-toluoyl-protected chloroglycoside (S.2) in 1 mL THF in another round-bottom flask under argon atmosphere. Add all of the prepared Grignard reagent (step 3) to this solution and stir 14 hr at room temperature. 5. During the 14-hr incubation, prepare another batch of Grignard reagent (repeat steps 1 to 3). At the end of the 14-hr incubation, add another 100 µL Grignard reagent to the reaction mixture (step 4) and continue to stir 1 hr at room temperature. Synthesis of S.10 uses 1.0 eq S.9, 0.992 eq magnesium metal, 0.150 eq S.2, and 17.16 eq THF. Development of a Universal Nucleobase and Unnatural Nucleobases

Work up and purify S.10 6. Add 10 mL of ethyl acetate and 10 mL of saturated aqueous NH4Cl and transfer to a 125-mL separatory funnel. Rinse the reaction vessel twice, each time with 5 mL ethyl acetate, and add to the funnel.



7. Extract separate, and retain the organic layer. Extract the aqueous layer twice, each time with 15 mL of ethyl acetate, and combine these organic layers with the original organic layer. 8. Filter and dry the organic layer (see Basic Protocol, steps 11 to 12). Perform flash chromatography (see Basic Protocol 1, steps 13 to 19). Elute with a gradient of 95:5 to 85:15 (v/v) hexanes/ethyl acetate. Perform TLC (see Basic Protocol 1, step 8) using 9:1 hexanes/ethyl acetate. For S.10, Rf = ∼0.30.

9. Evaporate eluate and analyze product by NMR (see Basic Protocol 1, steps 20 to 22). Approximately 13 mg (9% yield) of purified β-product (S.10) and 48 mg (32%) of α-product should be obtained.

Deprotect 3′- and 5′-OH groups 10. Equip a 100-mL two-neck flask with two rubber septa and a stir bar. Keep the vessel under argon using an 18-G syringe needle threaded through a septum. 11. Add 264 mg (0.576 mmol) S.10 (from step 9) and 10 mL anhydrous methanol to the flask. 12. Add, in a dropwise fashion, 2 mL of 1 M sodium methoxide (2.0 mmol) in methanol to the reaction mixture and keep stirring for 45 min. Synthesis of S.11 uses 1.0 eq S.10, 3.5 eq sodium methoxide, and 514 eq methanol.

Work up and purify S.11 13. Add 100 mg NH4Cl to quench the reaction and continue to stir for 10 min. 14. Evaporate solvents and purify the crude product by flash chromatography (see Basic Protocol 1, steps 13 to 19) using a 1.5 × 10–cm silica gel column packed with 20 g silica gel. Elute with a gradient of 1:99 to 5:95 (v/v) methanol/CH2Cl2. Perform TLC (see Basic Protocol, step 8) using 5:95 methanol/CH2Cl2. For S.11, Rf = ∼0.30.

15. Combine appropriate fractions based on TLC and evaporate to dryness in a rotary evaporator. Flash chromatography gives approximately 105 mg (82% yield) of purified product (S.11) as a white solid.

Tritylate 5′-OH 16. Equip a 100-mL two-neck flask with two rubber septa and a stir bar. Keep the vessel under argon using an 18-G syringe needle threaded through one of the septa. 17. Add 90 mg (0.405 mmol) of S.11 (from step 15), 2 mL anhydrous pyridine, and 2 mL CH2Cl2 to the flask. 18. Add, in a dropwise fashion, 0.30 mL (2.15 mmol) TEA, followed by 175 mg (0.516 mmol) 4,4′-DMTr-Cl in two portions over a 30-min period, and continue to stir for 2 hr at room temperature. Tritylation uses 1.0 eq S.11, 1.27 eq DMTr-Cl, 5.31 eq triethylamine, 61 eq pyridine, and 77 eq dichloromethane.

Work up and purify tritylated product 19. Add 20 mL ethyl acetate and 10 mL saturated NaHCO3 and transfer the mixture to a 125-mL separatory funnel. Rinse the reaction vessel twice, each time with 5 mL ethyl acetate, and add to the funnel.



Supplement 12

20. Extract separate, and retain the organic layer. Extract the aqueous layer twice, each time with 20 mL ethyl acetate. 21. Combine the organic layers and dry over anhydrous Na2SO4. 22. Filter through filter paper into a 300-mL round-bottom flask and evaporate under reduced pressure in a rotary evaporator. 23. Purify the crude product by flash chromatography (see Basic Protocol 1, steps 13 to 19) using a 1.5 × 10–cm silica gel column packed with 20 g silica gel. Elute with a gradient of 1:1 to 4:1 (v/v) hexanes/ethyl acetate. Perform TLC (see Basic Protocol, step 8) using 1:1 hexanes/ethyl acetate. For the tritylated S.11, Rf = ∼0.3.

24. Combine the appropriate fractions based on TLC analysis and evaporate to dryness in a rotary evaporator. Flash chromatography affords 185 mg (87% yield) of purified DMTr-protected product as a white foam.

Prepare phosphoramidite S.12 by phosphitylation 25. Equip a 100-mL two-neck flask with two rubber septa and a stir bar. Keep the vessel under argon using an 18-G syringe needle threaded through one of the septa. 26. Add 185 mg (0.353 mmol) tritylated product (from step 24) and 3.5 mL anhydrous CH2Cl2 to the flask. 27. Add a catalytic amount (2 mg, 0.016 mmol) of DMAP, followed by 0.350 mL (2.512 mmol) triethylamine and 0.160 mL (0.718 mmol) 2-cyanoethyl diisopropylaminochlorophosphoramidite. 28. Continue to stir for 30 min at room temperature. Synthesis of S.12 uses 1.0 eq tritylated nucleoside, 0.045 eq DMAP, 7.1 eq TEA, 2.0 eq 2-cyanoethyl diisopropylaminochlorophosphoramidite, and 155 eq dichloromethane.

Work up and purify S.12 29. Add 20 mL ethyl acetate and 20 mL saturated NaHCO3 and repeat steps 19 to 22 of this protocol. 30. Purify the crude product by flash chromatography (see Basic Protocol 1, steps 13 to 19). Elute with a gradient of 1:9 to 3:7 (v/v) ethyl acetate/hexane containing 5% triethylamine. Perform TLC using 3:7 ethyl acetate/hexane. For S.12, Rf = ∼0.83.

31. Combine appropriate fractions based on TLC and evaporate to dryness. Flash chromatography affords 238 mg (93% yield) of phosphoramidite product (S.12) as a white foam.

Prepare triphosphate S.13 by phosphorylation of S.11 32. Equip a 25-mL two-neck flask with two rubber septa and a stir bar. Keep the vessel under argon using an 18-G needle threaded through the septum. 33. Add 16 mg (0.072 mmol) free nucleoside (S.11, from step 15) to the flask. 34. Add 0.36 mL (3.08 mmol) trimethyl phosphate and 23 mg (0.107 mmol) ProtonSponge. Cool to 0°C. Development of a Universal Nucleobase and Unnatural Nucleobases

35. Add 8 µL (0.090 mmol) POCl3 in a dropwise fashion and continue to stir for 2 hr at 0°C.



36. Add 105 µL (0.441 mmol) tributylamine, followed by a solution of 62 mg (0.171 mmol) TBAP in 0.8 mL DMF, and stir for 1 min. Synthesis of S.13 uses 1.0 eq S.11, 43 eq trimethyl phosphate, 1.5 eq Proton-Sponge, 1.3 eq POCl3, 6.1 eq tributylamine, 2.4 eq TBAP, and 143 eq DMF.

Work up and purify S.13 37. Add 7 mL of 1 M TEAB. Dilute the resulting crude solution ∼10 fold with water and lyophilize. 38. Dissolve in 200 mL of 1:1 (v/v) DMSO/isopropanol and purify by HPLC (UNIT 10.5) using the following conditions: Column: Rainin C18-Dynamax 60 Å column Buffer A: 0.1 M TEAB, pH 7.5 Buffer B: Acetonitrile Gradient: 96% to 70% buffer A from 0 to 30 min Flow rate: 10 mL/min. Approximately 1 mg of triphosphate (S.13) is obtained as a white solid after lyophilization.

SYNTHESIS OF 1,4-DIMETHYLNAPHTHALENE-C-NUCLEOSIDE This protocol describes the synthesis of 1,4-dimethylnaphthalene-C-nucleoside (DMN, see Fig. 1.5.6B), as derived from the condensation of the aryllithium species onto an aldehyde followed by a ring-closing reaction, shown in Figure 1.5.7 (Ogawa et al., 2000a). The steps below detail the synthesis of the free nucleoside (S.18). The preparation of the phosphoramidite and triphosphate are performed as described in Basic Protocol 3. A similar strategy may be used to synthesize the 1-methyl-3-naphthalene nucleoside (3MN; Fig. 1.5.6B), starting from 3-bromo-1-methylnaphthalene (Ogawa, et al., 2000b).

BASIC PROTOCOL 4

Materials Argon (see recipe) 2-Bromo-1,4-dimethylnaphthalene (S.14; Aldrich; Sharma, 1993) Tetrahydrofuran (THF), anhydrous n-Butyllithium (Aldrich) Cyclohexane Protected aldehyde of sugar precursor (S.15; Eaton and Millican, 1988) Ethyl acetate Saturated sodium bicarbonate (NaHCO3) solution Sodium sulfate (Na2SO4), anhydrous Silica gel, 200-400 mesh, 60 Å Hexane Ethyl acetate Pyridine, anhydrous Triethylamine (TEA), freshly distilled from calcium Methanesulfonyl chloride Acetic acid, glacial Methanol, anhydrous Trifluoroacetic acid Dichloromethane (CH2Cl2), freshly distilled from calcium hydride 2,3-Dichloro-5,6-dicyano-1,4-benzoquinone (DDQ; Aldrich) 25-mL, 50-mL, and 100-mL two-neck round bottom flasks with 14/20 joints, oven-dried Rubber septa 18-G needles



Supplement 10

Me

A

Me

4

Me

Me

3

5

2 6

Me

1

TM (trimethylbenzene)

DM (dimethylbenzene)

2Nap (2-naphthalene)

6 7

5

B

Me

4

Me

8

3 2

1

Me

3MN (1-methyl-3-naphthalene)

DMN (dimethylnaphthalene)

C

Me

2MN (3-methyl-2-naphthalene)

Figure 1.5.6 Unnatural hydrophobic C-nucleosides. Synthetic strategies are presented in (A) Basic Protocol 3, (B) Basic Protocol 4, and (C) Alternate Protocol.

OPMB

Me

DMTrO CHO OPMB 15

Me Br 14

a,b

Me

Me Me

DMTrO

c

Me

HO

O OPMB

O 16

OPMB

17

d

Me Me

HO O OH


18

Figure 1.5.7 Synthesis of 1,4-dimethylnaphthalene-C-nucleoside (see Basic Protocol 4). Reagents: (A) n-butyllithium, −78°C, then S.15, THF (steps 1 to 8); (B) methanesulfonyl chloride, triethylamine, pyridine, 0°C (steps 9 to 14); (C) acetic acid (steps 15 to 20); (D) DDQ (steps 21 to 26).



Dry ice/acetone bath (−78°C) Cotton Glass funnel Rotary evaporator (Büchi) equipped with a dry ice condenser and a vacuum system Heavy-walled glass columns (1.5-cm i.d.; 10-, 15-, and 20-cm length) with glass adapters attached to compressed air or nitrogen source (see flash chromatography steps in APPENDIX 3E) Sea sand Additional reagents and equipment for purification and analysis (see Basic Protocol 1) NOTE: n-Butyllithium is flammable and sensitive to moisture. It should be handled carefully and quickly. Perform condensation of 2-bromo-1,4-dimethylnaphthalene (S.14) with protected aldehyde of sugar precursor (S.15) 1. Equip a 100-mL two-neck flask with two rubber septa and a stir bar. Keep the vessel under argon using an 18-G needle threaded through one of the septa. 2. Add 640 mg (2.72 mmol) 2-bromo-1,4-dimethylnaphthalene (S.14) and 15 mL THF to the flask under argon atmosphere and cool to −78°C in a dry ice/acetone bath. 3. Slowly add dropwise 2.0 mL of 2 M n-butyllithium (4.0 mmol) in cyclohexane and continue to stir for 15 min at −78°C. 4. Add, in a dropwise fashion, a solution of 1.214 g (1.801 mmol) of the protected aldehyde of the sugar precursor (S.15) in 5 mL THF and continue to stir for 1 hr at −78°C. 5. Remove the flask from the bath and warm to room temperature over 1.5 hr. Synthesis of the 1,4-dimethylnaphthalene derivative uses 1.0 eq S.14, 1.47 eq n-butyllithium, 0.66 eq S.15, and 90 eq THF.

Work up and purify product 6. Add 10 mL ethyl acetate and 10 mL saturated NaHCO3 to partition. Filter and dry the organic layer (see Basic Protocol 1, steps 11 and 12). 7. Perform flash chromatography (see Basic Protocol 1, steps 13 to 19). Elute with a gradient of 9:1 to 8:2 (v/v) hexane/ethyl acetate. Perform TLC (see Basic Protocol 1, step 8) using 7:3 hexane/ethyl acetate. For this product, Rf ∼0.23.

8. Combine the appropriate fractions based on TLC and evaporate to dryness with a rotary evaporator. Flash chromatography gives the C1 S-isomer (712 mg, 48% yield) and the C1 R-isomer (625 mg, 42% yield). S and R isomers were assigned based on the α or β anomer that was formed after cyclization.

Perform in situ cyclization of mesylate 9. Equip a 25-mL two-neck flask with two rubber septa and a stir bar. Keep the vessel under argon using an 18-G syringe needle threaded through one of the septa. 10. Add 712 mg (0.857 mmol) of the S-diastereomer (step 8) and 9 mL anhydrous pyridine to the flask and cool to 0°C.



Supplement 10

11. Add 0.57 mL (4.09 mmol) triethylamine followed by 90 µL (1.17 mmol) methanesulfonyl chloride, and warm slowly to room temperature over 1.5 hr with stirring. Synthesis of S.16 uses 1.0 eq S-diastereomer, 4.8 eq triethylamine, 1.37 eq methanesulfonyl chloride, and 130 eq pyridine.

Work up and purify S.16 12. Add 2 mL saturated NaHCO3 to quench the reaction. 13. Perform flash chromatography (see Basic Protocol 1, steps 13 to 19). Elute with a gradient of 95:5 to 85:15 (v/v) hexane/ethyl acetate. Perform TLC using 7:3 hexane/ethyl acetate. For S.16, Rf = ∼0.68.

14. Collect appropriate fractions based on TLC and evaporate to dryness with a rotary evaporator. Flash chromatography yields 176 mg (30% yield) of purified product (S.16).

Detritylate 5′-OH group 15. Equip a 25-mL two-neck flask with two rubber septa and a stir bar. Keep the vessel under argon using an 18-G syringe needle threaded through one of the septa. 16. Add 190 mg (0.274 mmol) S.16, 6 mL acetic acid, and 1 mL methanol to the flask. 17. Add 1 mL trifluoroacetic acid in a dropwise fashion (∼10 drops) and continue stirring for 20 min at room temperature. Detritylation uses 1.0 eq S.16, 383 eq acetic acid, 90 eq methanol, and 47.4 eq trifluoroacetic acid.

Work up and purify S.17 18. Concentrate the reaction solution with a rotary evaporator. 19. Remove solvents (see Basic Protocol, step 12) and perform flash chromatography (see Basic Protocol 1, steps 13 to 19). Elute with a gradient of 4:1 to 1:1 (v/v) hexane/ethyl acetate. Perform TLC using ethyl acetate. For S.17, Rf = ∼0.75.

20. Combine appropriate fractions based on TLC and evaporate to dryness with a rotary evaporator. Flash chromatography affords 93 mg (86% yield) of purified product (S.17).

Remove 3′-PMB group 21. Equip a 50-mL two-neck flask with two rubber septa and a stir bar. Keep the vessel under argon using an 18-G syringe needle threaded through one of the septa. 22. Add 8 mg, 0.020 mmol S.17, 1 mL CH2Cl2, and 1 drop H2O to the flask. 23. Add 7 mg (0.031 mmol) DDQ and continue to stir for 1 hr at room temperature. Deprotection uses 1.0 eq S.17, 780 eq CH2Cl2, 277.5 eq H2O, and 1.55 eq DDQ.


Work up and purify S.18 24. Add 15 mL ethyl acetate and 15 mL saturated NaHCO3 to partition. Filter and dry the organic layer (see Basic Protocol 1, steps 11 and 12).



25. Perform flash chromatography (see Basic Protocol 1, steps 13 to 19). Elute with a gradient of 1:99 to 5:95 (v/v) methanol/CH2Cl2. Perform TLC using 3:97 methanol/CH2Cl2. For S.18, Rf = ∼0.3.

26. Combine appropriate fractions based on TLC and evaporate to dryness using rotary evaporator. Flash chromatography affords 3 mg (55% yield) of pure product (S.18).

SYNTHESIS OF 3-METHYL-2-NAPHTHALENE-C-NUCLEOSIDE This protocol describes the synthesis of the 3-methyl-2-naphthalene-C-nucleoside (Fig. 1.5.6C), which was derived from the condensation of the aryllithium species onto another aldehyde followed by in situ cyclization as shown in Figure 1.5.8 (Ogawa et al., 2000a). The preparation of phosphoramidite and triphosphate follows the general procedures described in Basic Protocol 3.

ALTERNATE PROTOCOL

Additional Materials (also see Basic Protocols 3 and 4) 2-Bromo-3-methylnaphthalene (S.19; prepared according to Lambert et al., 1979) Protected aldehyde of sugar precursor (S.20, prepared according to Solomon and Hopkins, 1993) Perform condensation of 2-bromo-3-methylnaphthalene (S.19) with the protected aldehyde of the sugar precursor (S.20) 1. Place a magnetic stir bar in a dry 100-mL two-neck flask with two rubber septa, and place the flask on top of a magnetic stir plate. 2. Evacuate the reaction flask on vacuum line and then flush it with argon. Repeat this procedure three times and attach the flask to an argon line on the manifold. 3. Add 595 mg (2.70 mmol) 2-bromo-3-methylnaphthalene (S.19) and 8.9 mL THF to the flask under argon atmosphere and cool to −78°C in a dry ice/acetone bath. 4. Add dropwise 1.50 mL of 2 M n-butyllithium (3.0 mmol) in cyclohexane over 10 min and continue to stir 15 min at −78°C. 5. Add dropwise a solution of 502 mg (1.74 mmol) aldehyde (S.20) in 4.4 mL THF over 5 min and continue to stir 30 min at −78°C. 6. Remove the flask from the bath and warm to room temperature. During this time the reaction color changes from deep green to brown. Condensation uses 1.0 eq S.19, 1.1 eq n-butyllithium, 0.64 eq S.20, and 61 eq THF.

O

a,b,c

O Me Br 19

CHO OTBS 20

Me

HO O

21 OH

Figure 1.5.8 Synthesis of 3-methyl-2-naphthalene-C-nucleoside (see Alternate Protocol). Reagents: (A) n-butyllithium, −78°C, then S.2, THF (steps 1 to 9); (B) methanesulfonyl chloride, triethylamine, pyridine, 0°C (steps 10 to 16); (C) TFA (steps 17 to 22).



Supplement 10

Work up and purify product 7. Add 10 mL ethyl acetate and 10 mL saturated NaHCO3 to partition. Filter and dry the organic layer (see Basic Protocol 1, steps 11 and 12). 8. Perform flash chromatography (see Basic Protocol 1, steps 13 to 19). Elute with a gradient of 9:1 to 5:1 (v/v) hexane/ethyl acetate. Perform TLC (see Basic Protocol 1, step 8) using 9:1 hexane/ethyl acetate. For this product, Rf = ∼0.30.

9. Combine appropriate fractions based on TLC and evaporate to dryness with a rotary evaporator. Flash chromatography gives a crude mixture of both diastereomers (532 mg, 1.24 mmol).

Mesylate hydroxy group 10. Equip a 100-mL two-neck flask with two rubber septa and a stir bar. Keep the vessel under argon using an 18-G syringe needle threaded through one of the septa. 11. Add 510 mg (1.19 mmol) product (step 9, both diastereomers) and 32 mL CH2Cl2 to the flask and cool to 0°C. 12. Add 0.332 mL (2.38 mmol) triethylamine, followed by 0.120 mL (1.55 mmol) methanesulfonyl chloride, and warm slowly to room temperature over 30 min with stirring. Relative to S.19 (step 3), mesylation uses 0.88 eq triethylamine, 0.57 eq methanesulfonyl chloride, and 185 eq CH2Cl2.

Work up and purify product 13. Add 30 mL saturated aqueous NaHCO3 and transfer to a 125-mL separatory funnel. Rinse the reaction vessel twice, each time with 5 mL ethyl acetate and add to the funnel. 14. Separate the organic layer and extract the aqueous layer twice, each time with 15 mL ethyl acetate. 15. Filter and dry the organic layer (see Basic Protocol 1, steps 11 and 12), and perform flash chromatography (see Basic Protocol 1, steps 13 to 19). Elute with a gradient of 95:5 to 85:15 (v/v) hexanes/ethyl acetate. Perform TLC using 9:1 hexanes/ethyl acetate. For this product, Rf = ∼0.30.

16. Evaporate eluate and analyze product by NMR (see Basic Protocol 1, steps 20 to 22). The resulting crude product is pure enough for use in the next step.

Perform in situ cyclization of mesylate 17. Set up the reaction flask as in steps 1 and 2 above. 18. Add 605 mg (1.19 mmol) crude product (step 16) and 56 mL of 4:1 (v/v) TFA/methanol and stir 20 min at room temperature. Relative to S.19 (step 3), cyclization uses 216 eq trifluoroacetic acid and 103 eq methanol.


Work up and purify S.21 19. Concentrate the solution with a rotary evaporator and then neutralize the residual TFA with a minimum volume of saturated NaHCO3. 20. Transfer to a separatory funnel and extract four times, each time with 20 mL CH2Cl2.



21. Filter and dry the organic layer (see Basic Protocol 1, steps 11 and 12), and perform flash chromatography (see Basic Protocol 1, steps 13 to 19). Elute with 3:97 (v/v) methanol/CH2Cl2. Perform TLC using the same solvent For S.21, Rf = ∼0.2.

22. Collect the appropriate fractions and evaporate to dryness with a rotary evaporator. Flash chromatography affords the desired β-anomer (S.21) at 108 mg (0.418 mmol, 24% yield over 3 steps).

PURIFICATION OF DNA OLIGONUCLEOTIDES Oligonucleotides containing an unnatural base are synthesized using an ABI 392 DNA/RNA synthesizer with automatic removal of the 5′-trityl protecting group (so-called trityl-off procedure; APPENDIX 3C). The oligonucleotides are then deprotected and cleaved from the CPG support according to the instrument instruction manual. Figure 1.5.9 shows the sequences of four oligonucleotides that are used in the thermodynamic and kinetic analyses described in Basic Protocols 6 and 7. dC-CPG (500Å, 1.0 µmol) is used for S.22, dG-CPG (500 Å, 1.0 µmol) is used for S.23, dA-CPG (500 Å, 1.0 µmol) is used for S.24, the universal support (500Å, 1.0 µmol) is used for S.25 and dA-CPG (500Å, 1.0 µmol) is used for S.26. To be suitable for use in Basic Protocols 6 and 7, the oligonucleotides are purified by preparative denaturing PAGE (UNIT 10.4 and APPENDIX 3B), excised from the gel, electroeluted using a Schleicher & Schuell Elutrap electrophoresis chamber, and ethanol precipitated to concentrate and desalt the sample. The oligonucleotides are stored at −20 °C.

BASIC PROTOCOL 5

Materials CPG-bound DNA oligonucleotide (APPENDIX 3C) Concentrated ammonium hydroxide 95% (v/v) formamide in 10 mM Tris⋅Cl, pH 8.5 (APPENDIX 2A), with and without 0.05% (w/v) xylene cyanol and bromphenol blue TBE buffer (see APPENDIX 2A) Absolute ethanol, prechilled to –20°C 5 M NaCl, ice cold 80% ethanol, ice cold 1.5-mL screw-cap vial 60° and 80°C Dri-baths (Thermolyne) Speed-Vac evaporator (Savant) 50-mL polypropylene centrifuge tubes UV/Vis spectrophotometer Electroelution unit (Schleicher and Schuell) Silica gel–coated preparative TLC plate UV lamp (hand held) Refrigerated centrifuge and rotor appropriate for sample size Additional reagents and equipment for denaturing PAGE (UNIT 10.4 and APPENDIX 3B) Deprotect oligonucleotide and cleave from support 1. Transfer the CPG-bound DNA oligonucleotide to a 1.5-mL screw-cap vial. 2. Add 1.5 mL concentrated ammonium hydroxide, tighten the cap, and incubate 12 hr at 80°C (for 24-mer) or 60°C (for 13-mer or 45-mer). 3. Cool to 0°C and evaporate to dryness in a Speed-Vac evaporator.



Supplement 10

22 23 24 25 26

5 ′-dCGCATGXGTACGC 5 ′-dGCGTACXCATGCG 5 ′-dTAATACGACTCACTATAGGGAGA 5 ′-dTAATACGACTCACTATAGGGAGAX 5 ′-dCGCTAGGACGGCATTGGATCG XTCTCCCTATAGTGAGTCGTATTA

Figure 1.5.9 Oligonucleotides used for characterization of unnatural nucleobases.

Purify crude oligonucleotide 4. Add 300 µL water and 300 µL of 95% formamide/10 mM Tris⋅Cl, pH 8.5 (without xylene cyanol and bromphenol blue) and mix thoroughly. 5. Prepare a 12-well denaturing PAGE gel (UNIT according to the length of the oligonucleotide:

10.4

and

APPENDIX 3B)

as follows,

13-mer: 15% acrylamide; 42 cm long × 33 cm wide × 1.5 mm thick 24-mer: 20% acrylamide; 42 cm long × 33 cm wide × 1.5 mm thick 45-mer: 10% acrylamide; 42 cm long × 33 cm wide × 1.5 mm thick. 6. Load 60 µL of 95% formamide/10 mM Tris⋅Cl, pH 8.5, with 0.05% (w/v) xylene cyanol and bromphenol blue in each of the two outermost wells; load 60 µL of the sample in each of the ten innermost wells. 7. Purify the DNA to single-nucleotide resolution by denaturing PAGE (UNIT 10.4 and APPENDIX 3B) at 65 W until the full-length oligonucleotide has migrated two-thirds of the way down the gel, as indicated by the dye markers. 8. Excise the full-length DNA product by UV shadowing (UNIT 10.4). Isolate, desalt, and concentrate pure oligonucleotide 9. Place the gel slice in a 50-mL polypropylene centrifuge tube and crush thoroughly with a glass stirring rod. 10. Add 1× TBE electrophoresis buffer mix, and pour the gel slurry into the sample chamber of a properly assembled and oriented electroelution apparatus, according to the instrument instruction manual. 11. Electroelute the sample and monitor its progress by withdrawing small samples and measuring absorbance at 260 nm (A260). 12. Combine the most concentrated fractions. 13. Add 3 vol absolute ethanol (prechilled to –20°C) and 1⁄20 vol 5 M NaCl and freeze at −20°C for at least 30 min. 14. Centrifuge for at least 20 min at >15,000 × g, 4°C. 15. Carefully decant the supernatant, wash the DNA pellet twice with ice cold 80% ethanol, and allow it to air dry. 16. Dissolve DNA in an appropriate volume of sterile deionized water for determination of concentration and downstream applications. Development of a Universal Nucleobase and Unnatural Nucleobases



DETERMINATION OF THERMODYNAMIC STABILITY OF UNNATURAL BASE PAIRS Thermal stability or thermodynamic stability refers to the melting temperature (Tm) of a DNA duplex that contains the unnatural base pair. Thermodynamic selectivity refers to the difference in Tm between the desired unnatural base pair and the most stable mispair of an unnatural base with a natural base. Oligonucleotides S.22 and S.23 (Fig. 1.5.9) provide the sequence context for determination of the stability and selectivity of the unnatural nucleosides. Duplex oligonucleotide samples are prepared with the unnatural nucleoside in position X of, for example, S.22 and paired against S.23 containing each natural base as well as the same unnatural base (self-pair) and/or any other unnatural base to be tested. The complete set of pairs is also prepared with the unnatural nucleoside in S.23 to determine if there are sequence context effects (e.g., CXT versus TXC). The samples are heated and evaluated for temperature-dependent changes in absorption using a Cary 300 Bio UV-Vis spectrophotometer. Duplex melting curves are collected in triplicate, and melting temperatures are determined by the derivative method contained in the Cary WinUV-Vis software. For a more detailed description of Tm experiments, see UNIT 7.3.

BASIC PROTOCOL 6

Materials Complementary 13-mer oligonucleotides (e.g., S.22 and S.23 in water; see Basic Protocol 5) 2× Tm buffer (see recipe) Variable-temperature double-beam spectrophotometer (Cary 300 BIO UV-Vis spectrophotometer) with optically matched cuvettes 1. Prepare sample cells with 3 µM of each complementary oligonucleotide in 1× Tm buffer. Prepare reference cells with 1× Tm buffer alone. 2. Arrange cells in the spectrophotometer and monitor the sample absorbance at 260 nm while heating at a rate of 0.5°C/min from 20° to 80°C. 3. Collect three curves for each sample and analyze data by the first-derivative method provided with the instrument software. KINETIC ANALYSIS OF UNNATURAL BASE PAIR INCORPORATION, SELECTIVITY AND REPLICABILITY A rapid gel kinetic assay is used for the determination of steady-state parameters for the incorporation and extension of the unnatural base pairs, as well as to evaluate the selectivity, or fidelity of these steps. Initial velocities are determined by radiolabeled primer extension reactions at various concentrations of 5′-nucleoside triphosphate (natural or unnatural). To ensure single completed hit conditions, a final ratio of 30:1 primertemplate to enzyme is used, and less than 20% of the primer is extended. An example of a primer-template set for determining incorporation kinetics is shown in Figure 1.5.9 with S.24 as the primer and S.26 as the template. The reactions can be rapidly initiated and quenched by hand and analyzed by denaturing PAGE; a PhosphorImager (Molecular Dynamics) is used to quantify gel band intensities corresponding to the extended primer. Velocities are plotted against triphosphate concentration and fit to a rectangular hyperbola. Preliminary determination of KM is followed by kinetic analysis using triphosphate concentrations evenly distributed about the KM. Unnatural nucleosides are characterized both in the template strand and as the incoming triphosphate. Kinetic data are collected in triplicate. In general, these kinetic determinations follow the experimental protocols developed by Goodman and co-workers (Creighton et al., 1995).

BASIC PROTOCOL 7



Supplement 10

Materials 4 µM DNA oligonucleotide primer (e.g., S.24; see Basic Protocol 5) 10 U/µL T4 polynucleotide kinase and 10× buffer (New England Biolabs) 10 mCi/mL [γ-33P]ATP (2500 Ci/mmol) QIAquick Nucleotide Removal Kit (Qiagen) 1 µM DNA oligonucleotide template (e.g., S.26; see Basic Protocol 5) 10 U/µL exonuclease-free Klenow fragment (Amersham Pharmacia Biotech) and 10× random-prime buffer (see recipe) Unnatural triphosphates (see Basic Protocols 1 to 4 and Alternate Protocol) 100 mM stock of each natural dNTP (Amersham Pharmacia Biotech; sequencing grade) Enzyme dilution buffer (see recipe) Quench solution (see recipe) 1× TBE buffer (APPENDIX 2A) 37°, 25°, and 100°C Dri-Baths (Thermolyne) with heating blocks that accommodate 1.5- or 0.5-mL tubes Thin putty knife Geiger counter Chromatography paper (Whatman 3MM CHR, 35 × 45 cm) Gel dryer (Bio-Rad Model 583) Storage Phosphor Screen (Kodak or Molecular Dynamics, 35 × 43 cm) PhosphorImager (e.g., STORM Model 860; Molecular Dynamics) ImageQuant and compatible spreadsheet (MS Excel) and graphing (Kaleidagraph) software Additional reagents and equipment for denaturing PAGE (UNIT 10.4 and APPENDIX 3B) NOTE: Use autoclaved deionized water for all reagents. Radiolabel primer 1. Combine 25 µL of 4 µM oligonucleotide primer with 3.1 µl of 10× T4 polynucleotide kinase buffer, 2 µL of [γ-33P]ATP stock, and 1 µL of T4 polynucleotide kinase. Mix and incubate at 37°C for at least 30 min. 2. Remove excess nucleotide with the QIAquick Nucleotide Removal Kit, according to the manual, and observing proper radioactive waste disposal. 3. Dilute buffer EB (from QIAquick kit) five-fold and use 100 µL to elute the DNA. Store at 4°C. Allow the wetted column to sit for a few minutes before the final spin. Assuming 80% recovery of the labeled primer, this protocol yields 100 ìL of 800 nM 5′-radiolabeled primer.

Prepare analytical gel 5. Pour a 15% denaturing polyacrylamide gel (UNIT 10.4 and APPENDIX 3B) that is 42 cm long, 33 cm wide, and 0.5 mm thick, and has 34 wells. Make sure there are no trapped air bubbles. Allow the gel at least 1 hr to fully polymerize before use.


Anneal primer and template 6. Combine 25 µL of water, 10 µL of 10× random-prime buffer, 8 µL of 1 µM template oligonucleotide, and 5 µL of 5′-radiolabeled 800 nM primer oligonucleotide (from step 1) in a 1.5-mL microcentrifuge tube, mix well, then microcentrifuge briefly at maximum speed to bring all the solution to the bottom of the tube. This gives a two-fold excess of template over primer, necessary to ensure that greater than 95% of the primer is annealed to the template. The use of different primer or template sequences may require that this ratio be optimized.



7. Place the solution in a Dri-bath preheated to 100°C and immediately turn off the heating element. Allow the primer-template to anneal by slow cooling to room temperature in the Dri-bath (∼2 hr). This yields a final volume of 48 ìL, which, when combined with 2 ìL of enzyme, will yield sufficient volume for nine reactions of different triphosphate concentrations. Although it is best to use freshly annealed primer-template, annealed primer-templates can be stored at 4°C overnight. Warm to room temperature before use.

Perform reactions 8. Prepare 5-µL aliquots of 2× concentrations of the desired triphosphate range by combining the appropriate volume of 1 µM, 10 µM, 100 µM, 1 mM, or 10 mM of the previously diluted dNTP and water to a final volume of 5 µL. For the incorporation experiments, the unnatural triphosphate is used (in conjunction with the primer-template complex S.24-S.26) to determine the rate of unnatural base pair synthesis. To measure incorporation selectivity, each of the four natural triphosphates is used in conjunction with this primer-template. For the extension experiments, including both extension efficiency as well as extension fidelity, each of the four natural triphosphates is used in conjunction with primer-template S.25-S.26. It is a good idea to make a reasonably large number of small-volume aliquots of the various concentrations of dNTP to reduce the number of freeze-thaw cycles to which the dNTP is subjected. It is generally desirable to run a control for each experiment with only 5 ìL of water and no triphosphate.

9. Add 2 µL of the appropriate concentration of Klenow fragment freshly diluted in enzyme dilution buffer to cooled annealed primer-template. Mix thoroughly, but do not vortex. This yields 50 ìL of a 2× reaction solution that is 80 nM in primer-template, 2× in all buffer components, and 0.04× of the concentration of diluted enzyme used. This 2× reaction solution will be diluted to 1× when combined with the triphosphate aliquots to initiate the reaction. To ensure single completed hit conditions, make certain that the final enzyme concentration is no more than 1⁄30 that of the final primer-template concentration, final concentration referring to that achieved in step 10.

10. Add 5 µL enzyme-primer-template solution (from step 9) to each 5-µL aliquot of 2× triphosphate (from step 8), mix by pipetting up and down, close the tube, and place in a 25°C Dri-bath. Initiate all reactions in turn as quickly as possible (≤10 sec between starts). 11. Quench the reactions in turn with 20 µL quench solution precisely when the timer reaches the desired length for the experiment. Perform electrophoresis 12. Warm gel (prepared in step 5) by running it at 65 W in 1× TBE buffer until it reaches a temperature of at least 37°C. Wash the wells by pipetting 1× TBE into each a couple of times. 13. Load 8 µL quench solution into the three outermost lanes as well as any unused lanes. Load 8 µL of each quenched reaction into up to 28 of the innermost lanes. Run at 65 W. Up to 28 reactions can be loaded in one row of the gel. Four rows (112 reactions) can be run in one gel by allowing the first row to run at 65W for 30 min, stopping the gel and loading another row. The second row is run for 25 min, the third row for 20 min, and the fourth row until the dark blue dye band of the second row exits the gel. The wells are washed prior to loading each of the four rows.



Supplement 10

14. Remove the gel from the apparatus and cool it under running water. Separate the plates by using a thin putty knife, taking care to notice which plate the gel will remain stuck to and allowing it to do so. 15. Find the region of the gel containing the extended primers by surveying with a Geiger counter. 16. Place a large piece of plastic wrap over the gel, smooth it out, and cut the relevant area. 17. Carefully transfer the gel (covered on one side with plastic wrap) to a 35 × 45–cm sheet of chromatography paper, gel-side down, and smooth it out. Dry the gel completely (time may vary with apparatus) and allow it to cool nearly to room temperature. 18. Cut the chromatography paper down to size, if necessary, and transfer the gel to a storage phosphor screen, plastic-wrap side facing the screen. Allow image to develop overnight. Analyze data 19. Remove the gel from the storage screen and read the plate with a PhosphorImager. 20. Using ImageQuant software, draw rectangles around primer and extended primer and obtain a volume report of the quantitative data from the gel quantitative data. Use the control reaction (no triphosphate) as the background to be subtracted in each series of triphosphate concentrations. 21. Calculate the fraction of primer extended at each triphosphate concentration as: v = Ip/(I0 + Ip) where I0 is the primer band count and Ip is the product band count. 22. Subtract the value obtained for the control reaction from each reaction. This will force the data set to the origin. Make sure that 120°C) oven before use. Use anhydrous solvents. Deprotection of toluoyl group. There are no particular difficulties with this reaction; it will work effectively. Nucleophilic addition by aryllithium reagents. The reaction is moisture-sensitive. Anhydrous solvent is important for the reaction. Do not use n-butyllithium reagent with a milky precipitate, because such a reagent is decomposed in the presence of water. Removal of protecting group of hydroxy group. Difficulties are seldom encountered in this step.

Anticipated Results


An overall good yield of the free nucleoside (48% to 68%) of unnatural hydrophobic nucleobases can be achieved based on the strategies described in this unit. The yields of the nucleosidic glycosylation are generally good eno ug h f or th e pr ep ar atio n of phosphoramidites and triphosphates, although stereoselectivity of the desired β-anomer versus α-anomer may be mediocre in some cases.

The routine deprotection, phosphitylation, and phosphorylation are generally high-yielding. Although the Grignard addition described in Basic Protocol 3 generally favors the αanomer, acid treatment of the corresponding benzylic ethers in refluxing xylene will result in epimerization at C1′ to yield the β-anomer (Ren et al., 1996). Conversely, the aryllithium addition described in Basic Protocol 4 and the Alternate Protocol generally affords the desired product in a good yield. The product is a mixture of diastereomers, but is separable by column chromatography.

Time Considerations The estimated time to complete Basic Protocol 1 in its entirely is 3 weeks. The estimated time required to complete Basic Protocol 2 is 2 days. Basic Protocol 3 should take ∼6 days to complete. For Basic Protocol 4 and Alternate Protocol, allow ∼1 day for the Grignard reaction and several hours for all of the other reactions. For Basic Protocol 5, 2 to 3 days should be sufficient time to prepare the oligonucleotide. On day 1, the oligonucleotide is synthesized and the deprotection should be done overnight. Several hours should be allowed for gel pouring and polymerization; this step may also be done on day 1 or in the morning of day 2. The gel must be kept hydrated, regardless, by wrapping with plastic wrap. The electrophoresis step can take several hours. After separation, the excised gel slice may be stored at −20°C for further purification at a later date. To perform the electroelution, allow 95%. The synthesis is continued using either diethoxymethylacetate in dimethylformamide (DMF) to give 12C at the 8 position (steps 16a to 26a) or [13C]sodium ethyl xanthate to give 13C at the 8 position (steps 16b to 22b).

Perform ring closure For [7-15N]-2-thioxohypoxanthine (S.4a): 16a. Scrape out most of S.3 from the funnel into a 100-mL round-bottom flask. Rinse the funnel with portions of 96% formic acid and add them to the flask to give a final volume of 25 mL.

Table 1.6.1 Molecular Weights, TLC and HPLC Mobilities, and UV λmax of 15N-Labeled Adenosine and Guanosine Intermediatesa

Compound

mol. wt. (Da)

TLC Rfb

HPLC retention time (min)c

UV λmax (nm)

S.2 S.3 S.4a S.4b S.5a/b S.6a/b S.7a/b S.8a/b S.9a/b S.12a/b S.13a/b

173 159 169 202 137/138 155/156 287/288 269/270 285/286 316 287

0.2 0.0 0.2 0.0 0.1 0.3 0.2 0.1 0.0 0.1 0.0

1.2 0.8 0.7 1.4 0.8 2.4 4.5 2.6 0.9 3.3 1.4

360 301 280 299 250 265 264 259 295 280 253

aAbbreviations: HPLC, high-performance liquid chromatography; mol. wt., molecular weight; TLC, thin-layer chroma-

tography.

Synthesis of Specifically 15 N-Labeled Adenosine and Guanosine

bR values determined with 10:90 (v/v) CH OH/CH Cl . f 3 2 2 cGradient of 2:98 to 40:60 (v/v) acetonitrile/0.1 M triethylammonium acetate, pH 6.8, over 5 min, on a Waters NovaPak

C18 column.

1.6.4 Supplement 10


17a. Add a stir bar, attach a condenser, and reflux the solution for 1 hr to make the formate salt. 18a. Concentrate to dryness using a rotary evaporator and scrape down the sides of the flask with a spatula, if necessary. 19a. Insert a rubber septum and displace the air with nitrogen. 20a. Use syringes to add the following through the septum: 20 mL anhydrous DMF 1.63 mL DEMA (10 mmol, 2 eq) 0.24 mL of 96% formic acid (6 mmol, 1.2 eq). 21a. Heat the mixture for 3 hr in an oil bath set at 130°C. Follow the reaction by HPLC. The flask is lifted from the oil bath and allowed to cool briefly, and then a small syringe with a long, dry needle is used to get a sample for HPLC.

22a. Cool the flask, concentrate the solution to a solid using a rotary evaporator, and loosen it with a spatula if necessary. 23a. Add 15 mL acetonitrile to the flask, attach a condenser, and reflux it for 10 min using the 130°C oil bath. 24a. Cool the flask to room temperature, add 10 mL acetonitrile, and then chill it in an ice bath. 25a. Collect S.4a by vacuum filtration and wash it twice with 5 mL cold acetonitrile. 26a. Without removing it from the funnel, dry S.4a over P2O5 in the vacuum desiccator overnight. Proceed to step 27. The yield is usually >95%.

For [8-13C-7-15N]-2,8-dithioxohypoxanthine (S.4b): 16b. Scrape out most of S.3 from the funnel into a 100-mL round-bottom flask and add 0.80 g (5.5 mmol) [13C]NaSCSOEt. 17b. Insert a condenser into the flask, attach a nitrogen line and vent needle, and displace the air for 5 min. 18b. Add 15 mL DMF and reflux the mixture under nitrogen for ∼3 hr, using HPLC to monitor the reaction. 19b. Cool the mixture in an ice bath and add 50 mL cold acetonitrile to precipitate S.4b. 20b. Collect the solid S.4b by vacuum filtration and wash it twice with 5 mL cold acetonitrile. Save the filtrate and both washes. 21b. Concentrate the filtrate and washes, and purify this portion of S.4b by preparative reversed-phase chromatography. 22b. Dry the combined portions of S.4b over P2O5 in a vacuum desiccator overnight. Continue with step 27. The yield is usually >95%.



Supplement 10

Synthesize [7-15N]hypoxanthine and [8-13C-7-15N]hypoxanthine (S.5a/b) 27. Scrape out most of S.4a/b from the funnel into a 100-mL round-bottom flask. 28. Rinse the funnel with portions of water and add them to the flask to give a final volume of 30 mL. 29. Add a stir bar and 2 mL of 96% formic acid. 30. Weigh 4.5 g of 50% aqueous RaNi slurry into a small glass vial or beaker. CAUTION: RaNi is pyrophoric (will spontaneously burst into flames) if it is allowed to dry out. To weigh it, shake the bottle and immediately transfer some of the suspended RaNi to the vial. Continue adding the suspension (shaking the bottle each time) until 4.5 g of the 50:50 mixture of RaNi/water has been measured out.

31. Using a dropper, transfer the RaNi suspension over 5 min to the reaction flask, and add 1.5 g dipotassium salt of EDTA. Traces of remaining RaNi in the vial and dropper should be destroyed with 6 M HCl.

32. Connect a condenser to the flask. Using the 130°C oil bath, reflux for ∼2 hr while monitoring the reaction for completeness by HPLC. 33. Remove the flask from the oil bath and allow it to cool only briefly. 34. Remove the condenser from the flask and carefully filter the hot reaction mixture to remove the RaNi. Rinse the flask and then the funnel with three 10-mL portions of boiling water, adding these washes to the filtrate. To destroy the RaNi remaining in the funnel, the funnel should be transferred to a large beaker, and portions of 6 M HCl should be slowly added until no black particles can be observed.

35. Concentrate the filtrate and washes to dryness in a 100-mL round-bottom flask using the rotary evaporator. 36. Dry S.5a/b over P2O5 in the vacuum desiccator overnight. The yield is usually >95%. S.5a/b can be purified by reversed-phase chromatography if desired.

Synthesize [7-15N]- and [8-13C-7-15N]-6-chloropurine (S.6a/b) 37. Weigh 0.69 g (5.0 mmol) S.5a/b into a very dry 100-mL round-bottom flask. 38. Add 20 mL (215 mmol, 43 eq) POCl3 and 2 mL (16 mmol) N,N-dimethylaniline. Use great care with POCl3; it is very reactive.

39. Attach a condenser and reflux 20 min under nitrogen. The resulting solution should be black and homogeneous. It is essential for this reaction to remain anhydrous.

40. Monitor the reaction for completeness by HPLC and continue refluxing the mixture for ≤30 min more. 41. Concentrate the mixture to a very small volume using a rotary evaporator, first with an aspirator and then with a vacuum pump protected with a dry-ice trap. Add 10 mL N,N-dimethylaniline and continue the evaporation. Synthesis of Specifically N-Labeled Adenosine and Guanosine 15

It is necessary to remove all traces of POCl3, or localized heating from later neutralization with NH3 may cause the reaction to reverse. Be extremely careful to dry the evaporator condenser and trap prior to use, to keep dry ice in the traps throughout this process, and then to pour the collected POCl3 into a plastic container of ice to destroy it.

1.6.6 Supplement 10


42. Cool the flask in an ice bath and very slowly add 30 mL of 5% NH3 to dissolve the black gum. As a neutral compound, 6-chloropurine is insoluble in water, but under basic conditions it will ionize and therefore dissolve.

43. Make sure the pH of the solution is >10 (add more NH3 if necessary) and then pour it into a separatory funnel. Wash it first with 30 mL ethyl acetate and then twice with 30 mL ethyl ether. Check the layers by HPLC. 44. Combine all organic layers that contain traces of product in a separatory funnel, backwash with 5% NH3, and add this aqueous layer to the main reaction mixture. 45. Concentrate this aqueous solution to dryness to remove all the NH3. It is necessary to remove all traces of NH3, or localized heating from later neutralization with HCl may cause the reaction to reverse. As the NH3 evaporates, it is very likely to bump, so a large enough flask should be used.

46. Add 20 mL water and chill the flask in an ice bath. Slowly acidify the solution to pH 2 using 1 M HCl. The mixture will turn cloudy.

47. Set up a continuous extraction apparatus for solvents lighter than water. Pour the aqueous layer into the extractor and add ethyl ether until the level is just under the side arm. 48. Fill a 250-mL round-bottom flask with ethyl ether, add a stir bar, connect to the extractor, and place it in an oil bath. Attach a condenser to the top of the extractor, and start heating the oil bath to 45°C. 49. Continue the extraction for 3 to 4 days, using fresh ether each day. Check both the aqueous and ether layers each day by HPLC. Verify that the pH of the aqueous layer is still 13). However, many N-acyl protecting groups are readily removed by ammonolysis. This observation forms the basis for the widespread use of 28% NH4OH as a deprotection reagent when N-acylprotected nucleosides are employed in oligonucleotide synthesis. A salient feature of N-acyl protecting groups is that their stability in alkaline pH can be modulated by the steric and electronic characteristics of specific acyl groups. A study comparing the stability of various N-acyl nucleosides in alkaline medium (0.2 N NaOH/MeOH) has been reported (Köster et al., 1981). Importantly, the stability of the acyl function towards alkaline hydrolysis is determined by the nature of the heterocyclic base. For example, the rate of deacylation of N-acyl derivatives of deoxycytidine is faster than that of

deoxyadenosine or deoxyguanosine. The hydrolytic lability is also determined by inductive, resonance, and steric effects. For example, in a series of N-acyl nucleosides, N-benzoyl nucleosides are hydrolyzed sixteen times faster than N-(2,4-dimethyl)benzoyl nucleosides, presumably because of steric effects. Similarly, N-(2,4dimethoxy)benzoyl nucleoside is hydrolyzed eight times faster than N-(4-dimethylamino)benzoyl nucleoside, perhaps due to a combination of inductive and resonance effects. The choice of a particular N-acyl protecting group also depends on the type of coupling chemistry that is employed. For example, when phosphodiester and phosphotriester chemistries are used in oligonucleotide synthesis, it is necessary to select sturdy N-acyl protecting groups that can withstand the harsh reagents and conditions employed during synthesis. This requirement is met by the benzoyl group for adenine and cytosine, and the isobutyryl group for guanine. However, removal of these protecting groups requires prolonged heating

Protection of Nucleosides for Oligonucleotide Synthesis


(12 to 14 hr) with 28% NH4OH at 55°C. In spite of this limitation, these protecting groups have remained popular even with the advent of automated solid-phase oligonucleotide synthesis using phosphoramidite and H-phosphonate chemistries.

mechanism for depurination of deoxyadenosine (Fig. 2.1.8) involves initial protonation of the nucleoside to produce the N1-protonated form S.11, and then equilibration (prototropic shift) to the N7-protonated species S.12 or S.13, followed by cleavage of the glycosidic bond to give S.15 via the oxonium S.14 (Zoltewicz et al., 1970; Zoltewicz and Clark, 1972). It is also conceivable that, at lower pH, depurination can occur via protonation of the purine nucleobase at both N1 and N7. Presence of the 2′-OH has a significant effect on the nucleoside’s susceptibility to depurination. For example, guanosine and adenosine are more resistant to depurination compared with deoxyadenosine and deoxyguanosine. Deoxyadenosine itself depurinates 1200 times faster than adenosine (York, 1981). Interestingly, N-acyl-protected purine nucleosides (particularly deoxyadenosine) are more prone to depurination than unprotected nucleosides. Among N-acyl-protected deoxyadenosines, protection at N6 with αphenylcinnamoyl, naphthaloyl, 3-methoxy-4phenoxybenzoyl, 9-fluorenylmethoxycarbonyl (FMOC), and tert-butylphenoxyacetyl (t-PAC) groups (Fig. 2.1.7) provides greater resistance to depurination than with N6-benzoyl (reviewed in Beaucage and Iyer, 1992). It is believed that in the case of acyl-protected

PROTECTION OF PURINE NUCLEOBASES: THE PROBLEM OF DEPURINATION The development of suitable protecting groups for purine nucleobases has been an area of considerable interest because purine nucleosides rapidly depurinate under acidic conditions. The problem is compounded by the acidlabile DMTr group used for protection of the 5′-OH in solid-phase oligonucleotide synthesis. Prior to each coupling step in the synthesis cycle, the DMTr group is removed by exposure to a strong acid such as 2% dichloroacetic acid in dichloromethane. Consequently, the growing oligonucleotide chain is repeatedly exposed to strongly acidic conditions, potentially resulting in depurination and reduced yield of the desired “full-length” product. The kinetics and mechanisms of nucleoside depurination have been investigated by several research groups (Romero et al., 1978; Oivanen et al., 1987; Suzuki et al., 1994). The presumed

7

N

NH2 6

NH2 N

N1

9

HO

O

N

H

N

HO

3

HO

N

O

NH N

HO 11

H N HO

O HO 15

HO OH

HO

O

HO oxonium ion 14

O

N

NH2 N N

HO 12

H N HO

O

N

O NH N

NH2

HO 13

Nucleobase Protection of Deoxyribo- and Ribonucleosides

Figure 2.1.8 Scheme showing the proposed depurination mechanism for 2′-deoxyguanosine and 2′-deoxyadenosine catalyzed by protic acids. Modified from Iyer and Beaucage (1999) with permission from Elsevier Science Publishing.


Cl

Cl

Cl

OCOPh OCOPh

Cl Ph

O

N

RO

O

O

N

N

O

N

N

RO

RO

O

O

O

N

N

N

N

O

HN N

N N

RO

RO 16

N

RO

O

N

N

N

Ph

N

RO

O

N

OCOPh

N

RO 18

17

C

19

Figure 2.1.9 Examples of N6-protecting groups for 2′-deoxyadenosine derivatives that reduce depurination. Reprinted from Iyer and Beaucage (1999) with permission from Elsevier Science Publishing. R, DMTr.

purine nucleosides under acidic conditions, the initial site of protonation is N7, rendering the protonated species more prone to glycosidic cleavage. Naturally, caution should be exercised in the synthesis of oligonucleotides whose sequence contains deoxyadenosines at the 3′ terminus. Bis-acylation has also been studied as a strategy to reduce depurination. Imide protecting groups S.16 and S.17, as well as the diamide protecting group S.18, have been investigated (Fig. 2.1.9; Kume et al., 1982, 1984). However, these groups were labile to aqueous pyridine, a

Protecting group

solvent used during the oxidation step in solidphase oligonucleotide synthesis. Thus, alternate oxidants have to be employed for the oxidation step. Since N-acyl-protected purine nucleosides are sensitive to deacylation, alternate protecting groups have been investigated. Prominent among these are the amidine protecting groups, which are introduced using an exchange reaction with appropriate amidine acetals. Interestingly, N-amidine-protected nucleosides (Fig. 2.1.10) resist depurination 20-fold better than the corresponding N-benzoyl nucleosides

Nucleobase (protection site)

Deprotection conditions

References

G (N2), A (N6)

NH4OH, heat

Smrt and Sorm, 1967 Vu et al., 1990 Sproat et al., 1991 Caruthers et al., 1985 McBride et al., 1986

G (N2), A (N6)

NH4OH, heat; 0.5 M NH2NH2·H2O/pyr/AcOH

McBride et al., 1986 Froehler and Matteucci, 1983

A (N6)

NH4OH/10% NH4OAc/heat; 0.5 M NH2NH2·H2O/pyr/AcOH

Froehler and Matteucci, 1983

G (N2), A (N6)

NH4OH, heat

McBride et al., 1986

A (N6)

NH4OH/10% NH4OAc/heat; 0.5 M NH2NH2·H2O/pyr/AcOH

Froehler and Matteucci, 1983

C (N4)

NH4OH, heat

McBride et al., 1986

H Me2N (dimethylamino)methylene

H

n-Bu2N (di-n-butylamino)methylene

H

i-Pr2N (diisopropylamino)methylene

Me2N 1-(dimethylamino)ethylidene

H

i-Bu2N (diisobutylamino)methylene

N Me N-methylpyrrolidin-2-ylidene

Figure 2.1.10

Amidine protecting groups for exocyclic amino functions. Pyr, pyridine.



(Smrt and Sorm, 1967; Holy and Zemlicka, 1969, and references therein; Froehler and Matteucci, 1983; Caruthers et al., 1985; Vu et al., 1990; Sproat et al., 1991). It is presumed that the protonation sites of amidine-protected nucleosides are N1 and N6 instead of N1 and N7, respectively, resulting in a slower rate of depurination. Indeed, amidineprotected nucleoside phosphoramidites are frequently employed in solid-phase oligonucleotide synthesis. However, changes are required in the oxidation step to avoid nucleobase modifications (Mullah et al., 1995). Nucleosides protected with O-nitrophenylsulfonyl and tris(benzoyloxy)trityl (S.19) groups also appear to be more resistant to depurination (Shimidzu and Letsinger, 1968; Honda et al., 1984, and references therein). However, the derived nucleoside phosphoramidites couple less efficiently than the N-acyl-protected nucleoside phosphoramidites (Sekine et al., 1985). Depurination is also influenced by other factors such as the nature of the solid support (controlled-pore glass versus polystyrene; see UNIT 3.1), the composition of the deblocking solution and deblocking time, and the washing solvent and washing time that are employed in solid-phase synthesis of oligonucleotides (Paul and Royappa, 1996). Depurination is faster at terminal sites than at internal sites in an oligonucleotide chain (Suzuki et al., 1994). Interestingly, a solution of 15% dichloroacetic acid (DCA) in methylene chloride was ideal as a detritylating reagent that induced minimal depurination compared to the traditionally used 2% DCA/methylene chloride. It is pertinent that with modern DNA synthesizers the pulsed delivery of reagents to the synthesis columns, in conjunction with optimized synthesis programs, results in short contact times and has greatly minimized the depurination problem. Thus, N-acyl-protected nucleoside (benzoyl for dA and dC, and isobutyryl for dG) phosphoramidites and H-phosphonates can be used for the efficient synthesis of oligonucleotides.

RECENT TRENDS IN NUCLEOBASE PROTECTION

Nucleobase Protection of Deoxyribo- and Ribonucleosides

Over the past few years, new applications of oligonucleotides in diagnostics and therapeutics have emerged, necessitating the expeditious synthesis of large numbers of oligonucleotides. In order to speed the synthesis process, more “labile” N-acyl protecting groups for nucleosides have been sought. As the simplest

member of the family of N-acyl protecting groups, the N4-acetyl of cytosine has been used for “ultrafast” DNA synthesis using the phosphoramidite approach. Rapid deprotection rates were achieved using methylamine/ammonia (Reddy et al., 1994). This group is unsuitable for use in phosphodiester and phosphotriester chemistries, however. The synthesis of oligonucleotide analogs carrying sensitive backbones requires protecting groups that (1) withstand the synthetic rigors of chain assembly, and (2) can be removed chemoselectively under mild conditions. Nucleobases protected by phenoxyacetyl (Singh and Misra, 1988; Chaix et al., 1989; Sproat et al., 1991), and their derivatives such as t-PAC, show accelerated deacylation (Köster et al., 1981; Sinha et al., 1993) under mildly basic conditions. It is presumed that the inductive effect of the phenoxy group renders the amide carbonyl group more susceptible to nucleophilic attack, facilitating rapid base-catalyzed hydrolysis. Thus, PAC- and t-PAC-phosphoramidites have been employed for rapid synthesis of oligonucleotides and certain analogs. As a rule, following oligonucleotide assembly on solid support, the PAC and t-PAC groups are removed under milder conditions (28% NH4OH, room temperature; Sinha et al., 1993) or, according to a recent report, using gaseous amines under pressure (Boal et al., 1996). However, the use of t-PAC-protected nucleoside phosphoramidites results in trans acylation of the t-PAC protecting groups during the capping step when acetic anhydride is employed as a capping reagent. Thus, tert-butylphenoxyacetic anhydride should be used for the capping step in solid-phase oligonucleotide synthesis (Sinha et al., 1993). The concept of neighboring-group participation has also been used to design acyl protecting groups and to accelerate the deprotection step (Dreef-Tromp et al., 1990; Kuijpers et al., 1990). For example, removal of the 2-(acetoxymethyl)benzoyl group from nucleosides under basic conditions is accelerated by intramolecular participation of the deacylated hydroxymethyl group. New protecting groups have also been introduced that can be chemoselectively removed under neutral or mildly basic conditions. Indeed, the allyloxycarbonyl protecting group (Hayakawa et al., 1986, 1990) is chemoselectively removed using Pd(0), whereas the (p-nitrophenyl)ethoxycarbonyl group (Trichtinger et al., 1983; Pfleiderer et al., 1985; Pfister et al., 1988) and the 2-dansylethoxy cabonyl group


(Wagner and Pfleiderer, 1997) are selectively removed using DBU. Consequently, supportbound oligonucleotides can be prepared using building blocks that carry these protecting groups. Recently, the N-pent-4-enoyl (PNT) group has been introduced as a new acyl protecting group for situations where multiple deprotection protocols are used (Iyer et al., 1997; see also references therein). PNT-protected nucleoside phosphoramidites can potentially be used for the rapid synthesis of oligonucleotides and oligonucleotide analogs, as well as for the preparation of support-bound oligonucleotides. Several groups have been reported for nucleobase protection and may be particularly valuable in the synthesis of oligonucleotides intended for specific applications. Figure 2.1.7 shows a partial list of protecting groups. A complete list of such groups is covered in other reviews (Sonveaux, 1986; Beaucage and Iyer, 1992; Iyer and Beaucage, 1999). Nonetheless, it should be noted that many of these still need to be evaluated in routine synthesis. Because of various issues associated with nucleobase protection, oligonucleotide synthesis has been evaluated using building blocks bearing unprotected nucleobases. However, these efforts have met with only limited success (Narang et al., 1972; Fourrey and Varenne, 1985; Gryaznov and Letsinger, 1991; Uchiyama et al., 1993, and references therein). Clearly, more work is necessary in this area.

CONCLUSION Evidently, a nucleobase protecting group should meet several criteria before it can be adapted for routine oligonucleotide synthesis. It is imperative, therefore, that the potential for side reactions be closely examined when designing new reagents, evaluating new protecting groups, and implementing modifications of established protocols during oligonucleotide synthesis, as well as during the manufacture of oligonucleotides and their analogs. The development of nucleobase protecting groups and deprotection protocols has been crucial for the successful synthesis of oligonucleotides, functionalized oligonucleotides, oligonucleotide analogs, ribonucleotides, and phosphorylated biomolecules (reviewed in Beaucage and Iyer, 1992, 1993a,b,c). As in the past, oligonucleotides are expected to play a dominant role in fostering advances in functional genomics, proteomics, diagnostics, and therapeutics. Consequently, continued demand exists for simultaneous synthesis of large numbers of oligonucleotides in miniature formats,

and for the synthesis and manufacture of novel analogs. It is hoped that the present commentary will serve as a framework for developing new protecting group strategies for oligonucleotide synthesis that could meet these challenges.

ACKNOWLEDGMENT The author wishes to thank Dr. WenQiang Zhou for his help with the artwork in this manuscript.

LITERATURE CITED Andrus, A. and Beaucage, S.L. 1988. 2-Mercaptobenzothiazole—an improved reagent for the removal of methylphosphate protecting groups from oligodeoxynucleotide phosphotriesters. Tetrahedron Lett. 29:5479-5482. Balgobin, N., Josephson, S., and Chattopadhyaya, J.B. 1981. A general approach to the chemical synthesis of oligodeoxyribonucleotides. Acta. Chem. Scand. B35:201-212. Beaucage, S.L. and Caruthers, M.H. 1981. Deoxynucleoside phosphoramidites—A new class of key intermediates for deoxypolynucleotide synthesis. Tetrahedron Lett. 22:1859-1862. Beaucage, S.L. and Iyer, R.P. 1992. Advances in the synthesis of oligonucleotides by the phosphoramidite approach. Tetrahedron 48:22232311. Beaucage, S.L. and Iyer, R.P. 1993a. The functionalization of oligonucleotides via phosphoramidite derivatives. Tetrahedron 49:19251963. Beaucage, S.L. and Iyer, R.P. 1993b. The synthesis of modified oligonucleotides by the phosphoramidite approach and their applications. Tetrahedron 49:6123-6194. Beaucage, S.L. and Iyer, R.P. 1993c. The synthesis of specific ribonucleotides and unrelated phosphorylated biomolecules by the phosphoramidite method. Tetrahedron 49:10441-10488. Bhat, V., Ugarkar, B.G., Sayeed, V.A., Grimm, K., Kosora, N., and Domenico, P. 1989. A simple and convenient method for the selective N-acylations of cytosine nucleosides. Nucleosides Nucleotides 8:179-183. Boal, J.H., Wilk, A., Harindranath, N., Max, E.E., Kempe, T., and Beaucage, S.L. 1996. Cleavage of oligodeoxyribonucleotides from controlledpore glass supports and their rapid deprotection by gaseous amines. Nucl. Acids Res. 24:31153117. Bridson, P.K., Markiewicz, W., and Reese, C.B. 1977. Acylation of 2′,3′,5′-tri-O-acetylguanosine. J. Chem. Soc., Chem. Commun. 791-792. Brown, J.M., Christodoulou, C., Modak, A.S., Reese, C.B., and Serafinowska, H.T. 1989. Synthesis of the 3′-terminal half of yeast alanine transfer ribonucleic acid (tRNAala) by the phosphotriester approach in solution. Part 2. J. Chem. Soc. Perkin Trans. 1:1751-1767.



Büchi, H. and Khorana, H.G. 1972. CV. Total synthesis of the structural gene for an alanine transfer ribonucleic acid from yeast. Chemical synthesis of an icosadeoxyribonucleotide corresponding to the nucleotide sequence 31 to 50. J. Mol. Biol. 72:251-288. Caruthers, M.H., McBride, L.J., Bracco, L.P., and Dubendorff, J.W. 1985. Studies on nucleotide chemistry 15. Synthesis of oligodeoxynucleotides using amidine protected nucleosides. Nucleosides Nucleotides 4:95-105. Chaix, C., Molko, D., and Téoule, R. 1989. The use of labile base protecting groups in oligoribonucleotide synthesis. Tetrahedron Lett. 30:71-74. Clauwaert, J. and Stockx, J. 1986. Interactions of polynucleotides and their components. I. Dissociation constants of the bases and their derivatives. Z. Naturforsch. B. 23:25-30. den Hartog, J.A.J., Willie, G., Scheublin, R.A., and van Boom, J.H. 1982. Chemical synthesis of a messenger ribonucleic acid fragment: AUGUUCUUCUUCUUCUUC. Biochemistry 21:10091018. Dikshit, A., Chaddha, M., Singh, R.K., and Misra, K. 1988. Naphthaloyl group: A new selective amino protecting group for deoxynucleosides in oligonucleotide synthesis. Can. J. Chem. 66:2989-2994. Divakar, K.J. and Reese, C.B. 1982. 4-(1,2,4-Triazol-1-yl)- and 4-(3-nitro-1,2,4-triazol-1-yl)-1(β-D-2,3,5-tri-O-acetylarabinofuranosyl)pyrim idin-2(1H)-ones. Valuable intermediates in the synthesis of derivatives of 1-(β-D-arabinofuranosyl)cytosine (Ara-C). J. Chem. Soc. Perkin Trans. 1:1171-1176. Dreef-Tromp, C.M., van Dam, E.M.A., van den Elst, H., van der Marel, G.A., and van Boom, J.H. 1990. Solid-phase synthesis of H-Phe-Tyr(pATAT)-NH2: A nucleopeptide fragment from the nucleoprotein of bacteriophage φX174. Nucl. Acids Res. 18:6491-6495. Dunn, D.B., and Hall, R.H. 1975. Purines, pyrimidines, nucleosides and nucleotides: Physical constants and spectral properties. In Handbook of Biochemistry and Molecular Biology, 3rd ed., Vol. 1: Nucleic Acids (G.D. Fasman, ed.) pp. 65-125. CRC Press, Boca Raton, Fla. Engels, J.W. and Mag, M. 1982. Amide protection in oligodeoxynucleotide synthesis. Nucleosides Nucleotides 6:473-475. Fourrey, J.-L. and Varenne, J. 1985. Preparation and phosphorylation reactivity of N-nonacylated nucleoside phosphoramidites. Tetrahedron Lett. 26:2663-2666. Francois, P., Hamoir, G., Sonveaux, E., Vermeersch, H., and Ma, Y. 1985. On the phosphorylation of deoxyribonucleosides and the protection of deoxyguanosine. Bull. Soc. Chim. Belg. 94:821823. Nucleobase Protection of Deoxyribo- and Ribonucleosides

Froehler, B.C. and Matteucci, M.D. 1983. Dialkylformamidines: Depurination resistant N6-protecting group for deoxyadenosine. Nucl. Acids Res. 11:8031-8036.

Froehler, B.C., Ng, P.G., and Matteucci, M.D. 1986. Synthesis of DNA via deoxynucleoside H-phosphonate intermediates. Nucl. Acids Res. 14:5399-5407. Fujii, M., Yamakage, S., Takaku, H., and Hata, T. 1987. (Butylthio)carbonyl group: A new protecting group for the guanine residue in oligoribonucleotide synthesis. Tetrahedron Lett. 28:57135716. Gaffney, B.L. and Jones, R.A. 1982. A new strategy for the protection of deoxyguanosine during oligonucleotide synthesis. Tetrahedron Lett. 23:2257-2260. Garegg, P.J., Lindh, I., Regberg, T., Stawinski, J., Strömberg, R., and Henrichson, C. 1986. Nucleoside H-phosphonates. IV. Automated solid phase synthesis of oligoribonucleotides by the hydrogenphosphonate approach. Tetrahedron Lett. 27:4055-4058. Gilham, P.T. and Khorana, H.G. 1958. Studies on polynucleotides. I. A new and general method for the chemical synthesis of the C5′-C3′ internucleotidic linkage. Syntheses of deoxyribo-dinucleotides. J. Am. Chem. Soc. 80:6212-6222. Gryaznov, S.M. and Letsinger, R.L. 1991. Synthesis of oligonucleotides via monomers with unprotected bases. J. Am. Chem. Soc. 113:5876-5877. Hagen, M.D. and Chládek, S. 1989. General synthesis of 2′(3′)-O-aminoacyl oligoribonucleotides. The protecion of the guanine moiety. J. Org. Chem. 54:3189-3195. Hall, R.H., Todd, A.R., and Webb, R.F. 1957. Nucleotides. Part XLI. Mixed anhydrides as intermediates in the synthesis of dinucleoside phosphates. J. Chem. Soc. 3291-3296. Hayakawa, Y., Kato, H., Uchiyama, M., Kajino, H., and Noyori, R. 1986. Allyloxycarbonyl group: A versatile blocking group for nucleotide synthesis. J. Org. Chem. 51:2400-2402. Hayakawa, Y., Wakabayashi, S., Kato, H., and Noyori, R. 1990. The allylic protection method in solid-phase oligonucleotide synthesis. An efficient preparation of solid-anchored DNA oligomers. J. Am. Chem. Soc. 112:1691-1696. Hayakawa, Y., Hirose, M., and Noyori, R. 1993. O-Allyl protection of guanine and thymine residues in oligodeoxyribonucleotides. J. Org. Chem. 58:5551-5555. Heikkilä, J. and Chattopadhyaya, J. 1983. The 9fluorenylmethoxycarbonyl (Fmoc) group for the protection of amino functions of cytidine, adenosine, guanosine and their 2′-deoxysugar derivatives. Acta Chem. Scand. B37:263-265. Himmelsbach, F., Schulz, B.S., Trichtinger, T., Charubala, R., and Pfleiderer, W. 1984. The pnitrophenylethyl (NPE) group. A versatile new blocking group for phosphate and aglycone protection in nucleosides and nucleotides. Tetrahedron 40:59-72.


Holy, A. and Zemlicka, J. 1969. Oligonucleotidic compounds. XXXIII. A study on hydrolysis of N-dimethylaminomethylenecytidine, -adenosine, -guanosine, and related 2′-deoxy compounds. Collect. Czech. Chem. Commun. 34:2449-2458. Honda, S., Urakami, K., Koura, K., Terada, K., Sato, Y., Kohno, K., Sekine, M., and Hata, T. 1984. Synthesis of oligoribonucleotides by use of S,Sdiphenyl N-monomethoxytrityl ribonucleoside 3′-phosphorodithioates. Tetrahedron 40:153163. Huynh-Dinh, T., Langlois d’Estaintot, B., Allard, P., and Igolen, J. 1985. Synthèse simplifiée de sondes mixtes avec des triazolo-nucléosides. Tetrahedron Lett. 26:431-434. Igolen, J. and Morin, C. 1980. Rapid synthesis of protected 2′-deoxycytidine derivatives. J. Org. Chem. 45:4802-4804. Ito, T., Ued a, S., and Takaku, H. 19 86. (Methoxyethoxy)methyl group: New amide and hydroxyl protecting groups of uridine in oligonucleotide synthesis. J. Org. Chem. 51:931933. Iyer, R.P. and Beaucage, S.L. 1999. Oligonucleotide synthesis. In Comprehensive Natural Products Chemistry, Vol. 7: DNA and Aspects of Molecular Biology (E.T. Kool, ed.) pp. 105-152. Elsevier Science Publishing, New York. Iyer, R.P., Yu, D., Habus, I., Ho, N.H., Johnson, S., Devlin, T., Jiang, Z., Zhou, W., Xie, J., and Agrawal, S. 1997. N-Pent-4-enoyl (PNT) group as a universal nucleobase protector: Applications in the rapid and facile synthesis of oligonucleotides, analogs and conjugates. Tetrahedron 53:2731-2750. Jones, S.S., Reese, C.B., Sibanda, S., and Ubasawa, A. 1981. The protection of uracil and guanine residues in oligonucleotide synthesis. Tetrahedron Lett. 22:4755-4758.

Krecmerová, M., Hrebabecky, H., and Holy, A. 1990. Synthesis of 5′-O-phosphonomethyl derivatives of pyridine 2′-deoxynucleosides. Collect. Czech. Chem. Commun. 55:2521-2536. Kuijpers, W.H.A., Huskens, J., and van Boeckel, C.A.A. 1990. The 2-(acetoxymethyl)benzoyl (AMB) group as a new base-protecting group, designed for the protection of phosphate modified oligonucleotides. Tetrahedron Lett. 31:6729-6732. Kume, A., Sekine, M., and Hata, T. 1982. Phthaloyl group: A new amino protecting group of deoxyadenosine in oligonucleotide synthesis. Tetrahedron Lett. 23:4365-4368. Kume, A., Iwase, R., Sekine, M., and Hata, T. 1984. Cyclic diacyl groups for protection of the N6amino group of deoxyadenosine in oligodeoxynucleotide synthesis. Nucl. Acids Res. 12:8525-8538. Letsinger, R.L. and Ogilvie, K.K. 1969. Synthesis of oligothymidylates via phosphotriester intermediates. J. Am. Chem. Soc. 91:3350-3355. Letsinger, R.L. and Lunsford, W.B. 1976. Synthesis of thymidine oligonucleotides by phosphite triester intermediates. J. Am. Chem. Soc. 98:36553661. Li, B.F.L., Reese, C.B., and Swann, P.F. 1987. Synthesis and characterization of oligodeoxynucleotides containing 4-O-methylthymine. Biochemistry 26:1086-1093. MacMillan, A.M. and Verdine, G.L. 1991. Engineering tethered DNA molecules by the convertible nucleoside approach. Tetrahedron 47:26032616. Mag, M. and Engels, J.W. 1988. Synthesis and structure assignments of amide protected nucleosides and their use as phosphoramidites in deoxyoligonucleotide synthesis. Nucl. Acids Res. 16:3525-3543. Marugg, J.E., Tromp, M., Jhurani, P., Hoyng, C.F., van der Marel, G.A., and van Boom, J.H. 1984. Synthesis of DNA fragments by the hydroxybenzotriazole phosphodiester approach. Tetrahedron 40:73-78.

Kamaike, K., Hasegawa, Y., and Ishido, Y. 1988. A simple, preparative procedure for N3-anisoyluridine and O6-diphenylcarbamoylguanosine 2′O-(tetrahydropyran-2-yl) derivatives via the corresponding 3′,5′-dibenzoates. Nucleosides Nucleotides 7:37-43.

Matteucci, M.D. and Caruthers, M.H. 1980. The synthesis of oligodeoxypyrimidines on a polymer support. Tetrahedron Lett. 21:719-722.

Kamimura, T., Masegi, T., Urakami, K., Honda, S., Sekine, M., and Hata, T. 1983a. A new protecting tactics for the uracil residue in oligoribonucleotide synthesis. Chem. Lett. 1051-1054.

McBride, L.J., Kierzek, R., Beaucage, S.L., and Caruthers, M.H. 1986. Amidine protecting groups for oligonucleotide synthesis. J. Am. Chem. Soc. 108:2040-2048.

Kamimura, T., Tsuchiya, M., Koura, K., Sekine, M., and Hata, T. 1983b. Diphenylcarbamoyl and propionyl groups: A new combination of protecting groups on the guanine residue. Tetrahedron Lett. 24:2775-2778.

McBride, L.J., Eadie, J.S., Efcavitch, J.W., and Andrus, A. 1987. Base modification and the phosphoramidite approach. Nucleosides Nucleotides 6:297-300.

Khorana, H.G. 1979. Total synthesis of a gene. Science 203:614-625.

Michelson, A.M. and Todd, A.R. 1955. Nucleotides. Part XXXII. Synthesis of a dithymidine dinucleotide containing a 3′:5′-internucleotidic linkage. J. Chem. Soc. 2632-2638.

Köster, H., Kulikowski, K., Liese, T., Heikens, W., and Kohli, V. 1981. N-acyl protecting groups for deoxynucleosides. A quantitative and comparative study. Tetrahedron 37:363-369.

Mishra, R.K. and Misra, K. 1986. Improved synthesis of oligodeoxyribonucleotide using 3methoxy-4-phenoxybenzoyl group for amino protection. Nucl. Acids Res. 14:6197-6213.



Mullah, B., Andrus, A., Zhao, H., and Jones, R.A. 1995. Oxidative conversion of N-dimethylformamidine nucleosides to N-cyano nucleosides. Tetrahedron Lett. 36:4373-4376.

Prasad, A.K. and Wengel, J. 1996. Enzyme-mediated protecting group chemistry on the hydroxyl groups of nucleosides. Nucleosides Nucleotides 15:1347-1359.

Nagaich, A.K. and Misra, K. 1989. Highly efficient synthesis of oligodeoxyribonucleotides using αphenyl cinnamoyl group for selective amino protection. Nucl. Acids Res. 17:5125-5134.

Rao, T.S., Reese, C.B., Serafinowska, H.T., Takaku, H., and Zappia, G. 1987. Solid-phase synthesis of the 3′-terminal nonadecaribonucleoside octadecaphosphate sequence of yeast alanine transfer ribonucleic acid. Tetrahedron Lett. 28:48974900.

Narang, S.A., Itakura, K., and Wightman, R.H. 1972. A simplification in the synthesis of deoxyribooligonucleotides. Can. J. Chem. 50:769770. Nielsen, J., Dahl, O., Remaud, G., and Chattopadhyaya, J. 1987. Phosphitylation of guanine or inosine bases during the preparation of nucleoside phosphoramidites. Isolation of model products as thiophosphoric amide derivatives and structure elucidation by 15N NMR spectroscopy. Acta. Chem. Scand. B41:633-639. Nyilas, A., Zhou, X.-X., Welch, C.J., and Chattopadhyaya, J. 1987. A versatile strategy for the O4protection and modification of the lactam function of uridine and uridylic acid. Nucl. Acids Res. Symp. Ser. 18:157-160. Nyilas, A., Földesi, A., and Chattopadhyaya, J. 1988. Arenesulfonylethoxycarbonyl—A set of amino protecting groups for DNA and RNA synthesis. Nucleosides Nucleotides 7:787-793. Ogilvie, K.K., Nemer, M.J., Hakimelahi, G.H., Proba, Z.A., and Lucas, M. 1982. N-Levulination of nucleosides. Tetrahedron Lett. 23:26152618. Oivanen, M., Lönnberg, H., Zhou, X.X., and Chattopadhyaya, J. 1987. Acidic hydrolysis of 6-substituted 9-(2-deoxy-β-D-erythro-pentofuranosyl)purines and their 9-(1-alkoxyethyl) counterparts: Kinetics and mechanism. Tetrahedron 43:1133-1140. Paul, C.H. and Royappa, A.T. 1996. Acid binding and detritylation during oligonucleotide synthesis. Nucl. Acids Res. 24:3048-3052. Pfister, M. and Pfleiderer, W. 1989. New results in oligoribonucleotide synthesis. Nucleosides Nucleotides 8:1001-1006. Pfister, M., Farkas, S., Charubala, R., and Pfleiderer, W. 1988. Recent progress in oligoribonucleotide synthesis. Nucleosides Nucleotides 7:595-600. Pfleiderer, W., Himmelsbach, F., Charubala, R., Schirmeister, H., Beiter, A., Schulz, B., and Trichtinger, T. 1985. The p-nitrophenylethyl group—An universal blocking group in nucleoside and nucleotide chemistry. Nucleosides Nucleotides 4:81-94. Pon, R.T., Damha, M.J., and Ogilvie, K.K. 1985a. Necessary protection of the O6-position of guanosine during the solid phase synthesis of oligonucleotides by the phosphoramidite approach. Tetrahedron Lett. 26:2525-2528. Nucleobase Protection of Deoxyribo- and Ribonucleosides

Pon, R.T., Damha, M.J., and Ogilvie, K.K. 1985b. Modification of guanine bases by nucleoside phosphoramidite reagents during the solid phase synthesis of oligonucleotides. Nucl. Acids Res. 13:6447-6465.

Reddy, M.P., Hanna, N.B., and Farooqui, F. 1994. Fast cleavage and deprotection of oligonucleotides. Tetrahedron Lett. 35:4311-4314. Reese, C.B. 1978. The chemical synthesis of oligoand poly-nucleotides by the phosphotriester approach. Tetrahedron 34:3143-3179. Reese, C.B. and Ubasawa, A. 1980. Reaction between 1-arenesulphonyl-3-nitro-1,2,4-triazoles and nucleoside residues. Elucidation of the nature of side-reactions during oligonucleotide synthesis. Tetrahedron Lett. 21:2265-2268. Reese, C.B. and Skone, P.A. 1984. The protection of thymine and guanine residues in oligodeoxyribonucleotide synthesis. J. Chem. Soc. Perkin Trans. 1:1263. Romero, R., Stein, R., Bull, H.G., and Cordes, E.H. 1978. Secondary deuterium isotope effects for acid-catalyzed hydrolysis of inosine and adenosine. J. Am. Chem. Soc. 100:7620-7624. Saenger, W. 1984. Principles of Nucleic Acids Structure. Springer-Verlag, New York. Scalfi-Happ, C., Happ, E., and Chládek, S. 1987. New approach to the synthesis of 2′(3′)-O-aminoacyl-oligoribonucleotides related to the 3′-terminus of aminoacyl transfer ribonucleic acid. Nucleosides Nucleotides 6:345-348. Schaller, H., Weimann, G., Lerch, W.B., and Khorana, H.G. 1963. Studies on polynucleotides. XXIV. The stepwise synthesis of specific deoxyribopolynucleotides (4). Protected derivatives of deoxyribonucleosides and new syntheses of deoxyribonucleoside-3′ phosphates. J. Am. Chem. Soc. 85:3821-3827. Schulhof, J.C., Molko, D., and Teoule, R. 1987. Facile removal of new base protecting groups useful in oligonucleotide synthesis. Tetrahedron Lett. 28:51-54. Seela, F. and Driller, H. 1989. 7-Deaza-2′-deoxyO6-methylguanosine: Selective N2-formylation via a formamidine, phosphoramidite. Synthesis and properties of oligonucleotides. Nucleosides Nucleotides 8:1-21. Sekine, M. 1989. General method for the preparation of N3- and O4-substituted uridine derivatives by phase-transfer reactions. J. Org. Chem. 54:2321-2326. Sekine, M., Masuda, N., and Hata, T. 1985. Introduction of the 4,4′,4′′-tris(benzoyloxy)trityl group into the exo amino groups of deoxyribonucleosides and its properties. Tetrahedron 41:5445-5453.


Shabarova, Z. and Bogdanov, A. 1994. Advanced organic chemistry of nucleic acids. VCH Publishers, New York. Shimidzu, T. and Letsinger, R.L. 1968. Synthesis of deoxyguanylyl-deoxyguanosine on an insoluble polymer support. J. Org. Chem. 33:708-711. Singh, R.K. and Misra, K. 1988. Improvements in oligodeoxyribonucleotide synthesis using phenoxyacetyl as amino protecting group. Indian J. Chem. 27B:409-417. Sinha, N.D., Davis, P., Usman, N., Pérez, J., Hodge, R., Kremsky, J., and Casale, R. 1993. Labile exocyclic amine protection of nucleosides in DNA, RNA and oligonucleotide analog synthesis facilitating N-deacylation, minimizing depurination and chain degradation. Biochimie 75:1323. Smrt, J. and Sorm, F. 1967. Oligonucleotidic compounds. XVIII. Synthesis of guanylyl-(3′-5′)uridylyl-(3′-5′)-arabinofuranosyluracil and guanylyl-(3′-5′)-uridylyl-(3′-5′)-arabinofurano -sylcytosine. Collect. Czech. Chem. Commun. 32:3169-3176. Sonveaux, E. 1986. The organic chemistry underlying DNA synthesis. Bioorganic Chem. 14:274325. Sproat, B.S., Iribarren, A.M., Guimil Garcia, R., and Beijer, B. 1991. New synthetic routes to synthons suitable for 2′-O-allyloligoribonucleotide assembly. Nucl. Acids Res. 19:733-738. Sung, W.L. 1982. Synthesis of 4-(1,2,4-triazol-1yl)pyrimidin-2(1H)-one-ribonucleotide and its application in synthesis of oligoribonucleotides. J. Org. Chem. 47:3623-3628. Suzuki, T., Ohsumi, S., and Makino, K. 1994. Mechanistic studies on depurination and apurinic site chain breakage in oligodeoxyribonucletides. Nucl. Acids Res. 22:4997-5003. Takaku, H., Imai, K., and Nagai, M. 1988. Triphenylmethanesulfenyl group. A new protecting group for the uracil residue in oligoribonucleotide synthesis. Chem. Lett. 857-860. Tanimura, H., Fukazawa, T., Sekine, M., Hata, T., Efcavitch, J.W., and Zon, G. 1988. The practical synthesis of RNA fragments in the solid phase approach. Tetrahedron Lett. 29:577-578. Ti, G.S., Gaffney, B.L., and Jones, R.A. 1982. Transient protection: Efficient one-flask synthesis of protected deoxynucleosides. J. Am. Chem. Soc. 104:1316-1319. Trichtinger, T., Charubala, R., and Pfleiderer, W. 19 83. Synth esis of O6-p-nitrophenylethyl guanosine and 2′-deoxyguanosine derivatives. Tetrahedron Lett. 24:711-714. Uchiyama, M., Aso, Y., and Noyori, R. 1993. O-Selective phosphorylation of nucleosides without N-protection. J. Org. Chem. 58:373-379.

Urdea, M.S., Ku, L., Horn, T., Gee, Y.G., and Warner, B.D. 1986. Base modification and cloning efficiency of oligodeoxyribonucleotides synthesized by the phosphoramidite method: Methyl versus cyanoethyl phosphorous protection. Nucl. Acids Res. Symp. Ser. 16:257-260. Uznanski, B., Grajkowski, A., and Wilk, A. 1989. The isopropoxyacetic group for convenient base protection during solid-support synthesis of oligodeoxyribonucleotides and their triester analogs. Nucl. Acids Res. 17:4863-4871. Van Aerschot, A., Herdewijn, P., Janssen, G., and Vanderhaeghe, H. 1988. Protection of the lactam function of 2′-deoxyinosine with a 2-(4-nitrophenyl)-ethyl moiety. Nucleosides Nucleotides 7:519-536. Vu, H., McCollum, C., Jacobson, K., Theisen, P., Vinayak, R., Spiess, E., and Andrus, A. 1990. Fast oligonucleotide deprotection phosphoramidite chemistry for DNA synthesis. Tetrahedron Lett. 31:7269-7272. Wagner, T. and Pfleiderer, W. 1997. Aglycone protection by the (2-dansylethoxy)carbonyl (= {2{[5-(dimethylamino)naphthalen-1-yl] sulfonyl}ethoxy}carbonyl; dnseoc) group—A new variation in oligodeoxyribonucleotide synthesis. Helv. Chim. Acta 80:200-212. Watanabe, K.A. and Fox, J.J. 1966. A simple method for selective acylation of cytidine on the 4-amino group. Angew. Chem. Intl. Ed. Engl. 5:579-580. Welch, C.J., Bazin, H., Heikkilä, J., and Chattopadhyaya, J. 1985. Synthesis of C-5 and N-3 arenesulfenyl uridines. Preparation and properties of a new class of uracil protecting group. Acta Chem. Scand. B39:203-212. York, J.L. 1981. Effect of structure of the aglycon on the acid-catalyzed hydrolysis of adenine nucleosides. J. Org. Chem. 46:2171-2173. Zhou, X.-X. and Chattopadhyaya, J. 1986. Site-specific modification of the pyrimidine residue during the deprotection of the fully-protected diuridylic acid. Tetrahedron 42:5149-5156. Zhou, X.X., Sandström, A., and Chattopadhyaya, J. 1986. A convenient preparation of 2-N-(4-tbutylbenzoyl)-6-O-(2-nitrophenyl)guanosine and its application in the synthesis of 5′(GpGpGpU)3′ constituting the 3′-anticodon stem of E.coli tRNAIle. Chem. Scr. 26:241-249. Zoltewicz, J.A. and Clark, D.F. 1972. Kinetics and mechanism of the hydrolysis of guanosine and 7-methylguanosine nucleosides in perchloric acid. J. Org. Chem. 37:1193-1197. Zoltewicz, J.A., Clark, D.F., Sharpless, T.W., and Grahe, G. 1970. Kinetics and mechanism of the hydrolysis of some purine nucleosides. J. Am. Chem. Soc. 92:1741-1750.

Contributed by Radhakrishnan P. Iyer OriGenix Technologies Laval, Quebec, Canada



Protection of 2′-Hydroxy Functions of Ribonucleosides The methods used to protect 2′-hydroxy functions of ribonucleosides have recently been reviewed (Beaucage and Iyer, 1992; Sonveaux, 1994; Beaucage and Caruthers, 1996). In addition, there have been earlier brief reviews (Ohtsuka and Iwai, 1987; Reese, 1989). The main purpose of this article is to discuss 2′-protection in the context of effective oligoribonucleotide synthesis. For this reason, emphasis will be placed on what are now, or are likely to become, the 2′-protecting groups of choice in the synthesis of oligo- and poly-ribonucleotides (RNA sequences). As a result, only some of the protecting groups that have been suggested for this purpose are considered in detail here, and some interesting chemistry has necessarily been omitted.

CONSIDERATIONS FOR 2′-PROTECTING GROUPS IN OLIGORIBONUCLEOTIDE SYNTHESIS There are three main general criteria that all protecting groups should fulfill (Reese, 1978). (1) They should be easy to introduce and, as part of this criterion, the reagents involved in their introduction should be readily accessible. (2) They should be stable and remain intact until it is appropriate to remove them. (3) They should be removable at the appropriate time using conditions under which the desired product is completely stable. In the case of chiral substrates such as ribonucleosides, achiral protecting groups are desirable for analytical (e.g., NMR, TLC, and HPLC) purposes. In the case of all substrates, it is desirable that the introduction of protecting groups should not result in unduly complex NMR spectra.

O

The successful chemical synthesis of polynucleotides (including RNA sequences) depends on the choice of suitable protecting groups and effective phosphorylation procedures. Arguably the most crucial single decision that has to be made in oligoribonucleotide synthesis is the choice of the protecting group (R; see S.1) for the 2′-hydroxy functions (Reese, 1978). This protecting group must remain intact until the very last step of the synthesis (Fig. 2.2.1), and must then be removable under conditions that are mild enough to avoid subsequent attack of the released 2′-hydroxy functions (see S.2) on the vicinal phosphodiester internucleotide linkages, thereby leading to their cleavage or migration. Protecting groups are often removed hydrolytically under either basic or acidic conditions. Cleavage of interribonucleotide linkages can occur under relatively mild basic conditions (Järvinen et al., 1991; Kuusela and Lönnberg, 1994). This process, illustrated in Figure 2.2.2A, essentially involves an ester exchange reaction between the 2′-hydroxy function of the 3′-linked nucleoside residue and the 5′-hydroxy function of the 5′-linked nucleoside residue, leading to a (2′,3′)-cyclic phosphate (S.3). This intermediate then undergoes further basecatalyzed hydrolysis to give a mixture of isomeric 2′- and 3′-phosphates (S.5 and S.6, respectively). Under acidic conditions (Fig. 2.2.2B), internucleotide cleavage and migration can both occur (Griffin et al., 1968). These processes are both believed to proceed via a phosphorane intermediate (S.7; Järvinen et al., 1991). If the P-O(2′) bond is then cleaved, starting material S.2 is regenerated. If the PO(3′) bond is cleaved, the isomeric product S.8

B

O

O

O P O O

O O

B'

Figure 2.2.1

OH

O P O O

OR

1

B

O O

OR

O

UNIT 2.2

O O

B'

OH

2

Scheme showing protected 2′-hydroxy functions. B and B′ are bases.

Contributed by Colin B. Reese Current Protocols in Nucleic Acid Chemistry (2000) 2.2.1-2.2.24 Copyright © 2000 by John Wiley & Sons, Inc.


2.2.1

O O

B

O

O O

O

OH

B

O

A

P

O

O

HO

O

O

O P O

3

HO

B

O

OH

5

HO

O P O O

O O

B' HO

B'

O

O

B

O

OH O

2

OH

O

OH

O P O

4

OH 6

B O

B

O 3' 2' O HO

O

B

O O

OH

P

O

O OH

5'

O

O

H3O

B' O OH

H3O

7

O P O O

O O

B

O HO

O O P O

B'

O H3O

OH

O O

B'

OH

8

2

O

B

O O O

P

HO O O

3

O O

B'

OH 4

Figure 2.2.2 Scheme showing cleavage of interribonucleotide linkages under (A) basic and (B) acidic conditions. Although only shown in panel A, the hydrolysis of S.3 can yield either the 2′-phosphate (S.5) or the 3′-phosphate (S.6) under either basic or acidic conditions.

Protection of 2′Hydroxy Functions of Ribonucleosides

with the migrated internucleotide linkage is obtained. Finally, if the P-O(5′) bond is cleaved, the (2′,3′)-cyclic phosphate S.3 is obtained. The cyclic phosphate S.3 undergoes further hydrolysis to give an isomeric mixture of the corresponding 2′- and 3′-phosphates (S.5 and S.6, respectively) under acidic as well as under basic conditions. The significance of these reactions in the context of oligo- and poly-ribonucleotide synthesis will be considered later. However, it is clearly of crucial importance that the 2′-protecting group should strictly satisfy the above criteria (1) and (2). As will become apparent, this is a very demanding requirement, as the 2′-pro-

tecting group must also be fully compatible with the groups that are used to protect the 5′-terminal hydroxy function, the base residues, and the internucleotide linkages. It is therefore appropriate to consider these other protecting group requirements at the outset.

Protection of the 5′-Terminal Hydroxy Function Figure 2.2.3 illustrates a number of groups (R′ in S.9) used to protect the 5′-terminal hydroxy function. Although a good deal of work has been carried out on the synthesis of oligoribonucleotides in solution, most of the recent studies in this area have been concerned with


R'O

B

O O

Ph

Ph C

R

OMe O

OR

9

10a, R = H (MMTr) 10b, R = OMe (DMTr)

CHBr2

O O

11 (Px)

O O

O

O 12 (Fmoc)

Figure 2.2.3

13 (Lev)

14 (Dbmb)

S

O

15 (Ptmt)

Several protecting groups for 5′-terminal hydroxy functions.

solid-phase synthesis. The 5′-terminal protecting group (R′) that has been used most widely for this purpose is the (di-p-anisyl)phenylmethyl group (also known as 4,4′-dimethoxytrityl or DMTr; S.10b; Schaller et al., 1963; UNIT 2.3). The 9-phenylxanthen-9-yl (pixyl or Px, S.11) group has very similar properties to the DMTr group and is equally suitable (Chattopadhyaya and Reese, 1978). The somewhat less labile p-anisyl(diphenyl)methyl group (4-monomethoxytrityl or MMTr; S.10a; Schaller et al., 1963) has also been used, but its greater stability to acid makes it generally less suitable. The great advantage of the DMTr (S.10b) and Px (S.11) protecting groups in solid-phase synthesis, and perhaps also in solution-phase synthesis, is that they can be rapidly and quantitatively removed by treatment with acids, such as diand tri-chloroacetic acids, in anhydrous dichloromethane solution (Sproat and Gait, 1984). A further advantage shared by all three of these protecting groups is that with acid treatment they give rise to colored carbocations that can easily be assayed spectrophotometrically. This permits coupling efficiencies to be monitored. Clearly, if one of these three protecting groups is used in oligoribonucleotide synthesis, the 2′-protecting group (R in formula S.9) must be completely stable under the acidic conditions required for 5′-deprotection. Numerous other 5′-protecting groups have been suggested (Sonveaux, 1994; UNIT 2.3), some of which are removable under mildly basic or virtually neutral conditions. Protecting groups in this latter category include 9-fluorenylmethoxycarbonyl (Fmoc; S.12; Pathak and

Chattopadhyaya, 1985), levulinyl (Lev; S.13; van Boom and Burgers, 1976), 2-(dibromomethyl)benzoyl (Dbmb; S.14; Chattopadhyaya et al., 1979), and 2-(isopropylthiomethoxymethyl)benzoyl (Ptmt; S.15; Brown et al., 1989a). None of these protecting groups has found widespread use in the solidphase synthesis of RNA sequences, but some have proved to be useful in solution-phase synthesis.

Protection of Base Residues The protection of base residues is illustrated in Figure 2.2.4. In the solid-phase synthesis of RNA sequences (Rao et al., 1993), adenine, cytosine, and guanine residues are generally protected by N-acylation (as in S.16, S.18, and S.19, respectively), while uracil residues are left unprotected (as in S.23; UNIT 2.1). The N-acyl protecting groups are usually removed by ammonolysis in the step before the removal of the 2′-protecting groups. As RNA can undergo internucleotide cleavage (Fig. 2.2.2A) under ammonolytic conditions, the base-protecting groups must be removable using conditions under which the 2′-protecting groups are completely stable. Thus, the choice of an N-acyl protecting group for a particular base residue is, to some extent, dependent on the 2′-protecting group used. The dimethylaminomethylene protecting group, which is also removable under ammonolytic conditions, has been recommended for the protection of adenine and guanine residues (as in S.17 and S.20, respectively; Vinayak et al., 1992). Particularly in the solution-phase synthesis of RNA sequences, it may be desir-



O

O HN N N

N

R N

N

N

N

HN

NMe2

N

N

N O

N

NH

N

N

N

18

17

16

O

R

O

N H

R

19 O

O N N

NH N

N

N

N

NMe2

OAr

HN

N N

N H

N

R

21

O

O

N

O

N

20

O

O R

N

N H

N

R

O NO2

N O

24

O

N

22

N

N N

23

NPh2

O

OAr N

N

OH

26

25

OH

27

Figure 2.2.4 Several protecting groups for base residues (A: S.16 and S.17; C: S.18; G: S.19 to S.22; U: S.23 to S.25). S.26 and S.27 are used in oximate treatment for the removal of aryl (Ar) groups.

able to protect guanine residues on O6 as well as on N2 (as in S.21 and S.22). Aryl protecting groups are particularly suitable for this purpose (as in S.21; Ar = 2-nitrophenyl, 3-chlorophenyl, and 3,5-dichlorophenyl; Jones et al., 1981; Reese and Skone, 1984; Brown et al., 1989a); they may readily be removed by treatment with the N1,N1,N3,N3-tetramethylguanidinium salt of (E)-2-nitrobenzaldoxime S.26 or of (E)pyridine-2-carboxaldoxime S.27 (oximate treatment; Reese and Zard, 1981) before the ammonolytic removal of the N-acyl protecting groups. The N,N-diphenylcarbamoyl group (as in S.22; Kamimura et al., 1984) is removable by ammonolysis (UNIT 2.1). In the solution-phase synthesis of RNA sequences, it may also be desirable to protect

O

NC

B

O

O O


28

Figure 2.2.5 group.

Protection of Internucleotide Linkages Virtually all of the groups commonly used to protect the internucleotide linkages in both solid- and solution-phase oligo- and poly-ribonucleotide synthesis are removed under basic conditions (Fig. 2.2.5). The 2-cyanoethyl group (as in S.28; Sinha et al., 1983) is by far

O

O OR O P O O

uracil residues on O4 with an aryl group (as in S.24; Ar = 2,4-dimethylphenyl; Jones et al., 1981) or on N3 with an acyl group (as in S.25; R = 4-MeO.C6H4; Kamimura et al., 1984). O4-Aryl and N3-acyl protecting groups may be removed from uracil residues by oximate treatment and by ammonolysis, respectively (UNIT 2.1). It should be noted that the ammonolytic and oxime treatment conditions are both basic.

O

OR

O P O O

B'

O

Cl OR

O 29

O

B

O

B'

B

O

O OR ArS P O O O O

OR

B'

OR

30

Several protecting groups for internucleotide linkages. Ar is phenyl or another aryl


the most commonly used protecting group for the internucleotide linkages in the solid-phase synthesis of RNA sequences, and the 2-chlorophenyl group (as in S.29; Reese, 1970) has been widely used for this purpose in solutionphase synthesis. Another approach to the synthesis of oligoribonucleotides was pioneered by Hata and co-workers (Honda et al., 1984) and involves intermediate S-aryl phosphorothioates (S.30, Ar = Ph). 2-Cyanoethyl protecting groups are usually removed at the same time as N-acyl base-protecting groups by treatment with ammonia (Sinha et al., 1983). 2-Chlorophenyl-protected oligo- and poly-ribonucleotides are best unblocked by treatment with the conjugate base of (E)-2-nitrobenzaldoxime S.26 or (E)-pyridine-2-carboxaldoxime S.27 (Reese et al., 1978; Reese and Zard, 1981). S-Aryl phosphorothioates (S.30), which are masked phosphodiesters, may also be unblocked by oximate treatment (Kamimura et al., 1984). In order to avoid internucleotide cleavage (Fig. 2.2.2A), it is necessary that the 2′-protecting group (R in S.28, S.29, and S.30) should be completely stable under the basic conditions used in the unblocking of the internucleotide linkages.

tions required to remove the 5′-terminal DMTr protecting group S.10b. The 2′-protecting groups must also be stable under the basic conditions (i.e., concentrated aqueous ammonia and oximate ions) required to unblock the base residues and the internucleotide linkages. It is further desirable that 2′-protecting groups not be excessively bulky and thereby impede the coupling process. For the successful removal of the 2′-protecting groups, it must always be borne in mind that RNA is a very sensitive material that is unstable under both acidic and basic conditions and in the presence of various hydrolytic enzymes. It is therefore desirable that manipulation should be kept to a minimum in the isolation of fully unblocked RNA.

Ether Protecting Groups The protection of the 2′-hydroxy functions as readily cleavable ether groups would appear at first sight to be an attractive proposition. Indeed, the possibility of using the benzyl protecting group was first examined over 30 years ago (Griffin et al., 1966). Uridine was converted into its 2′-O-benzyl derivative S.31, which was then successfully converted via the protected dinucleoside phosphate S.32 into uridylyl(3′→5′)-uridine (S.33; Fig. 2.2.6). The benzyl protecting group, which is stable both to acidand base-catalyzed hydrolysis, was removed by catalytic hydrogenolysis in the presence of palladized charcoal. However, it was subsequently reported that concomitant hydrogenation of the uracil 5,6-double bond can occur (Reitz and Pfleiderer, 1975). There is also a danger that the total removal of all of the 2′-protecting groups of per-2′-O-benzylated RNA sequences may not always be possible. The 2-nitrobenzyl group (as in S.34; Fig. 2.2.7), which was introduced by Ikehara and

PROTECTION OF THE 2′-HYDROXY FUNCTION IN OLIGORIBONUCLEOTIDE SYNTHESIS It is clear from the above discussion that the requirements for a 2′-protecting group in oligoand polyribonucleotide synthesis are very demanding indeed. With regard to solid-phase synthesis, in addition to meeting the above general criteria for protecting groups, it is crucially important that 2′-protecting groups be stable to repeated exposure to the acidic condi-

Me3C O

N

O HO

Ura

O

HO

O

O

OBn

O P O O

O

Ura

O

O

H

OMe

32

31

OH

O

O

Ura

O

O NH

HO

O

O P O O

O HO

Ura

OH

33

Ura = uracil-1-yl Bn = PhCH2

Figure 2.2.6 uridine.

Scheme showing the preparation of uridylyl-(3′→5′)-uridine from 2′-O -benzyl (Bn)-



O

B

O O O O P O O

O

O

B OMe

O O O P O

NO2

R

O 35a, R = H 35b, R = OMe

34

Figure 2.2.7 The 2-nitrobenzyl (S.34), 4-methoxybenzyl (S.35a), and 3,4-dimethoxybenzyl (S.35b) protecting groups.

co-workers (Ohtsuka et al., 1978), is potentially a more useful 2′-protecting group. Like the benzyl group, it is stable both to acid- and base-catalyzed hydrolysis. However, it may be cleaved photochemically by irradiation with ultraviolet light (λ > 280 nm). It was later reported that photolytic cleavage of the 2-nitrobenzyl protecting group proceeds more efficiently in slightly acidic (pH 3.5) 0.1 mol dm−3 ammonium formate solution (Hayes et al., 1985). A serious drawback to the use of the 2-nitrobenzyl protecting group is that the photolytic cleavage reaction does not always proceed quantitatively (Ohtsuka and Iwai, 1987), especially in the unblocking of relatively highmolecular-weight RNA sequences. Takaku and co-workers have used the 4methoxybenzyl (as in S.35a; Takaku and Kamaike, 1982; Takaku et al., 1984) and 3,4dimethoxybenzyl (S.35b; Takaku et al., 1986) groups to protect 2′-hydroxy functions in solution-phase oligoribonucleotide synthesis. The 4-methoxybenzyl protecting groups were removed from a hexaribonucleoside pentaphosphate (Takaku et al., 1984) by treatment for 3 hr at room temperature with a reagent prepared by adding triphenylmethyl tetrafluoroborate (∼0.10 mmol/mL) to acetonitrile/water (4:1 v/v). However, Takaku et al. (1986) reported that incomplete unblocking and some cleavage of the glycosidic linkages can occur under these presumably rather acidic conditions. Some cleavage and migration of the internucleotide linkages might also be expected to occur. The 3,4-dimethoxybenzyl protecting group may be

removed (Takaku et al., 1986) under somewhat milder conditions by treatment with 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) in wet dichloromethane; this group would appear to be potentially more promising than the 4-methoxybenzyl group for the protection of the 2′-hydroxy functions in oligoribonucleotide synthesis.

The tert-Butyldimethylsilyl Protecting Group The tert-butyldimethylsilyl group (TBDMS; as in S.37 and S.38; Fig. 2.2.8) was originally suggested by Stork and Hudrlik (1968) for the protection of enols, and was first used by Corey and Venkateswarlu (1972) for the protection of alcoholic hydroxy functions. Ogilvie and coworkers (Ogilvie et al., 1974) then introduced it as a protecting group for the 2′-hydroxy functions of ribonucleoside building blocks. The TBDMS group is at present the most widely used 2′-protecting group in solid-phase oligoribonucleotide synthesis (Damha and Ogilvie, 1993). It meets some but by no means all of the above general requirements for protecting groups. It may be readily introduced (Usman et al., 1987), for example, by treating a 5′-O-DMTr-ribonucleoside derivative (S.36) with tert-butylchlorodimethylsilane and imidazole in N,N-dimethylformamide (DMF) solution (Fig.2.2.8). Although the regiochemistry of the silylation reaction can be controlled to some extent (Hakimelahi et al., 1982), a mixture of 2′- and 3′-isomers (S.37 and S.38, respectively) is in-

DMTrO DMTrO

O

B

HO


36

Figure 2.2.8 group.

OH

DMTrO

O 3'

2'

Me2Si(t-Bu)Cl imidazole

B

O

HO O Me Si Me t-Bu 37

B

OH O Me Si Me t-Bu 38

Scheme showing introduction of the tert-butyldimethylsilyl (TBDMS) protecting


AcO

O

Ade

AcO

O

Ade

OH O Me Si Me t-Bu

HO

O Me Si Me t-Bu 39

40 Ade = adenin-9-yl

Figure 2.2.9 Scheme showing the interconversion of the 2′-O - (S.39) and 3′-O - (S.40) TBDMS adenosine derivatives.

variably obtained. Fortunately, such isomeric mixtures can usually be separated by chromatography on silica gel. However, great care has to be taken in the purification and isolation of the 2′-protected ribonucleoside building blocks (S.37), as the TBDMS group readily migrates from the 2′- to the 3′-hydroxy function and vice versa. Interconversion between the adenosine derivatives S.39 and S.40 (Fig. 2.2.9) was found to be a base-catalyzed first-order equilibration reaction (Jones and Reese, 1979). Equilibration rates were observed to be the same in both directions, and the equilibrium constant was estimated to be 1.0. The half time (t1/2) for equilibration in anhydrous pyridine solution at 36°C was 19 hr. The equilibration rate was increased by a factor of 3.0 when 0.1 mol equiv. (with respect to substrate) of benzylamine (pKa 9.34) was added. Equilibration was faster still (t1/2 = ∼1 hr at 36°C) in methanol-d4 solution without added base. When 0.1 mol equiv. of triethylamine (pKa 10.87) was added to the methanol-d4 solution at 20°C, equilibration was complete within ∼5 min. Precautions must be taken to avoid migration of the TBDMS protecting group during the purification and isolation of 2′-O-TBDMS-5′O-DMTr-ribonucleoside derivatives (S.37) and during the course of their conversion into the required monomeric building blocks. Otherwise, the resulting synthetic RNA sequences will be contaminated with material containing

DMTrO

O HO

(2′→5′)-internucleotide linkages. Thus, in the preparation of 3′-phosphoramidite building blocks (S.41; Fig. 2.2.10), it is advisable that the presence of a strong base such as diisopropylethylamine be avoided. Usman and coworkers (Scaringe et al., 1990) have recommended that a mixture of 2,4,6-collidine and 1-methylimidazole be used. Although the presence of contaminating isomeric 2′-phosphoramidites (S.42) can be detected above a certain level by 31P NMR spectroscopy, these impurities cannot readily be removed. It was recently reported that the oligoribonucleotide r[(Up)20U], prepared by treating its per-2′-O-TBDMS derivative with tetra-n-butylammonium fluoride, contained an average of 1.3% (2′→5′)-internucleotide linkages (Morgan et. al., 1995). As acid was not used either during or after the unblocking process, a reasonable explanation for this observation is that the phosphoramidite S.41 (B = uracil-1-yl) used in its synthesis was contaminated with 1.3% of its 2′-isomer (S.42; B = uracil-1-yl). Using an analytical procedure similar to that described by Morgan et al., it was later concluded that some commercially supplied 2′-OTBDMS-protected ribonucleoside phosphoramidites (S.41) were contaminated with comparable amounts (i.e., >1%) of isomeric 2′-phosphoramidites (S.42; Reese et al., unpub. observ.). However, some other batches of commercially supplied material were estimated to contain smaller quantities of the corresponding

DMTrO

B

OTBDMS

NC

B

O

i O

OTBDMS

B

O

TBDMSO

O

O P N(i-Pr)2

37

DMTrO

41

P O

CN

(i-Pr)2N 42

Figure 2.2.10 Scheme showing conversion of 2′-O -TBDMS-5′-O -DMTr-ribonucleoside into its corresponding 3′-phosphoramidite (S.41) and the structure of the possibly contaminating isomeric 2′-phosphoramidite (S.42). Reagents (i): NCCH2CH2OPN(i-Pr)2Cl, base.



2′-isomer (S.42; Reese et al., unpub. observ.). It would therefore appear that migration of the TBDMS group can to a large extent be controlled if careful manufacturing protocols are observed. The use of the TBDMS protecting group in the solid-phase synthesis of RNA sequences can lead to long coupling times and unsatisfactory coupling efficiencies, possibly due to the bulk of this protecting group. However, the use of 5-ethylthio-1H-tetrazole (S.43b; Fig. 2.2.11) instead of 1H-tetrazole (S.43a) as the phosphoramidite activator can result in shorter coupling times and higher quality products (Sproat et al., 1995; UNIT 3.5). One important advantage of the TBDMS protecting group is that it appears to be stable under the acidic conditions used to remove the 5′-terminal DMTr protecting group in solid-phase synthesis. Problems have arisen in the unblocking of 2′-O-TBDMS-protected RNA sequences. The standard unblocking procedure used in the solid-phase synthesis of DNA sequences involves heating the fully loaded solid support with concentrated aqueous ammonia at 55°C overnight (Brown and Brown, 1991). These ammonolytic conditions lead to the release of the oligodeoxyribonucleotides from the solid support, the removal of the 2-cyanoethyl protecting groups from the internucleotide linkages, and the removal of N-acyl protecting groups from the base residues (in oligodeoxyribonucleotide synthesis, the adenine and cytosine residues are usually protected with benzoyl groups as in S.16 and S.18, R = Ph, and the guanine residues are usually protected with isobutyryl groups as in S.19, R = Me2CH). If these ammonolysis conditions are also used in the unblocking of 2′-O-TBDMS-protected oligoribonucleotides, appreciable loss of the 2′O-TBDMS protecting groups and concomitant cleavage of the internucleotide linkages are likely to occur (Stawinski et al., 1988). This problem of premature removal of 2′-OTBDMS protecting groups has been largely

N N N N H 43a, R = H 43b, R = SEt R


Et3N • 3HF

44

Figure 2.2.11 Phosphoramidite activators 5ethylthio-1H-tetrazole (S.43b) and 1H-tetrazole (S.43a), and the unblocking reagent triethylamine trihydrofluoride (S.44).

overcome by protecting the base residues with more labile acyl groups (Chaix et al., 1989), and by replacing concentrated aqueous ammonia with a more selective reagent such as 35% aqueous ammonia/ethanol (3:1 v/v) ammonia/ethanol (Mullah and Andrus, 1996), anhydrous ethanolic ammonia (Goodwin et al., 1994), or aqueous methylamine (Wincott et al., 1995). Téoule and co-workers (Chaix et al., 1989) have recommended that adenine and guanine residues be protected with phenoxyacetyl groups (as in S.16 and S.19, R = CH2OPh) and that cytosine residues be acetylated (as in S.18, R = Me). These workers found that the half times for removal of the latter protecting groups in aqueous ammonia/ethanol (1:1 v/v) at room temperature ranged from 10 to 15 min. It should therefore be possible completely to unblock base residues that are protected in this way without any significant loss of the 2′-O-TBDMS protecting groups and without internucleotide cleavage. In the final unblocking step, the TBDMS protecting groups are removed from the 2′-hydroxy functions of the synthetic RNA sequences. Until recently, a solution of tetra-nbutylammonium fluoride in tetrahydrofuran (Damha and Ogilvie, 1993) was almost always used for this unblocking process. However, it has been reported that the use of the latter reagent results in an inconvenient work-up procedure and can lead to incomplete unblocking (Sproat et al., 1995). More recently, it has been suggested that triethylamine trihydrofluoride (S.44; Fig. 2.2.11; Gasparutto et al., 1992; Westman and Strömberg, 1994) is a more suitable reagent for this purpose. Both the neat reagent S.44, which is slightly acidic as evidenced by the concomitant loss of 5′-O-DMTr protecting groups (Mullah and Andrus, 1996), and a solution of S.44 and triethylamine in 1-methylpyrrolidone (Wincott et al., 1995) have been used. There now seems to be little doubt that, if the above precautions are taken and the most suitable base-protecting groups and reagents are used, the TBDMS group may be used effectively for the protection of the 2′-hydroxy functions in the solid-phase synthesis of RNA sequences. 2′-O-TBDMS-ribonucleoside 3′H-phosphonate building blocks (S.45; Fig. 2.2.12) have also been used successfully in the solid-phase synthesis of oligoribonucleotides (Rozners et al., 1994). It is reasonable to assume that the same precautions and considerations that apply to solid-phase synthesis based on phosphoramidite building blocks (S.41) should


DMTrO

sequences containing one or more (2′→5′)-internucleotide linkages.

B

O

O OTBDMS O P O H 45

Figure 2.2.12 2′-O -TBDMS-ribonucleoside 3′-H-phosphonate.

be taken into account if good quality RNA sequences are to be obtained from the corresponding H-phosphonates (S.45).

Acetal Protecting Groups In general, acetal groups have several distinct advantages over the TBDMS group as far as the protection of the 2′-hydroxy functions in oligoribonucleotide synthesis is concerned. First, acetal protecting groups can usually be placed regiospecifically on the 2′-hydroxy functions (see below) and, once in position, they cannot migrate. Secondly, they are completely stable under the basic conditions that normally obtain during the unblocking of internucleotide linkages and base residues. Thirdly, the 2′-protected RNA sequences obtained after the removal of the other protecting groups (Rao et al., 1993) can be purified under neutral or basic conditions without any danger of endonuclease-promoted digestion. However, there is one important drawback to the use of acetal protecting groups in that they are generally removed by acid-catalyzed hydrolysis. Unless the acidic conditions used are particularly mild, both cleavage and migration of the internucleotide linkages can occur (Griffin et al., 1968; Capaldi and Reese, 1994; Fig. 2.2.2B). While cleavage of internucleotide linkages is clearly highly undesirable, migration is a very much more serious matter as it is virtually impossible to free even a relatively low-molecular-weight RNA sequence from contaminating isomeric

AcO

O

B

HO

B

O

i, ii OH

AcO

The tetrahydropyran-2-yl (Thp) group In the 1960s, the use of the 2′-O-tetrahydropyran-2-yl protecting group (Thp) in oligoribonucleotide synthesis was examined (Smith et al., 1962; Smrt and Šorm, 1962; Griffin and Reese, 1964). Pure 2′-O-Thp derivatives of uridine and adenosine (S.47a and S.47b, respectively) were prepared according to the procedure indicated in Figure 2.2.13, and were converted into dinucleoside phosphates by the methods then available (Griffin and Reese, 1964; Griffin et al., 1968). Careful unblocking studies were carried out in 0.01 mol dm−3 hydrochloric acid (pH 2.0) at 24°C (Griffin et al., 1968), and the half time (t1/2) for the conversion of 2′-O-Thp-UpU (S.49; Fig. 2.2.14) into completely unprotected uridylyl(3′→5′)-uridine (UpU; S.50) was found to be 29 min. It can therefore be estimated that >99.9% removal of the Thp group would occur in 99.9% removal of the Thp protecting group from 2′-O-ThpUpU, not more than 0.02% phosphoryl migration and 0.01% internucleotide cleavage would be expected to occur. Thus, it seemed reasonable to conclude from the data then available that the Thp group was suitable for the protection of the 2′-hydroxy functions in oligoribonucleotide synthesis.

HO

O

O

O (Thp)

46

47

48

a, B = uracil-1-yl b, B = adenin-9-yl

Figure 2.2.13 Scheme showing preparation of 2′-O -Thp derivatives of uridine and adenosine. Reagents: (i) 3,4-dihydro-2H-pyran (S.48), toluene-4-sulfonic acid (TsOH), dioxane; (ii) NaOMe, MeOH.



HO

O

Ura

HO

O OThp O P O O

Ura

HO

O

O HO

OH

49

Ura

O HO

O OH O P O

H3O Ura

O HO

O

Ura

O O P O O

OH

HO

O

Ura

OH

51

50

Ura = uracil-1-yl

Figure 2.2.14 Scheme showing conversion of 2′-O -Thp-UpU into unprotected uridylyl-(3′→5′)uridine (S.50) and the structure of its (2′→5′)-isomer (S.51).

A particular disadvantage of the Thp group is that it is chiral, and therefore its use in the protection of ribonucleoside derivatives and other chiral compounds leads to mixtures of diastereoisomers. Thus, two diastereoisomers each of 2′-O-Thp-uridine (S.47a) and 2′-OThp-adenosine (S.47b) were obtained (Fig. 2.2.13; Griffin et al., 1968). Although both pairs of diastereoisomers were easily separable and all four compounds were obtained as pure crystalline solids, this is clearly an undesirable complication. The 4-methoxytetrahydropyran-4-yl (Mthp) group A search for an achiral alternative to the Thp protecting group led to the introduction of the 4-methoxytetrahydropyran-4-yl group (Mthp; Reese et al., 1967; 1970). 2′-O-Mthp derivatives of ribonucleosides (S.54) were first prepared from 3′,5′-di-O-acyl-ribonucleosides (S.53; Fig. 2.2.15). However, they are more

R

conveniently prepared from the corresponding 3′,5′-O-(1,1,3,3-tetraisopropyldisiloxan-1,3diyl) derivatives (S.56; Brown et al., 1989a). 2′-O-Mthp derivatives (S.54) are usually obtained as pure crystalline solids in satisfactory to good yields (Reese et al., 1970). The half times for the hydrolysis of 2′-OMthp-uridine and 2′-O-Mthp-adenosine (S.54, B = uracil-1-yl and adenosine-9-yl, respectively) in 0.01 mol dm−3 hydrochloric acid at 22°C were found to be 18.7 and 34 min, respectively (Norman et al., 1984). It is interesting to note that the removal of the Mthp protecting group from 2′-O-Mthp-uridylyl-(3′→5′)-uridine (S.57a) and 2′-O-Mthp-adenylyl-(3′→5′)adenosine (S.57b; Fig. 2.2.16) under the same conditions was found to proceed at significantly faster rates (t1/2 = 6.1 and 19.9 min, respectively; Norman et al., 1984). The rate of removal of the Thp protecting group from 2′O-Thp-UpU (S.49) is also faster than from 2′-O-Thp-uridine (S.47a; Griffin et al., 1968).

B

O

O

O R'

O

O HO

OH

i, ii

O 53

B

HO

OH

B

HO

HO

iii

OMe

O O

OMe O

i, iv

52

O (i-Pr)2Si O (i-Pr)2Si O

B

O

55

O 54

(Mthp)

OH

56


Figure 2.2.15 Scheme showing preparation of 2′-O -Mthp ribonucleoside derivatives (S.54) via 3′,5′-di-O-acyl-ribonucleosides (S.53) or 3′,5′-O-(1,1,3,3-tetraisopropyldisiloxan-1,3-diyl) derivatives (S.56). Reagents: (i) 4-methoxy-5,6-dihydro-2 H-pyran (S.55), toluene-4-sulfonic acid (TsOH), dioxane; (ii) NH3, MeOH; (iii) (i-Pr)2Si(Cl)OSi(Cl)(i-Pr)2, imidazole, MeCN; (iv) Et4NF, MeCN.


HO

O

B

HO

O

O OMthp O P O O

O HO

B

O

OH

O P O

H3O B

O

OH

O HO

57

B

OH

58

a, B = uracil-1-yl b, B = adenin-9-yl

Figure 2.2.16

Scheme illustrating the removal of the Mthp protecting group.

The fact that the presence of a vicinal phosphodiester internucleotide linkage appears to facilitate the acid-catalyzed unblocking of a 2′-O-Mthp- or 2′-O-Thp-protected hydroxy function is clearly advantageous if migration and cleavage of the internucleotide linkages (Fig. 2.2.2B) in the final unblocking step of oligoribonucleotide synthesis are to be kept to a minimum. As well as being achiral, Mthp has an additional advantage over Thp in that it is more labile to acidic hydrolysis. The Mthp and Thp protecting groups have been used in both solution- and solid-phase synthesis of RNA sequences. The Mthp group was introduced particularly for solution-phase synthesis, and it has been used successfully in the preparation of the 3′-terminal decamer, nonadecamer, and heptatriacontamer (37-mer) sequences (r[UpCpGpUpCpCpApCpCpA], r[ApUpUpCpCpGpGpApCpUpCpGpUpCpCpApCpCpA], and r[GpGpApGpApGpGpUpCpUpCpCp GpGpTpψpCpGpApUpUpCpCpGpGpApCpUpCpGpUpCpCpApCpCpA], respectively) of yeast alanine transfer RNA (tRNAAla; Jones et al., 1980, 1983; Brown et al., 1989a,b). This work has already been reviewed (Reese, 1989). The tetrahydrofuran-2-yl (Thf) and 1,5dimethoxycarbonyl-3-methoxypentan-3-yl (Mdmp) groups The above approach to the solution-phase synthesis of RNA sequences was successful largely because treatment with acid was completely avoided until the final unblocking step. However, other workers (Ohtsuka et al., 1984) reported a solution-phase block synthesis of a tritriacontamer (33-mer) sequence of E. coli tRNA2Gly using tetrahydrofuran-2-yl (Thf; S.59; Fig. 2.2.17) and DMTr groups for the protection of the 2′- and 5′-hydroxy functions, respectively. The 5′-terminal DMTr protecting groups were removed by treatment with zinc

bromide in dry dichloromethane/isopropanol solution rather than with a protic acid. Although Thf is more labile than Thp (and probably also Mthp) to acid-catalyzed hydrolysis (Kruse et al., 1979), the latter combination of protecting groups was apparently effective. There are a number of reports in the literature relating to solid-phase RNA synthesis in which the 2′-hydroxy functions are protected by Thp, Mthp, or Thf groups and the 5′-hydroxy functions are also protected with acid-labile groups (Tanaka et al., 1986; Kierzek et al., 1986; Iwai et al., 1987; Tanimura et al., 1989; Tanimura and Imada, 1990). Such acid-labile groups include DMTr (S.10b), 9-phenylxanthen-9-yl (S.11), and 9-(4-methoxyphenyl)xanthen-9-yl (S.60; Fig. 2.2.17; UNIT 2.3). Although some sequences appear to have been prepared successfully in this way, other reports suggest that this is an unsound strategy, particularly for the synthesis of comparatively high-molecular-weight RNA sequences (Reese and Skone, 1985; Christodoulou et al., 1986; Kierzek, 1994). Even when precautions are taken to maintain stringently anhydrous conditions, the repeated exposure of the growing protected oligoribonucleotide to di- or tri-chloroacetic acid in order to remove the 5′-protecting group in each synthetic cycle is likely to lead to some loss of such relatively labile 2′-

MeO

O

O

59

60

Figure 2.2.17 Tetrahydrofuran-2-yl (S.59) and 9-(4-methoxyphenyl)xanthen-9-yl (S.60) protecting groups.



O O (Lev)

B

O

HO

B

O

O NC

O

OThf

i

NC

O P O O

B

O O

O O

(Fmoc) NC

O

OThf

O

61

O

OThf

O P O B

OThf

62

O

B

O

HO

O

B

O

ii O

OMthp

NC

O P O O

O O

B

OMthp

63

O

OMthp

O P O O

O O

B

OMthp

64

Figure 2.2.18 Scheme showing removal of levulinyl (top) and Fmoc (bottom) protecting groups. Reagents: (i) N2H4⋅H2O, C5H5N, AcOH; (ii) 0.1 M 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU), MeCN.


protecting groups, resulting both in cleavage and migration of internucleotide linkages (Pathak and Chattopadhyaya, 1985; Reese and Skone, 1985; Kierzek, 1994). If Thp, Mthp, or Thf groups are to be used to protect the 2′-hydroxy functions, it would appear to be a better strategy in solid-phase synthesis to protect the 5′-terminal hydroxy function with a group that is readily removable under virtually neutral or mildly basic conditions. Thus, following van Boom’s use of the levulinyl group (as in S.61; Fig. 2.2.18) for the protection of the 5′-hydroxy functions in solution-phase synthesis (den Hartog et al., 1981), other workers (Iwai and Ohtsuka, 1988) successfully used Lev in conjunction with Thf in solid-phase oligoribonucleotide synthesis. The Lev group was removed (Fig. 2.2.18) in the usual way (den Hartog et al., 1981) by treatment with hydrazine hydrate in pyridine/acetic acid solution. A number of RNA sequences, including a heneicosamer (21mer), were prepared in this way. In another study (Lehmann et al., 1989), the Fmoc group (S.12) was used to protect the 5′-hydroxy functions in solid-phase oligoribonucleotide synthesis. These workers protected the 2′-hydroxy functions with Mthp groups (as in S.63) and removed the Fmoc group with 1,8- diazabicyclo[5.4.0]undec-7-ene (DBU) in acetonitrile solution (Fig. 2.2.18). As the authors (Lehmann

et al., 1989) pointed out, it is very likely that some concomitant loss of the 2-cyanoethyl protecting groups from the partially protected oligoribonucleotides (S.63) occurs during DBU treatment. A nonadecamer and an icosamer (20-mer) RNA sequence were prepared successfully in this way. Although the above approach using either 2′-O-Thf and 5′-O-Lev protection or 2′-OMthp and 5′-O-Fmoc protection (Fig. 2.2.18) was successful, it is much more convenient to use the acid-labile DMTr (S.10b) or Px (S.11) groups to protect the 5′-hydroxy functions in solid-phase RNA synthesis. The latter modified trityl groups can be rapidly and quantitatively removed under anhydrous conditions and the released carbocations can easily be assayed spectrophotometrically (Brown and Brown, 1991). For this reason, attempts have been made to develop somewhat more sophisticated acetal protecting groups that are stable under normal “detritylation” conditions and are also sufficiently labile to acidic hydrolysis in the final unblocking step for cleavage and migration of the internucleotide linkages (Fig. 2.2.2B) to be avoided. Chattopadhyaya and co-workers (Sandström et al., 1985) showed that the 1,5-dimethoxycarbonyl-3-methoxypentan-3-yl (Mdmp) group (as in S.65; Fig. 2.2.19), derived from dimethyl 4-ketopimelate, was


HO

B

O HO

HO

OMe

O

CO2Me

concentrated aqueous NH3

B

O HO

O

OMe

CO2Me

CONH 2 CONH 2

(Mdmp) 66

65

X

O

OMe O

O

B

O

Thy O

O

OMe

HO 67a, X = O 67b, X = S 67c, X = SO2

Figure 2.2.19

O

B

O

H O

O

N

NAr 68

OMe H Ar

69

Thy = thymin-1-yl Ar = aryl

Acetal protecting groups labile to acidic hydrolysis.

converted under the standard ammonolytic unblocking conditions used in solid-phase synthesis into the corresponding bis-amide (S.66), which was seventeen times more labile to acidic hydrolysis than the bis-ester (S.65). However, the Mdmp protecting group itself is unlikely to find application in solid-phase synthesis of oligoribonucleotides as it undergoes hydrolysis in 4:1 (v/v) acetic acid/water solution even more rapidly than the Mthp group.

The 1-Aryl-4-methoxypiperidin-4-yl (Ctmp and Fpmp) groups Acetal hydrolysis is a second-order reaction; its rate, which is proportional to the concentrations both of substrate and hydrogen ions, is very sensitive to inductive effects (Kreevoy and Taft, 1955). Thus 5′-O-(4methoxytetrahydrothiopyran-4-yl)-thymidine (S.67b; Fig. 2.2.19) was found to be ∼5 times more labile to acidic hydrolysis than the corresponding Mthp derivative (S.67a) and was estimated to be >2000 times more labile than the corresponding sulfone (S.67c; van Boom et al., 1972). It seemed possible that, if the aryl substituent Ar were selected carefully, a 1-aryl-4methoxypiperidin-4-yl protecting group (as in S.68) could be identified that would be almost fully protonated (as in S.69) under detritylation conditions (i.e., in dichloromethane containing, for instance, 2% to 3% trichloroacetic

acid), but would be virtually unprotonated (as in S.68) under the milder conditions of acidic hydrolysis obtaining in the final unblocking step of oligoribonucleotide synthesis. Although it would, of course, depend on the aryl substituent, it seemed possible that the rate of hydrolysis of the unprotonated and protonated piperidinyl species might correspond approximately to those of the Mthp (S.67a) and sulfone (S.67c) derivatives, respectively. In the overall rate expression for the hydrolysis of a 1-aryl4-methoxypiperidin-4-yl derivative, it is reasonable to assume that the component relating to the hydrolysis of the conjugate acid S.69 is likely to be negligible in comparison with that relating to the unprotonated S.68 and that, as a first approximation, it can be ignored. If this is the case, the observed rate of hydrolysis of the 1-aryl-4-methoxypiperidin-4-yl acetal system should be pH independent. For example, if the pH of the hydrolytic medium is lowered by one unit, the concentration of the unprotonated acetal S.68 will decrease by an order of magnitude, and at the same time the rate of hydrolysis of the remaining unprotonated acetal will increase by an order of magnitude. Despite the very limited synthetic methodology available at the outset, it was possible to prepare 2′-O-[1-(2-chloro-4-methylphenyl)-4methoxypiperidin-4-yl] (Ctmp) ribonucleoside derivatives (S.71a; Fig. 2.2.20A) and



A B

O

HO

O

(i-Pr)2Si O (i-Pr)2Si O

OH

B

O

i, ii HO

O

OMe R

1

N

56

2

R 71

B O Cl

O

iii Cl

Cl

Cl

72

MeO

OMe

OMe

73 iv

N

v R

N

1

1

R

NH2 R

1 2

R

2

2

R

R

75

70

74 1

2

a, R = Cl, R = Me 1 2 b, R = F, R = H

Figure 2.2.20 (A) Scheme showing preparation of Ctmp (S.71a) and Fpmp (S.71a) ribonucleoside derivatives. (B) Preparation of the 1-aryl-4-methoxy-1,2,5,6-tetrahydropyridines (S.70) required in (A). Reagents (i) S.70, CF3CO2H, CH2Cl2; (ii) Et4NF, MeCN; (iii) ethylene, AlCl3, CH2Cl2; (iv) toluene-4-sulfonic acid monohydrate (TsOH⋅H2O), MeOH, and reflux followed by (MeO)3CH; (v) (i-Pr)2NEt, Et2O→BF3, CH2Cl2, 0°C.


show that the Ctmp protecting group had the desired properties (Reese et al., 1986). Thus it can be seen from Figure 2.2.21 that the rate of hydrolysis of 2′-O-Ctmp-uridine (S.71a, B = uracil-1-yl) at 30°C is only 1.75 times faster at pH 0.5 than it is at pH 2.5. At 25°C, 2′-O-Ctmpuridine is ∼40 times more stable than 2′-OMthp-uridine (S.54, B = uracil-1-yl) at pH 1.0 (Reese et al., 1986), but it is nearly 1.6 times more labile than 2′-O-Mthp-uridine at pH 3.0. The first general criterion that all protecting groups should meet (see above) is that they should be easy to introduce, and an important part of this criterion is that the reagent required should be readily accessible. The Ctmp and related piperidine-derived protecting groups are easy to introduce, but until recently the preparation of the enol ether reagents (such as S.70) involved a number of steps. However, these 1-aryl-4-methoxy-1,2,5,6-tetrahydropyridine derivatives can now be readily prepared (Fig. 2.2.20B) in two steps and in good overall yields (Faja et al., 1997) from 1,5-dichloropentan-3-one (S.73; Owen and Reese, 1970) and the appropriate primary aromatic amine (S.74). The procedure for the preparation of 2′-O-(1-aryl-4-methoxypiperidin-4-yl) ribonucleoside derivatives (such as S.71; Fig.

Figure 2.2.21 Dependence of half times (t1/2) on pH for hydrolysis of 2′-O -Ctmp-uridine (S.71a) and 2′-O -Fpmp-uridine (S.71b) at 30°C.


PxO

O

NC

B

O

DMTrO

OCtmp

N(i-Pr)2

77

Ctmp-protected phosphoramidite (S.76) and H-phosphonate (S.77).

2.2.20A; Rao et al., 1987, 1993) is closely similar to that used in the preparation of the corresponding 2′-O-Mthp derivatives (S.54; Fig. 2.2.15), except that a much smaller excess of the enol ether reagent S.70 is needed. The 2′-O-Ctmp protecting group was used in conjunction with the 5′-O-Px protecting group (Rao et al., 1987) or the 5′-O-DMTr protecting group (Sakatsume et al., 1989) in the solid-phase synthesis of oligoribonucleotides. Phosphoramidite building blocks (S.76; Fig. 2.2.22) were successfully used in the preparation of the 3′-terminal nonadecamer sequence r[ApUpUpCpCpGpGpApCpUp CpGpUpCpCpApCpCpA] of yeast tRNA Ala (Rao et al., 1987), and H-phosphonate building blocks (S.77) were used successfully in the preparation of the octadecamer sequence, r[ApGpUpApUpApApGpApGpGpApCpApUp ApUpG] (Sakatsume et al., 1989). However, the required enol ether reagent (S.70) was difficult to prepare by the original procedure (Reese et al., 1986), and its preparation by the improved protocol (Fig. 2.2.20B; Faja et al., 1997) involves either the use of an expensive aromatic amine (S.74a) or an additional chlorination step. It was later found that several other 1-aryl4-methoxypiperidin-4-yl groups were also suitable for the protection of the 2′-hydroxy functions in solid-phase oligoribonucleotide synthesis. Among these is the 1-(2fluorophenyl)-4-methoxypiperidin-4-yl (Fpmp) protecting group (as in S.71b; Reese and Thompson, 1988). The Fpmp protecting group has two distinct advantages over the Ctmp group. First, the enol ether reagent S.70b,

PxO

NC

B

OCtmp O O P O Et3NH H

O P 76

Figure 2.2.22

O

B

O O

OFpmp

O P

which is a low-melting solid, is readily prepared (Faja et al., 1997) from 2-fluoroaniline (S.74b), which is an inexpensive starting material. Secondly, the Fpmp group is somewhat more stable than the Ctmp group to acidic hydrolysis in the pH range of 0.5 to 1.0 (∼1.4 times at 30°C; Fig. 2.2.21), and therefore the risk of concomitant 2′-unblocking in the detritylation steps is even smaller. However, removal of the 2′-O-Fpmp protecting group occurs more slowly than removal of the 2′-OCtmp group in the final unblocking step of oligoribonucleotide synthesis, and this can be disadvantageous (see below). It can be seen from Figure 2.2.21 that at 30°C the rate of hydrolysis of 2′-O-Fpmp-uridine (S.71b; B = uracil-1-yl) is only about twice as fast at pH 0.5 as at pH 2.5. The 2′-O-Fpmp protecting group has been widely used in solid-phase oligoribonucleotide synthesis (Beijer et al., 1990; Rao et al., 1993; Capaldi and Reese, 1994; Pieles et al., 1994; Sproat et al., 1994; Rao and Macfarlane, 1995; McGregor et al., 1996). The 5′-O-Px2′-O-Fpmp phosphoramidite building blocks S.78 (Fig. 2.2.23) were used successfully in the synthesis of r[UpCpGpUpCpCpApCpCpA], r[ApUpUpCpCpGpGpApUpCpGpUpCpCp ApCpCpA], and r[GpGpApGpApGpGpUp CpUpCp CpGpGpUpUpCpGpApUpUpCpCpGp GpApCpUpCpGpUpCpCpApCpCpA], the 3′-decamer, nonadecamer, and heptatriacontamer (37-mer) sequences, respectively, of unmodified yeast tRNAAla (Rao et al., 1993). Sproat and co-workers (Pieles et al., 1994) carried out the solid-phase synthesis of some modified oligoribonucleotides containing

DMTrO

NC

O

OFpmp

O P

DMTrO

NC

79

B

O O

OMe

O P

Ni-Pr2

Ni-Pr2 78

B

O

Ni-Pr2 80

Figure 2.2.23 5′-O -Px-2′-O -Fpmp phosphoramidite (S.78), 5′-O -DMTr-2′-O -Fpmp phosphoramidite (S.79), and 5′-O -DMTr-2′-O -methyl-ribonucleoside phosphoramidite (S.80).



O

B

O

O O

O

B

O

O

CO2R

O P O

R

O

1

O

O

O

O P O O R 1

RO2C

2

2

81a, R = NO2, R = H 1 2 81b, R = H, R = NO2

82a, R = Me 82b, R = H

Figure 2.2.24 Additional acetal protecting groups: (2-nitrobenzyloxy)methyl (S.81a), (4-nitrobenzyloxy)methyl (S.81b), (2,6-dimethoxycarbonyl)phenoxymethyl (S.82a), and (2,6-dicarboxy)phenoxymethyl (S.82b).


pseudouridine, 2′-O-methylpseudouridine, and some other 2′-O-methyl-ribonucleoside residues starting from the appropriate 5′-ODMTr-2′-O-Fpmp and 5′-O-DMTr-2′-Omethyl-ribonucleoside phosphoramidites (S.79 and S.80, respectively). In all of the early work (Rao et al., 1993), the 2′-O-Fpmp and 5′-terminal Px (or DMTr) protecting groups were removed by treatment with 0.01 mol dm−3 hydrochloric acid (pH ∼2) at room temperature. However, it soon became clear that the susceptibility of the internucleotide linkages of oligoribonucleotides to acid-catalyzed cleavage and migration was sequence dependent, and that certain sequences were unstable at pH 2 and room temperature (Capaldi and Reese, 1994). Thus, despite the relative stability of uridylyl(3′→5′)-uridine (S.50; Griffin et al., 1968) at pH 2 and room temperature, r[(Up)9U] and r[(Up)19U] both underwent virtually complete degradation in the course of the removal of the 2′-O-Fpmp protecting groups under the same conditions (Capaldi and Reese, 1994). However, when unblocking was carried out at room temperature above pH 3.0, virtually no internucleotide cleavage or migration could be detected (see Conclusions). Other workers subsequently reported that no cleavage or migration of the internucleotide linkages could be detected after r[(Up)20U] had been allowed to stand at pH 3.25 in 0.5 M sodium acetate buffer solution at room temperature for 96 hr (Rao and Macfarlane, 1995), which is very much more than the time required to remove the 2′-O-Fpmp protecting groups. These workers went on to recommend that 2′-O-Fpmp protecting groups be removed at pH 3.25 and 30°C in 0.5 M sodium acetate buffer solution. They successfully unblocked RNA sequences containing up to 50 nucleoside residues under these conditions, and obtained oligoribonucleotides that

were active as ribozymes and ribozyme substrates. Other acetal groups Three other interesting and potentially useful acetal groups have recently been suggested for the protection of the 2′-hydroxy functions in solid-phase oligoribonucleotide synthesis. Like the 2-nitrobenzyl group (as in S.34, see above), the (2-nitrobenzyloxy)methyl group (as in S.81a; Schwartz et al., 1992; Fig. 2.2.24) is removable photochemically; however, possibly for steric reasons, its use leads to faster and more efficient coupling reactions. The (2-nitrobenzyloxy)methyl protecting group has been used successfully in the solid-phase synthesis of a number of RNA sequences including a dodecamer, a hexadecamer, and a tritriacontamer (33-mer) sequence that are all components of ribozyme structures. The related (4-nitrobenzyloxy)methyl protecting group (as in S.81b; Gough et al., 1996), which has also been used successfully in solid-phase oligoribonucleotide synthesis, is removable by treatment with tetra-n-butylammonium fluoride in THF solution. Finally the (2,6-dimethoxycarbonyl)phenoxymethyl protecting group (as in S.82a; Rastogi and Usher, 1995), which has been used in the solid-phase synthesis of two dinucleoside phosphates, is extremely (over 100 times more than the Fpmp group) stable under standard detritylation conditions. After the assembly of the desired RNA sequences, the two methoxycarbonyl groups are saponified by treatment with aqueous sodium hydroxide, which also releases the product from the solid support and removes base-labile protecting groups. The resulting (2,6-dicarboxy)phenoxymethyl acetal system (as in S.82b) is estimated to be >1300 times more labile to acidic hydrolysis at pH 3.0 than the original (2,6-di-


R2O

HO

R2O

B

O O

O

O

R1

O

OH

R1

83

Figure 2.2.25 derivatives.

B

O

84

Scheme showing interconversion of isomeric 2′- and 3′-O -acyl-ribonucleoside

methoxycarbonyl)phenoxymethyl acetal system; however, it is still ∼2.3 times more stable at pH 3.0 than the Fpmp protecting group.

with 2′,3′-di-O-acetyluridine 5′-phosphate (S.86) by the now obsolete phosphodiester approach in solution to give, after deprotection, guanylyl-(3′→5′)-uridine (S.87; B = guanin-9yl) and cytidylyl-(3′→5′)-uridine (S.87; B = cytosin-1-yl), respectively. Both of the latter dinucleoside phosphates were apparently free from their (2′→5′)-isomers. Two later studies relating to the use of 2′-Oacyl protecting groups in solid-phase oligoribonucleotide synthesis are also of interest. In one study (Kempe et al., 1982), oligoribonucleotides and chimeric RNA:DNA sequences were prepared from 2′-O-benzoyl-protected phosphoramidites (S.88; Fig. 2.2.27). However, as these phosphoramidites (S.88) were contaminated with 1% to 3% of the isomeric 2′-phosphoramidites, the integrity of the internucleotide linkages in the target RNA sequences was to some extent compromised. The other study (Rozners et al., 1992) described the solid-phase synthesis of oligoribonucleotides from 2′-O-(2-chlorobenzoyl)3′-H-phosphonate building blocks (S.89). This is a more promising approach for two reasons. First, it was possible to separate the isomeric

Ester Protecting Groups It has been known for many years that isomeric 2′- and 3′-O-acyl-ribonucleoside derivatives (S.83 and S.84, respectively; Fig. 2.2.25) interconvert under mildly basic conditions, and that the equilibrium mixture eventually obtained is generally somewhat richer in the 3′isomer (Reese and Trentham, 1965). Unlike corresponding mixtures of 2′- and 3′-OTBDMS derivatives (e.g., S.39 and S.40; Fig. 2.2.9), it is usually very difficult or even impossible to separate isomeric mixtures of 2′- and 3′-esters (S.83 and S.84) by standard chromatographic methods. Furthermore, acyl migration can occur during chromatography. For these reasons, 2′-O-acyl protecting groups have only very rarely been used in oligoribonucleotide synthesis. However, in an early study (Fromageot et al., 1968), N2,O2′,O5′-tribenzoylguanosine (S.85a; Fig. 2.2.26) and N4,O2′,O5′triacetylcytidine (S.85b), two pure crystalline compounds, were both successfully coupled

R

O

B

O

O HO

O

HO

O

O

B

R 85 i, ii O HO P O

O

O AcO

O OH O P O O O

Ura

OAc

HO

Ura

OH

87

86 Ura = uracil-1-yl a, B = 2-N-benzoylguanin-9-yl, R = Ph b, B = 4-N-acetylcytosin-1-yl, R = Me

Figure 2.2.26 Scheme showing preparation of guanylyl-(3′→5′)-uridine and cytidylyl-(3′→5′)uridine using a 2′-O -acyl protecting group. Reagents: (i) mesitylene-2-sulfonyl chloride, C5H5N; (ii) MeNH2, EtOH, or NH3, MeOH.



DMTrO DMTrO

O

B

O

B

O

O O O P O H

OBz

MeO P NMe2

DMTrO

O

O Cl

O

Cl

89

88

B

O O P H

90

Figure 2.2.27 2′-O -Benzoyl-protected 3′-phosphoramidite (S.88), 2′-O -(2-chlorobenzoyl)-protected 3′-H-phosphonate (S.89), and the isomeric 2′-H-phosphonate (S.90).

2′- and 3′-H-phosphonates (S.90 and S.89, respectively) by chromatography on silica gel, and thereby obtain isomerically pure building blocks (S.89). Secondly, after the desired RNA sequences had been assembled, the 2′-protecting groups could be removed by ammonolysis under conditions that were mild enough to avoid cleavage of the internucleotide linkages. A number of RNA sequences of moderate length were prepared successfully by this approach.

CONCLUSIONS At present, the TBDMS group (as in S.37; Fig. 2.2.28) is the most widely used protecting group for 2′-hydroxy functions in solid-phase oligoribonucleotide synthesis. The Fpmp (as in S.71b; Fig. 2.2.28) group is also widely used. Both 2′-O-TBDMS- and 2′-O-Fpmp-protected phosphoramidites (S.41 and S.79, respectively) are commercially available. So far, there is no report in the literature that constitutes a thorough and definitive comparison of these two protecting groups. However, it is worthwhile discussing how they meet the necessary criteria

DMTrO

O

HO

B

B

O HO

for protecting groups in oligoribonucleotide synthesis, and also whether improvements could be made by modifying them. With regard to the introduction of these two protecting groups, the reagents required, namely tert-butylchlorodimethylsilane and 1(2-fluorophenyl)-4-methoxy-1,2,5,6-tetrahydropyridine (S.70b; Fig. 2.2.28) are both readily available. However, as far as the introduction of these protecting groups is concerned, the Fpmp group has the edge over the TBDMS group inasmuch as it can be introduced regiospecifically (Fig. 2.2.20A) and cannot then migrate. As indicated above, great care has to be exercised in the preparation of TBDMS-protected building blocks (S.41) in order to avoid contamination with the isomeric 2′-phosphoramidites (S.42; Fig. 2.2.10), the presence of which will inevitably lead to (2′→5′)-internucleotide linkages in the final product. In solid-phase synthesis involving phosphoramidite building blocks, it seems clear that coupling rates and efficiencies are generally lower when ribonucleoside rather than 2′-deoxyribonucleoside building blocks are used

DMTrO O

O

OMe F

HO O Me Si Me t-Bu

NC

N

O

B

OTBDMS

O P N(i-Pr)2 41

71b

37

OMe DMTrO

NC

O O

B

OFpmp

N F

O P N(i-Pr)2


79

70b

Figure 2.2.28 Structures relating to a discussion of the relative merits of the TBDMS and Fpmp protecting groups.


DMTrO

B

O O

HO

O

OFpmp

B

O

OH

O P O

O P O O

B

O O

O

OFpmp

n-2

O HO

O

n-2

OH

O P O

O P O O

B

O

H3O

B

O

HO

OH

91

O

B

OH

92

Figure 2.2.29 Unblocking of 2′-O -Fpmp-protected oligoribonucleotides under mild conditions of acidic hydrolysis.

(Hayakawa et al., 1996). It is not yet clear whether TBDMS-protected or Fpmp-protected phosphoramidites (S.41 or S.79; Fig. 2.2.28) are the more hindered. Although TBDMS ethers (Kawahara et al., 1996) and Fpmp acetals are both susceptible to acid-catalyzed hydrolysis, the available evidence suggests that both groups remain intact under the anhydrous acidic conditions used during the detritylation steps. The Fpmp group has advantages over the TBDMS protecting group in the ammonolytic unblocking step at the end of the synthesis. First, in the Fpmp approach, adenine, cytosine, and guanine base residues are protected with relatively stable acyl groups (as in S.16, R = Me3C, S.18, R = Ph, and S.19, R = PhCH2, respectively; Fig. 2.2.4; Rao et al., 1993). However, it is advisable to use much more labile acyl protecting groups in the TBDMS approach (Chaix et al., 1989). More importantly, as the Fpmp protecting group is completely stable under the ammonolytic conditions, “Fpmp-on” RNA sequences (S.91; Fig. 2.2.29) are obtained (Rao et al., 1993). Such “Fpmp-on” oligoribonucleotides are stable to endonucleases and base, and may be conveniently purified and stored. On treatment with aqueous acid under very mild conditions (see below), they are readily converted into unprotected RNA sequences (S.92). “TBDMS-on” RNA sequences do not appear to have been purified and isolated in this way. The one clear advantage that the TBDMS approach has over the Fpmp approach is that removal of the TBDMS protecting group in the final unblocking step does not normally involve acidic hydrolysis, and therefore cannot lead to migration of the internucleotide linkages. However, such migration in the Fpmp approach can be virtually eliminated by carefully controlling

the unblocking conditions. Hecht and co-workers (Morgan et al., 1995) reported that when S.91 (B = uracil-1-yl, n = 21; Fig.2.2.29) was unblocked in 0.5 mol dm−3 sodium acetate buffer, pH 3.25, at 25°C for 20 hr, analysis of the resulting r[(Up)20U] (S.92; B = uracil-1-yl, n = 21) revealed that an average of 0.40% migration per internucleotide linkage had occurred. However, Reese et al. (unpub. observ.) have found that under somewhat milder unblocking conditions (0.5 mol dm−3 sodium acetate buffer, pH 4.0, at 35°C), unblocking of S.91 (B = uracil-1-yl, n = 20) was complete after 9 hr and no migration of internucleotide linkages could be detected in the resulting r[(Up)19U] (S.92; B = uracil-1-yl, n = 20). As has been suggested before (Capaldi and Reese, 1994), it cannot be concluded from the results obtained by Strömberg and co-workers (Rozners et al., 1994) in connection with the use of 2′-O-Ctmp5′-O-DMTr-uridine 3′-H-phosphonate (S.77; B = uracil-1-yl; Fig. 2.2.22) and the corresponding Fpmp-protected H-phosphonate building block in the synthesis of r[(Up)11U] and r[(Up)11A] that Ctmp and Fpmp are unsuitable protecting groups for the 2′-hydroxy functions in the H-phosphonate approach to the solid-phase RNA synthesis. A much more likely explanation for Strömberg’s observations is that r[(Up)11U] and r[(Up)11A], like r[(Up)9U] and r[(Up)19U] (Capaldi and Reese 1994), are particularly labile at pH 2.0 and room temperature. Although it is clear that the solid-phase synthesis of relatively high-molecular-weight RNA sequences using TBDMS, Fpmp, or other groups to protect the 2′-hydroxy functions is now a feasible proposition, it is likely that even better protecting groups will be identified in the



future. The next generation of 2′-protecting groups could include modifications of TBDMS, Fpmp, and some of the other groups described above, and it could also include completely different groups. Any alternative silyl protecting group (S.93; Fig. 2.2.30) would need to be bulky to be sufficiently stable, and so far there is no evidence that any such group is likely to have superior properties to those of the TBDMS group itself. However, modification of the Fpmp group could well lead to improvements. A choice is already available between the Fpmp group, which has the advantage of being more stable at low pH (Fig. 2.2.21) during the detritylation steps, and the Ctmp group, which has the advantage of being more labile at high pH during the final unblocking step. It might well be possible, by a careful choice of R1 and R2, to identify a 1-aryl-4-alkoxypiperidin-4-yl protecting group (S.94; Fig. 2.2.30) that is as stable as (or perhaps even more stable than) the Fpmp group at low pH and as labile as (or perhaps even more labile than) the Ctmp group at high pH. Most of the above discussion has been concerned with the small-scale synthesis of RNA sequences on a solid support. In the light of recent developments in the possible use of oligonucleotide analogs in chemotherapy, a demand has arisen for the development of methods for large-scale synthesis. This may well involve a shift from solid-phase to solutionphase methodology. While this need not necessarily affect the strategy of 2′-protection, the cost of the requisite monomeric building blocks is likely to become a matter of crucial importance. Therefore, particular emphasis will need to be laid on the first general criterion for protecting groups–that they should be easy to introduce and that the reagents involved should be readily accessible. It is not envisaged that this will present a problem for the Fpmp and most other related 1-aryl-4-alkoxypiperidin-4yl protecting groups (S.94). Apart from the practical problems associated with the prepara-

N

R3


93

94

ACKNOWLEDGMENT The author would like to acknowledge the huge contributions that his co-workers have made over a period of more than 30 years to studies on the chemical synthesis of oligo- and poly-ribonucleotides, and especially to those studies relating to the problem of 2′-protection. Some of their names appear in the references below; they are all owed an enormous debt of gratitude.

LITERATURE CITED Beaucage, S.L. and Caruthers, M.H. 1996. The chemical synthesis of DNA/RNA. In Bioorganic Chemistry: Nucleic Acids (S.M. Hecht, ed.) pp. 36-74. Oxford University Press, New York and Oxford. Beaucage, S.L. and Iyer, R.P. 1992. Advances in the synthesis of oligonucleotides by the phosphoramidite approach. Tetrahedron 48:22232311. Beijer, B., Sulston, I., Sproat, B.S., Rider, P., Lamond, A.I., and Neuner, P. 1990. Synthesis and applicatons of oligoribonucleotides with selected 2′-O-methylation using the 2′-O-[1-(2fluorophenyl)-4-methoxypiperidin-4-yl] protecting group. Nucl. Acids Res. 18:5143-5151. Brown, T. and Brown, D.J.S. 1991. Modern machine-aided methods of oligoribonucleotide synthesis. In Oligonucleotides and Analogues. A Practical Approach (F. Eckstein, ed.) pp. 1-24. IRL Press, Oxford. Brown, J.M., Christodoulou, C., Jones, S.S., Modak, A.S., Reese, C.B., Sibanda, S., and Ubasawa, A. 1989a. Synthesis of the 3′-terminal half of yeast alanine transfer ribonucleic acid (tRNAAla) by the phosphotriester approach in solution. Part 1. Preparation of nucleoside building blocks. J. Chem. Soc. Perkin Trans. 1 1735-1750. Brown, J.M., Christodoulou, C., Modak, A.S., Reese, C.B., and Serafinowska, H.T. 1989b. Synthesis of the 3′-terminal half of yeast alanine transfer ribonucleic acid (tRNAAla) by the phosphotriester approach in solution. Part 2. J. Chem. Soc. Perkin Trans. 1 1751-1767.

R1O R2 R1 Si

tion of very large quantities of 2′-O-TBDMS5′-O-DMTr-protected 3′-phosphoramidites (S.41) that are free from their 2′-isomers (S.42), there is no obvious reason why the TBDMS group should not also be used to protect 2′-hydroxy functions in the large-scale synthesis of RNA sequences in solution.

R2

Figure 2.2.30 Substituted silyl and 1-aryl-4alkoxypiperidin-4-yl protecting groups.

Capaldi, D.C. and Reese, C.B. 1994. Use of the 1-(2-fluorophenyl)-4-methoxypiperidin-4-yl (Fpmp) and related protecting groups in oligoribonucleotide synthesis: Stability of internucleotide linkages to aqueous acid. Nucl. Acids Res. 22:2209-2216.


Chaix, C., Molko, D., and Téoule, R. 1989. The use of labile base protecting groups in oligoribonucleotide synthesis. Tetrahedron Lett. 30:71-74. Chattopadhyaya, J.B. and Reese, C.B. 1978. The 9-phenylxanthen-9-yl protecting group. J.Chem. Soc., Chem. Commun. 639-640. Chattopadhyaya, J.B., Reese, C.B., and Todd, A.H. 1979. 2-Dibromobenzoyl: An acyl protecting group removable under exceptionally mild conditions. J. Chem. Soc., Chem. Commun. 987988. Christodoulou, C., Agrawal, S., and Gait, M.J. 1986. Incompatibility of acid-labile 2′ and 5′ protecting groups for solid-phase synthesis of oligoribonucleotides. Tetrahedron Lett. 27:1521-1522. Corey, E.J. and Venkateswarlu, A. 1972. Protection of hydroxyl groups as tert-butyldimethylsilyl derivatives. J. Am. Chem. Soc. 94:6190-6191. Damha, M.J. and Ogilvie, K.K. 1993. Oligoribonucleotide synthesis. The silyl-phosphoramidite method. In Protocols for Oligonucleotides and Analogs (S. Agrawal, ed.) pp. 81-114. Humana Press, Totowa, N.J. den Hartog, J.A.J., Wille, G., and van Boom, J.H. 1981. Synthesis of oligoribonucleotides with sequences identical to the nucleation region of tobacco mosaic virus RNA: Preparation of AAG, AAGAAG and AAGAAGUUG via phosphotriester methods. Rec. Trav. Chim. 100:320-330. Faja, M., Reese, C.B., Song, Q., and Zhang, P.-Z. 1997. Facile preparation of acetals and enol ethers derived from 1-arylpiperidin-4-ones. J. Chem. Soc. Perkin Trans. 1 191-194. Fromageot, H.P.M., Reese, C.B., and Sulston, J.E. 1968. The synthesis of oligoribonucleotides. VI. 2′-O-Acyl ribonucleoside derivatives as intermediates in the synthesis of dinucleoside phosphates. Tetrahedron 24:3533-3540. Gasparutto, D., Livache, T., Bazin, H., Duplaa, A.M., Guy, A., Khorlin, A., Molko, D., Roget, A., and Téoule, R. 1992. Chemical synthesis of a biologically active natural RNA with its minor bases. Nucl. Acids Res. 20:5159-5166. Goodwin, J.T., Stanick, W.A., and Glick, G.D. 1994. Improved solid-phase synthesis of long oligoribonucleotides. Application to tRNAPhe and tRNAGly. J. Org. Chem. 59:7941-7943. Gough, G.R., Miller, T.J., and Mantick, N.A. 1996. p-Nitrobenzyloxymethyl: A new fluoride-removable protecting group for ribonucleoside 2′hydroxyls. Tetrahedron Lett. 37:981-982. Griffin, B.E. and Reese, C.B. 1964. Oligoribonucleotide synthesis via 2,5-protected ribonucleoside derivatives. Tetrahedron Lett. 29252931. Griffin, B.E., Reese, C.B., Stephenson, G.F., and Trentham, D.R. 1966. Oligoribonucleotide synthesis from nucleoside 2′-O-benzyl ethers. Tetrahedron Lett. 4349-4354.

Griffin, B.E., Jarman, M., and Reese, C.B. 1968. The synthesis of oligoribonucleotides. IV. Preparation of dinucleoside phosphates from 2′,5′-protected ribonucleoside derivatives. Tetrahedron 24:639-662. Hakimelahi, G.H., Proba, Z.A., and Ogilvie, K.K. 1982. New catalysts and procedures for the dimethoxytritylation and selective silylation of ribonucleosides. Can. J. Chem. 60:1106-1113. Hayakawa, Y., Kataoka, M., and Noyori, R. 1996. Benzimidazolium triflate as an efficient promoter for nucleotide synthesis via the phosphoramidite method. J. Org. Chem. 61:79967997. Hayes, J.A., Brunden, M.J., Gilham, P.T., and Gough, G.R. 1985. High-yield synthesis of oligoribonucleotides using o-nitrobenzyl protection of 2′-hydroxyls. Tetrahedron Lett. 26:24072410. Honda, S., Urakami, K., Koura, K., Terada, K., Sato, Y., Kohno, K., Sekine, M., and Hata, T. 1984. Synthesis of oligoribonucleotides by the use of S,S-diphenyl N-monomethoxytrityl ribonucleoside 3′-phosphorodithioates. Tetrahedron 40:153-163. Iwai, S. and Ohtsuka, E. 1988. 5′-Levulinyl and 2′-tetrahydrofuranyl protection for the synthesis of oligoribonucleotides by the phosphoramidite approach. Nucl. Acids Res. 16:9443-9456. Iwai, S., Yamada, E., Asaka, M., Hayasa, Y., Inone, H., and Ohtsuka, E. 1987. A new solid phase synthesis of oligoribonucleotides by the phosphoro-p-anisidate method using tetrahydrofuranyl protection of 2′-hydroxyl groups. Nucl. Acids Res. 15:3761-3772. Järvinen, P., Oivanen, M., and Lönnberg, H. 1991. Interconversion and phosphoester hydrolysis of 2′,5′ and 3′,5′-dinucleoside monophosphates: Kinetics and mechanisms. J. Org. Chem. 56:5396-5401. Jones, S.S. and Reese, C.B. 1979. Migration of t-butyldimethylsilyl protecting groups. J. Chem. Soc. Perkin Trans. 1 2762-2764. Jones, S.S., Rayner, B., Reese, C.B., Ubasawa, A., and Ubasawa, M. 1980. Synthesis of the 3′-terminal decaribonucleoside nonaphosphate of yeast alanine transfer ribonucleic acid. Tetrahedron 36:3075-3085. Jones, S.S., Reese, C.B., Sibanda, S., and Ubasawa, A. 1981. The protection of uracil and guanine residues in oligonucleotide synthesis. Tetrahedron Lett. 22:4755-4758. Jones, S.S., Reese, C.B., and Sibanda, S. 1983. Studies directed towards the synthesis of yeast alanine tRNA. In Current Trends in Organic Synthesis (H. Nozaki, ed.) pp. 71-81. Pergamon Press, Oxford. Kamimura, T., Tsuchiya, M., Urakami, K., Koura, K., Sekine, M., Shinozaki, K., Miura, K., and Hata, T. 1984. Synthesis of a dodecaribonucleotide GUAUCAAUAAUG by use of fully protected ribonucleotide building blocks. J. Am. Chem. Soc. 106:4552-4557.



Kawahara, S., Wada, T., and Sekine, M. 1996. Unprecedented mild acid-catalyzed desilylation of the 2′-O-tert-butyldimethylsilyl group from chemically synthesized oligoribonucleotide intermediates via neighbouring group participation of the internucleotide phosphate residue. J. Am. Chem. Soc. 118:9461-9468. Kempe, T., Chow, F., Sundquist, W.I., Nardi, T.J., Paulson, B., and Peterson, S.M. 1982. Selective 2′-benzoylation at the cis 2′,3′-diols of protected ribonucleosides. New solid phase synthesis of RNA and DNA-RNA mixtures. Nucl. Acids Res. 10:6695-6714. Kierzek, R. 1994. The stability of trisubstituted internucleotide bond in the presence of vicinal 2′hydroxyl. Chemical synthesis of uridylyl(2′phosphate)-(3′→5′)-uridine. Nucleosides Nucleotides 13:1757-1768. Kierzek, R., Caruthers, M.H., Longfellow, C.E., Swinton, D., Turner, D.H., and Freier, S.M. 1986. Polymer-supported RNA synthesis and its application to test the nearest-neighbour model for duplex stability. Biochemistry 25:7840-7846. Kreevoy, M.M. and Taft, R.W. Jr. 1955. The evaluation of inductive and resonance effects on reactivity. I. Hydrolysis rates of acetals of non-conjugated aldehydes and ketones. J. Am.Chem. Soc. 77:5590-5595. Kruse, C.G., Jonkers, F.L., Dert, V., and van der Gen, A. 1979. Synthetic applications of 2-chlorotetrahydrofuran: Protection of alcohols as tetrahydro2-furanyl (THF) ethers. Rec.Trav. Chim. 98:371380. Kuusela, S. and Lönnberg, H. 1994. Hydrolysis and isomerisation of the internucleosidic phosphodiester bonds of polyuridylic acids: Kinetics and mechanism. J. Chem. Soc. Perkin Trans. 2 21092113. Lehmann, C., Xu, Y.-Z., Christodoulou, C., Tan, Z.-K., and Gait, M.J. 1989. Solid-phase synthesis of oligoribonucleotides using 9-fluorenylmethoxycarbonyl (Fmoc) for 5′-hydroxyl protection. Nucl. Acids Res. 17:2379-2390. McGregor, A., Rao, M.V., Duckworth, G., Stockley, P.G., and Connolly, B.A. 1996. Preparation of oligoribonucleotides containing 4-thiouridine using Fpmp chemistry. Photo crosslinking to RNA bridging proteins using 350 nm irradiation. Nucl. Acids Res. 24 : 3173-3180. Morgan, M.A., Kazakov, S.A., and Hecht, S.M. 1995. Phosphoryl migration during the chemical synthesis of RNA. Nucl. Acids Res. 23:3949-3953. Mullah, B. and Andrus, A. 1996. Purification of 5′-O-trityl-on oligoribonucleotides. Investigation of phosphate migration during purification and detritylation. Nucleosides Nucleotides 15:419-430. Protection of 2′Hydroxy Functions of Ribonucleosides

Norman, D.G., Reese, C.B., and Serafinowska, H.T. 1984. The protection of 2′-hydroxy functions in oligoribonucleotide synthesis. Tetrahedron Lett. 25:3015-3018.

Ogilvie, K.K., Sadana, K.L., Thompson, A.E., Quillian, M.A., and Westmore, J.B. 1974. The use of silyl groups in protecting the hydroxyl functions of ribonucleosides. Tetrahedron Lett. 28612863. Ohtsuka, E. and Iwai, S. 1987. Chemical synthesis of RNA. In Synthesis and Applications of DNA and RNA (S.A. Narang, ed.) pp. 115-136. Academic Press, San Diego. Ohtsuka, E., Tanaka, S., and Ikehara, M. 1978. Synthesis of the heptanucleotide corresponding to a eukaryotic initiator tRNA loop sequence. J. Am. Chem. Soc. 100:8210-8213. Ohtsuka, E., Yamane, A., Doi, T., and Ikehara, M. 1984. Chemical synthesis of the 5′-half molecule of E.coli tRNA2Gly. Tetrahedron 40:47-57. Owen, G.R. and Reese, C.B. 1970. A convenient preparation of tetrahydro-4H-pyran-4-one. J. Chem. Soc. C 2401-2403. Pathak, T. and Chattopadhyaya, J. 1985. The 2′-hydroxy function assisted cleavage of the internucleotide phosphotriester bond of a ribonucleotide under acidic conditions. Acta Chem. Scand. B 39:799-806. Pieles, U., Beijer, B., Bohmann, K., Weston, S., O’Loughlin, S., Adam, V., and Sproat, B.S. 1994. New and convenient protection system for pseudouridine, highly suitable for solid phase oligoribonucleotide synthesis. J. Chem. Soc. Perkin Trans. 1 3423-3429. Rao, M.V. and Macfarlane, K. 1995. Improvements to the chemical synthesis of biologically-active RNA using 2′-O-Fpmp chemistry. Nucleosides Nucleotides 14:911-915. Rao, T.S., Reese, C.B., Serafinowska, H.T., Takaku, H., and Zappia, G. 1987. Solid phase synthesis of the 3′-terminal nonadecaribonucleoside octadecaphosphate sequence of yeast alanine transfer ribonucleic acid. Tetrahedron Lett. 28:48974900. Rao, M.V., Reese, C.B., Schehlmann, V., and Yu, P.S. 1993. Use of the 1-(2-fluorophenyl)-4methoxypiperidin-4-yl (Fpmp) protecting group in the solid phase synthesis of oligo- and polyribonucleotides. J. Chem. Soc. Perkin Trans. 1 43-55. Rastogi, H. and Usher, D.A. 1995. A new 2′-hydroxyl protecting group for the automated synthesis of oligoribonucleotides. Nucl. Acids Res. 23:4872-4877. Reese, C.B. 1970. A systematic approach to oligoribonucleotide synthesis. Colloq. Int. Cent. Natl. Rech. Sci. 182:319-328. Reese, C.B. 1978. The chemical synthesis of oligoand poly-nucleotides by the phosphotriester approach. Tetrahedron 34:3143-3179. Reese, C.B. 1989. The chemical synthesis of oligoand poly-ribonucleotides. In Nucleic Acids and Molecular Biology, Vol. 3 (F. Eckstein and D.M.J. Lilley, ed.) pp. 164-181. Springer-Verlag, Berlin.


Reese, C.B. and Skone, P.A. 1984. The protection of thymine and guanine residues in oligodeoxyribonucleotide synthesis. J. Chem. Soc. Perkin Trans. 1 1263-1271. Reese, C.B. and Skone, P.A. 1985. Action of acid on oligoribonucleotide phosphotriester intermediates. Effects of released vicinal hydroxy functions. Nucl. Acids Res. 13:5215-5231. Reese, C.B. and Thompson, E.A. 1988. A new synthesis of 1-arylpiperidin-4-ols. J. Chem. Soc. Perkin Trans. 1 2881-2885. Reese, C.B. and Trentham, D.R. 1965. Acyl migration in ribonucleoside derivatives. TetrahedronLett. 2467-2472. Reese, C.B. and Zard, L. 1981. Some observations relating to oximate ion promoted unblocking of oligonucleotide aryl esters. Nucl. Acids Res. 9:4611-4626. Reese, C.B., Saffhill, R., and Sulston, J.E. 1967. A symmetrical alternative to the tetrahydropyranyl protecting group. J. Am. Chem. Soc. 89:33663368. Reese, C.B., Saffhill, R., and Sulston, J.E. 1970. 4-Methoxytetrahydropyran-4-yl. A symmetrical alternative to the tetrahydropyranyl protecting group. Tetrahedron 26:1023-1030. Reese, C.B., Titmas, R.C., and Yau, L. 1978. Oximate ion promoted unblocking of oligonucleotide phosphotriester intermediates. Tetrahedron Lett. 30:2727-2730. Reese, C.B., Serafinowska, H.T., and Zappia, G. 1986. An acetal group suitable for the protection of 2′-hydroxy functions in rapid oligoribonucleotide synthesis. Tetrahedron Lett. 27:22912294. Reitz, G. and Pfleiderer, W. 1975. Synthese und Eigenschaften von O-benzyl substituierten Diuridylphosphaten. Chem. Ber. 108:2878-2894. Rozners, E., Renhofa, R., Petrova, M., Popelis, J. Kumpins, V., and Bizdena, E., 1992. Synthesis of oligoribonucleotides by the H-phosphonate approach using base labile 2′-O-protecting groups. V. Recent progress in development of the method. Nucleosides Nucleotides 11:579-1593. Rozners, E., Westman, W., and Strömberg, R. 1994. Evaluation of 2’-hydroxyl protection in RNA synthesis using the H-phosphonate approach. Nucl. Acids Res. 22:94-99. Sakatsume, O., Ohtsuki, M., Takaku, H., and Reese, C.B. 1989. Solid phase synthesis of oligoribonucleotides using the 1-[(2-chloro-4-methyl)]-4methoxypiperidin-4-yl (Ctmp) group for the protection of the 2′-hydroxy functions and the H-phosphonate approach. Nucl. Acids Res. 17:3689-3697. Sandström, A., Kwiatkowski, M., and Chattopadhyaya, J. 1985. Chemical synthesis of a pentaribonucleoside tetraphosphate constituting the 3′acceptor stem sequence of E. coli tRNAIle using 2′-O-(3-methoxy-1,5-dicarbomethoxypentan3-yl)-ribonucleoside building blocks. Acta Chem. Scand. B 39:273-290.

Scaringe, S.A., Francklyn, C., and Usman, N. 1990. Chemical synthesis of biologically active oligoribonucleotides using β-cyanoethyl protected ribonucleoside phosphoramidites. Nucl.Acids Res. 18:5433-5441. Schaller, H., Weimann, G., Lerch, B., and Khorana, H.G. 1963. Protected derivatives of deoxyribonucleosides and new syntheses of deoxyribonucleoside-3′ phosphates. J. Am.Chem. Soc. 85:3821-3827. Schwartz, M.E., Breaker, R.R., Asteriadis, G.T., deBear, J.S., and Gough, G.R. 1992. Rapid synthesis of oligoribonucleotides using 2′-O-(o-nitrobenzyloxymethyl)-protected monomers. Bioorg. Med. Chem. Lett. 2:1019-1024. Sinha, N.O., Biernat, J., and Köster, H. 1983. β-Cyanoethyl N,N-dialkylamino/N-morpholinomonochlorophosphoramidites, new phosphitylating agents facilitating ease of deprotection and work-up of synthesized oligonucleotides. Tetrahedron Lett. 24:5843-5846. Smith, M., Rammler, D.H., Goldberg, I.H., and Khorana, H.G. 1962. Studies on polynucleotides XIV. Specific synthesis of the C-3′-C-5′ inter ribonucleotide linkage. Synthesis of uridylyl(3′→5′)-uridine and uridylyl-(3′→5′)-adenosine. J. Am. Chem. Soc. 84:430-440. Smrt, J. and Šorm, F. 1962. Oligonucleotidic compounds I. The direct blocking of 2′-hydroxyl in ribonucleoside-3′ phosphates. The synthesis of 6-azauridylyl-(5′→3′)-uridine. Coll. Czech. Chem. Commun. 27:73-86. Sonveaux, E. 1994. Protecting groups in oligonucleotide synthesis. In Protocols for Oligonucleotide Conjugates: Synthesis and Analytical Techniques (S. Agrawal, ed.) pp. 1-71. Humana Press, Totowa, N.J. Sproat, B.S. and Gait, M.J. 1984. Solid-phase synthesis of oligodeoxyribonucleotides by the phosphotriester method. In Oligonucleotide Synthesis. A Practical Approach (M.J. Gait, ed.) pp. 83-115. IRL Press, Oxford. Sproat, B.S., Beijer, B., Groetli, M., Ryder, U., Morand, K.L., and Lamond, A.I. 1994. Novel solid-phase synthesis of branched oligoribonucleotides including a substrate for RNA debranching enzyme. J. Chem. Soc. Perkin Trans. 1 419-431. Sproat, B.S., Calonna, F., Mullah, B., Tsou, D., Andrus, A., Hampel, A., and Vinayak, R. 1995. An efficient method for the isolation and purification of oligoribonucleotides. Nucleosides Nucleotides 14:255-273. Stawinski, J., Strömberg, R., Thelin, M., and Westman, E. 1988. Studies on the t-butyldimethylsilyl group as 2′-O-protection in oligoribonucleotide synthesis via the H-phosphonate approach. Nucl. Acids Res. 16:9285-9298. Stork, G. and Hudrlik, P.F. 1968. Isolation of ketone enolates as trialkylsilyl ethers. J. Am.Chem. Soc. 90:4462-4464.



Takaku, H. and Kamaike, K. 1982. Synthesis of oligoribonucleotides using 4-methoxybenzyl group as a new protecting group of the 2′-hydroxyl group of adenosine. Chem. Lett. 189-192.

van Boom, J.H. and Burgers, P.M.J. 1976. Use of levulinic acid in the protection of oligonucleotides via the modified phosphotriester method: Synthesis of the decaribonucleotide UAUAUAUAUA. Tetrahedron Lett. 4875-4878.

Takaku, H., Kamaike, K., and Tsuchiya, H. 1984. Synthesis of ribooligonucleotides using the 4methoxybenzyl group as a new protecting group for the 2′-hydroxyl group. J. Org. Chem. 49:5156.

van Boom, J.H., van Deursen, P., Meeuse, J., and Reese, C.B. 1972. Two sulphur-containing protecting groups for alcoholic hydroxyl functions. J. Chem. Soc., Chem. Commun. 766-767.

Takaku, H., Ito, T. and Iwai, K. 1986. Use of the 3,4-dimethoxybenzyl group as a protecting group for the 2′-hydroxyl group in the synthesis of oligoribonucleotides. Chem. Lett. 1005-1008.

Vinayak, R., Anderson, P., McCollum, C., and Hampel, A. 1992. Chemical synthesis of RNA using fast oligonucleotide deprotection chemistry. Nucl. Acids Res. 20:1265-1269.

Tanaka, T., Fujino, K., Tamatsukuri, S., and Ikehara, M. 1986. Synthesis of oligoribonucleotides via the phosphite triester approach on a solid support. Chem. Pharm. Bull. Jpn. 34:4126-4132.

Westman, E. and Strömberg, R. 1994. Removal of t-butyldimethylsilyl protection in RNA synthesis. Triethylamine trihydrofluoride (TEA,3HF) is a more reliable alternative to tetrabutylammonium fluoride (TBAF). Nucl. Acids Res. 22:2430-2431.

Tanimura, H. and Imada, T. 1990. The utility of 2′-Thp group in the synthesis of the relatively long RNA fragments on the solid support. Chem. Lett. 2081-2084. Tanimura, H., Mieda, M., Fukazawa, T., Sekine, M., and Hata, T. 1989. Chemical synthesis of the 24 RNA fragments corresponding to hop stunt viroid. Nucl. Acids Res. 17:8135-8147. Usman, N., Ogilvie, K.K., Jiang, M.Y., and Cedergren, R.J. 1987. Automated chemical synthesis of long oligoribonucleotides using 2′-O-silylated ribonucleoside 3′-phosphoramidites on a controlled pore glass support: Synthesis of a 43-nucleotide sequence similar to the 3′-half molecule of Escherichia coli formylmethionine tRNA. J. Am. Chem. Soc.109:7845-7854.

Wincott, F., DiRenzo, A., Shaffer, C., Grimm, S., Tracz, D., Workman, C., Sweedler, D., Gonzalez, C., Scaringe, S., and Usman, N. 1995. Synthesis, deprotection, analysis and purification of RNA and ribozymes. Nucl. Acids Res. 23:2677-2684.

Contributed by Colin B. Reese King’s College London London, United Kingdom



Protection of 5′-Hydroxy Functions of Nucleosides The 5′-OH group is the primary hydroxy group of nucleosides. It is the least influenced by the electron-withdrawing effects of the other substituents on the sugar moiety. Moreover, it is the least sterically hindered hydroxy function, and shows the highest reactivity of all nucleoside hydroxy groups in nucleophilic substitutions. Although nucleobases can eventually be left unprotected, and nucleosides with free 3′-hydroxy groups have been used in some triester syntheses, it is mandatory to protect 5′ hydroxyls in all methods of oligonucleotide synthesis that require nucleoside synthons. For chemical oligonucleotide synthesis, the blocking groups for the 5′-hydroxy function must be integrated into an orthogonal protection system. Several such systems have been proposed and are in use for deoxyribo- and ribooligonucleotide synthesis as well as for the preparation of structurally modified oligonucleotide analogs. These protection schemes, details of which are described elsewhere (e.g., UNITS 2.1-2.5 and 3.1-3.4), can be distinguished by requiring, in principle, three alternative methods for 5′-deprotection: (1) acid conditions, (2) alkaline or ammoniacal conditions, or (3) selective deblocking reagents applied essentially in the absence of acid or base. Protecting groups for 5′-hydroxy functions can be broadly classified into these three categories, and the subsequent sections of this unit will follow this division. Additional criteria governing the choice of 5′-hydroxyl-protecting groups include (1) the direction of chain lengthening, (2) the use of polymer supports and/or other purification handles, and (3) additional features related to molecular instability or chemical reactivity in the case of oligonucleotides deviating from biological structure. In particular, the direction of chain extension determines whether a 5′-hydroxyl-protecting group will be permanent (i.e., remain attached to the growing oligonucleotide chain) or intermediary (i.e., will have to be removed prior to each chain extension). If lengthening occurs from the 5′ to the 3′ end, a permanent protecting group is used. If chain extension is done from the 3′ to the 5′ end, which is currently most common, the 5′-hydroxy function must be substituted by an intermediary protecting group. In polymer support synthesis, the poly-

UNIT 2.3

meric carrier assumes the role of a permanent protecting group and is now usually attached to the 3′-hydroxy end. However, there are a number of mostly earlier publications that describe carrier fixation through the 5′ end (see Miscellaneous Acid-Labile 5′-Substituents and see 5′-Hydroxyl-Protecting Groups Cleaved Under Nonacidic and Nonalkaline Conditions). In polymer support synthesis and in some solution methods, it is desirable to simplify the workup of the crude product obtained after oligonucleotide chain extensions. A variety of 5′-hydroxyl-protecting groups have been designed that serve as “purification handles” for this purpose (see Triaryl-methyl Groups as Affinity Ligands). The underlying idea is to single out the product chain from a complex admixture of truncated and failure sequences, although this may still be an unattainable goal. Nevertheless, such “handle” methods may significantly reduce the time and effort for oligonucleotide purification, especially on a preparative scale. A final point in these introductory remarks is that there is no protecting group exclusively in use for the 5′-hydroxy function. Only the conditions of the reaction determine whether the same group will serve for the protection of the 5′, the 3′, or the 2′ hydroxyl, or even of functional groups at nucleobases, because there are only subtle differences in the reactivity of all of these functions. To master these subtle differences is an art of regioselective substitution and comprises much of the challenge of oligonucleotide synthesis. Additionally, there are differences in the approach used for deoxyribo- versus ribooligonucleotide synthesis, and these will be discussed in the subsequent sections.

SCOPE OF THIS OVERVIEW This unit will deal with substituents for the 5′-hydroxy function that fulfill the following criteria of protecting groups. (1) The substituents can be affixed to/removed from the 5′hydroxy function of a growing oligonucleotide chain before/after chain elongation or in the context of other reactions of the oligonucleotide chain. (2) The substituents remain bound during chain elongation and do not interfere with other oligonucleotide reactions. Substituents that serve to permanently modify

Contributed by H. Seliger Current Protocols in Nucleic Acid Chemistry (2000) 2.3.1-2.3.34 Copyright © 2000 by John Wiley & Sons, Inc.


2.3.1

the 5′ position are not in the scope of this unit. This includes chemical modifications of the 5′-hydroxy function as well as the substitution of the 5′-hydroxy function by linkers, labels, spacers, and other groups that are meant to remain an integral part of the completed oligonucleotide chain (e.g., UNITS 4.2 & 4.3). Also, protecting groups used specifically for the preparation of certain oligonucleotide analogs will not be treated here, as such syntheses are discussed in further units (e.g., see Chapter 4). Protected nucleoside-5′-phosphates have played a role especially in the early days of oligonucleotide synthesis. Essentially, a 5′phosphate residue can be used for protection of the terminus, because enzymatic hydrolysis can easily reconvert to a 5′ hydroxyl. However, only those cases when this was the declared reason for 5′-phosphorylation will be mentioned in the following text (see examples given in Fig. 2.3.10). The topic of 5′-hydroxyl protection is essential to oligonucleotide synthesis and, therefore, is generally included in all textbooks on preparative nucleotide chemistry, as well as in books and articles dealing specifically with the field of protective groups. A few recent reviews have dealt more extensively with groups for protection of 5′ hydroxyls, including articles by Sonveaux (1986) and Beaucage and Iyer (1992, 1993). More extensive reviews in the earlier literature are given by Kössel and Seliger (1975) and Reese (1978).

ACID-LABILE PROTECTING GROUPS Triarylmethyl and Related Substituents

Protection of 5′-Hydroxy Functions of Nucleosides

Introduction of trityl and substituted trityl groups The triphenylmethyl (trityl or Tr) group (S.1; Fig. 2.3.1), a well-known sugar protecting group, was first used for nucleoside 5′-protection in Lord Todd’s laboratory (Andersen et al., 1954) during hydrogenolytic removal. However, this work had limited success. The breakthrough came from H.G. Khorana’s group, who discovered that detritylation in acidic medium is greatly facilitated if the trityl group is modified with one to three p-methoxy substituents (Gilham and Khorana, 1958; Smith et al., 1962). In 80% acetic acid, the rate of hydrolysis increased roughly 10-fold with each introduction of an additional methoxy group.

R

R'

C

R'' 1 2 3 4

R = R' = R'' = H R = R' = H; R'' = OCH3 R = H; R' = R'' = OCH3 R = R' = R'' = OCH3

Figure 2.3.1 Trityl and p-methoxy-substituted trityl protecting groups.

Since their introduction as acid-labile protecting groups for ribo- (Smith et al., 1962) and deoxyribonucleotide (Schaller et al., 1963) chemistry, the mono- and dimethoxytrityl groups (MMTr, S.2, and DMTr, S.3, respectively; Fig. 2.3.1) have become standard for the protection of the 5′-hydroxy function. The DMTr group has especially proven its value in automated solid-phase deoxyribooligonucleotide synthesis, for five main reasons (Sonveaux, 1986). (1) It can be introduced regiospecifically and in high yield at the 5′-hydroxy function of (base-protected) nucleosides. (2) It can be readily and quantitatively removed from the growing oligonucleotide chain by nonaqueous acid. (3) It is sufficiently stable to tetrazole, which is used as an activator in the chain extension step. (4) Its deprotection in nonaqueous acid gives an intense color reaction (ascribed to a cationic species), which can be monitored by spectroscopy to estimate yields of chain elongation. (5) A terminal 5′-DMTr group conveys a certain hydrophobicity to the longest oligonucleotide chain in the crude product released from the polymer support. This hydrophobicity is often used to isolate the target oligonucleotide from the mixture of truncated and failure chains. These considerations (1 to 5) will be elaborated on further throughout this unit. Other trityl protecting groups are less useful for automated synthesis. Unsubstituted trityl and MMTr require conditions that are too harsh for multistep removal with the complication of depurination (see below). The trimethoxytrityl group (TMTr; S.4; Fig. 2.3.1) is extremely sensitive. It can be introduced readily and in high yield at the 5′-position of deoxyribonucleosides and was found to be completely stable when stored at −20°C for ∼6 months; however, it is partially removed in a mixture of tetrazole


for tritylation. This method is no longer in use. The more elegant route B relies on transient silyl protection to apply acyl groups regioselectively to the nucleobases in a procedure that can be carried out in a single reaction tube (Ti et al., 1982; also see “silylation first” procedure in Fritz et al., 1982). The alternative route C uses initial 5′-tritylation followed by silylation and then acylation (“tritylation first” procedure, Fritz et al., 1982; for a recent report see Wada et al., 1998a,b). In most cases, pyridine serves both to dissolve the reactants and to neutralize the ensuing hydrochloric acid. If necessary, trityl groups can be substituted at the unprotected 5′-hydroxy function of oligonucleotide chains, either postsynthesis or in exchange for other protecting groups. This can be done by treating a support-bound oligonucleotide with 4,4′-DMTr chloride in pyridine/4-dimethylaminopyridine for 1 hr, which restores ∼95% of the previously removed DMTr (Kotschi, 1987; Reddy et al., 1987). This reaction could even be performed in the presence of unprotected internucleotidic bonds due to the lability of phosphoric acid trityl esters (Reddy et al., 1987). With support-bound nucleosides, however, the reaction was more sluggish and had to be

and acetonitrile (Kotschi, 1987). This detritylation occurred more readily with purine than with pyrimidine nucleosides. Also, an oligoadenylate prepared with TMTr-deoxyadenosine contained a significant admixture of longer chains, obviously arising from multiple monomer addition. The trityl protecting groups are generally introduced by treatment of nucleosides with the respective trityl chloride. The reactivity of these trityl chlorides increases with increasing number of p-methoxy substituents. If the reaction is run at room temperature and with not more than a slight excess of trityl chloride, the substitution will be highly regiospecific at the 5′ hydroxyl. If the reagent is in higher excess and the temperature is raised, the substitution also occurs on the more sterically hindered 3′ hydroxyl. Exocyclic amino groups are usually protected prior to tritylation, because they would otherwise react with trityl chlorides. Three routes to base-protected, 5′-tritylated nucleosides are described in Figure 2.3.2. The classical approach (route A) from Khorana’s laboratory uses per-acylation of all hydroxy and amino functions, followed by treatment with strong alkali, which selectively cleaves acyl esters and thus liberates 5′ and 3′ hydroxyls

route A AcylO

O

route B

B

Acyl

HO

AcylO

B

O

acylation

route C DMTrO

HO

B

OH 3',5'-silylation

Me3SiO

O

tritylation

B

O

3'-silylation DMTrO

O

B

alkaline deesterification Me3SiO

Me3SiO acylation

HO

Acyl

O

B

Acyl

Me3SiO

O

desilylation

OH

B

DMTrO

DMTrO

O

B

Acyl

Me3SiO

Me3SiO

tritylation

acylation

Acyl

O

B

desilylation

OH

Figure 2.3.2 Three routes to N-protected, 5′-O-tritylated 2′-deoxyribonucleosides (modified from Fritz et al., 1982, with permission from Verlag Chemie).



activated by addition of tetra-n-butylammonium nitrate and 2,4,6-collidine in dimethylformamide (DMF). The retritylation procedure has recently been applied to the preparation of oligonucleotide-polyamide conjugates (Tong et al., 1993) and to the postsynthetic introduction of the 4-(17-tetrabenzo(a,c,g,i)fluorenylmethyl)-4′,4′′-dimeth oxytrityl protecting group (TBF-DMTr; S.25; Fig. 2.3.8; Ramage and Wahl, 1993; see Triarylmethyl Groups as Affinity Ligands). Other methods of tritylation can be applied if this is required by the sensitivity of modified nucleoside or oligonucleotide reactants. Reports have described the application of powdered molecular sieves as acid scavengers (Kohli et al., 1980), and the use of 4-N,N-dimethylaminopyridine in a mixture of triethylamine and DMF as an alternative solvent (Chaudhary and Hernandez, 1979). As alternatives to trityl chlorides, N-tritylpyridinium fluoroborate (Fersht and Jencks, 1970) and DMTr tetrafluoroborate (Lakshman and Zajc, 1996) have been described.

Removal of trityl and substituted trityl groups The removal of trityl groups was initially performed with 80% aqueous acetic acid (e.g., Smith et al., 1962; Schaller et al., 1963), and such aqueous media are still in use if the deprotection is carried out in the last step of oligonucleotide workup. Nuclear magnetic resonance (NMR) studies have shown that the chemical shifts of all nucleoside protons change upon removal of 5′-trityl groups. On this basis, rate constants were determined for the hydrolysis of different methoxy-substituted trityl groups from the 5′ and 3′ positions of deoxythymidine in aqueous tetrahydrofuran solution brought to different pHs with HCl; a linear relationship between pH and hydrolysis rate was established (Regel et al., 1974).

If detritylations are done as intermediate steps during machine-aided polymer-support synthesis, nonaqueous acidic media are generally applied. The most common are solutions of strong protic acids, such as trichloro- or dichloroacetic acid (Adams et al., 1983; for a recent survey of detritylation conditions see Habus and Agrawal, 1994). Methylene chloride is routinely used as a solvent, although there is a tendency to avoid chlorinated hydrocarbons. For this reason, toluene has been recommended as an alternative for detritylation with dichloroacetic acid (Krotz et al., 1999; Table 2.3.1). This was the result of extensive kinetic studies that showed that the rate of detritylation decreases in the series DMTrdGi-Bu > DMTrdABz > DMTrdCBz > DMTrdT (where Bz is benzoyl), and is highest not only in haloaliphatic but also in aromatic solvents, but is slow in DMF, hexane, ethylacetate, tetrahydrofuran, or tertbutylmethyl ether. Unwanted retritylation can occur through reversal of the equilibrium generated by nonaqueous detritylation, between the colored cationic species and the DMTr-oligonucleotide (Fig. 2.3.3; Dellinger et al., 1998). The same problem was reported for the 9-phenylxanthen9-yl group (see discussion of pixyl and related protecting groups, below; Reese et al., 1986). This is usually not a problem in solid-support oligonucleotide synthesis, since this equilibrium is shifted by washing and filtration; however, it can lead to incomplete deblocking in solution-phase synthesis. Detritylation steps in solid-phase oligonucleotide synthesis are generally believed to be complete when appropriate treatment with nonaqueous acid and extensive washing leave resins and wash solutions colorless. Minute deviations from this scheme of quantitative deblocking, which can occur, for example, through trace impurities in the deblocking solution, are listed among the reasons for the occurrence of truncated sequences in the crude

OCH3

H3CO


C O

P

Figure 2.3.3

OCH3

O

B

CI2CHCOOH

H3CO

C

CI2CHCOO

HO

O

O

O

P

P

B

= polymer support

Detritylation equilibrium.


products released from the polymer support (e.g., Fearon et al., 1995). Whether this is a potential reason for failure cannot be decided without distinguishing between the efficiencies of the individual steps in the elongation cycle. The literature appears undecided about whether such failures occur statistically throughout all cycles (Fearon et al., 1995) or with higher probability during the initial chain elongation (Temsamani et al., 1995). Depurination as a side reaction during detritylation The most stringent problem, however, is the avoidance of depurination on removal of trityl groups. N-Acylated nucleosides and N-acylated units in oligonucleotides are especially susceptible to deglycosidation. This leads to the formation of apurinic sites, with subsequent chain cleavage during ammoniacal deprotection. This side reaction becomes more and more problematic with longer oligonucleotide chains or larger-scale preparations. The incentive to overcome the depurination problem has led to the development of a wide variety of finely tuned acid-deprotection conditions; examples are given in Table 2.3.1. Although most publications list only the optimum detritylation conditions without giving background data, recent studies, stimulated by large-scale oligonu-

cleotide support synthesis, have been accompanied by extensive analyses of detritylation versus depurination kinetics (Paul and Royappa, 1996; Septak, 1996). The essence of their findings is that haloacetic acids bind strongly to immobilized growing oligonucleotide chains. If, as usual, very dilute acid solutions are applied, detritylation is slowed by depletion of acid from the medium, whereas depurination is allowed to proceed through acid saturation. The authors, therefore, recommend using a short pulse of more concentrated acid (e.g., 15% dichloroacetic acid in methylene chloride) and avoiding any acetonitrile contamination. The length of the acid treatment must be adjusted to the length of the growing oligonucleotide chain. In some cases, ion exchange resins in the H+ form may be a good choice for detritylation (Patil et al., 1994), especially to substitute for acetic acid in large-scale preparations (Iyer et al., 1995). A long treatment with silica gel was found advantageous in the preparation of sensitive nucleosides (Rosowsky et al., 1989). Lewis acid deprotection of trityl and substituted trityl groups and miscellaneous detritylation methods Great expectations to solve the depurination problem had accompanied the introduction of

Table 2.3.1 Examples of Acidic Deprotection Conditions of 5′-Trityl Groupsa

Protecting group DMTr DMTr DMTr DMTr DMTr DMTr DMTr DMTr DMTr DMTr DMTr

Deprotection conditions

Application

Reference

Benzene sulfonic acid in 9:1 (v/v) DMF/DCM 3% (w/v) TCA in DCM

ODN solid phase

Patel et al. (1982)

3% (w/v) TCA in 95:5 (v/v) DCM/CH3OH 0.1 M p-toluene sulfonic acid in THF 0.1 M p-toluene sulfonic acid in acetonitrile 2% (v/v) DCA/0.1% (v/v) CH3OH in DCM 3% (v/v) DCA in 1,2-dichloroethane 3% (v/v) DCA in toluene 3% (w/v) TCA in 1% CH3OH/nitromethane 2% (w/v) benzene sulfonic acid in 7:3 (v/v) DCM/CH3OH 15% (v/v) DCA in DCM

A-rich ODN solid phase Tanaka and Oishi (1985) Protected dA Takaku et al. (1983) ODN solid phase

ODN solid phase

Matteucci and Caruthers (1981) Seliger et al. (1987); Septak (1996) Habus and Agrawal (1994) Sproat and Gait (1984)

ODN solid phase ODN solid phase

Krotz et al. (1999) Sinha et al. (1984)

ODN solution

Gaffney et al. (1984)

ODN solid phase

Habus and Agrawal (1994)

ODN solid phase ODN solid phase

aAbbreviations: DCA, dichloroacetic acid; DCM, dichloromethane; DMF, dimethylformamide; DMTr, dimethoxytrityl; ODN, conditions applied in oligodeoxynucleotide synthesis; TCA, trichloroacetic acid; THF, tetrahydrofuran.



OCH3

CH3O

MeOH

Ph

HN N

C O

ZnBr2

O

B

OH


N

N

HO

O HO

O ZnBr2

N

H

Figure 2.3.4 Proposed mechanism for the selective removal of 5′-O-DMTr groups by ZnBr2 in organic solvent (according to Matteucci and Caruthers, 1980).

Figure 2.3.5 Deacylation of nucleobases as a side reaction in the removal of 5′-O-trityl protecting groups with zinc bromide (Kierzek et al., 1981).

zinc bromide as a detritylating agent (Kohli et al., 1980; Matteucci and Caruthers, 1980). The mechanism, described for βmethoxyethoxymethyl ethers by Corey et al. (1976), involves the formation of a bidentate complex with O-C5′ and intracyclic oxygen (Fig. 2.3.4). This explains the selectivity for 5′-hydroxy over 3′-hydroxy detritylation, which was found to disappear upon addition of an alcohol (Waldmeier et al., 1982). This does not play a role in oligonucleotide support synthesis, since the 3′-hydroxy function, conventionally, is anchored to a support. Here the objective is to ensure the most efficient and quantitative deblocking of immobilized growing chains. Initially, a saturated (∼0.1 M) solution of zinc bromide in nitromethane was applied (Matteucci and Caruthers, 1980, 1981). Under these conditions, depurination of N6benzoyl-deoxyadenosine was found to be insignificant over a 24-hr period. However, the formation of an unreactive chain end (presumably a zinc compound) was observed under anhydrous conditions, requiring a hydrolytic wash after detritylation. Additionally, the time required for complete deblocking was relatively long (∼15 min for purine, ∼30 min for pyrimidine; Matteucci and Caruthers, 1981). Therefore, it was found advantageous to add 1% water (Seliger et al., 1982; Winnacker and Dörper, 1982) or 5% methanol (Caruthers, 1982) to the nitromethane solution, which serves to increase the zinc bromide concentration. Applying zinc bromide in a number of protic solvents, Itakura and colleagues (Kierzek et al., 1981) found that a 0.7 M solution of zinc bromide in 9:1 (v/v) chloroform/methanol would lead to complete detritylation within 1 min. Depurination was not detected; however, on prolonged treatment, N-acyl groups were found to be removed, especially from the adenine moiety (Fig. 2.3.5). Since this was also attributed to a nucleophilic attack on a chelate

of zinc bromide with the protected base (Kierzek et al., 1981; Beaucage and Iyer, 1992), sterically hindered alcohol components were found to suppress this side reaction. A 1 M solution of zinc bromide in 85:15 (v/v) dichloromethane/isopropanol was found to be an optimal deblocking reagent (Kierzek et al., 1981; Ito et al., 1982; Itakura et al., 1984). Addition of zinc bromide to amide groups linking the oligonucleotide chains to the support was also postulated (Ito et al., 1982; Adams et al., 1983). In spite of these extensive investigations, neither zinc bromide nor other Lewis acids— such as TiCl4, AlCl3 (Matteucci and Caruthers, 1981), diethyl and diisopropyl aluminum chloride (Köster and Sinha, 1982), or boron trifluoride, applied as etherate (Engels, 1979) or methanol complex (Mitchell et al., 1990)—are of importance in current automated solid-phase deoxyribooligonucleotide synthesis. One of the reasons may be the deposition or adsorption of reagents and by-products (e.g., zinc salts) within the oligonucleotide/solid support system, resulting in the necessity for extensive and time-consuming washes. Outside the mainstream, a number of studies have reported the application of unusual detritylation reagents. Examples are formic acid (Bessodes et al., 1986), 1,1,1,3,3,3-hexafluoro2-propanol (Leonard and Neelima, 1995), chlorine/chloroform solution (Fuentes et al., 1994), or diethyl oxomalonate/methanol solution (Sekine, 1994); some earlier reports are summarized in other reviews (e.g., Beaucage and Iyer, 1992). Other noteworthy detritylation alternatives, such as reductive cleavage with radical anion (Letsinger and Finnan, 1975) or electrochemical deblocking (Mairanovsky, 1976), may not be easily adaptable to automated oligonucleotide synthesis. In routine solid-phase oligonucleotide synthesis, the detritylation solution usually goes to waste after yield monitoring. However, this is


not tolerable for syntheses scaled up to kilogram dimensions and beyond, because the DMTr group comprises ∼35% of the total weight of constituent-protected nucleoside phosphoramidites. Recently, a process has been reported for the workup and neutralization of the detritylation solution, coupled with the reconversion of the ensuing dimethoxytrityl alcohol (DMTrOH) to DMTr chloride. This process resulted in the recycling of 89% of the weight of DMTr residues. At the same time, >90% of the hazardous solvent dichloromethane was recovered with >95% purity (Guo et al., 1998). Pixyl and related protecting groups As a structural analog to the trityl group, the 9-phenylxanthen-9-yl (pixyl or Px) protecting group (S.5, Fig. 2.3.6) has been described (Chattopadhyaya and Reese, 1978; Chattopadhyaya, 1980). This protecting group is removed by acid at approximately the same rate as the DMTr group; however, the pixyl derivatives of nucleosides can be more readily purified by crystallization. The preparation of monomeric (Christodoulou and Reese, 1983) and all sixteen dimeric (Balgobin et al., 1981b) building blocks for deoxyriboolignucleotide synthesis in solution has been done using the pixyl group, and the combination of such blocks to deoxyribooligonucleotides of biological interest has been described (Josephson and Chattopadhyaya, 1981; Balgobin and Chattopadhyaya, 1982b). A more acid-labile variant is the 9-(panisyl)xanthen-9-yl (MOX) group (S.6; Kwiatkowski et al., 1983; Kwiatkowski and Chattopadhyaya, 1984; Tanimura et al., 1988, 1989; Tanimura and Imada, 1990). Alternatively, the 9-phenylthioxanthen-9-yl (S-pixyl; S.7) and 9-phenyl-7-chlorothioxanthen-9-yl (S.8) groups were introduced to modulate deprotection (Balgobin and Chattopadhyaya, 1982a). The recent finding that the pixyl substituent is susceptible to photochemical cleav-

R

R' X 5 6 7 8

X = O; R = H; R' = H X = O; R = OCH3; R' = H X = S; R = H; R' = H X = S; R = H; R' = Cl

Figure 2.3.6 9-Phenylxanthen-9-yl and related 5′-hydroxyl-protecting groups.

age (Misetic and Boyd, 1998) may revive interest in this 5′-protecting group. Trityl and related groups for 5′-hydroxyl protection in oligoribonucleotide synthesis Somewhat different considerations apply to the use of trityl and related groups for 5′-hydroxyl protection in oligoribonucleotide synthesis. Depurination is not very problematic in this case, which allowed the application of MMTr as the preferred protecting group in earlier studies focusing on the preparation of relatively short sequences by solution methods (for reviews, see Reese, 1978, 1989; Ohtsuka and Iwai, 1987). Detailed procedures for the chemical preparation of small oligoribonucleotides were described by van Boom and Wreesmann (1984). In such oligoribonucleotide syntheses where the conditions of acid deprotection are not of great concern, essentially all of the previously described trityl-derived protecting groups should be applicable in RNA synthesis. This has, in fact, been shown in a number of publications cited earlier; however, for simplicity, most protecting groups have been tested first (and often only) in DNA chemistry. Nonetheless, modern strategies of oligoribonucleotide synthesis, in particular the preparation of long sequences on solid phase, require orthogonality of the complete set of protecting groups used for all functionalities of the RNA synthons. In particular, the intermediate protecting group for the 5′ hydroxyl must blend with the groups used for 2′-hydroxyl protection. The latter must be stable throughout all chain elongations, and their removal after completion of synthesis, in the presence of unprotected phosphodiester internucleotide bonds, should not be done in an alkaline medium for risk of isomerization (e.g., Beaucage and Iyer, 1992). Therefore, 2′-hydroxyl-protecting groups should be removed in acidic solution or by a specific reagent in near-neutral medium. Hence, acid-labile protecting groups of the trityl type can be used for 5′-protection of ribonucleotide synthons only under the following conditions. (1) If an acid-labile protecting group is used for the 2′ hydroxyl, there must be a significant difference in the deprotection rates for the 5′ and 2′ functionalities. (2) If the 2′-hydroxyl-protecting groups are stable to acid and released by a specific reagent in a close-to-neutral medium, there is free choice of acid-labile 5′-hydroxyl protection. (3) An acid-labile protecting group can be used for 2′-hydroxyl protection in combination with a substituent for




5′-hydroxyl protection that can be removed under nonacidic conditions. This alternative will be discussed below (see Blocking Groups Labile to Nonacidic Conditions). It is beyond the scope of this unit to give credit to all publications that have described solutions to the above protection alternatives (1) and (2); these are treated in detail in recent reviews (Beaucage and Iyer, 1992, 1993). As a general guideline, it can be said that the combination of the most common acid-labile protecting groups (i.e., DMTr or pixyl for 5′ hydroxyls, and tetrahydropyranyl or 4-methoxytetrahydropyranyl for 2′ hydroxyls) is not fully compatible with approach (1). This can be remedied by changing the conditions or properties of the 5′-hydroxyl-protecting groups or by changing the nature of the 2′-hydroxyl-protecting group. For instance, it was found that the 5′-DMTr group could be deprotected selectively at 0°C without cleavage of 2′-tetrahydropyranyl (Seliger et al., 1983; Caruthers et al., 1986). Since this condition does not lend itself to automated support synthesis, the use of the 4,4′,4′′-trimethoxytrityl protecting group was investigated for the 5′ hydroxyl (Seliger et al., 1986); however, the lability of this substituent caused substantial loss on chromatographic purification of the synthons. A number of 2′-hydroxyl-protecting groups with modulated or increased acid stability have been described that are compatible with 5′DMTr protection. Examples are the 1-(2chloro-4-methylphenyl)-4-methoxypiperidin -4-yl (CTMP) and 1-(2-fluorophenyl)-4methoxypiperidin-4-yl (FPMP) groups (Reese et al., 1986; Reese and Thompson, 1988), and the 1-(2-chloro-ethoxy)ethyl (Yamakage et al., 1989) and 3-methoxy-1,5dicarbomethoxypentan-3-yl (Sandström et al., 1985) groups. However, none of these approaches is of widespread application in current solid-support oligoribonucleotide synthesis. The approach that has currently won the competition for a DMTr-compatible system of 2′- and 5′-hydroxyl protection comes from alternative (2) above, namely the introduction of appropriate silyl groups at the 2′ hydroxyl. This approach, which was mainly developed in the laboratory of Ogilvie (Usman et al., 1987, and references therein; Ogilvie et al., 1988), has been extensively investigated during recent years (for discussion, see Beaucage and Iyer, 1992) and is at present the most commonly used method for automated solid-phase oligoribonucleotide preparations.

Triarylmethyl Groups as Affinity Ligands Up to now, the modification of trityl protecting groups has been discussed only with respect to steering the acid lability by introduction of one, two, or three p-methoxy substituents. In an attempt to simplify the workup of the crude product of polymer-support synthesis, it was first shown that MMTr or DMTr groups can serve to single out the target chains by hydrophobic chromatography if they remain bound to the last monomer unit after completion of chain elongations (“trityl-on purification”; Seliger et al., 1977a, 1978). Subsequent to purification, these groups are cleaved with 80% acetic acid to generate the unprotected oligonucleotide. Such groups are referred to as “purification handles.” More recently, a number of trityl groups substituted with longer alkyl chains, ranging from C8H17 to C16H33 (S.9 to S.13; Table 2.3.2), have been described as lipophilic protecting groups (Görtz and Seliger, 1981) and used for separation on reversed-phase columns (Seliger and Görtz, 1981). The structures of substituted trityl groups tailored for purification assistance and other special applications are summarized in Table 2.3.2. The 4-decyloxytrityl (C10Tr or DTr) group (S.10, Fig. 2.3.7; Seliger and Schmidt, 1987) was found to be especially useful for the purification of genes and gene fragments with lengths up to 147 bases (Seliger et al., 1987; Schmidt et al., 1988). The C16Tr group (S.13) has the highest lipophilic affinity; however, the solubility of C16Tr-nucleosides in acetonitrile and other solvents common for oligonucleotide synthesis is somewhat decreased. In order to minimize the risk of depurination during terminal acid deprotection, a series of more labile 4-methoxy-4′-alkoxytrityl groups (S.14 to S.20; Table 2.3.2) was proposed for purification assistance (Gupta et al., 1991). Of these, the 4-methoxy-4′-octyloxytrityl group (S.19; MOTr; Fig. 2.3.7) was found most suitable. The 4,4′-dianisyl-2′′-hexadecyloxyphenyl group (S.21; Table 2.3.2) has mainly been used for oligoribonucleotide purification (van Boom and Wreesman, 1984). The pixyl (S.5), MOX (S.6; Kwiatkowski et al., 1983; Kwiatkowski and Chattopadhyaya, 1984; Tanimura et al., 1988, 1989; Tanimura and Imada, 1990), and especially the 9-(4-octadecyloxyphenyl)xanthen-9-yl (C18Px; S.22; Fig. 2.3.7; Welch et al., 1986) groups have been used as purification handles in solution- and solid-phase oligonucleotide synthesis.


Table 2.3.2 5′-Hydroxyl-Protecting Groups (R1, R2, and R3) of the Substituted Triarylmethyl Type as Purification Handles or for Other Special Applicationsa,b

StrucR1 ture S.2 Phenyl S.3

Phenyl

p-Anisyl

p-Anisyl

S.5

Phenyl

Xanthen-9-yl

Xanthen-9-yl

AbbreviaDeprotection tion MMTr Acid, ZnBr2, etc. DMTr Acid, ZnBr2, etc. Px Acid

S.6

p-Anisyl

Xanthen-9-yl

Xanthen-9-yl

C1Px

Acid

S.9 Phenyl S.10 Phenyl

Phenyl Phenyl

4-Octyloxyphenyl 4-Decyloxyphenyl

Acid Acid

S.11 Phenyl

Phenyl

Acid

Seliger and Schmidt (1987)

S.12 Phenyl

Phenyl

C14Tr

Acid


S.13 Phenyl

Phenyl

C16Tr

Acid


S.14 S.15 S.16 S.17 S.18 S.19 S.20

p-Anisyl p-Anisyl p-Anisyl p-Anisyl p-Anisyl p-Anisyl p-Anisyl

4-Dodecyloxyphenyl 4-Tetradecyloxyphenyl 4-Hexadecyloxyphenyl 4-Propyloxyphenyl 4-Butyloxyphenyl 4-Pentyloxyphenyl 4-Hexyloxyphenyl 4-Heptyloxyphenyl 4-Octyloxyphenyl 4-Dodecyloxyphenyl 2-Hexadecyloxyphenyl Xanthen-9-yl

C8Tr C10Tr or DTr C12Tr

Chattopadhyaya and Reese (1978) Kwiatkowski and Chattopadhyaya (1984) Seliger and Schmidt (1987) Seliger and Schmidt (1987)

Acid Acid Acid Acid Acid Acid Acid

Gupta et al. (1991) Gupta et al. (1991) Gupta et al. (1991) Gupta et al. (1991) Gupta et al. (1991) Gupta et al. (1991) Gupta et al. (1991)

Acid

Phenyl Phenyl Phenyl Phenyl Phenyl Phenyl Phenyl

S.21 p-Anisyl

R2

R3

Phenyl

p-Anisyl

p-Anisyl

S.22 4-Octadecyl- Xanthen-9-yl oxyphenyl S.23 p-Anisyl p-Anisyl S.24 Phenyl Phenyl S.25 p-Anisyl

p-Anisyl

S.26 p-Anisyl

p-Anisyl

S.30 Phenyl S.31 o-Anisyl S.32 Phenyl S.33 Phenyl S.34 Phenyl S.35 p-Anisyl

p-Anisyl o-Anisyl p-Fluorophenyl Phenyl o-Anisyl p-Anisyl

S.36 p-Anisyl

p-Anisyl

Pyrenyl 4-(17-Tetrabenzo(a,c,g,i)fluorenylmethyl)phenyl 4-(17-Tetrabenzo(a,c,g,i)fluorenylmethyl)phenyl 4-[(Succinimidyl-Noxy)carbonyl]phenyl 1-Naphthyl 1-Naphthyl 1-Naphthyl p-Tolyl o-Anisyl 3-(Imidazolyl-1methyl)phenyl 3-(Imidazolyl-1ethylcarbamoyl) phenyl

MOTr

Reference Seliger et al. (1978) Seliger et al. (1978)

C18Px

Acid

BMPM TBF-Tr

Acid Acid

van Boom and Wreesman (1984) Kwiatkowski and Chattopadhyaya (1984) Fourrey et al. (1987) Ramage and Wahl (1993)

TBFDMTr

Acid

Ramage and Wahl (1993)

Acid

Gildea et al. (1990)

Acid, ZnBr2 Acid, ZnBr2 Acid, ZnBr2

Fisher and Caruthers (1983) Fisher and Caruthers (1983) Fisher and Caruthers (1983)

IDTr

Acid, ZnBr2 Acid, ZnBr2 Acid

Fisher and Caruthers (1983) Fisher and Caruthers (1983) Sekine and Hata (1987)

IETr

Acid

Sekine et al. (1993)



Table 2.3.2 Continued

StrucR1 ture S.37 p-Anisyl

R2

R3

p-Anisyl

S.38 p-Anisyl

p-Anisyl

S.39 p-Anisyl

p-Anisyl

S.40 p-Anisyl

p-Anisyl

3-(Imidazolyl-1propylcarbamoyl) phenyl 3-(Imidazolyl-1IBTr butylcarbamoyl)phenyl 3-(Imidazolyl-1IHTr hexylcarbamoyl)phenyl 3-(N-MethylimidIMTr azolyl-2-ethylcarbamoyl)phenyl p-Benzoyloxyphenyl

S.41 p-Benzoyl- p-Benzoyloxyl oxyphenyl S.42 p-(4,5p-(4,5-DichloroDichlorophthalimido) phthalimido) phenyl phenyl S.43 p-(Levulin p-(Levulin yloxy)phenyl yloxy)phenyl S.44 p-Anisyl

p-Anisyl

S.45 p-Anisyl

p-Anisyl

p-(4,5-Dichlorophthalimido) phenyl

AbbreviaDeprotection tion IPTr Acid

CPTr

p-(Levulinyloxy)phenyl p-(Fluorenyl-9-methoxycarbonyl)phenyl p-(Fluorenyl-9-methoxy carbonyl)aminophenyl

Reference Sekine et al. (1993)

Acid


Acid


Acid


Alkali

Sekine and Hata (1983)

Hydrazine in Sekine and Hata (1984); pyridine/acetic Happ and Scalfi-Happ acid (1988) Hydrazine in Sekine and Hata (1985) pyridine/acetic acid β- elimination Happ and Scalfi-Happ (1988) β-elimination Happ and Scalfi-Happ (1988)

aAbbreviations: BMPM, 1,1-bis(4-methoxyphenyl)-1-pyrenylmethyl; CPTr, tris-(4,5-dichlorophthalimido)trityl; DMTr, dimethoxytrityl; DTr, 4-decyloxytrityl; IBTr, (imidazolyl-1-butylcarbamoyl)dimethoxytrityl; IDTr, (imidazolyl-1-methylcarbamoyl)dimethoxytrityl; IETr, (imidazolyl-1-ethylcarbamoyl)dimethoxytrityl; IHTr, (imidazolyl-1-hexylcarbamoyl)dimethoxytrityl; IMTr, N-methylimidazolyl-(Z-ethylcarbamoyl)dimethoxytrityl; IPTr, (imidazolyl-1-propylcarbamoyl)dimethoxytrityl; MOTr, 4-methoxy-4′-octyloxytrityl; MMTr, monomethoxytrityl; Px, 9-phenylxanthen-9-yl (pixyl); Tr, triphenylmethyl (trityl). bOther applications might include purification, fluorescent labeling, visible absorption, biotin substitution, color coding, 3′-phosphate activation, and nonacidic cleavage.


The 1,1-bis(4-methoxyphenyl)-1-pyrenylmethyl (BMPM) group (S.23; Fig. 2.3.8; Table 2.3.2; Fourrey et al., 1987) not only allows the easy purification of target sequences from the solid-phase preparation of deoxyribooligonucleotides and their methyl phosphonate analogs, but also allows their detection by thinlayer chromatography (TLC) or gel electrophoresis down to the picomole level by virtue of the fluorescence of the 5′ substituent in the visible range. Ramage and Wahl (1993) have described the 4-(17-tetrabenzo(a,c,g,i)fluorenylmethyl)trityl (TBF-Tr; S.24) and 4-(17tetrabenzo(a,c,g,i)fluorenylmethyl)-4′,4′′dimethoxytrityl (TBF-DMTr; S.25; Fig. 2.3.8; Table 2.3.2) protecting groups. The latter group was especially found interesting for the purification of long oligonucleotides. Its acid hydrolysis is about twice as fast as that of DMTr, and the product is easily identified through the visible absorption of the substituent. The yields

of TBF-DMTr-protected synthons in phosphoramidite synthesis were only between 70% and 88%, but this problem could be circumvented by postsynthetic treatment of the 5′-deprotected, support-bound oligonucleotide with TBF-DMTrCl. Gildea et al. (1990) have described a DMTr group substituted with a hydroxysuccinimide active ester residue (S.26; Fig. 2.3.9). This linker allowed the addition of, for example, biotin, allowing the possibility of purifying a target oligonucleotide from a crude solid-phase product on a streptavidin-agarose column (Fig. 2.3.9). The 5′-hydroxyl-protecting group could subsequently be released by acid treatment to elute the unprotected oligonucleotide. This trityl-on purification scheme can be coupled with chemical 5′ phosphorylation, as demonstrated by Lönnberg and collaborators (see protecting group S.27 in Figure 2.3.10; Guzaev et al., 1995). Bannwarth and Wippler have de-


OC18H37

C10H21O

CH3O

C

C

O OC8H17

10 (DTr)

22 (C18Px)

19 (MOTr)

Figure 2.3.7 Examples of trityl or pixyl groups modified to serve as purification handles (for complete list see Table 2.3.2).

scribed an interesting approach to combined purification and phosphorylation by adding a protected uridine-5′-phosphate at the 5′ terminus of an oligonucleotide chain. The uridine base was modified with a DMTr-protected thioether residue (S.28 in Fig. 2.3.10). After detritylation, the target sequence was selectively retained and purified by disulfide formation with the activated thiol function of a resin. The oligonucleotide was released by periodate treatment with formation of a terminal 5′ phosphate (Bannwarth and Wippler, 1990). Of course, the “dual applications” described above for many modified or substituted trityl groups can be similarly found in oligoribonucleotide chemistry. The trityl-on purification, for instance, is adaptable in the ribonucleotide series, and a recent publication has established that the deprotection of DMTr does not cause migration of the internucleotidic linkages (Mullah and Andrus, 1996). The 4,4′-dimethoxy-2′′-hexadecyloxytrityl group (S.21; van Boom and Wreesmann, 1984) and the 9(4-octadecyloxyphenyl)xanthen-9-yl group (S.22; Welch et al., 1986) were alternatively introduced to promote the separation of RNA fragments. A problem inherent to polymer-support synthesis is that the product is always a more or less complex mixture of the desired chain with

CH3O

truncated and failure sequences (Földes-Papp et al., 1998, and references therein). In view of the efficiency and selectivity of affinity techniques, a solution-phase synthesis using “affinity protecting groups” would be an interesting alternative to solid-phase preparations (Seliger, 1993). An example of such an affinity separation–based solution synthesis was described (Seliger et al., 1977b) using a combination of the 5′-MMTr and 3′-lipoyl groups. Very recently, an approach to large-scale oligonucleotide synthesis was described in which the chain elongation was done in solution and the extended chain was intermediately anchored to a polymer via a Diels-Alder reaction, so as to allow filtration of educts and by-products (product-anchored sequential synthesis or PASS; Pieken, 1997). A purification-oriented oligonucleotide synthesis in solution has also been achieved through the use of a bridged bis-DMTr 5′-protecting group as soluble carrier (S.29 in Fig. 2.3.11; Biernat et al., 1983).

Color-Coded Triarylmethyl Groups and Triarylmethyl Groups with a Catalytic Function In order to modulate the visible absorption of the species obtained after nonaqueous acid deprotection, a variety of groups of the triarylmethyl type has been constructed (see S.30 to

OCH3

OCH3

C

C

OCH3

23 (Bmpm)

25 (Tbf-DMTr)

Figure 2.3.8 Examples of trityl groups tailored to serve as combined purification handles and visible-absorbing fluorescent markers.



OCH3

CH3O

C

O

O

N

O

O

26

OCH3

HN

C O oligonucleotide−OH

CH3O

O NH H N

S

N H

O

O

Figure 2.3.9 A hydroxysuccinimide-substituted DMTr group and its biotinyl derivative for affinity purification of solid-phase oligonucleotide products.

S.34 in Table 2.3.2). This should allow the “color coding” of different synthons (Fisher and Caruthers, 1983), a principle that has not found application in routine automated oli-

EtO2C CO2Et DMTrO

DMTrO

B

O

EtO2C CO2Et

O

O P O O

NC O

gonucleotide synthesis, but may be useful to monitor the mixed simultaneous addition of two or more nucleotides.

O

O O

DMTrO

B

O

O P O − O

NH3

O

B

OH

O 27

P

P AcOH / H2O 1. Cl2CHCO2H / CH2Cl2 2. NH3

−

EtO2C CO2Et

O

O P O − O

O

HO

B base

O

O P O − O

OH

O

B

OH

BzO O OBz O MMTrS

O P OR

N 11


O

N

O O

O

B

OH 28

Figure 2.3.10

5′-Phosphorylation via DMTr- or MMTr-protected protecting groups.


Cl3CCH2O Cl3CCH2O P O O O

B

O C O

O C O

O

B

O O P OCH2CCl3 OCH2CCl3

29

Figure 2.3.11 Example of modified trityl protection for a purification-oriented solution synthesis (Biernat et al., 1983).

tuted derivatives, other acid-labile residues do not play a role as protecting groups in current oligonucleotide chemistry. Acetal groups have, for example, been used in early syntheses (e.g., Grams and Letsinger, 1970), and 5′-O-tetrahydropyranyl or 5′-O-methoxytetrahydropyranyl derivatives have been mentioned (Reese, 1978). Although a thorough discussion is not in the scope of this unit, there are also approaches to solid-phase oligonucleotide synthesis that use a support anchored to the 5′-hydroxy function. A number of studies appeared, especially in the 1980s, in which support-bound trityl groups were used as anchors for oligonucleotide preparations in the 5′-to-3′ direction (Shabarova, 1980; Belagaje and Brush, 1982; BirchHirschfeld et al., 1983; Rosenthal et al., 1983). The availability of 3′-protected nucleoside-5′phosphoramidites or -H-phosphonates allows

Through another structural modification, namely the introduction of an imidazol-1-ylmethyl residue, the DMTr substituent could be transformed into a protecting group that serves to activate a protected 3′-phosphate moiety (S.35; Fig. 2.3.12; Sekine and Hata, 1987). This concept was recently exended to the introduction of a number of 3-imidazolylalkylcarbamoylphenyl-4,4′-dianisylmethyl substituents (S.36 to S.39, Fig. 2.3.12), as well as a corresponding N-methylimidazolyl derivative (S.40; Sekine et al., 1993; Wada et al., 1998b). In addition to accelerating the rate of internucleotide bond formation, a shift in the ratio of diastereomeric triester was observed in some cases.

Miscellaneous Acid-Labile 5′-Substituents In view of the multiple advantages of trityl groups, as well as their modified and substi-

OCH3

OCH3

CH3O

C

O

O

B

CH3O

O N

+

N P O

−

Cl

C

OCH3

O

O O

O

O P SPh

HN

SPh

n

SPh

CH3O

C

O

O

B

O

O

O P SPh

HN

SPh

N MeN

N 35

B

36 37 38 39

n=2 n=3 n=4 n=6

N

40

Figure 2.3.12 DMTr groups substituted with imidazolyl residues for 3′-phosphate activation (S.36 to S.40), and a proposed activated intermediate (S.35).



the use of all timely methods of solid-phase synthesis in both directions of chain growth. Finally, it may be noteworthy that Tr or DMTr groups present at the 5′ terminus of deoxyribooligonucleotides were found to enhance their anti-HIV activity (Furukawa et al., 1994; Hotoda et al., 1994), a fact that was attributed to the enhancement of membrane permeability through the remaining protecting group.

BLOCKING GROUPS LABILE TO NONACIDIC CONDITIONS Acyl Substituents and Base-Labile Triarylmethyl Groups


In spite of their obvious advantages, acidlabile protecting groups for the 5′-hydroxy function have some limitations to their application in oligonucleotide synthesis. In the deoxyribonucleotide series, even a small number of statistically distributed depurination sites may seriously impair the quality of a solid-phase or large-scale synthesis product. In the ribonucleotide series, multiple acid deprotection steps may interfere with the stability of 2′-hydroxylprotecting groups. This has stimulated the search for new orthogonal protection schemes involving 5′-hydroxyl-protecting groups that can be deblocked under nonacidic conditions. Of course, there are limitations to the use of simple acyl groups such as acetyl or benzoyl as long as the exocyclic amino groups of nucleobases are blocked by acyl residues as well. Thus, acetyl protecting groups are found in the recent literature only for syntheses at the monomer level—e.g., in the preparation of 15N-labeled (Kamaike et al., 1995, 1996) or 2H-labeled nucleosides (Kawashima et al., 1995, 1997), or of components for oligodeoxyribonucleotides with base-modified or mutagenic units (Matsuda et al., 1993; Ozaki et al., 1994; for earlier literature see, e.g., Kössel and Seliger, 1975; Reese, 1978; Sonveaux, 1986). From the preparative standpoint, methods for rapid O-acylation of nucleoside hydroxy groups through phase-transfer catalysis may be noteworthy (Sekine, 1993). Isobutyryl (Gaffney and Jones, 1982a,b) or methoxyacetyl groups (Reese and Skone, 1984) were introduced at sugar hydroxy functions in order to allow the additional protection of the 6-oxo group of deoxyguanosine as well as the 4-oxo group of thymidine. Acyl groups that can be removed under virtually neutral conditions are more attractive. Earlier examples, such as methoxy- and phenoxyacetyl (Reese and Stewart, 1968), did not

completely fulfill the expectations. A more interesting alternative was the o-bromomethylbenzoyl group (S.46; Fig. 2.3.13; Chattopadhyaya et al., 1979), which was used for the preparation of an SV40-specific deoxyribooligonucleotide (Chattopadhyaya and Reese, 1980). Derivatives of the 4-hydroxybutyryl and 2-hydroxymethylbenzoyl groups (Brown et al., 1984; van Boom and Wreesman, 1984; Reese, 1985; Brown et al., 1989a,b) are similarly of interest, since the neighboring participation of the hydroxy groups allows their hydrolysis under extremely mild conditions. Of course, the hydroxy groups have to be protected during chain elongation. In case of the 4-(methylthiomethoxy)butyryl (MTMB; S.51), 2-(methylthiomethoxymethyl)benzoyl (MTMT; S.47), and 2-(isopropylthiomethoxymethyl)benzoyl (DTMT; S.48) groups (Fig. 2.3.13), the deblocking of the methylthiomethyl residue was done with mercury(II) perchlorate and 2,4,6-collidine within 3 hr at room temperature, and the subsequent treatment with K2CO3 in tetrahydrofuran/water released the 2-hydroxymethylbenzoyl substituent within 30 sec (Fig. 2.3.14). As an alternative, the 2-(2,4-dinitrophenylsulfenyloxymethyl) benzoyl (DNBSB) group (S.49; Fig. 2.3.13) was initiated by removal of the sulfenyl residue with p-toluenethiol (Christodoulou et al., 1987a,b). Although these “protected protecting groups” are appealing from their deprotection conditions, difficulties in in-

O R'

R 46 R = Br; R' = H 47 R = OCH2SCH3; R' = H 48 R = OCH2SCH(CH3)2; R' = H 49 R = O S

NO2

R' = H

O2N 50 R = OCOCH2CH2COCH3; R' = NO2 O

R'' 51 R'' = OCH2SCH3

Figure 2.3.13 Acyl substituents for removal at close-to-neutral pH.


S O O

O

OH

Hg(CIO4)2

T

O

collidine

O

T

K2CO3 THF/H2O

HO

O

T

O

O

HO

HO

HO

Figure 2.3.14 Example of two-step removal of an ortho-substituted benzoyl protecting group (Brown et al., 1984; Reese, 1985).

troduction and, in particular, the two-step procedure of removal make them less practical for current automated oligonucleotide synthesis. The β-benzoylpropionyl (S.52; Letsinger et al., 1967) and the levulinyl (S.53; van Boom and Burgers, 1976; Iwai and Ohtsuka, 1988; Iwai et al., 1990) groups can be deprotected by hydrazine in a pyridine/acetic acid mixture (Fig. 2.3.15). However, a partial deprotection of acyl groups from the nucleobases was also observed under the conditions of hydrazinolysis of the β-benzoylpropionyl group (Letsinger and Miller, 1969). Especially mild conditions of hydrazinolytic deprotection apply to the 2-levulinyloxymethyl-5-nitrobenzoyl group (S.50; Fig. 2.3.13; Kamaike et al., 1997). Appropriate substitution with electrondonating substituents that stabilize a trityl cation may strongly modify the conditions of cleavage. Based upon the earlier finding that the 4-hydroxytrityl group is hydrolyzed much more easily than the corresponding 4-acetoxytrityl residue (Taunton-Rigby et al., 1972), the 4,4′,4′′-tris(benzoyloxy)trityl substituent (S.41; Table 2.3.2) was prepared as an acid-stable but base-labile “protected 5′-protecting group” (Sekine and Hata, 1983), the removal of which is shown in Figure 2.3.16. This principle was further extended to 4,4′,4′′-tris(4,5dichlorophthalimido)trityl (S.42; Sekine and Hata, 1984) and 4,4′,4′′-tris(levulinyloxy)trityl (S.43; Table 2.3.2; Sekine and Hata, 1985) as protecting groups labile to hydrazine treatment in pyridine/acetic acid. Based on the same general concept, the 4-(9-fluorenylmethoxycarbonyl)oxy-4′,4′′-dimethoxytrityl (S.44) and 4-(9-fluorenyl-

methoxycarbonyl)amino-4′,4′′-dimethoxytrityl) (S.45; Table 2.3.2) groups were introduced. Their release could be triggered through β-elimination (Scalfi-Happ et al., 1987; Happ and Scalfi-Happ, 1988). The usefulness of hydrazine-labile modified trityl groups such as 4,4′,4′′-tris(4,5-dichlorophthalimido)trityl (S.42; Sekine and Hata, 1984, 1986; ScalfiHapp et al., 1987) and the S.44 and S.45 groups for β-elimination-triggered deprotection (Happ and Scalfi-Happ, 1988) was also demonstrated for large-scale oligoribonucleotide synthesis and for the preparation of 2′(3′)-Oaminoacyl-oligoribonucleotides (Scalfi-Happ et al., 1987).

Carbonate-Type Protecting Groups Protecting groups of the carbonate type, most popular in peptide chemistry, have also received much attention in oligonucleotide synthesis. In earlier work, the isobutyloxycarbonyl group (S.54; Fig. 2.3.17; Ogilvie and Letsinger, 1967) and the p-nitrophenyloxycarbonyl (NPOC) group (S.55; Letsinger and Ogilvie, 1967) were shown to be introduced rather selectively at 5′-hydroxy functions en route to 3′-tritylated thymidine and uridine derivatives. However, their removal in dilute sodium hydroxide/dioxane would not be compatible with exocyclic acyl protection of nucleobases. p-Nitrophenyloxycarbonyl and a number of other carbonate protecting groups could also be introduced into partially protected thymidine and uridine via the corresponding 5′-chloroformates (Seliger, 1972). These studies later led to the introduction of the 5′-p-phenylazophenyl-

O O

R

O

B

O HO

NH2NH2

HO

N NH R

O

O

B

pyridine / acetic acid HO

52 R = phenyl 53 R = methyl

Figure 2.3.15 groups.

Hydazinolytic deprotection of β-benzoylpropionyl (S.52) and levulinyl (S.53)



O Ph

O

O

O

O

Ph O O

C X

OH

O

C X

C

HX HX

O Ph

O

O

X = nucleoside-5'-oxy

O 41

Figure 2.3.16 group.


4,4′,4′′-Tris(benzoyloxy)trityl as an acid-stable, base-labile 5′-hydroxyl-protecting

oxycarbonyl (PAPOC) group (S.56; Kössel and Seliger, 1975), which was introduced via pphenylazophenylchloroformate (Seliger and Kotschi, 1985; Seliger et al., 1986). This group could be released by a two-step treatment with β-cyanoethanol/triethylamine and diazabicycloundecene (DBU). This group retains some of the features of trityl- or pixyl-type substituents (i.e., its deprotection gives a visible-absorbing solution), and the PAPOC nucleosides are readily crystalline; however, the two-step deprotection procedure is less advantageous. More attention has been given to the 9-fluorenylmethoxycarbonyl (FMOC) group (S.57, Fig. 2.3.17), a standard in peptide protection. FMOC-protected synthons allowed an alternative protocol for solid-support deoxyribooligonucleotide synthesis, retaining the standard acyl protection for the bases and avoiding acid deblocking steps (Gioeli and Chattopadhyaya, 1982; Balgobin and Chattopadhyaya, 1987; Ma and Sonveaux, 1987, 1989). The potential of 5′-FMOC protection for oligoribonucleotide synthesis was demonstrated by Fukuda et al. (1988) and by the group of Gait (Lehmann et al., 1989). The FMOC group was also used in acetal-linked solid-support oligonucleotide synthesis (Palom et al., 1993). Recently, the p-chlorophenyloxycarbonyl group (S.57a, Fig. 2.3.17) was found to be removed in 1 atm.

27. Elute the column with the following methylene chloride/methanol solutions: 500 mL of 98:2 (v/v); 500 mL of 96:4 (v/v); 1000 mL of 9:1 (v/v); and 1000 mL of 8:2 (v/v). Combine the product-containing fractions and evaporate to a colorless glass. 28. Dissolve the residue in 25 mL chloroform, precipitate by dropwise addition to 2500 mL rapidly stirring pentane, and decant off most of the supernatant (>2000 mL). 29. Collect solids by centrifuging 10 min at 3000 rpm, 15°C, and decant the pentane. 30. Wash twice by resuspending in 250 mL fresh pentane and repeating the centrifugation. Dry the resulting white powder (S.1a) in vacuo. 31. Characterize the final product by TLC and 1H-NMR. 2′-O-MTHP-uridine (S.1a): yield 26.1 g (73% from S.6); Rf (silica, chloroform/methanol 9:1): 0.24; 1H-NMR (250 MHz, DMSO-d6): 1.50-1.90 (m×m, 4H; -CH2-(3,5) of MTHP), 2.95 (s, 3H; H3CO- of MTHP), 3.35-3-50 (m, 2H, -CH2-(5′)), 3.5-3.8 (m×m, 4H; -CH2-(2,6) of MTHP), 3.90 (m, 1H; -CH-(4′)), 3.96 (m, 1H, -CH-(3′)), 4.25-4.36 (m, 1H; -CH-(2′)), 5.10-5.25 (d×tbr, 2H, exchangeable with D2O; 3′/5′-OH), 5.73 (d, J = 8.1, 1H; uridyl-H5), 6.00 (d, J = 7.8, 1H; -CH-(1′)), 7.93 (d, 1H; uridyl-H6), 11.4 (sbr, 1H, exchangeable; uridyl-N3-H).

A Base-Labile Protecting Group


PREPARATION OF 2′-O-(4-METHOXYTETRAHYDROPYRAN-4-YL)4-N-BENZOYLCYTIDINE (S.1b) FROM CYTIDINE

SUPPORT PROTOCOL 2

In the first step of this procedure, cytidine (S.7; Fig. 2.4.4) is simultaneously protected on its 3′- and 5′-hydroxy groups by the bifunctional protecting reagent 1,3-dichloro1,1,3,3-disiloxane (Markiewicz and Wiewerowski, 1985). The derivative S.8 is formed exclusively, due to the higher reactivity of the sterically more accessible 5′-hydroxy group and the subsequently favorable cyclization to the 3′-hydroxy group. Selective benzoylation on N4 of the base is then achieved by treatment with the active ester 1-hydroxybenzotriazolyl benzoate to give S.9. The latter compound is ketalized at the 2′-hydroxy group upon reaction with 5,6-dihydro-4-methoxy-2H-pyran. Then a solution of n-tetrabutylammonium fluoride in acetonitrile removes the disiloxane protecting group to give 2′-OMTHP-4-N-benzoylcytidine (S.1b) in good yield (cf. Reese et al., 1970; van Boom and Wreesmann, 1984). Chromatography is required only for the last step of the sequence. Additional Materials (also see Basic Protocol and Support Protocol 1) Cytidine (S.7; Sigma or Fluka), dried before use for 2 hr at 50°C over phosphorus pentoxide in vacuo N,N-Dimethylformamide, dried by stirring overnight at room temperature with calcium hydride (5 g/L) and subsequent distillation under reduced pressure (b.p. 70° to 80°C, 20 to 30 mmHg) Dry pyridine (see Support Protocol 1 for drying procedure) 1,3-Dichloro-1,1,3,3-tetraisopropyldisiloxane (see Support Protocol 5) 2 M triethylammonium bicarbonate buffer (see recipe) Acetone (analytical grade) 1-Hydroxybenzotriazole (Fluka), dried before use for 72 hr at 50°C over phosphorus pentoxide in vacuo Triethylamine, dried by refluxing 2 hr over calcium hydride (5 g/L) followed by distillation Benzoyl chloride (Fluka, puriss.) Acetonitrile (analytical grade) 1 M n-tetrabutylammonium fluoride (Aldrich/Fluka) in acetonitrile CAUTION: 1-Hydroxybenzotriazole may explode at higher temperatures.

NH2 4

N N1

5'

O

4'

N 2

6

HO

NH2

3

5

steps 1-7

O

Si

1' 2'

3'

HO

N

O

steps 8-15

O

O

O Si O

OH

OH

7 O

O

8

HN

HN N

Si

N

O

O

N O

steps 16-30

N

HO

O

O

O Si

O

OH

HO

O

OCH3

O 9

1b

Figure 2.4.4 Scheme showing the preparation of 2′-O-MTHP-4-N-benzoylcytidine (S.1b) from cytidine (see Support Protocol 2).



Prepare 3′,5′-O-(tetraisopropyldisiloxane-1,3-diyl)cytidine (S.8) 1. Prepare and stir a suspension of the following: 12.12 g dry cytidine (S.7; 50 mmol) 200 mL dry N,N-dimethylformamide 40 mL dry pyridine (40 mL). 2. Add 18 mL of 1,3-dichloro-1,1,3,3-tetraisopropyldisiloxane (62.5 mmol) to 50 mL dry N,N-dimethylformamide and add dropwise to the stirred cytidine solution over a period of 15 min. Stir the reaction mixture for 1 hr at room temperature. 3. Check the reaction by TLC (see Basic Protocol, steps 4 to 6). TLC should indicate complete reaction of the starting material (Rf < 0.05) to S.8 (Rf = 0.28).

4. Neutralize with 75 mL of 2 M triethylammonium bicarbonate buffer and concentrate under reduced pressure to a small volume (100 mL) using a rotary evaporator. 5. Dissolve in 1000 mL methylene chloride and wash with 500 mL of 1 M aqueous sodium hydrogen carbonate followed by 500 mL water. 6. Dry the organic layer over 50 g magnesium sulfate and concentrate to a colorless oil. 7. Crystallize from 500 mL acetone to produce pure 3′,5′-O-(tetraisopropyldisiloxane1,3-diyl)cytidine. 3′,5′-O-(Tetraisopropyldisiloxane-1,3-diyl)cytidine (S.8): yield 19.5 g (85%), m.p. 226-228 (decomp.), Rf = 0.28. 1H-NMR (CDCl3/CD3OD): 8.00 (d, J = 7.5, 1H; cytidyl-H6), 7.58 (d, J = 7.5, 1H; cytidyl-H5), 5.64 (d, J = 5.5, 1H; -CH-(1′)); 13C-NMR (CDCl3/CD3OD): 166.2 (C4), 156.3 (C2), 140.6 (C6), 94.7 (C5), 91.5 (C1′), 81.6 (C4′), 75.1 (C3′), 68.2 (C2′), 60.0 (C5′), 17.4, 17.0 13.5, 13.0, 12.5 (TIPS).

Prepare 3′,5′-O-(tetraisopropyldisiloxane-1,3-diyl)-4-N-benzoyl-cytidine (S.9) 8. Prepare and stir a solution of the following: 10.0 g dry 1-hydroxybenzotriazole (75 mmol) 300 mL dry dioxane 21 mL dry triethylamine (150 mmol). 9. Combine 8.70 mL benzoyl chloride (75 mmol) and 50 mL dry dioxane, and add dropwise to the above mixture (step 8) over 15 min. Stir for 1 hr at room temperature. 10. Filter off the precipitated triethylammonium chloride salts under inert atmosphere using a glass filter crucible, and add the filtrate to a solution containing 24.25 g of 2′-O-(tetraisopropyldisiloxane-1,3-diyl)cytidine (S.8; 50 mmol) in 150 mL dry N,Ndimethylformamide. 11. Evaporate off a volume of ∼100 mL under reduced pressure and stir the residue at room temperature for 3 days. Check the reaction by TLC. TLC should indicate complete reaction of the starting material.

12. Add 5 mL water and concentrate the reaction mixture under reduced pressure to a small volume (100 mL). 13. Dissolve in 1000 mL methylene chloride and wash as in step 5. A Base-Labile Protecting Group

14. Dry the organic layer over 50 g magnesium sulfate and concentrate to a light brown oil.


15. Crystallize from a minimal amount of refluxing acetonitrile to yield analytically pure 3′,5′-O-(tetraisopropyldisiloxane-1,3-diyl)-4-N-benzoylcytidine (S.9). 3′,5′-O-(Tetraisopropyldisiloxane-1,3-diyl)-4-N-benzoylcytidine (S.9): yield 22.6 g (77% in first crop). Rf = 0.40 (silica, chloroform/methanol 9:1). Elemental analysis: calc.; C 57.04, H 7.30, N 7.26%; found: C 56.36, H 7.08, N 7.26%.

Prepare 2′-O-MTHP-4-N-benzoylcytidine (S.1b) 16. Dissolve 4.75 g toluene-p-sulfonic acid monohydrate (25 mmol) in 125 mL dry dioxane and evaporate under reduced pressure to a colorless oil to remove traces of water. 17. Under an inert atmosphere, add 29.45 g of 3′,5′-O-(tetraisopropyldisiloxane-1,3diyl)-4-N-benzoylcytidine (S.9; 50 mmol) followed by 250 mL dry dioxane. 18. To the resulting clear solution, add 29 mL of 5,6-dihydro-4-methoxy-2H-pyran (28.5 g; 250 mmol). Stir 20 hr at room temperature and check the reaction by TLC. TLC should show complete reaction of the starting material.

19. Neutralize the reaction with 3.5 mL half-saturated methanolic ammonia solution (check with moist pH indicator paper) and concentrate immediately under reduced pressure to a small volume (100 mL). 20. Dissolve in 1000 mL methylene chloride and wash as in step 5. 21. Dry the organic layer over 50 g magnesium sulfate and concentrate to a glass. 22. Add 250 mL of 1 M n-tetrabutylammonium fluoride in acetonitrile, stir for 1 hr at room temperature, and check the reaction by TLC. TLC should show complete removal of the silyl protecting group. n-Tetrabutylammonium fluoride in acetonitrile must be prepared by the experimenter. Commercially available solutions in tetrahydrofuran can also be used.

23. Concentrate under reduced pressure to a small volume (100 mL). Dissolve in methylene chloride, wash, and dry as in steps 20 and 21. 24. Dissolve with a minimal amount of dichloromethane and apply to a column of silica gel (300 g; 12 × 7 cm) packed in methylene chloride. Short-column chromatography is performed according to Hunt and Rigby (1967) and is similar to the flash chromatography described by Still et al. (1978). Do not use air pressure >1 atm.

25. Elute the column with the following methylene chloride/methanol solutions: 500 mL of 98:2 (v/v); 500 mL of 96:4 (v/v); and 1000 mL of 9:1 (v/v). 26. Collect the main product-containing fractions and evaporate to a glass. 27. Dissolve the residue in 25 mL chloroform, precipitate by dropwise addition to 2500 mL rapidly stirring pentane, and decant off most of the supernatant (>2000 mL). 28. Collect the solids by centrifuging 10 min at 3000 rpm g, 15°C, and decant the pentane. 29. Wash twice by resuspending in 250 mL fresh pentane and repeating the centrifugation. Dry the resulting white powder (S.1b) in vacuo. 30. Characterize the final product by TLC and 1H-NMR. 2′-O-MTHP-4-N-benzoylcytidine (S.1b): yield 16.1 g (70% from S.9); Rf (silica, chloroform/methanol 9:1): 0.43; 1H-NMR (250 MHz, DMSO-d6): 1.50-1.90 (m×m, 4H; -CH2-



(3,5) of MTHP), 2.93 (s, 3H; H3CO- of MTHP), 3.40-3.60 (m, 2H, -CH2-(5′)), 3.5-3.8 (m×m, 4H; -CH2-(2,6) of MTHP), 3.96 (m, 1H; -CH-(4′)), 4.02 (m, 1H, -CH-(3′)), 4.39 (m(d×d), 1H; -CH-(2′)), 5.19 (d, J = 6.5, 1H, exchangeable with D2O; 3′-OH), 5.25 (t, J = 2.7, 1H, exchangeable with D2O; 5′-OH), 6.14 (d, J = 6.8, 1H; -CH-(1′)), 7.36 (d, J = 7.7, 1H; cytidyl-H5), 7.4-8.1 (m×m, 5H; benzoyl-aromatic H), 8.43 (d, 1H; cytidyl-H6), 11.3 (sbr, 1H, exchangeable with D2O; cytidyl-N4-H). SUPPORT PROTOCOL 3

PREPARATION OF 2′-O-(4-METHOXYTETRAHYDROPYRAN-4-YL)6-N-BENZOYLADENOSINE (S.1c) FROM ADENOSINE In this protocol, adenosine (S.10, Fig. 2.4.5) is base protected prior to 3′,5′-disilylation by benzoylation of its crystalline 2′,3′,5′-tri-O-acetyl-derivative (S.11), yielding 6-N-benzoyladenosine (S.12) after deacetylation in situ. The chosen route (cf. Schaller et al., 1963; Reese et al., 1970; Büchi and Khorana, 1972; Jones, 1984; van Boom and Wreesmann, 1984; Markiewicz and Wiewerowski, 1985) is then comparable to the cytidine case (see Support Protocol 2): 3′,5′-disilylation to yield S.13, followed by acid-catalyzed ketalization of the 2′-OH by 5,6-dihydro-4-methoxy-2H-pyran, and cleavage of the silyl protecting group by fluoride ion to yield 2′-O-MTHP-6-N-benzoyladenosine (S.1c) in good chromatographically isolated yield. Additional Materials (also see Basic Protocol and Support Protocols 1 and 2) Adenosine (S.10; Sigma or Fluka), dried before use for 2 hr at 50°C over phosphorus pentoxide in vacuo Dry pyridine (see Support Protocol 1 for drying procedure) Acetic anhydride, fractionally distilled with 10% toluene to remove acetic acid 25% (w/v) sodium methoxide in methanol (pract., Aldrich, Fluka) Acetic acid (analytical grade) Diethyl ether (analytical grade) Imidazole (analytical grade) 0.1 M hydrochloric acid

O NH2

7

6

N 5 HO

N 4 N 9

5'

O

4' 3'

HO

1

N

N

8

steps 1-5

2

3

1'

HN

NH2

N

AcO

O

N

N steps 6-18

N

N

HO

O

N N

steps 19-27

2'

OH

AcO

10

HO

OAc

O

O HN

HN N O Si

N O

N

N N

OH

12

11

step 28

N

HO

O

N N

O Si O

OH 13

HO

O

OCH3

O 1c


Figure 2.4.5 Scheme showing the preparation of 2′-O-MTHP-6-N-benzoyladenosine (S.1c) from adenosine (see Support Protocol 3). The final step from S.13 to S.1c is performed as in steps 16 to 30 of Support Protocol 2 (see Fig. 2.4.4).


Prepare 2′,3′,5′-tri-O-acetyladenosine (S.11) 1. Prepare and stir a suspension of 10.0 g adenosine (S.10; 37.4 mmol) in 50 mL dry pyridine. 2. Add 23 mL acetic anhydride (0.24 mol) and stir overnight at room temperature. 3. Cool in an ice bath and quench with 40 mL methanol. 4. Remove from the ice bath and stir 1 hr at room temperature. 5. Concentrate to dryness, and recrystallize the white solid residue twice from 100 mL hot ethanol to yield 12.0 g (82%) tri-O-acetyladenosine (S.11). Check the product by TLC (see Basic Protocol, steps 4 to 6). The product is sufficiently pure (Rf = 0.31) for use in the subsequent steps.

Prepare 6-N-benzoyladenosine (S.12) 6. Dissolve crystalline S.11 in 50 mL pyridine, cool in an ice bath, and add 5.3 mL benzoyl chloride (45.8 mmol). 7. Let the reaction mixture warm to room temperature and stir for 2 hr. 8. Quench with 6 mL water and stir for another 30 min. 9. Concentrate to a syrup using a rotary evaporator. 10. Partition the residue between 100 mL chloroform and 100 mL saturated aqueous sodium hydrogen carbonate solution. Reextract the aqueous phase with two 50-mL portions of chloroform. 11. Dry the combined organic extracts over ~20 g anhydrous sodium sulfate and evaporate to a glass (Rf = 0.71). 12. Redissolve in 50 mL of 1:1 (v/v) methanol/pyridine, cool on ice, and then add 15 mL (∼3 eq) of 25% sodium methoxide in methanol. 13. Let the reaction mixture warm to room temperature and monitor by TLC. Deacetylation to the more polar product (Rf = 0.24) occurs in 1 atm pressure for elution (Hunt and Rigby, 1967; Still et al., 1978). 27. Combine the product-containing fractions, evaporate the solvent, and dry under high vacuum to yield 3.21 g (81%) chromatographically pure S.13 as a colorless foam. Check by TLC (Rf = 0.58). Prepare 2′-O-MTHP-6-N-benzoyladenosine (S.1c) 28. Introduce the MTHP protecting group and perform desilylation (see Support Protocol 2, steps 16 to 30), replacing S.9 (29.45 g; 50 mmol) with the corresponding adenosine derivative S.13 (30.84 g; 50 mmol) to yield S.1c as a dry white powder. 2′-O-MTHP-6-N-benzoyladenosine (S.1c): yield 18.9 g (78% from S.13); Rf (silica, chloroform/methanol 9:1): 0.30; 1H-NMR (250 MHz, DMSO-d6): 1.40-1.90 (m×m, 4H; -CH2(3,5) of MTHP), 2.55 (s, 3H; H3CO- of MTHP), 3.20-3.80 (m×m, 2H, -CH2-(5′)), 3.3-3.8 (m×m, 4H; -CH2-(2,6) of MTHP), 4.05 (m, 1H; -CH-(4′)), 4.18 (m, 1H, -CH-(3′)), 5.00 (m(d×d), 1H; -CH-(2′)), 5.27 (t, J = 5.5, 1H, exchangeable with D2O; 5′-OH), 5.33 (d, J = 4.5, 1H, exchangeable with D2O; 3′-OH), 6.19 (d, J = 7.4, 1H; -CH-(1′)), 7.5-8.1 (m×m, 5H; benzoyl-aromatic H), 8.77 (2s, 2H; adenyl-H2/8), 11.3 (sbr, 1H, exchangeable with D2O; adenyl-N6-H). SUPPORT PROTOCOL 4

PREPARATION OF 2′-O-(4-METHOXYTETRAHYDROPYRAN-4-YL)-2-NISOBUTYRYLGUANOSINE (S.1d) FROM GUANOSINE Guanosine (S.14; Fig. 2.4.6) is base protected on N2 (S.16) by reaction with isobutyryl chloride after transiently blocking the 2′-, 3′-, and 5′-hydroxy groups by persilylation (S.15). The chosen route is then comparable to the cytidine and adenosine cases: 3′,5′-disilylation to S.17, followed by acid-catalyzed ketalization of the 2′-OH by 5,6-dihydro-4-methoxy-2H-pyran and cleavage of the silyl protecting group by fluoride ion to yield 2′-O-MTHP-2-N-isobutyrylguanosine (S.1d) in satisfactory isolated yield. Generally, the higher basicity and polarity of guanosine derivatives renders them more difficult to handle, and the yields are frequently compromised by loss of material in aqueous extraction media or on polar silica gel phases. Nevertheless, the proposed sequence is a viable route that is based on well-described literature procedures (Reese et al., 1970; Ti et al., 1982; Jones, 1984; van Boom and Wreesmann, 1984; Markiewicz and Wiewerowski, 1985; McLaughlin et al., 1985).


Additional Materials (also see Basic Protocol and Support Protocols 1 and 2) Guanosine (S.14; Sigma or Fluka), dried before use for 2 hr at 50°C over phosphorus pentoxide in vacuo Dry pyridine (see Support Protocol 1 for drying procedure)


O

O 7 N 5

6 1

N 4 N 9

5'

HO

O

4' 3'

HO

1'

N

NH

8 3

2

NH2

steps 1-2

N

(H3C)3SiO

O

O N

NH N

NH2

steps 3-9

NH

N

HO

N

O

O

N H

steps 10-16

2'

(H3C)3SiO

OH

OSi(CH3)3

HO

15

14

O

O N O Si

N O

NH N

OH

16

N H

N

O steps 17-33

N

HO

O

NH N

O

N H

O Si O

OH

HO

O

OCH3

17 O 1d

Figure 2.4.6 Scheme showing the preparation of 2′-O-MTHP-2-N-isobutyrylguanosine (S.1d) from guanosine (see Support Protocol 4).

Trimethylsilyl chloride (analytical grade) Isobutyryl chloride (analytical grade) Concentrated aqueous ammonia (~25%) Methylene chloride (analytical grade) Phosphorus pentoxide Prepare 2-N-isobutyrylguanosine (S.16) 1. Suspend 14.15 g dry guanosine (S.14; 50 mmol) in 200 mL dry pyridine. 2. Add 47.5 mL trimethylsilyl chloride (375 mmol) and stir for 3 hr at ambient temperature to produce S.15. 3. Cool to 0°C and add 15.6 mL isobutyryl chloride (150 mmol), dropwise, over a period of 30 min. 4. Allow to warm to room temperature and stir overnight. 5. Cool again to 0°C and quench by adding 50 mL water. Stir for another 10 min at room temperature. 6. Add 100 mL concentrated aqueous ammonia to the clear solution and stir for an additional 30 min. 7. Pour the mixture into 800 mL water and extract with 150 mL methylene chloride. 8. Evaporate the aqueous phase to dryness and crystallize three times from 50 to 100 mL boiling water. 9. Dry the resulting solids for 2 days under high vacuum over phosphorus pentoxide to obtain 7.25 g (41%) 2-N-isobutyrylguanosine (S.16) as a grayish-white, chromatographically homogenous powder. Check by TLC (see Basic Protocol, steps 4 to 6; Rf = 0.074). Prepare 3′,5′-O-(tetraisopropyldisiloxane-1,3-diyl)-2-N-isobutyrylguanosine (S.17) 10. Dissolve 7.05 g of 2-N-isobutyrylguanosine (S.16; 20.0 mmol) in 125 mL dry pyridine. 11. Add 6.6 mL of 1,3-dichloro-1,1,3,3-tetraisopropyldisiloxane (24 mmol) and stir overnight. Check the reaction by TLC. TLC shows complete reaction of the starting material.



12. Evaporate the suspension and partition between 250 mL chloroform and 250 mL saturated aqueous sodium hydrogen carbonate solution. 13. Reextract the aqueous layer twice with 125 mL chloroform, dry the organic layer over ~30 g sodium sulfate, and evaporate to a glass using a rotary evaporator. 14. Remove traces of pyridine by azeotropic rotoevaporation from 100 mL toluene (twice), 100 mL ethanol (once), and 100 mL chloroform (once). 15. Purify the crude product by short-column chromatography using 200 g silica gel (see Basic Protocol, step 12). Elute with chloroform until the product appears, then with chloroform containing 2% (v/v) ethanol (∼1500 mL total in 30-mL fractions). Do not use >1 atm pressure for elution (Hunt and Rigby, 1967; Still et al., 1978). 16. Combine the product-containing fractions, evaporate the solvent, and dry under high vacuum to give 7.25 g (61%) chromatographically pure S.17 as a colorless foam. Check by TLC (Rf = 0.55). Prepare 2′-O-MTHP-2-N-isobutyrylguanosine (S.1d) 17. Dissolve 0.95 g of toluene-p-sulfonic acid monohydrate (5 mmol) in 25 mL dry dioxane and evaporate under reduced pressure to a colorless oil to remove traces of water. 18. Under an inert atmosphere, add 5.95 g of 3′,5′-O-(tetraisopropyldisiloxane-1,3-diyl)2-N-isobutyrylguanosine (S.17; 10 mmol) followed by 50 mL dry dioxane. 19. To the resulting clear solution, add 5.8 mL of 5,6-dihydro-4-methoxy-2H-pyran (5.7 g, 50 mmol). Stir 20 hr at room temperature and check the reaction by TLC. TLC should show complete reaction of the starting material.

20. Neutralize the reaction with 0.8 mL half-saturated methanolic ammonia solution and concentrate immediately under reduced pressure to a small volume (20 mL). 21. Dissolve the residue in 200 mL chloroform and wash with 100 mL of 1 M aqueous sodium hydrogen carbonate solution. 22. Dry the organic layer over 50 g magnesium sulfate and concentrate to a glass. 23. Add 50 mL of 1 M n-tetrabutylammonium fluoride in acetonitrile, stir for 1 hr at room temperature, and check the reaction by TLC. TLC should show complete removal of the silyl group.

24. Concentrate under reduced pressure to a small volume (20 mL). Dissolve in chloroform and wash as in step 21. 25. Reextract the aqueous phase twice with 100 mL chloroform. 26. Dry the organic layer over 10 g anhydrous sodium sulfate and concentrate to a glass. 27. Dissolve the residue with minimal amount of chloroform and apply to a column of silica gel (100 g; 8 × 5 cm) packed in chloroform.


28. Elute the column with chloroform until the product appears, and then collect 20-mL fractions with the following chloroform/ethanol solutions: 300 mL of 98:2 (v/v); 300 mL of 96:4 (v/v), 600 mL of 9:1 (v/v). 29. Collect the main product-containing fractions and evaporate to a glass.


30. Dissolve the residue in 25 mL chloroform, precipitate by dropwise addition to 500 mL rapidly stirring pentane, and decant off most of the supernatant (>400 mL). 31. Collect the solids by centrifuging 10 min at 3000 rpm, 15°C, and decant the pentane. 32. Wash the solids twice by resuspending in 100 mL fresh pentane and repeating the centrifugation. Dry the resulting white powder (S.1d) in vacuo. 33. Characterize the final product by TLC and 1H-NMR. 2′-O-MTHP-2-N-isobutyrylguanosine (S.1d): yield 2.42 g (52% from S.17); Rf (silica, chloroform/methanol 9:1): 0.29; 1H-NMR (250 MHz, DMSO-d6): 1.12 (2d, J = 6.8, 6H; (H3C)2CH-)), 1.40-1.90 (m×m, 4H; -CH2-(3,5) of MTHP), 2.55 (s, 3H; H3CO- of MTHP), 2.75 (septet, J = 6.8, 1H; (H3C)2CH-)), 3.19-3.33 (m×m, 2H, -CH2-(5′)), 3.4-3.7 (m×m, 4H; -CH2-(2,6) of MTHP), 3.98 (m(t), 1H; -CH-(4′)), 4.10 (m(t), 1H, -CH-(3′)), 4.74 (m(d×d), 1H; -CH-(2′)), 5.77 (m(t), 1H, exchangeable with D2O; 5′-OH), 5.19 (m(d), 1H, exchangeable with D2O; 3′-OH), 6.98 (d, J = 7.9, 1H; -CH-(1′)), 7.5-8.1 (m×m, 5H; benzoyl-aromatic H), 8.31 (s, 1H; guanyl-H8), 11.7/12.1 (2sbr, 2H, exchangeable with D2O; guanyl-N1-H and N2-H).

PREPARATION OF 1,3-DICHLORO-1,1,3,3TETRAISOPROPYLDISILOXANE

SUPPORT PROTOCOL 5

Although this reagent is commercially available (e.g., Fluka), this protocol provides a convenient procedure for its preparation. Materials Magnesium curls Dry diethyl ether, distilled from phosphorus pentoxide (30 g/L) Isopropyl bromide, distilled from calcium hydride (5 g/L) Trichlorosilane, freshly distilled 0.1 N hydrochloric acid Magnesium sulfate Methylene chloride (analytical grade), dried by passage through activated basic alumina Chlorine gas, dried over concentrated sulfuric acid NaCl plates for infrared (IR) spectroscopy Additional reagents and equipment for IR spectroscopy Prepare 1,1,3,3-tetraisopropyldisiloxane 1. Combine 64 g magnesium curls and 200 mL dry diethyl ether. 2. Add a solution of 270 mL isopropyl bromide in 400 mL dry diethyl ether dropwise to the solution. 3. Stir the reaction mixture mechanically and heat under reflux for 3.5 hr. 4. Add a solution of 100 mL trichlorosilane and 400 mL dry diethyl ether dropwise to the stirred solution (step 3) and heat under reflux overnight. 5. Quench by adding 800 mL of 0.1 N hydrochloric acid in a dropwise fashion. 6. Stir the mixture and heat under reflux for another 3.5 hr. 7. Separate out the organic layer and extract the aqueous layer three times with 300 mL diethyl ether.



8. Dry the combined organic layers over ~50 g magnesium sulfate and concentrate to a colorless oil under reduced pressure. 9. Distill the residue to produce pure 1,1,3,3-tetraisopropyldisiloxane (100 g; b.p. 80° to 90°C, 10 mmHg) as a colorless oil. Prepare 1,3-dichloro-1,1,3,3-tetraisopropyldisiloxane 10. Dissolve the above product (100 g of 1,1,3,3-tetraisopropyldisiloxane) in 500 mL methylene chloride and pass a stream of dry chlorine gas through the solution. Protect the apparatus from atmospheric moisture using a paraffin oil bubbler. Calcium chloride drying tubes are entirely inefficient and will lead to reaction failure.

11. When the temperature rises to ∼27° to 30°C, cool the reaction mixture to 17° to 20°C by immersing it in an ice water bath while continuing the stream of chlorine gas. 12. After 2 hr, and then after every hour, withdraw a small sample of the reaction mixture and analyze by IR spectroscopy using NaCl plates. 13. Stop the chlorination when the IR spectrum indicates the disappearance of the absorption band at 2100 cm−1 (Si-H). 14. Evaporate off the volatile compounds and distill the residue under diminished pressure (b.p. 85° to 90°C, 2 mmHg) to give 1,3-dichloro-1,1,3,3-tetraisopropyldisiloxane (70 g) as a colorless oil. Store dry at 0°C for up to several months or even years. REAGENTS AND SOLUTIONS Use distilled, deionized water for all recipes and protocol steps. For common stock solutions, see APPENDIX 2A; for suppliers, see SUPPLIERS APPENDIX.

Anisaldehyde reagent 1 mL 4-methoxybenzaldehyde 2 mL concentrated sulfuric acid 10 mL glacial acetic acid 87 mL methanol (analytical grade) Store up to 6 months at room temperature All analytical-grade reagents are preferred if available.

Methanolic ammonia solution, half-saturated (∼8 M) Pass dry ammonia gas (passed over solid potassium hydroxide pellets) through 500 mL analytical-grade methanol at −20°C until saturation. Dilute with 500 mL analytical-grade methanol. Store up to 4 weeks at –18°C. Triethylammonium bicarbonate buffer, 2 M Combine 825 mL analytical-grade triethylamine and 2175 mL water. Saturate with carbon dioxide gas at 0°C until the pH of the clear solution reaches 7.5. Store up to 4 weeks at 4°C. COMMENTARY Background Information


Within the field of solid-phase peptide synthesis, the introduction of a base-labile protecting group for the nucleophilic (amino-) center has been very successful, exceeding by far the status of alternative and becoming a standard

technique in many laboratories (Atherton and Sheppard, 1989, and references cited therein). The base-sensitive amino-protecting 9-fluorenylmethoxycarbonyl (FMOC) group, first introduced by Carpino and Han in 1970 (for a review see Carpino, 1987), is strictly orthogo-


nal to the acid-labile protecting groups used for the side chains. Orthogonality is fulfilled when permanent side-chain protecting groups remain completely stable during peptide assembly while being repeatedly exposed to the conditions applied for removal of the temporary terminal protecting group. For solid-phase oligonucleotide assemblies, the use of 5′-OFMOC chemistry may provide an opportunity for such orthogonality, and also avoids the problem of depurination occasionally observed during repeated acidolytic steps required for 5′-O-detritylation. Compared to solid-phase peptide synthesis, functional protection is less demanding in oligonucleotide synthesis due to the smaller number of building blocks. In the deoxyribonucleotide series, very satisfactory protecting groups have resulted from the pioneering efforts in the chemical synthesis of genes by the group of Khorana (e.g., Schaller et al., 1963; Büchi and Khorana, 1972). For example, permanent protection of the exocyclic amino functions as N-acyl and N-aroyl derivatives, respectively, is compatible with nonnucleophilic conditions, but the protecting groups may be cleaved at the very end of the synthesis by hydrolysis in aqueous ammonium hydroxide at elevated temperature. On the other hand, an adequate treatment of the protecting group strategy in RNA synthesis requires the differential handling of the 2′-, 3′-, and 5′-hydroxy groups. If phosphoramidite chemistry is chosen for chain extension, the central problem is the selection of a suitable combination of the permanent 2′-hydroxylprotecting group together with a temporary 5′-hydroxyl-protecting group. The former must remain intact until the very end of the synthesis, including base deprotection, whereafter it must be removed very cleanly and under conditions that leave the phosphate-ribose backbone in complete constitutional and regiochemical integrity. This was convincingly shown both for the acetal (tetrahydropyranyl or THP) as well as the presently applied 4-methoxytetrahydropyranyl (MTHP) ketal-type of 2′-hydroxylprotecting groups some years ago by the research group of Reese (Norman et al., 1984). It was demonstrated by HPLC analysis that diuridine and diadenosine phosphates are stable in 0.01 N hydrochloric acid for time periods over ten times longer than that required for complete 2′-deprotection. Moreover, the authors (Christodoulou et al., 1986) and others (Reese and Skone, 1985) showed that if a conventional acid-labile group such as di-O-(p-an-

isyl)phenylmethyl (dimethoxytrityl or DMTr) or 9-phenylxanthen-9-yl (pixyl or Px) is used to protect the 5′ position, concomitant cleavage of the 2′-O-THP or 2′-O-MTHP takes place to an unacceptable extent. For orthogonal protection of the 5′-hydroxy group, the authors therefore chose the base-labile FMOC protecting group (Fig. 2.4.1 and Fig. 2.4.7). FMOC was first proposed for 5′-O-protection in the deoxyribonucleotide series for solidphase synthesis of an octathymidylic acid fragment (Gioeli and Chattopadhyaya, 1982), and was later used in the same laboratory for the regioselective construction of a 2′-O-pixylprotected dinucleotide (Pathak and Chattopadhyaya, 1985). Later, the results reported for oligothymidylic acids were extended to deoxyribonucleotide sequences containing 4-Nbenzoyl-2′-deoxycytidine and 6-N-benzoyl-2′deoxyadenosine, whereby the above procedure for the regioselective introduction of the FMOC group was again confirmed (Ma and Sonveaux, 1987; Balgobin and Chattopadhyaya, 1987). An approach for a solution-phase RNA synthesis combining an acid-labile acetal group (1ethoxyethyl) for 2′-protection with 5′-OFMOC was proposed by Fukuda et al. (1988). However, the drawback of using chiral protecting groups on chiral nucleoside building blocks has been commented on with crystallographic data (Lehmann et al., 1991). Further efforts to incorporate the FMOC strategy within the framework of solid-phase

B = U; CBz; ABz; Gi-Bu

O [Fmoc] temporary base-labile 5'-O-protection

O O RO

B

5'

O 3'

2'

O O P N(i-Pr)2

OCH3

O [Mthp]

permanent acid-labile 2'-O-protection

phosphate protection R = CH2CH2CN R = CH3 R = CH2CHCH2

Figure 2.4.7 Orthogonal protection scheme for solid-phase RNA synthesis applying a temporary base-labile (FMOC) 5′-O-protecting group and a permanent acid-labile (MTHP) 2′O-protecting group within the phosphoramidite strategy. Compatibility with current variants of phoshate internucleotide protection as well as anchorage to the solid support are discussed in the text.




oligonucleotide chemistry were aimed at avoiding unwanted side reactions during deprotection. Primarily, it was reported (Gao et al., 1985) that, under strongly basic deprotection conditions (5% 1,8-diazabicyclo-[5.4.0]undec7-ene [DBU] in acetonitrile), methyl group transfer to the N3 position of thymine (and presumably uracil) occurs via an intramolecular SN2-reaction from the phosphotriester linkage if the common methyl phosphoramidite chemistry is applied. It was not possible to use H-phosphonates, since the latter are decomposed immediately under the action of base (Lehmann et al., 1989). The authors therefore first applied cyanoethyl phosphoramidites, presently in use for DNA synthesis, conscious of the fact that probably some or all of the phosphate protection is lost during FMOC cleavage. Nevertheless, the results were remarkably encouraging, and it was possible to assemble oligomers of up to 20 residues containing all four ribonucleosides in good yield and in regiochemically homogenous form. The fact that phosphoramidite building blocks are of a significantly basic nature and are prone to hydrolysis to the corresponding H-phosphonates on normal silica gel surfaces demands deactivation of such surfaces by the addition of base to the eluent. If the kind of base (which should be nonnucleophilic and of a low pKa) is chosen inadequately or the 5′-OFMOC-nucleoside-3′-O-phosphoramidite is exposed to higher concentrations of base at elevated temperatures, some loss of the protecting group may occur during isolation, which may lead to a danger of double couplings. The authors therefore recommend (Lehmann et al., 1989) repeated coevaporation of the derived phosphoramidite fractions from toluene under high vacuum followed by immediate precipitation of the product. An evaluation of the less basic N-morpholino phosphoramidite derivatives (McBride and Caruthers, 1983) may be advantageous in the present context. From the more recent literature, two improvements to the FMOC strategy emerge as particularly noteworthy. First, it was possible to apply a safer protection of the phosphate moiety by use of the allyl group as described for a variety of solid-phase protocols by Hayakawa et al. (1990). Moreover it was found (Bergmann et al., 1995) that the final deprotection conditions normally used for removal of the base amino protecting groups (concentrated ammonia, several hours at 55°C) simultaneously take off the allyl groups on the internucleotide phosphate linkage. A second improve-

ment consists of the development of a different linker unit to the solid support, which is of the acetal type and hence is acid labile (2′,3′-Omethylidene-4-phenoxyacetyl; Palom et al., 1993). This allows cleavage of the FMOC group under nucleophilic conditions (10% piperidine in N,N-dimethylformamide) that are significantly less basic and are therefore compatible with an oligonucleotide assembly via methyl phosphoramidites. In addition, two recently introduced alternatives to the FMOC group (BSMOC and MSPOC; Carpino and Mansour, 1999), which are cleaved with lower concentrations of weak nucleophilic bases such as piperidine or morpholine, may be particularly well suited for this oligonucleotide assembly scheme. The linker arm to the support previously used by the authors (succinyl-sarcosyl; Brown et al., 1989) is stable under strongly basic conditions (up to 5% DBU in CH3CN), but is cleaved by exposure to nucleophiles. Finally, the authors would like to encourage the exploitation of new chemistries potentially compatible with the FMOC strategy. For instance, it might be possible to combine the photolabile o-nitrobenzyloxymethyl group (Schwartz et al., 1992; Pitsch, 1997) with the base-labile FMOC group to further increase the scope of this remarkably versatile tool in the chemical synthesis of biooligomers.

Critical Parameters and Troubleshooting The synthesis outlined in the Basic Protocol is fairly short and straightforward. Other than careful attention to basic organic synthesis techniques, little troubleshooting advice needs to be offered. Critical to achieving the expected yields is strict adherence to the given reaction and isolation conditions. The syntheses of the starting materials are considerably more involved, although the expected yields are quite good. Success is dependent upon careful operational planning inherent to multistep syntheses.

Anticipated Results Following the synthesis strategy outlined in the Basic Protocol, yields of between 52% and 65% can be achieved for ribonucleosides protected at the 5′ hydroxyl by FMOC, and 76% can be achieved for 5′-O-FMOC-2′-deoxythymidine. Similarly, good to excellent yields are reported in the literature for deoxyguanosine, deoxyadenosine, and deoxycytidine derivatives (Balgobin and Chattopadhyaya, 1987; Ma and Sonveaux, 1987). The four Support


Protocols for the preparation of the starting materials (2′-O-MTHP-protected ribonucleosides) provide yields of ∼70%, except for the synthesis of 2′-O-MTHP-2-N-isobutyrylguanosine, which provide a yield of ∼52%.

Time Considerations The procedure described in the Basic Protocol may be carried out within 1 to 2 days starting from 2′-O-MTHP-protected ribonucleosides. The preparation of the MTHP-protected starting intermediates can take 2 to 3 weeks per nucleoside derivative. The authors suggest starting with uridine (the simplest case) and allowing time for crystallization of the 3′,5′diacetyl intermediate.

Literature Cited Atherton, E. and Sheppard, R.C. 1989. Solid-Phase Peptide Synthesis: A Practical Approach. IRL Press, Oxford. Balgobin, N. and Chattopadhyaya, J.B. 1987. Solid phase synthesis of DNA under a non-depurinating condition with a base labile 5′-protecting group (Fmoc) using phosphite-amidite approach. Nucleosides Nucleotides 6:461-463. Bergmann, F., Kueng, E., Iaiza, P., and Bannwarth, W. 1995. Allyl as internucleotide protecting group in DNA synthesis to be cleaved off by ammonia. Tetrahedron 51:6971-6976. Brown, T., Pritchard, C.E., Turner, G., and Salisbury, S.A. 1989. A new base-stable linker for solidphase oligonucleotide synthesis. J. Chem. Soc., Chem. Commun. 891-893. Büchi, H. and Khorana, H.G. 1972. Total synthesis of the structural gene for an alanine transfer ribonucleic acid from yeast. Chemical synthesis of an icosadeoxyribonucleotide corresponding to the nucleotide sequence 31 to 50. J. Mol. Biol. 72:251-288. Carpino, L.A. 1987. The 9-fluorenylmethoxycarbonyl family of base-sensitive amino-protecting groups. Acc. Chem. Res. 20:401-407.

Fukuda, T., Hamana, T., and Marumoto, R. 1988. Synthesis of RNA oligomers using 9-fluorenylmethoxycarbonyl (Fmoc) group for 5′-hydroxyl protection. Nucl. Acids Res. 16:13-16. Gao, X., Gaffney, B.L., Senior, M., Riddle, R.R., and Jones, R.A. 1985. Methylation of thymine residues during oligonucleotide synthesis. Nucl. Acids Res. 13:573-584. Gioeli, C. and Chattopadhyaya, J.B. 1982. The fluoren-9-ylmethoxycarbonyl group for the protection of hydroxy-groups; its application in the synthesis of an octathymidylic acid fragment. J. Chem. Soc. Chem. Commun. 1982:672-674. Hayakawa, Y., Uchiyama, M., Kato, H., and Noyori, R. 1990. Allylic protection of internucleotide linkage. Tetrahedron Lett. 26:6505-6508. Hunt, B.J. and Ribgy, W. 1967. Short column chromatography. Chem. Ind. 1967:1868-1869. Jones, R.A. 1984. Preparation of protected deoxyribonucleotides. In Oligonucleotide Synthesis: A Practical Approach (M.J. Gait, ed.) pp. 23-34. IRL Press, Oxford. Jork, H., Funk, W., Fischer, W., and Wimmer, H. 1990. Thin-Layer Chromatography, Vol. Ia: Reagents and Detection Methods, pp. 195-198. VCH-Verlagsgesellschaft, Weinheim. Lehmann, C., Xu, Y.-Z., Christodoulou, C., Tan, Z.-K., and Gait, M.J. 1989. Solid-phase synthesis of oligoribonucleotides using 9-fluorenylmethoxy-carbonyl (Fmoc) for 5′-hydroxyl protection. Nucl. Acids Res. 17:2379-2390. Lehmann, C., Xu, Y.-Z., Christodoulou, C., Gait, M.J., Van Meervelt, L., Moore, M., and Kennard, O. 1991. 3′/5′-Regioselectivity of introduction of the 9-fluorenylmethoxy-carbonyl group to 2′O-tetrahydropyran-2-yl and 2′-O-(4methoxytetrahydropyran-4-yl)-nucleosides: Useful intermediates for solid-phase RNA synthesis. Nucleosides Nucleotides 10:1599-1614. Ma, Y. and Sonveaux, E. 1987. The 9-fluorenylmethyloxycarbonyl (Fmoc) group as a 5′-O base labile protecting group in solid supported oligonucleotide synthesis. Nucleosides Nucleotides 6:491-493.

Carpino, L.A. and Han, G.Y. 1970. the 9-fluorenylmethoxycarbonyl function, a new base-sensitive amino-protecting group. J. Am. Chem. Soc. 92:5748-5749.

Markiewicz, W.T. and Wiewerowski, M. 1985. Simultaneous protection of 3′- and 5′-hydroxyl groups of nucleosides. In Nucleic Acid Chemistry, Section III: Nucleosides (L.B. Townsend and R.S. Tipson, eds.) pp. 229-231. John Wiley & Sons, New York.

Carpino, L.A. and Mansour, E.M.E. 1999. The 2methylsulfonyl-3-phenyl-1-prop-2-enyloxycarbonyl (MSPOC) amino-protecting group. J. Org. Chem. 64:8399-8401.

McBride, L.J. and Caruthers, M.H. 1983. An investigation of several deoxynucleoside phosphoramidites useful for synthesizing deoxyoligonucleotides. Tetrahedron Lett. 24:245-248.

Christodoulou, C., Agrawal, S., and Gait, M.J. 1986. Incompatibility of acid-labile 2′ and 5′ protecting groups for solid-phase synthesis of oligoribonucleotides. Tetrahedron Lett. 27:1521-1522.

McLaughlin, L.W., Piel, N., and Hellmann, T. 1985. Preparation of protected ribonucleotides suitable for chemical oligoribonucleotide synthesis. Synthesis 1985:322-323.

Fromageot, W.P.M., Griffin, B.E., Reese, C.B., and Sulston, J.E. 1967. Monoacylation of ribonucleosides and derivatives via orthoester exchange. Tetrahedron 23:2315-2331.

Norman, D.G., Reese, C.B., and Serafinowska, H.T. 1984. The protection of 2′-hydroxy functions in oligoribonucleotide synthesis. Tetrahedron Lett. 25:3015-3018.



Palom, Y., Alazzouzi, E-M., Gordillo, F., Grandas, A., and Pedroso, E. 1993. An acid-labile linker for solid-phase oligoribonucleotide synthesis using Fmoc group for 5′-hydroxyl protection. Tetrahedron Lett. 34:2195-2198. Pathak, T. and Chattopadhyaya, J. 1985. The 2′-hydroxyl function assisted cleavage of the internucleotide phosphotriester bond of a ribonucleotide under acidic conditions. Acta Chem. Scand. B39:799-806.

van Boom, J.H. and Wreesmann, C.T.J. 1984. Chemical synthesis of small oligoribonucleotides in solution. In Oligonucleotide Synthesis: A Practical Approach (M.J. Gait, ed.) pp. 153-183. IRL Press, Oxford.

Key-References Blackburn, M. and Gait, M.J. (eds.) 1996. Nucleic Acids in Chemistry and Biology. Oxford University Press, New York.

Pitsch, S. 1997. An efficient synthesis of enantiomeric ribonucleic acid from D-glucose. Helv. Chim. Acta 80:2286-2314.

In particular, chapter 3 on chemical synthesis is very recommendable as an illustrative overview to the present topics.

Reese, C.B. and Skone, P.A. 1985. Action of acid on oligoribonucleotide phosphotriester intermediates. Effect of released vicinal hydroxy functions. Nucl. Acids Res. 13:5215-5231.

Gait, M.J. (ed.) 1984. Oligonucleotide Synthesis: A Practical Approach, IRL Press, Oxford.

Reese, C.B., Saffhill, R., and Sulston, J. 1970. 4Methoxytetrahydropyran-4-yl: A symmetrical alternative to the tetrahydropyranyl group. Tetrahedron 26:1023-1030. Schaller, H., Weiman, G., Lerch, B., and Khorana, H.G. 1963. Protected derivatives of deoxyribonucleotides and new syntheses of deoxyribonucleotide-3′ phosphates. J. Am. Chem. Soc. 85:3821-3827. Schwartz, M.E., Breaker, R.R., Asteriadis, G.T., deBear, J.S., and Gough, G.R. 1992. Rapid synthesis of oligoribonucleotides using 2′-O-(o-nitrobenzyloxymethyl)-protected monomers. Bioorg. Med. Chem. Lett. 2:1019-1024. Still, W.C., Kahn, M., and Mitra, A. 1978. Rapid chromatographic technique for preparative separations with moderate resolution. J. Org. Chem. 43:2923-2925. Ti, G.S., Gaffney, B.L., and Jones, R.A. 1982. Transient protection: Efficient one-flask synthesis of protected deoxynucleosides. J. Am. Chem. Soc. 104:1316-1319.

Basic principles of oligonucleotide synthesis are illustrated by practical advice through further stepby-step protocols; some of the chapters may be regarded as primers for units in the present volume. Lehmann et al., 1989. See above. The procedure in the Basic Protocol is first described for the four 2′-O-MTHP-protected ribonucleosides. For more background information on particular complications arising when the chiral 2′-O-tetrahydropyranyl (THP) group is used, see Lehmann et al., 1991.

Contributed by Michael J. Gait MRC Laboratory of Molecular Biology Cambridge, United Kingdom Christian Lehmann Institute of Organic Chemistry University of Lausanne Lausanne, Switzerland



2′-Hydroxyl-Protecting Groups That Are Either Photochemically Labile or Sensitive to Fluoride Ions

UNIT 2.5

Automated chemical synthesis of RNA, like that of DNA, uses protected nucleotide monomers for construction of oligonucleotide chains. RNA monomers, however, are more difficult to obtain than their deoxy counterparts, because of the necessity of protecting their 2′-hydroxyl functions. This unit describes the stepwise preparation, starting from uridine, cytidine, adenosine, and guanosine, of some suitably 2′-protected ribonucleosides. In addition, details are given for protecting the 5′-hydroxyl and the nucleobase, affording ribonucleosides that can be easily converted into either phosphoramidite or phosphonate derivatives, ready to be used in a synthesizer for making RNA. Two alternative sets of protected ribonucleosides are represented here. They are distinguished by the differing conditions required for removal of their 2′-protecting groups. The first consists of uridine, cytidine, adenosine, and guanosine derivatives carrying 2-nitrobenzyloxymethyl (NBOM) groups on their 2′-hydroxyls. The synthesis of these four ribonucleosides is described in Basic Protocols 1 to 4. Oligoribonucleotides synthesized from these components are deprotected by exposure to long-wave UV light (UNIT 3.7). The second set of nucleosides has its 2′-hydroxyls protected with tert-butyldimethylsilyl (TBDMS) groups; these can be removed from product oligoribonucleotides by treatment with tetra-n-butylammonium fluoride. Their syntheses are described in Alternate Protocols 1 to 4. CAUTION: The syntheses in these protocols involve the use of chemicals that have various degrees and kinds of toxicity. Avoid skin contact and inhalation of dusts or vapors. Most operations should be carried out in a well-vented fume hood. NOTE: These reactions should be carried out under strictly anhydrous conditions, using anhydrous solvents and reagents. All glassware should be dried in an oven prior to use. All connections to atmospheric pressure should be through a drying tower containing a desiccant. NOTE: For general information regarding thin-layer chromatography (TLC) or column chromatography, see APPENDIX 3D and APPENDIX 3E, respectively. PREPARATION OF N-PROTECTED 5′-O-(4,4′-DIMETHOXYTRITYL)2′-O-(2-NITROBENZYLOXYMETHYL) NUCLEOSIDES These protocols describe the preparation of ribonucleoside derivatives that incorporate 2-nitrobenzyloxymethyl groups for protection of their 2′-hydroxyls (Figure 2.5.1). In all cases, the 2-nitrobenzyloxymethyl group is introduced into the nucleosides by means of an alkylation reaction utilizing the reagent 2-nitrobenzyl chloromethyl ether, which is prepared as required from its precursor, 2-nitrobenzyl methylthiomethyl ether (see Support Protocol 1). Before any of the nucleoside syntheses described below are carried out, an adequate stock of the methylthiomethyl ether should be accumulated. In addition to the synthesis details, data pertaining to the Rf values from TLC (Table 2.5.1), proton nuclear magnetic resonance (1H NMR; Tables 2.5.2 and 2.5.3), and 13C NMR (Table 2.5.4) of the various 2-nitrobenzyloxymethyl-protected nucleosides are shown.

Contributed by Tod J. Miller, Miriam E. Schwartz, and Geoffrey R. Gough Current Protocols in Nucleic Acid Chemistry (2000) 2.5.1-2.5.36 Copyright © 2000 by John Wiley & Sons, Inc.


2.5.1 Supplement 3

CH3O DMTrO

B

O 3'

2'

DMTr = OH

O

C

O NO2 CH3O O

O NH

B= N

O

HN

O

Uracil-1-yl

HN N

N N

N

O

N4-benzoylcytosin-1-yl

O N

N

N

N

N6-benzoyladenin-9-yl

NH N

O

N H

N2-isobutyrylguanin-9-yl

Figure 2.5.1 The four 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl) ribonucleosides. The general structure of these ribonucleosides is in the upper-left corner and the four bases (B) are shown below. DMTr is the 4,4′-dimethoxytrityl group.

These data should prove useful in confirming the identities of the products and their synthetic intermediates. NOTE: 2-Nitrobenzyloxymethyl groups are designed to be removable by irradiation with long-wave UV light. Sunlight, or even the light emitted by standard overhead fluorescent bulbs, contains enough UV light to cause slow loss of these protecting groups. Minimize exposure of sensitive compounds by carrying out operations in an area, such as a fume hood, fitted with yellow fluorescent tubes (GE or Sylvania Golds). As an added precaution, wrap flasks and chromatography columns in aluminum foil.

Table 2.5.1 Rf Values of 2-Nitrobenzyloxymethyl-Protected Ribonucleosides on Merck Silica Gel 60 F254 TLC Plates

2′-OH-Protecting Groups That Are Photochemically Labile or Sensitive to Fluoride Ions

Ribonucleosidea

Rf valueb

Solvent system

U2′NBOM U3′NBOM A2′NBOM A3′NBOM ABz2′NBOM Gi-Bu2′NBOM 5′DMTrU2′NBOM 5′DMTrU3′NBOM 5′DMTrC2′NBOM 5′DMTrCBz2′NBOM 5′DMTrABz2′NBOM 5′DMTrGi-Bu2′NBOM

0.31 0.31 0.42 0.33 0.56 0.24 0.51 0.29 0.05 0.42 0.56 0.38

9:1 (v/v) chloroform/methanol 9:1 (v/v) chloroform/methanol 9:1 (v/v) chloroform/methanol 9:1 (v/v) chloroform/methanol 9:1 (v/v) chloroform/methanol 9:1 (v/v) chloroform/methanol 2:1 (v/v) ethyl acetate/hexane 2:1 (v/v) ethyl acetate/hexane 2:1 (v/v) ethyl acetate/hexane 2:1 (v/v) ethyl acetate/hexane 2:1 (v/v) ethyl acetate/hexane 2:1 (v/v) ethyl acetate/hexane

a Abbreviations: A, adenosine; C, cytidine; G, guanosine; U, uridine; Bz, benzoyl; DMTr, 4,4′-dimethoxytrityl; i-Bu, isobutyryl; NBOM, 2-nitrobenzyloxymethyl. b TLC Rf values are notoriously irreproducible unless measured under strictly controlled conditions. However, the relative mobilities of substances run on the same plate in the same solvent are valuable guides to compound identification.

2.5.2 Supplement 3


Table 2.5.2 Derivativesb

1H-NMR

Chemical Shiftsa of Nontritylated 2-Nitrobenzyloxymethyl Ribonucleoside

Hydrogen(s)

A2′NBOM

A3′NBOM

ABz2′NBOM

Gi-Bu2′NBOM

H1′ H2′ H3′ H4′ H5′ H2 or H5 H6 or H8 -CH2-O -CH2-Ar Ar(NBOM) 2′- or 3′-OH 5′-OH N6-NH2 Ar (Bz)

6.07 d 4.87 m 4.34 bs 4.03 bs 3.67 m 8.02 s 8.35 s 4.75 s 4.56 s 7.3-8.0 m 5.39 d 5.53 bs 7.29 bs NA

5.96 d 4.85 m 4.34 bs 4.12 bs 3.65 m 8.20 s 8.42 s 5.04 s 5.01 s 7.6-8.1 m 5.71 d 5.63 bs 7.46 bs NA

6.12 d 5.26 m 4.59 d 4.40 bs 3.98 d, 3.78 d 8.60 s 8.69 s 4.73 d 4.90 m 7.4-8.1 m ND ND 7.26 bs 7.3-8.1 m

5.97 d 5.16 m 4.30 bs 3.97 bs 3.61 d 8.27 s ND 4.87 m 4.71 bs 7.2-8.1 m ND ND NA NA

The internal reference for 1H-NMR spectra was tetramethylsilane at 0 ppm. The solvents were dimethy sulfoxide-d6 for the A derivatives and acetonitrile-d3 for the G derivative. All shifts are measured in ppm. b Abbreviations: A, adenosine; G, guanosine; Ar, aromatic; Bz, benzoyl; i-Bu, isobutyryl; NBOM, 2-nitrobenzyloxymethyl; bs, broad singlet; d, doublet; m, multiplet; s, singlet; NA, not applicable; ND, not determined. a

Table 2.5.3 1H-NMR Chemical Shiftsa of 5′-Dimethoxytritylated 2-Nitrobenzyloxymethyl Ribonucleoside Derivativesb

Hydrogen(s)

5′DMTrU2′ NBOM

5′DMTrC2′ NBOM

5′DMTrCBz2′ 5′DMTrABz2′ 5′DMTrGi-Bu NBOM NBOM 2′NBOM

H1′ H2′ H3′ H4′ H5′ H2 or H5 H6 or H8 -OCH3 Ar (DMTr) -CH2-O-CH2-Ar Ar (NBOM) 2′- or 3′-OH N1- or N3-H N4- or N6-H Ar (Bz) i-Bu-CH(CH3)2

6.06 d 4.40 dd 4.51 m 4.11 bs 3.56 bs 5.21 d 7.96 d 3.82 s 6.80-7.40 m 5.00-5.15 bs 5.00-5.15 bs 7.40-8.41 m 5.19 bs 10.00 s NA NA NA

5.87 d 4.15 d 4.26 m 4.00 d 3.28 bs 5.44 d 7.79 d 3.73 s 6.87-7.40 m 4.85-5.05 m 4.85-5.05 m 7.57-8.08 m 5.30 d NA ND NA NA

6.10 bs 4.58 m 4.43 d 4.19 d 3.64 bs 5.46 d 7.93 d 3.86 s 6.80-7.50 m 5.12 m 5.12 m 7.50-8.20 m ND NA ND 7.10-8.70 m NA

5.99 d 5.26 m 4.78 d 4.40 bs 3.91 m 8.72 s 8.68 bs 3.80 s 6.75-7.40 m 4.60 m 4.83 m 7.40-8.10 m ND NA 10.00 s 7.10-8.10 m NA

5.99 d 5.20 m 4.60 dd 4.26 d 3.48 d, 3.21 dd 8.65 bs ND 3.79 s 6.70-7.40 m 4.81 m 4.90 m 7.40-8.70 m ND 9.28 s NA NA 0.92 dd

a The internal reference for 1H-NMR spectra was tetramethylsilane at 0 ppm. The solvent was chloroform-d. All shifts are measured in ppm. b Abbreviations: A, adenosine; C, cytidine; G, guanosine; U, uridine; Ar, aromatic; Bz, benzoyl; DMTr, 4,4′-dimethoxytrityl; i-Bu, isobutyryl; NBOM, 2-nitrobenzyloxymethyl; bs, broad singlet; d, doublet; dd, doublet of doublets; m, multiplet; s, singlet; NA, not applicable; ND, not determined.



Supplement 3

Table 2.5.4

13C-NMR

Chemical Shiftsa of 2-Nitrobenzyloxymethyl Ribonucleoside Derivativesb

Carbon

A2′NBOM

U2′NBOM

Gi-Bu2′NBOM

C1′ C2′ C3′ C4′ C5′ C2 C4 C5 C6 C8 -OCH2-OCH2-Ar NBOM (C1) NBOM (C2) NBOM (C3) NBOM (C4) NBOM (C5) NBOM (C6) i-Bu (CH3) i-Bu (CH3) i-Bu (CH) i-Bu C=O

86.5 78.3 69.3 86.2 61.5 152.2 148.7 119.2 156.0 139.7 65.6 94.0 133.6 146.6 133.3 128.1 124.3 128.3 NA NA NA NA

86.4 78.5 68.7 85.3 60.6 150.2 163.1 102.0 140.4 NA 65.9 94.1 133.9 147.1 133.7 128.6 124.6 128.8 NA NA NA NA

86.7 81.0 70.3 85.0 61.8 149.0 148.4 120.3 154.9 137.7 66.2 95.2 134.3 146.5 134.0 127.9 124.7 128.7 19.1 19.3 35.1 180.4

a The internal reference for 13C-NMR spectra was dimethyl sulfoxide-d6 at 39.7 ppm. The solvent was dimethyl sulfoxide-d6. While not exhaustive, these useful and commonly observed resonances are listed with tentative assignments. All shifts are measured in ppm. b Abbreviations: A, adenosine; G, guanosine; U, uridine; Ar, aromatic; i-Bu, isobutyryl; NBOM, 2-nitrobenzyloxymethyl; NA, not applicable.

BASIC PROTOCOL 1


Synthesis of 5′-O-(4,4′-Dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)uridine In this protocol, uridine is converted into its dibutylstannylene derivative, then alkylated using 2-nitrobenzyl chloromethyl ether. The resulting mixture of 2′- and 3′-O-(2-nitrobenzyloxymethyl)uridine is treated with 4,4′-dimethoxytrityl chloride. The dimethoxytrityl derivatives are separated by silica-gel chromatography, affording pure 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)uridine. The overall reaction sequence is illustrated in Figure 2.5.2. Materials Methanol Uridine Dibutyltin oxide Phosphorus pentoxide Anhydrous tetra-n-butylammonium bromide Anhydrous dimethylformamide, store over 4A molecular sieves 2-Nitrobenzyl chloromethyl ether, make fresh (see Support Protocol 1) Anhydrous pyridine, store over coarse granules of calcium hydride 9:1 and 95:5 (v/v) chloroform/methanol 66% (v/v) aqueous pyridine Silica gel 60, 70 to 230 mesh ASTM (e.g., EM Science)

2.5.4 Supplement 3


O

O

NH

NH HO

N

O

O

HO MeOH

OH

OH

N

O

Dibutyltin oxide O

O 2-Nitrobenzyl chloromethyl ether Bu4NBr/DMF

O Sn

Bu

Bu

O

O NH

HO

N

O OH

NH

O

DMTrO 1. DMTrCl/Pyridine

O

2. Separation of isomers

O NO2

+

N

O OH

O

O

O NO2

3'-isomer

Figure 2.5.2 Preparation of 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)uridine from uridine. Bu is n-butyl.

2.5 × 30–cm and 3 × 60–cm glass chromatography columns packed with silica gel 60, 70 to 230 mesh ASTM, in chloroform to a bed height of 25 and 50 cm, respectively (see Support Protocol 3) Chloroform Anhydrous triethylamine, store over coarse granules of calcium hydride 4,4′-Dimethoxytrityl chloride Ethyl acetate 1 M NaHCO3 2 M NaCl Anhydrous sodium sulfate 2:1 (v/v) ethyl acetate/hexane 0% to 1% (v/v) methanol in chloroform, containing 0.25% (v/v) pyridine 250-mL and 2-L round-bottom flasks Boiling chips Water-cooled reflux condenser Heating mantle Rotary evaporator connected interchangeably to water aspirator and vacuum pump 50-mL pressure-equalizing dropping funnel Large petri plate Filter paper, coarse porosity and fast flow rate (e.g., Quantitative Q8, Fisher) Additional reagents and equipment for TLC and column chromatography (see Support Protocol 3) Prepare 2′,3′-O-(dibutylstannylene)uridine 1. Add a stir-bar to 1 L methanol in a 2-L round-bottom flask and begin stirring. Add 4.88 g (20 mmol) uridine and 5.00 g (20 mmol) dibutyltin oxide.



Supplement 3

2. Remove stir-bar, add boiling chips, and fit flask with a water-cooled reflux condenser. Heat mixture to a boil using a heating mantle, reflux 45 min to 1 hr, and let cool to room temperature. 3. Decant solution from boiling chips into a second 2-L flask and remove methanol using a rotary evaporator connected to a water aspirator. 4. Dry crystalline residue to constant weight in a vacuum desiccator over phosphorus pentoxide. A yield of 9 g 2′,3′-O-(dibutylstannylene)uridine (with a melting point of 232° to 234°C) is expected. If desired, this intermediate can be stored at –20°C in a sealed, dry container (stable for many months).

Alkylate to form 2′- and 3′-O-(2-nitrobenzyloxymethyl)uridine 5. Add 5.10 g (10.7 mmol) dry 2′,3′-O-(dibutylstannylene)uridine and 1.72 g (5.35 mmol) anhydrous tetra-n-butylammonium bromide to 40 mL anhydrous dimethylformamide in a 250-mL round-bottom flask, while stirring. Fit flask with a 50-mL pressure-equalizing dropping funnel. 6. Use dropping funnel to add a freshly prepared solution of 2-nitrobenzyl chloromethyl ether (from 16 mmol 2-nitrobenzyl methylthiomethyl ether; see Support Protocol 1) dropwise over a 5-min period while stirring, then seal flask and stir 2 hr at room temperature. 7. Add 5 mL anhydrous pyridine followed by 2 mL water and continue stirring 20 min at room temperature. 8. Check reaction by TLC (see Support Protocol 3) using 9:1 chloroform/methanol. The unresolved 1:1 mixture of 2′- and 3′-isomers will be visible under UV light as a dark band about a third of the way up the plate (Table 2.5.1).

9. Remove volatile solvents using the rotary evaporator connected to water aspirator, then continue evaporation under high vacuum (i.e., with a vacuum pump) to remove as much of the dimethylformamide as possible. If desired, this intermediate can be stored at –20°C in a sealed, dry container (stable for many months).

Purify 2′- and 3′-O-(2-nitrobenzyloxymethyl)uridine 10. Dissolve residue in the minimum volume of 66% aqueous pyridine and, using a Pasteur pipet, transfer it dropwise onto 10 g silica gel 60 in a large petri plate. Allow wet silica to dry in the moving airstream inside a fume hood. An overnight drying time should be adequate if most of the dimethylformamide was removed and the minimum amount of aqueous pyridine was used.

11. Pulverize the dry, coated silica. Divide the powdery material into halves and layer them onto two 2.5 × 30–cm glass chromatography columns. 12. Wash each column with 100 mL chloroform, elute with 95:5 chloroform/methanol, and collect fractions as described (see Support Protocol 3). 2′-OH-Protecting Groups That Are Photochemically Labile or Sensitive to Fluoride Ions

The volume of 95:5 chloroform/methanol should be determined from TLC results.

13. Monitor fractions by TLC using 9:1 chloroform/methanol, comparing selected fractions against samples of the original reaction mixture to locate the products.

2.5.6 Supplement 3


14. Pool and evaporate the appropriate fractions using the rotary evaporator connected to a water aspirator. A yield of 4.2 g of a 1:1 mixture of the 2′- and 3′-isomers of (2-nitrobenzyloxymethyl)uridine is expected. If desired, this intermediate can be stored at –20°C in a sealed, dry container (stable for many months).

Prepare 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)uridine 15. Dry 4.2 g (10.3 mmol) mixed 2′- and 3′-O-(2-nitrobenzyloxymethyl)uridine isomers by three co-evaporations with 30 mL anhydrous pyridine using the rotary evaporator connected to a vacuum pump. 16. Dissolve residue in 114 mL anhydrous pyridine, then add 1.60 mL (11.4 mmol) anhydrous triethylamine and 3.87 g (11.4 mmol) 4,4′-dimethoxytrityl chloride. Stir 3 hr at room temperature. 17. Add 5 mL methanol, allow to stand 10 min, then remove solvents on the rotary evaporator connected to a vacuum pump. 18. Dissolve residue in 250 mL ethyl acetate. Extract three times with 100 mL of 1 M NaHCO3 and once with 100 mL of 2 M NaCl. CAUTION: Carbon dioxide is released during the sodium bicarbonate extractions.

19. Dry ethyl acetate layer by addition of 5 g anhydrous sodium sulfate, filter through filter paper to remove salt, and rotary evaporate filtrate to a yellow foam using a water aspirator. 20. Analyze this material by TLC in 2:1 ethyl acetate/hexane. Two major bands should be visible. The upper (faster-running) one is the 2′-isomer (Table 2.5.1). If desired, this intermediate can be stored at –20°C in a sealed, dry container (stable for many months).

Purify 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)uridine 21. Dissolve foam (∼8 g) in 5 to 10 mL chloroform and add it to the top of a 3 × 60–cm silica gel column. Elute products from column using a 1.2-L stepwise gradient of 0% to 1% methanol in chloroform containing 0.25% pyridine as described. 22. Monitor fractions by TLC using 2:1 ethyl acetate/hexane. Any mixed fractions containing both isomers can be rechromatographed to obtain more pure 2′-isomer. Gradients of ethyl acetate in hexane may also be used for purification of these dimethoxytrityl nucleosides, as preliminary results with this alternative solvent mixture have been encouraging.

23 Combine those fractions containing the pure 2′-isomer, and rotary evaporate them to a foam using a water aspirator. A yield of 40% 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)uridine, based on the starting quantity of uridine, is expected.

24. Store protected nucleoside in the dark at −20°C in a sealed flask containing a trace of pyridine (stable for many months).



Supplement 3

BASIC PROTOCOL 2

Synthesis of 5′-O-(4,4′-Dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)-N4benzoylcytidine In this protocol, 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)uridine (see Basic Protocol 1) is converted into its corresponding cytidine derivative, then selectively benzoylated with pentafluorophenyl benzoate at the N4 position. The reaction sequence is illustrated in Figure 2.5.3. Materials 5′-O-(4,4′-Dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)uridine (see Basic Protocol 1) Anhydrous pyridine, stored over coarse granules of calcium hydride Acetic anhydride Ethyl acetate 1 M NaHCO3 2 M NaCl Anhydrous sodium sulfate 1,2,4-Triazole 4-Chlorophenyl phosphorodichloridate 3:1 (v/v) pyridine/concentrated ammonium hydroxide 2.5 × 30–cm glass chromatography column packed with silica gel 60, 70 to 230 mesh ASTM, in chloroform to a bed height of 25 cm (see Support Protocol 3) 0% to 1% and 0% to 4% (v/v) methanol in chloroform, containing 0.25% (v/v) pyridine 2:1 (v/v) ethyl acetate/hexane Pentafluorophenyl benzoate (see Support Protocol 2)

O

O NH

DMTrO

N

O OH

NH

O

DMTrO

O

O

N

O

Acetic anhydride/Pyridine AcO

O

O

O

NO2

NO2 O O P Cl Cl

1. Cl

2. NH4OH

Triazole/Pyridine O NH2

HN

N

N DMTrO


N

O OH

O

DMTrO Pentafluorophenyl benzoate

O

Pyridine

O NO2

N

O OH

O

O

O NO2

Figure 2.5.3 Preparation of 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)-N4-benzoylcytidine from 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)uridine.

2.5.8 Supplement 3


250- and 500-mL separatory funnels Filter paper, coarse porosity and fast flow rate (e.g., Quantitative Q8, Fisher) Rotary evaporator connected interchangeably to water aspirator and vacuum pump 10- and 25-mL round-bottom flasks Additional reagents and equipment for TLC and column chromatography (see Support Protocol 3) Protect 3′-hydroxyl of 5′-O-(4,4′-dimethoxytrityl)-2′-O(2-nitrobenzyloxymethyl)uridine 1. Add 580 mg (0.82 mmol) 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)uridine to a flask containing 3 mL anhydrous pyridine and stir to dissolve. 2. Add 1.64 mL (17 mmol) acetic anhydride and allow reaction mixture to stand 2 hr at room temperature. 3. Cool flask in an ice bath and add 2.5 mL anhydrous pyridine, followed by dropwise addition over a 5-min period of 4 mL water, while stirring and cooling continuously. 4. Add an additional 1 mL anhydrous pyridine to redissolve any precipitate, and allow solution to stand 18 hr at room temperature in the dark. 5. Transfer mixture to a 250-mL separatory funnel using 50 mL ethyl acetate and extract the organic layer three times with 50 mL of 1 M NaHCO3 and three times with 50 mL of 2 M NaCl. CAUTION: Carbon dioxide is released during the sodium bicarbonate extractions. The 3′-hydroxyl is temporarily protected as an acetate ester.

Prepare 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)cytidine 6. Dry ethyl acetate layer over 3 g anhydrous sodium sulfate, filter through filter paper to remove salt, and evaporate filtrate to a foam using a rotary evaporator connected to a water aspirator. 7. Co-evaporate the product three times with 10 mL anhydrous pyridine in a 25-mL round-bottom flask using the rotary evaporator connected to a vacuum pump. Dissolve final residue in 2.7 mL anhydrous pyridine. 8. Add 314 mg (4.54 mmol) 1,2,4-triazole, stir to dissolve (∼1 hr), then add 0.36 mL (2.2 mmol) 4-chlorophenyl phosphorodichloridate. Stir 1 hr and then let stand 72 hr at room temperature in the dark. 9. Cool flask in an ice bath and add 0.2 mL water while stirring. Stir 30 min at 0°C, then add 3 mL anhydrous pyridine. 10. Use a disposable Pasteur pipet to add the resulting solution dropwise to 80 mL of 3:1 pyridine/concentrated ammonium hydroxide while stirring. Loosely stopper the container and allow reaction mixture to stand 48 hr at room temperature in the dark. 11. Concentrate solution to a slurry by rotary evaporation using a water aspirator and transfer to a 500-mL separatory funnel using 150 mL ethyl acetate. 12. Extract three times with 100 mL of 1 M NaHCO3, then three times with 100 mL of 2 M NaCl. 13. Repeat step 6.



Supplement 3

14. Layer residue on a 2.5 × 30–cm silica gel column and elute products using a 1-L stepwise gradient of 0% to 4% methanol in chloroform containing 0.25% pyridine as described (see Support Protocol 3). 15. Monitor fractions by TLC (see Support Protocol 3) using 2:1 ethyl acetate/hexane. 16. Evaporate appropriate fractions using the rotary evaporator connected to a water aspirator to produce the pure product, 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)cytidine. A 71% yield, based on the starting quantity of 5′-O-(4,4’-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)uridine, is expected. If desired, this intermediate can be stored at –20°C in a sealed, dry container (stable for many months).

Prepare 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)-N4benzoylcytidine 17. Dissolve 412 mg (0.58 mmol) 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)cytidine in 1 mL anhydrous pyridine in a 10-mL round-bottom flask. 18. Add 251 mg (0.87 mmol) pentafluorophenyl benzoate and swirl flask to dissolve. Seal flask and allow solution to stand 6 days at room temperature in the dark. 19. Examine mixture by TLC using 2:1 ethyl acetate/hexane, comparing it to the starting material to confirm complete conversion to the N-benzoyl derivative. 20. Transfer reaction mixture to a 250-mL separatory funnel using 100 mL ethyl acetate, then extract as in step 5. 21. Repeat step 6. 22. Repeat steps 14 to 16, using a 1.2-L stepwise gradient of 0% to 1% methanol in chloroform containing 0.25% pyridine in step 14. A 91% yield based on the starting quantity of 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)cytidine is expected.

23. Store protected nucleoside in the dark at −20°C in a sealed flask containing a trace of pyridine (stable for many months). BASIC PROTOCOL 3

Synthesis of 5′-O-(4,4′-Dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)-N6-benzoyladenosine In this protocol, adenosine is converted into its dibutylstannylene derivative, then alkylated with 2-nitrobenzyl chloromethyl ether, and the resulting 2′- and 3′-O-(2-nitrobenzyloxymethyl)adenosines are separated by column chromatography. Benzoylation at the N6 position (to protect the nucleobase), followed by dimethoxytritylation (to protect the 5′-hydroxyl), completes the synthesis. The reaction sequence is illustrated in Figure 2.5.4.


Materials Methanol Adenosine Dibutyltin oxide Phosphorous pentoxide Anhydrous tetra-n-butylammonium bromide Anhydrous dimethylformamide, stored over 4A molecular sieves 2-Nitrobenzyl chloromethyl ether, make fresh (see Support Protocol 1)

2.5.10 Supplement 3


NH2 N HO

NH2 N

N

N

N

N

HO

O

O

Dibutyltin oxide MeOH

OH OH

N N 2-Nitrobenzyl chloromethyl ether

O

Bu4NBr/DMF

O Sn Bu

Bu

O NH2 N HO

O

HN N

N

N

HO

N 1. Separation of isomers

OH O

2. Me3SiCl/Pyridine 3. Benzoyl chloride 4. NH4OH

O NO2

+

O

N

N

OH O

N

O NO2

3'-isomer

O DMTrCl/Pyridine HN N DMTrO

O OH

N

N

O

N

O NO2

Figure 2.5.4 Preparation of 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)-N6-benzoyladenosine from adenosine. Bu is n-butyl.

Anhydrous pyridine, store over coarse granules of calcium hydride 9:1 and 95:5 (v/v) chloroform/methanol 66% (v/v) aqueous pyridine Silica gel 60, 70 to 230 mesh ASTM (e.g., EM Science) 2.5 × 25–cm, 2.5 × 30–cm, and 3 × 60–cm glass chromatography columns packed with silica gel 60, 70 to 230 mesh ASTM, in chloroform (see Support Protocol 3) 1% to 5% (v/v) methanol in chloroform 80% (v/v) aqueous acetonitrile Acetonitrile Trimethylchlorosilane Benzoyl chloride Concentrated ammonium hydroxide Ethyl acetate 1 M NaHCO3 2 M NaCl Anhydrous sodium sulfate Chloroform Triethylamine, stored over coarse granules of calcium hydride 4,4′-Dimethoxytrityl chloride 0% to 1% (v/v) methanol in chloroform containing 0.25% (v/v) pyridine 2:1 (v/v) ethyl acetate/hexane 100-mL, 250-mL, and 2-L round-bottom flasks Boiling chips Water-cooled reflux condenser Heating mantle



Supplement 3

Rotary evaporator connected interchangeably to water aspirator and vacuum pump Oil bath, 60°C 50-mL pressure-equalizing dropping funnel 500-mL separatory funnel Filter paper, coarse porosity and fast flow rate (e.g., Quantitative Q8, Fisher) Additional reagents and equipment for TLC and column chromatography (see Support Protocol 3) Prepare 2′,3′-O-(dibutylstannylene)adenosine 1. Add a stir-bar to 700 mL methanol in a 2-L round-bottom flask and begin stirring. Add 6.7 g (25 mmol) adenosine and 6.25 g (25 mmol) dibutyltin oxide. 2. Remove stir-bar, add boiling chips, and fit flask with a water-cooled reflux condenser. Heat mixture to boiling using a heating mantle, reflux 45 min, and let cool to room temperature. 3. Decant solution from boiling chips into a second 2-L flask and remove methanol using a rotary evaporator connected to a water aspirator. 4. Dry residue to constant weight in a vacuum desiccator over phosphorus pentoxide. A yield of 10.5 g is expected. If desired, this intermediate can be stored at –20°C in a sealed, dry container (stable for many months).

Alkylate to form 2′- and 3′-O-(2-nitrobenzyloxymethyl)adenosine 5. Add 8 g (16 mmol) dry 2′,3′-O-(dibutylstannylene)adenosine and 7.7 g (16 mmol) anhydrous tetra-n-butylammonium bromide to 65 mL anhydrous dimethylformamide in a 250-mL round-bottom flask while stirring. 6. Heat to 60°C in an oil bath and fit flask with a 50-mL pressure-equalizing dropping funnel. 7. Use the dropping funnel to add freshly prepared 2-nitrobenzyl chloromethyl ether (from 16 mmol 2-nitrobenzyl methylthiomethyl ether; see Support Protocol 1) dropwise over a 5-min period while stirring at 60°C. Seal flask and stir 1 hr at 60°C. 8. Add 20 mL anhydrous pyridine and stir 25 min at room temperature. 9. Examine reaction mixture by TLC (see Support Protocol 3) in 9:1 chloroform/methanol. The 2′- and 3′-isomers are visible as two closely spaced bands in the lower half of the plate. The upper band of this pair is the 2′-isomer (Table 2.5.1).

10. Remove solvents by rotary evaporation, using high vacuum (i.e., with a vacuum pump) to remove as much dimethylformamide as possible. If desired, this intermediate can be stored at –20°C in a sealed, dry container (stable for many months).


Purify 2′-O-(2-nitrobenzyloxymethyl)adenosine 11. Dissolve residue in the minimum volume of 66% aqueous pyridine and, using a Pasteur pipet, transfer it dropwise onto 15 g silica gel 60 in a large petri plate. Allow wet silica to dry in the moving stream of air inside a fume hood.

2.5.12 Supplement 3


An overnight drying time should be adequate if most of the dimethylformamide was removed and the minimum amount of aqueous pyridine was used.

12. Pulverize the dry, coated silica. Layer powder onto a 3 × 60–cm glass chromatography column and elute with 95:5 chloroform/methanol, collecting fractions as described (see Support Protocol 3). The volume of 95:5 chloroform/methanol should be determined from TLC results.

13. Monitor fractions by TLC in 9:1 chloroform/methanol. 14. Combine all fractions containing the 2′-isomer and rotary evaporate to dryness using a water aspirator. If desired, this intermediate can be stored at –20°C in a sealed, dry container (stable for many months).

15. Dry material on 6 g silica gel 60 as in step 11 and repeat step 12 using a 1-L stepwise gradient of 1% to 5% methanol in chloroform (beginning with 200 mL of 1% methanol, followed by 200 mL of 2% methanol, and so on). 16. Repeat steps 13 and 14. 2′-O-(2-Nitrobenzyloxymethyl)adenosine contaminated with a trace of the 3′-isomer is left after solvent removal. If desired, this intermediate can be stored at –20°C in a sealed, dry container (stable for many months).

17. Dissolve residue in a minimal amount of 80% aqueous acetonitrile with heating, then use the rotary evaporator connected to water aspirator to repeatedly co-evaporate with acetonitrile to a final volume of ∼5 mL until crystallization of the 2′-isomer occurs. A 24% yield of white crystals is obtained, based on the starting quantity of 2′,3′-O-(dibutylstannylene)adenosine. This material may be further recrystallized from water. The melting point of the product is 175° to 177°C. If desired, this intermediate can be stored at –20°C in a sealed, dry container (stable for many months).

Prepare 2′-O-(2-nitrobenzyloxymethyl)-N6-benzoyladenosine 18. Add 1.08 g (2.5 mmol) 2′-O-(2-nitrobenzyloxymethyl)adenosine to 10 mL anhydrous pyridine in a 100-mL round-bottom flask while stirring to dissolve. Add 1.59 mL (12.5 mmol) trimethylchlorosilane, seal flask, and stir 2 hr at room temperature. 19. Add 1.45 mL (12.5 mmol) benzoyl chloride and reseal flask. Stir 2 hr, then cool to 0°C and add 2.5 mL water. Stir 10 min, then add 5 mL concentrated ammonium hydroxide. Cover (do not seal) flask and stir 30 min at room temperature. Ammonium hydroxide from a newly opened bottle should be used, and the reaction should be checked with pH paper to make certain it remains alkaline throughout the 30 min. More ammonium hydroxide should be added in 1-mL aliquots, if necessary, to maintain alkalinity.

20. Remove ammonia and solvents by rotary evaporation using a water aspirator. 21. Transfer reaction mixture to a 500-mL separatory funnel using 150 mL ethyl acetate and extract three times with 100 mL of 1 M NaHCO3, then three times with 100 mL of 2 M NaCl. 22. Dry ethyl acetate layer over 5 g anhydrous sodium sulfate, filter through filter paper to remove salt, and rotary evaporate filtrate using a water aspirator.



Supplement 3

If desired, this intermediate can be stored at –20°C in a sealed, dry container (stable for many months).

23. Layer residue on a 2.5 × 25–cm silica gel column and elute with 100 mL chloroform, followed by 300 mL of 1% methanol in chloroform and 800 mL of 2% methanol in chloroform as described (see Support Protocol 3). 24. Monitor fractions by TLC using 9:1 chloroform/methanol. 25. Use the rotary evaporator connected to a water aspirator to evaporate appropriate fractions and produce 2′-O-(2-nitrobenzyloxymethyl)-N 6-benzoyladenosine. A yield of 88%, based on the starting quantity of 2′-O-(2-nitrobenzyloxymethyl)adenosine, is expected. If desired, this intermediate can be stored at –20°C in a sealed, dry container (stable for many months).

Prepare 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)-N6benzoyladenosine 26. Dry 1.05 g (1.96 mmol) 2′-O-(2-nitrobenzyloxymethyl)-N6-benzoyladenosine by three co-evaporations with 30 mL anhydrous pyridine using the rotary evaporator connected to a vacuum pump. 27. Dissolve residue in 30 mL anhydrous pyridine, then add 0.34 mL (2.4 mmol) triethylamine and 0.80 g (2.4 mmol) 4,4′-dimethoxytrityl chloride. Stir 3 hr at room temperature. 28. Add 1 mL methanol and, after 10 min, remove solvents by rotary evaporation using a vacuum pump. 29. Repeat steps 21 to 25 with the following exceptions: a. Extract three times with NaHCO3 but only once with NaCl in step 21. b. Use a 2.5 × 30–cm silica gel column and a 1.2-L stepwise gradient of 0% to 1% methanol in chloroform containing 0.25% pyridine in step 23. c. Use 2:1 ethyl acetate/hexane for TLC in step 24. The yield of 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)-N6-benzoyladenosine is expected to be 91% based on the starting quantity of 2′-O-(2-nitrobenzyloxymethyl)-N6-benzoyladenosine.

30. Store 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)-N6-benzoyladenosine in the dark at −20°C in a sealed container containing a trace of pyridine (stable for many months). BASIC PROTOCOL 4


Synthesis of 5′-O-(4,4′-Dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)-N2isobutyrylguanosine In this protocol, guanosine is protected at its N2 position with the isobutyryl group, then converted into its dibutylstannylene derivative and alkylated with 2-nitrobenzyl chloromethyl ether. The resulting 2′-O-(2-nitrobenzyloxymethyl) nucleoside is purified by silica-gel chromatography, then crystallized from an acetonitrile/water mixture. Dimethoxytritylation for protection of the 5′-hydroxyl is the final step in the synthesis. The reaction sequence is illustrated in Figure 2.5.5.

2.5.14 Supplement 3


O N HO

O N

NH

N

N

O

HO

NH2

O

1. Isobutyryl chloride/Pyridine 2. Aq. NaOH in ethanol

OH OH

N

NH N

O

N H

OH OH

Dibutyltin oxide MeOH

O

O N HO

O OH

O

O

NH

N

N

N

N H

HO

2-Nitrobenzyl chloromethyl ether

O

O O

N

NH N

O

N H

O Sn

NO2

Bu

Bu

O DMTrCl/Pyridine

N DMTrO

N

O OH

NH

O

N

O

N H

O NO2

2 Figure 2.5.5 Preparation of 5′-O-(4,4′-dimethoxytrityl)-2′-O-(2-nitrobenzyloxymethyl)-N -isobutyrylguanosine from guanosine. Bu is n-butyl.

Materials Guanosine hydrate Anhydrous pyridine, stored over coarse granules of calcium hydride Isobutyryl chloride Ethyl acetate 1 M NaHCO3 2 M NaCl Anhydrous sodium sulfate Ethanol 2 N NaOH, ice cold Dowex AG50W-X8 ion exchange resin (pyridinium form) 5 × 60–cm glass chromatography column filled with 50 mL Dowex AG50W-X8 ion-exchange resin (pyridinium form) in water 4:1 (v/v) chloroform/methanol Phosphorus pentoxide Dibutyltin oxide Methanol Anhydrous dimethylformamide, stored over 4A molecular sieves 2-Nitrobenzyl chloromethyl ether, make fresh (see Support Protocol 1) Silica gel 60, 70 to 230 mesh ASTM (e.g., EM Science) 2.5 × 25–cm and 2.5 × 30–cm glass chromatography columns packed with silica gel 60, 70 to 230 mesh ASTM, in chloroform to a bed height of 20 and 25 cm, respectively (see Support Protocol 3) Chloroform 2%, 4%, 6%, and 8% (v/v) methanol in chloroform



Supplement 4

9:1 (v/v) chloroform/methanol 80% (v/v) aqueous acetonitrile Acetonitrile Triethylamine, stored over coarse granules of calcium hydride 4,4′-Dimethoxytrityl chloride 0% to 1% (v/v) methanol in chloroform, containing 0.25% (v/v) pyridine 2:1 (v/v) ethyl acetate/hexane Oven, 130°C 500-mL and 2-L round-bottom flasks 100-mL pressure-equalizing dropping funnel Rotary evaporator connected interchangeably to water aspirator and vacuum pump 1-L separatory funnel Filter paper, coarse porosity and fast flow rate (e.g., Quantitative QB, Fisher) Water-cooled reflux condenser Heating mantle Additional reagents and equipment for TLC and column chromatography (see Support Protocol 3) NOTE: The pyridinium form of Dowex AG50W-X8 ion-exchange resin is made by washing the hydrogen form with several changes of 20% aqueous pyridine. Prepare tetraisobutyrylguanosine 1. Dry guanosine hydrate to constant weight in an oven at 130°C. A couple of hours drying is sufficient to drive off the water of crystallization.

2. Add 28.3 g (100 mmol) dry guanosine to 1 L anhydrous pyridine in a 2-L round-bottom flask while stirring. Fit flask with a 100-mL pressure-equalizing dropping funnel. The guanosine will form a suspension in the pyridine.

3. Cool mixture to 0°C in an ice bath and use the dropping funnel to add 100 mL isobutyryl chloride dropwise over a 10-min period with vigorous stirring at 0°C. Stir 2 hr at 0°C. 4. Use the dropping funnel to add 100 mL water dropwise to the stirring solution, slowly enough that the temperature is maintained 9:1), the bicyclic phosphoramidite S.36 produced the same dinucleotides with only moderate stereoselectivity (RP:SP ≈ 3:1) (Guo et al., 1998). The bicyclic phosphoramidites S.33-S.35 a r e nonetheless promising candidates for the preparation of P-diastereomerically enriched oligonucleoside phosphorothioates. Reagents other than 1H-tetrazole have also been used to activate deoxyribonucleoside phosphoramidites, including N-methylimidazolium trifluoromethanesulfonate (Arnold et al., 1989); N-methylimidazole hydrochloride (Hering et al., 1985); pyridinium tetrafluoroborate (Brill et al., 1991); pyridinium chloride, pyridinium bromide, and pyridinium 4-methylbenzinesulfonate (Beier and Pfleiderer, 1999); N-methylanilinium trifluoroacetate (Fourrey and Varenne, 1984); N-methylanilinium trichloroacetate (Fourrey et al., 1987); benzimidazolium triflate (Hayakawa et al., 1996); imidazolium triflate (Hayakawa and Kataoka, 1998); pyridine hydrochloride/imidazole (Gryaznov and Letsinger, 1992); 5-trifluoromethyl-1H-tetrazole (Hering et al., 1985); 5-(4-nitrophenyl)-1H-tetrazole (Froehler and Matteucci, 1983); 5-(2-nitrophenyl)1H-tetrazole (Pon, 1987; Montserrat et al., 1994); 1-hydroxybenzotriazole (Claesen et al., 1984); 2,4,5-tribromo- and 2-nitro-imidazoles

(Xin and Just, 1996); benzotriazole and 5-chlorobenzotriazoles (Xin and Just, 1996); and 4,5dichloro-, 2-bromo-4,5-dicyano-, and 4,5-dicyano-imidazoles (Xin and Just, 1996). 5Ethylthio- and 5-benzylthio-1H-tetrazoles are also potent in the activation of phosphoramidites; these activators have been particularly useful in RNA synthesis (Wincott et al., 1995; Wu and Pitsch, 1998). It is important to note, however, that 1H-tetrazole and those more acidic activators (pKa < 4.8) can cleave the 5′-dimethoxytrityl (DMTr) group of deoxyribonucleoside phosphoramidites to a small extent and trigger the formation of activated dimers. The coupling of these dimers during chain extension produced oligonucleotides longer (n + 1) than the expected size (n) (Krotz et al., 1997a). The rates of DMTr deprotection by an activator depend on its acidity, exposure time, and nature of the nucleobase carrying the DMTr group (purines deprotect faster than pyrimidines). Typically, when the activation of deoxyribonucleoside phosphoramidites is performed by 1H-tetrazole for a period of 100 sec, ∼0.3%–0.9% of (n + 1)-mers is observed; however, when 1H-tetrazole is replaced by 5ethylthio-1H-tetrazole under similar conditions, ∼1.0%–4.3% of (n + 1)-mers are generated. It should, therefore, be understood that extended coupling times and the use of more acidic activators during solid-phase oligode-

B

DMTrO

O

O

O P

OEt OEt

Oligodeoxyribonucleotide Synthesis Using the Phosphoramidite Method

31

N N N

OEt N P OEt

O

X N

S

i-Pr2N P

B

DMTrO

33 X = H 34 X = OMe

O

O P

X N

R

35 X = H 36 X = OMe

32

Figure 3.3.11 Model compounds used in the study of phosphoramidite activation by 1Htetrazole.

Figure 3.3.12 Nucleoside bicyclic phosphoramidites for the preparation of P-diastereomerically enriched oligonucleoside phosphorothioates.


oxyribonucleotide synthesis will result in lower recovery of full-length oligomers, because longer than full-length oligonucleotides will be produced. These observations prompted an extensive search for less-acidic, more-nucleophilic activators. It has been reported that 4,5-dicyanoimidazole (Xin and Just, 1996) is less acidic (pKa 5.2), more soluble, and more nucleophilic than 1H-tetrazole (Vargeese et al., 1998). Moreover, the usefulness of 4,5-dicyanoimidazole relates not only to efficient synthesis of oligonucleotides but also to the preparation of nucleoside phosphoramidites from phosphorodiamidite S.23. Even though 1H-tetrazole is still very popular as an activator for deoxyribonucleoside phosphoramidites, 4,5dicyanoimidazole is an attractive option for the activation of deoxyribonucleoside and ribonucleoside phosphoramites in solid-phase oligonucleotide synthesis.

pling yields averaging 45% on a CPG support (Sekine et al., 1986). Under similar conditions, activated S.38 required a coupling time of 60 min to produce yields of 65%–70% (Casale and McLaughlin, 1990). Such coupling yields are far below those obtained with conventional deoxyribonucleoside phosphoramidites (∼99%) and thus underscore the importance of steric hindrance when designing nucleobase protecting groups and/or modified nucleobases toward the synthesis of oligodeoxyribonucleotides and their analogues. One effective approach to lessening steric interferences is to increase the distance between the bulky entity and the phosphoramidite function by the use of flexible linkers. For example, unlike S.38, the deoxyribonucleoside phosphoramidite S.39 (Fig. 3.3.13) has been efficiently incorporated into oligonucleotides under the conditions used for standard 2-cyanoethyl deoxyribonucleoside phosphoramidites (Bergstrom and Gerry, 1994). Other factors influencing the coupling rates of activated deoxyribonucleoside phosphoramidites have been investigated by Dahl and his colleagues (1987). In a systematic study, it was observed that condensation rates varied with the nature of phosphoramidite Oalkyl and N,N-dialkylamino groups. Typically, coupling rates decreased according to the following order: O-methyl > O-(2-cyanoethyl) > O-(1-methyl-2-cyanoethyl) > O-(1,1-dimethyl-2-cyanoethyl) and N,N-diethylamino >

FACTORS AFFECTING THE CONDENSATION RATES OF DEOXYRIBONUCLEOSIDE PHOSPHORAMIDITES The steric bulk of specific guanine N-2 functional groups has been shown to affect significantly condensation rates and coupling efficiency of these deoxyribonucleoside phosphoramidite derivatives. Two classic examples illustrating this fact are the activation of phosphoramidites S.37 and S.38 with 1H-tetrazole (Fig. 3.3.13). In the case of activated S.37, a condensation time of 10 min generated cou-

O

O N N

O O

O

NH

N

DMTrO

N

P

N OBz

N

PixO

NH

O

NH N

NH

O

BzO OBz

OMe

i-Pr2N

P

OCH2CH2CN 38

37 4-NO2PhCH2CH2O N N

DMTrO

O

N N

N H

O

N N

O

i-Pr2N

P

Bz, benzoyl 4-NO2Ph, 4-nitrophenyl Pix, 9-phenylxanthen-9-yl

OCH2CH2CN 39

Figure 3.3.13 Deoxyribonucleoside phosphoramidites functionalized with nucleobase bulky groups.

Synthesis of Unmodified Oligonucleotides


N,N-diisopropylamino > N-morpholino > Nmethylanilino. The effect of steric hindrance on coupling rates is further illustrated by activation of the 2′-substituted nucleoside phosphoramidites S.40-S.43 (Fig. 3.3.14) and their competitive condensations with thymidine covalently attached to a solid support (Kierzek et al., 1987). Dimer formation was quantitated and correlated with the condensation rates of S.40-S.43. The amount of dimers formed decreased when the groups at C-2′ increased in sizes; thus 2′-H > 2′-O-methyl > 2′-O-tetrahydropyranyl > 2′O-tert-butyldimethylsilyl (Kierzek et al., 1987). In agreement with these findings, the deoxyribonucleoside phosphoramidites S.44 (Polushin, 1996) and S.45 (Jørgensen et al., 1994) have also exhibited significantly lower condensation rates and coupling efficiency because of steric factors (Fig. 3.3.14). Like sterically hindered groups attached to nucleobases (vide supra), it is also possible to decrease the steric demand of 2′-O-bulky protecting groups by increasing the distance between these groups and the phosphoramidite function. Specifically, the ribonucleoside 2′-O-triisopropylsilyloxymethyl phosphoramidite S.46 (Weiss, 1998; Wu and Pitsch, 1998; Fig. 3.3.15) allows much faster coupling reactions (2 min) in solid-phase oligoribonucleotide synthesis than do ribonucleoside 2′-O-tertbutyldimethysilyl phosphoramidites (5–8 min) under essentially identical conditions (see UNIT 3.4). Similarly, coupling reactions of t h e r i b o n u c l e o s i d e 2′-O-(o-nitrobenzyloxymethyl) phosphoramidite S.47 (Fig. 3.3.15) are faster (2 min) than those effected b y r i b o n u c l e o s i d e 2′-O-(o-nitrobenzyl) phosphoramidites (10 min) under the same conditions (deBear et al., 1987; Schwartz et al., 1992). Thus functional groups generating

B

DMTrO

i-Pr2N

O

P

O

O

O

Si

NCCH2CH2O 46

B

DMTrO

O O

i-Pr2N

P

O2N O

O

OCH2CH2CN 47

Figure 3.3.15 Efficient ribonucleoside phosphoramidites for solid-phase RNA synthesis.

steric bulk near the nucleosidic phosphoramidite moiety are likely to interfere with coupling rates and should be given consideration when developing novel phosphoramidite monomers.

SIGNIFICANCE OF THE “CAPPING” REACTION IN THE CHEMICAL SYNTHESIS OF OLIGODEOXYRIBONUCLEOTIDES The phosphoramidite approach to oligodeoxyribonucleotide synthesis is renowned for its high coupling efficiency. Nonetheless, oligonucleotide chain extension does not occur quantitatively even under optimum conditions. As a result, the desired n-mer oligodeoxyribonucleotide is contaminated in the final product with a population of shorter (n − 1)-oligomers. Separation of the n-mer oligonucleotide from (n − 1)-mers can be challenging; however, this problem is almost completely eliminated by acetylation of the remaining unphosphitylated oligonucleotides after each condensation step. This “capping” reaction terminates the elongation of the unphosphitylated oligomers

NHBz N N

DMTrO

O

O

DMTrO

Ura

Thy

DMTrO TBDMSO

O

O

O

i-Pr2N

O P

40 41 42 43


R OCH2CH2CN R=H R = OMe R = OThp R = OTBDMS

O i-Pr2N P NCCH2CH2O

HN

OMe O

44

i-Pr2N

O P

OCH2CH2CN 45

Thp, tetrahydropyran-2-yl TBDMS, tert-butyldimethylsilyl

Figure 3.3.14 Nucleoside phosphoramidites functionalized with 2′- or 3′-sterically demanding groups.


that would otherwise occur during the next coupling step. A very effective capping reagent is a solution of acetic anhydride, 2,6-lutidine, and N-methylimidazole in tetrahydrofuran (Farrance et al., 1989). Such a capping formulation not only prevents extension of unphosphitylated oligomers but also efficiently reduces the concentration of O6-phosphitylated guanine residues that are generated during the condensation step (Pon et al., 1986; Eadie and Davidson, 1987). The capping formulation is rich in acetate ions; these nucleophiles efficiently cleave O6-phosphitylated guanine adducts by attacking tricoordinated “enol phosphites” (like S.48; Fig. 3.3.16) and releasing unmodified guanine residues. Adducts such as S.48 or its oxidized form S.49, if not destroyed, can serve as secondary sites for oligonucleotide synthesis and lead to the formation of a complex mixture of branched oligodeoxyribonucleotides. The generation of these adducts is most efficiently minimized when the capping reaction is performed before the oxidation step. O,O-Diethyl-N,N-diisopropyl phosphoramidite has also been reported as an improved capping reagent in oligonucleotide synthesis (Yu et al., 1994). On the basis of the data presented, this phosphoramidite exhibited a capping efficiency that was only modestly superior to that of the standard acetic anhydride/N-methylimidazole/2,6-lutidine capping formulation. The phosphoramidite capping reagent was also claimed to not produce nucleobase modification; supporting data were, however, not shown. Interestingly, recent use of the lipophilic O-(2-cyanoethyl),O-octyl-N,N-diisopropyl phosphoramidite as a capping re-

B O O O

O

B

DMTrO

O

O O P

N

O

B

DMTrO

O

O P OCH2CH2CN O

OCH2CH2CN

N

O P OR N

Oxidation of a newly generated phosphite triester linkage to the corresponding phosphate triester function is an essential step in automated synthesis of oligodeoxyribonucleotides by the phosphoramidite approach. Without oxidation, an internucleoside phosphite triester function decomposes under the acidic conditions required for cleavage of the 5′-dimethoxytrityl group (Matteucci and Caruthers, 1981). Thus oxidation of phosphite triesters is absolutely necessary to ensure consistent highyielding oligodeoxyribonucleotide syntheses. An aqueous solution of iodine (0.05–0.1 M)

O

N

O

THE OXIDATION REACTION IN THE SYNTHESIS OF OLIGODEOXYRIBONUCLEOTIDES ACCORDING TO THE PHOSPHORAMIDITE METHOD

B

DMTrO

DMTrO

agent in solid-phase oligonucleotide synthesis has allowed facile separation of capped failure sequences from trityl-off full-length oligonucleotides by reverse-phase HPLC (RP-HPLC; Natt and Häner, 1997). This capping method simplified RP-HPLC purification of synthetic oligonucleotides and resulted in higher isolated yields. Phosphoramidite capping reagents may, however, like nucleoside phosphoramidites, phosphitylate nucleobases (especially at O-6 of guanosines) and eventually lead to the formation of, for example, 2,6-diaminopurine residues (Eadie and Davidson, 1987). This potential problem has not, as yet, been thoroughly investigated. Until this issue is resolved, use of the standard and well-studied acetic anhydride/N-methylimidazole/2,6-lutidine capping formulation during solid-phase oligonucleotide synthesis is recommended.

N

N

O P OR NHi-Bu

O

O

O

O

P

P 48 R = CH2CH2CN

N

N

NHi-Bu

49 R = CH2CH2CN

i-Bu, isobutyryl

6

Figure 3.3.16 Postulated O -guanine adducts generated during the chain extension step of the synthesis cycle according to the phosphoramidite method.



and 2,6-lutidine (Letsinger and Lunsford, 1976) or pyridine (Usman et al., 1985) in tetrahydrofuran is generally used for this task; this formulation is stable and provides rapid-reaction kinetics, usually without formation of side products. When the N-2 amino function of guanine is protected with a N,N-dimethylformamidine group (Zemlicka and Holy, 1967; McBride et al., 1986; Vu et al., 1990) during automated oligonucleotide synthesis, however, use of the traditional iodine formulation as oxidant led to cyanation of guanine at N-2 (Mullah et al., 1995). It has also been shown that this side reaction is completely eliminated by the use of a lower-concentration (0.02 M) iodine oxidation reagent without losing speed and efficiency in the conversion of internucleoside phosphite triesters to phosphate triesters (Mullah et al., 1995). Thus the latter aqueous iodine formulation is recommended for standard oligonucleotide synthesis. For specific applications, however, nonaqueous oxidizing reagents may advantageously offer an alternative to aqueous iodine for the oxidation of oligodeoxyribonucleoside phosphite triesters. For example, m-chloroperbenzoic acid (Tanaka and Letsinger, 1982); iodobenzene diacetate and tetra-n-butylammonium periodate (Fourrey and Varenne, 1985); tert-butyl hydroperoxide (Hayakawa et al., 1986; Hayakawa and Kataoka, 1998); ditert-butyl hydroperoxide; cumene hydroperoxide; hydrogen peroxide; bis-trimethylsilyl peroxide, and catalytic amounts of trimethylsilyl triflate (Hayakawa et al., 1986); dinitrogen tetroxide and molecular oxygen in the presence of 2,2′-azobis(2-methylpropionitrile) under thermal or photochemical conditions (Bentrude et al., 1989); and (1S)-(+)-(10-camphorsulfonyl) oxaziridine (S.50; Ugi et al., 1988) have been effective. The oxaziridine S.50 (Fig. 3.3.17) is particularly useful for the synthesis of oligonucleotide containing multiple 7deaza-2′-deoxyguanosine residues. Incorpora-

Thy

DMTrO

N S O O O 50


i-Pr2N

O O P

CH3 O 51

Figure 3.3.17 An oxaziridine derivative as a useful oxidant in the synthesis of oligonucleotides containing iodine sensitive residues, and a benzylic deoxyribonucleoside phosphoramidite suitable for the preparation of oligonucleotide analogues.

tion of this modified 2′-deoxyguanosine into oligonucleotides via the phosphoramidite approach is sensitive to iodine-containing solutions regardless of iodine concentration (Anonymous, 1996). In this context, it should be noted that when applied to oligonucleotide synthesis, the benzylic deoxyribonucleoside phosphoramidite S.51 (Fig. 3.3.17) generated internucleoside o-methylbenzyl phosphite esters that were sensitive to aqueous iodine oxidation. This sensitivity to iodine resulted in the loss of benzylic phosphate protection (Caruthers et al., 1987b). The absence of phosphate protecting groups did not, however, impair subsequent additions of S.51 to the DNA chain. In fact, an oligothymidylic acid (20-mer) was prepared by the iterative incorporation of S.51 with an average coupling efficiency of 96%. It was speculated that phosphate-phosphite mixed anhydrides could have been generated from the interaction of phosphate diesters with activated S.51 and then cleaved by excess 1H-tetrazole to regenerate the deoxyribonucleoside phosphorotetrazolide intermediates needed for chain extension. Because of the inherent hazards involved with handling peroxides, the use of oxaziridine S.50 is, therefore, recommended for the oxidation of phosphite triesters of those modified oligonucleotides that are reactive to iodine and/or necessitate rigorously anhydrous conditions. Moreover, the use of S.50 in oligonucleotide synthesis does not lead to detectable nucleobase modifications (Anonymous, 1996). It has also been shown that oxidation of the dinucleoside 2-cyano-1,1-dimethylethyl phosphite triester S.52 (Fig. 3.3.18) with iodine in the presence of water, alcohols, and amines produced the corresponding dinucleoside phosphate S.54, phosphate triester S.55, and phosphoramidate S.56, respectively (Nielsen and Caruthers, 1988). It is postulated that under these oxidative Arbuzov-type conditions, elimination of the 2-cyano-1,1-dimethylethyl group led to the dinucleoside phosphoryl iodide intermediate S.53. The formation of S.53 is supported by 31P-NMR data and thus provides a versatile pathway to the synthesis of oligodeoxyribonucleotide analogues from deoxynucleoside 3′-O-(2-cyano-1,1-dimethylethyl) or o-methylbenzyl phosphoramidites. Another example of the importance of P(III) oxidation in oligodeoxyribonucleotide synthesis according to the phosphoramidite approach is the incorporation of internucleotide phosphorothioates linkages into these bio-


P

O

I2

O P O

O

Thy

THF

H2O I

O

AcO

O

Thy

or ROH

O O

O P OR O

O

Thy

AcO

AcO

52

Thy

DMTrO

O

O O

NCCH2C(CH3)2O

Thy

DMTrO

Thy

DMTrO

53

54 R = H 55 R = CH3

R'NH2

Thy

DMTrO

O O

O P NHR' O

O

Thy

AcO Ac, acetyl

56 R' = n-butyl

Figure 3.3.18 Access to oligodeoxyribonucleotide analogues from deoxynucleoside (2-cyano1,1-dimethylethyl) phosphoramidites.

molecules. Oligonucleotides carrying internucleotide phosphorothioate diesters display enhanced resistance to hydrolysis catalyzed by nucleases (Eckstein, 1985). Because of this property, oligodeoxyribonucleoside phosphorothioates have been extensively used as antisense molecules in the inhibition of gene expression. Automated synthesis of these modified oligonucleotides via the phosphoramidite method consists of replacing the aqueous iodine oxidation step by a sulfurization reaction that had originally been effected by elemental sulfur. Given the poor solubility of elemental sulfur in organic solvents, its use in automated systems has been difficult. This problem was eliminated when phenylacetyl disulfide (S.57; Kamer et al. 1989; Roelen et al., 1991) and 3H-1,2-benzodithiol-3-one 1,1-dioxide (S.58; Iyer et al., 1990; Regan et al., 1992) were employed as sulfurization reagents (Fig. 3.3.19). These compounds are soluble in organic solvents and produce efficient and rapid sulfurization kinetics. For example, S.58 converted the dinucleoside phosphite triester S.59 to the corresponding phosphorothioate dimer S.60 in yields exceeding 99% within 30 sec at 25°C (Iyer et al., 1990; Regan et al., 1992). Deprotection of S.60 afforded the dinucleoside phosphorothioate S.61 (Fig. 3.3.19). Thus the sulfur-transfer reagent S.58 has enabled reliable automated synthesis of phosphorothioated oligomers carrying either exclusively or a predetermined number of phos-

phorothioate groups (Iyer et al., 1990). Given the biological significance of oligonucleoside phosphorothioates, application of S.58 to the synthesis of these modified oligonucleotides has spurred interest in the development of additional sulfurizing reagents. The most notable sulfur-transfer agents that have been reported during this decade include N,N,N′,N′tetraethylthiuram disulfide (Vu and Hirschbein, 1991), dibenzoyl tetrasulfide (Rao et al., 1992), bis-(O,O-diisopropoxyphosphinothioyl) disulfide (Stec et al., 1993), benzyltriethylammonium tetrathiomolybdate (Rao and Macfarlane, 1994), bis(p-toluenesulfonyl)disulfide (Efimov et al., 1995), 3-ethoxy1,2,4-dithiazoline-5-one (Xu et al., 1996), thiiranes (Arterburn and Perry, 1997), bis(ethoxythiocarbonyl)tetrasulfide (Zhang et al., 1998), and 3-methyl-1,2,4-dithiazoline-5one (Zhang et al., 1999). Out of these sulfurtransfer reagents, 3H-1,2-benzodithiol-3-one 1,1-dioxide and 3-ethoxy-1,2,4-dithiazoline-5one are currently the most extensively used in solid-phase synthesis of oligonucleoside phosphorothioates.

STRATEGIES IN THE DEPROTECTION OF SYNTHETIC OLIGODEOXYRIBONUCLEOTIDES The efficiency of the phosphoramidite method for solid-phase synthesis of oligodeoxyribonucleotides is such that oligonucleotides up to 50 bases long can be synthesized



O

O S

S

S S

O

O

57

B

DMTrO O

S O

O

O

58

O

O O

P − + S

O S

P OCH2CH2CN

O

B

DMTrO

O O

S

O

58

B

O

OCH2CH2CN

O

O

O

O

P

P

B

59

B

HO

O

−

deprotection

O P S O

B

DMTrO

O

O O

S P OCH2CH2CN O

B

OH

O

O

B

O

61 P 60

Figure 3.3.19 Preparation of oligodeoxyribonucleoside phosphorothioates according to the solidphase phosphoramidite method.


within a few hours. While the cleavage of these oligonucleotides from solid supports is normally accomplished by treatment with concentrated ammonium hydroxide for ∼1 hr at ambient temperature, it will take ∼10 hr at elevated temperature (55°C) to deprotect the N-isobutyryl group of guanines, and N-benzoyl group of cytosines and adenines (see UNIT 2.1). This time-consuming deprotection step clashed with the urgent demand for synthetic oligodeoxyribonucleotides and thus provided an incentive to improve the chemistry involved with postsynthesis oligonucleotide processing. Specifically, methods for rapid removal of oligonucleotide protecting groups have attracted considerable attention and motivated the development of novel base-labile blocking groups for nucleobases (Schulhof et al., 1987; Uznanski et al., 1989; Kuijpers et al., 1990; Vu et al., 1990; Beaucage and Iyer, 1992; Sinha et al., 1993; Iyer et al., 1997; see also UNIT 2.1). Concentrated solutions of ammonia in water, ethanol, or methanol have been used for the cleavage of these groups. Alternatively, an aqueous

solution of methylamine and ammonium hydroxide has been employed for the deprotection of oligonucleotides carrying N-acetyl cytosines, N-benzoyl adenines, and N-isobutyryl guanines (Reddy et al., 1994). With this reagent, oligonucleotides were cleaved from solid supports in 5 min at ambient temperature, and complete deprotection was accomplished in 5 min at 65°C. It should, however, be emphasized that an aqueous solution of methylamine and ammonium hydroxide cannot be used for the deprotection of oligonucleotides bearing conventional N-benzoyl cytosines because primary amines have been reported to attack N4-anisoyl- or N4-benzoyl-2′-deoxycytidine at C-4 to produce N4-alkylated 2′-deoxycytidine derivatives (Weber and Khorana, 1972; Reddy et al., 1997). Gaseous amines such as ammonia or methylamine have also been employed under pressure to achieve mild and rapid deprotection conditions (Boal et al., 1996). For example, oligodeoxyribonucleotides having cytosines, adenines, and guanines N-protected with a tert-


butylphenoxyacetyl group were released from CPG supports and fully deprotected at 25°C by pressurized ammonia or methylamine within 35 or 2 min, respectively. It has also been shown that when the N-benzoyl group is used for protection of cytosines and adenosines, and N-isobutyryl for guanines, complete deprotection of oligodeoxyribonucleotides by ammonia gas will take ∼7 hr at 25°C. At that temperature, it would take ∼36 hr for concentrated aqueous ammonium hydroxide to accomplish the same task (Boal et al., 1996). The use of ammonia or methylamine gas allows the simultaneous deprotection of a large number of oligodeoxyribonucleotides. In fact, the number of oligonucleotides or CPG columns that can be deprotected is limited only by the size of the pressure vessel employed. Because no water is present during deprotection, fully deblocked oligonucleotides remain adsorbed to CPG and thus prevent cross-contamination between columns. Oligonucleotides can then be eluted from individual columns with a minimum amount of water for further purification, if desired. This deprotection procedure eliminates hazards inherent to the handling and heating of aqueous amine solutions in glass vials and, more important, the time-consuming evaporation of these solutions. The gas-phase deprotection methodology is recommended when oligonucleotides carrying base-sensitive nucleobases demand mild deprotection conditions or when rapid deprotection is needed to accelerate the production of synthetic oligonucleotides.

ALTERNATE STRATEGIES TO THE SYNTHESIS OF OLIGODEOXYRIBONUCLEOTIDES ACCORDING TO THE PHOSPHORAMIDITE METHOD The versatility of the phosphoramidite approach to oligodeoxyribonucleotide synthesis has been further demonstrated by the use of deoxyribonucleoside phosphoramidites with unprotected nucleobases. The success of this strategy depends on a modified synthesis cycle protocol that involves treatment of the solid support with an equimolar solution (0.1 M) of pyridine hydrochloride and aniline in acetonitrile (Gryaznov and Letsinger, 1991) or benzimidazolium triflate in methanol (Hayakawa and Kataoka, 1998) immediately after each condensation reaction. This treatment destroys nucleobase adducts that are forming on the oligonucleotidic chain during each coupling step. This procedure should facilitate the syn-

thesis of oligonucleotides bearing base-sensitive functional groups because treatment with concentrated ammonium hydroxide at elevated temperature will no longer be required for oligonucleotide deprotection. Furthermore, depurination of adenine and guanine residues under the acidic conditions required for the removal of the 5′-O-DMTr group will become even less likely. More data are still needed to assess whether the synthesis of oligodeoxyribonucleotides according to the phosphoramidite method without nucleobase protection is trouble-free. The method is promising in that it may significantly expedite the production of synthetic oligonucleotides by shortening postsynthesis oligonucleotide processing time. Another strategy toward the preparation of oligodeoxyribonucleotides entails the stepwise condensation of dinucleotide phosphoramidite blocks such as S.62-S.65 (Fig. 3.3.20) instead of conventional monomeric deoxyribonucleoside phosphoramidites for chain extension. Activation of S.62 with 1H-tetrazole produced coupling yields (∼99%) similar to those generated by monomeric phosphoramidites (Kumar and Poonian, 1984). The incorporation of S.63 into oligonucleotides allowed syntheses of randomized DNA sequences containing the 20 codons corresponding to all natural amino acids (Neuner et al., 1998). The efficiency of dinucleotide phosphoramidites to solid-phase oligonucleotide synthesis has been further demonstrated by the preparation of a large oligomer (101-mer) through repetitive condensations of the dimeric phosphoramidite S.64 (Wolter et al., 1986). Furthermore, the impurity profile of oligonucleoside phosphorothioates synthesized by iterative coupling of the thioated dinucleotide phosphoramidite S.65 (Krotz et al., 1997b) showed at least 70% reduction of the (n − 1)-mers and a ∼50% reduction of phosphodiester formation when compared to profiles obtained by standard monomer phosphoramidite couplings. The use of dimeric phosphoramidites in the synthesis of unmodified oligodeoxyribonucleotides has not been widely adopted, probably because a library of up to 16 combinatorial dimers had to be prepared to accomplish the synthesis of one oligonucleotide. Conversely, the application of dimeric phosphoramidites to oligonucleotide analogue synthesis has been popular especially for the incorporation of modified internucleotide bridges. For example, the dimeric 5′-phosphonate–linked thymidine phosphoramidite S.66 (Zhao and Caruthers, 1996) and S.67 (Kofoed and Caruthers, 1996)



B

DMTrO

Thy

DMTrO

O

O O

O

X P OCH2CH2CN

O P OMe O

O

R2N

O P

B

O

i-Pr2N

OR'

Thy

O O P

OCH2CH2CN

64 X = O 65 X = S

62 R = R' = CH3 63 R = CH(CH3)2, R' = CH2CH2CN

Figure 3.3.20 Solid-phase oligonucleotide synthesis using dinucleotide phosphoramidite derivatives.

have been prepared and incorporated into oligodeoxyribonucleotides to assess the physicochemical and biochemical properties imparted by such modifications (Fig. 3.3.21). For similar purposes, and given the growing interest in the development of therapeutic oligonucleotides, a plethoric number of dimeric phosphoramidites structurally related to S.66-S.68 have been prepared in recent years. Because of the intense activity in this area of research, only a fraction of the work has, so far, been reviewed (see Beaucage and Iyer, 1993; Sanghvi and Cook, 1994; Agrawal and Iyer, 1995; Iyer et al., 1999). Oligodeoxyribonucleotides have also been prepared by the condensation of trinucleotide phosphoramidite blocks to enable oligonucleotide-directed mutagenesis. More and more, oligonucleotides of mixed composition are being used to generate combinatorial libraries of variants in the search for peptides and proteins with improved properties. The most direct route to controlled mutagenesis is indeed the use of trinucleotide synthons that correspond to the amino acid codons needed. The synthesis of trinucleotide phosphoramidites S.69 (Sondek

and Shortle, 1992; Virnekäs et al., 1994), S.70 (Lyttle et al., 1995), S.71 (Ono et al., 1995; Kayushin et al., 1996; Zehl et al., 1996), and S.72 (Gaytán et al., 1998; Fig. 3.3.22) representing the codons for all 20 amino acids has been achieved. The incorporation of S.69 into oligonucleotides was accomplished by allowing a coupling time of 1 min and performing the trinucleotide condensation step twice. Under these conditions, coupling yields averaged 96%–98.5% (Virnekäs et al., 1994). Considering that each trinucleotide condensation adds three nucleobases to the growing oligonucleotide chain, these coupling yields are equivalent to three individual monomeric phosphoramidite condensations, each with a coupling efficiency of 98%–99.5%. Incorporation of the trinucleotide phosphoramidites S.71 and S.72 into oligonucleotides via automated solidphase synthesis occurred in yields that varied with the sequence of the trinucleotide block used. Nonetheless, the incorporation of these trinucleotide phosphoramidite blocks into synthetic DNA in the controlled, codon-by-codon construction of combinatorial libraries of struc-

B

DMTrO DMTrO

Thy

O

O

1 O

O

2 O

O P OPhCl-2

O O P

O

Thy O

BnO O

i-Pr2N

O

Thy

DMTrO

P

OCH2CH2CN

i-Pr2N

O P

Thy

3 O 4 O

OCH2CH2CN

i-Pr2N 66


67

O P

O

B

OCH2CH2CN

68

Bn, benzyl PhCl-2, o-chlorophenyl

n , atom or group of atoms O

Figure 3.3.21 Solid-phase synthesis of oligonucleotide analogues from dimeric phosphoramidites carrying modified internucleotidic linkages.


tural genes will be invaluable in creating molecular diversity by mutagenesis.

CONCLUDING REMARKS Owing to the high performance of the phosphoramidite method, synthetic oligodeoxyribonucleotides became readily available and fueled the biotechnology revolution that has irreversibly changed biomedical research and the pharmaceutical industry. For example, without the ability to rapidly and efficiently synthesize DNA oligonucleotides, the development of the polymerase chain reaction (PCR) and its multiple applications would have been difficult, if not impossible, because this technology completely depends on the use of DNA primers. Similarly, the phosphoramidite method has been instrumental in the development of automated DNA sequencing, which also requires rapid and efficient synthesis of fluorescent DNA primers. Another important biological application for oligodeoxyribonucleotides generated by the phosphoramidite method relates to site-specific mutagenesis of protein genes. Mutagenesis of this type has been used to study protein structure-function relationships and to alter the therapeutic spectrum of pharmaceutically active proteins. In addition, the phosphoramidite method has been particularly useful in the synthesis of modified oligonucleotides for diagnostic applications and as potential therapeutic drugs. Although the latter research area is relatively new, several oligonucleotide-based drugs have already reached the clinic, and others are under preclinical investigation to benefit public health and push further the frontiers of knowledge.

B

DMTrO

O

Adams, S.P., Kavka, K.S., Wykes, E.J., Holder, S.B., and Galluppi, G.R. 1983. Hindered dialkylamino nucleoside phosphite reagents in the synthesis of two DNA 51-mers. J. Am. Chem. Soc. 105:661663. Agrawal, S. and Iyer, R.P. 1995. Modified oligonucleotides as therapeutic and diagnostic agents. Curr. Opin. Biotechnol. 6:12-19. Andrus, A. and Beaucage, S.L. 1988. 2-Mercaptobenzothiazole—An improved reagent for the removal of methyl phosphate protecting groups from oligodeoxynucleotide phosphotriesters. Tetrahedron Lett. 29:5479-5482. Anonymous. 1996. Non-aqueous oxidation with 10camphorsulfonyl-oxaziridine. The Glen Report 9:8-9. Arnold, L., Tocik, Z., Bradkova, E., Hostomsky, Z., Paces, V., and Smrt, J. 1989. Automated chloridite and amidite synthesis of oligodeoxyribonucleotides on a long chain support using amidine protected purine nucleosides. Collect. Czech. Chem. Commun. 54:523-532. Arterburn, J.B. and Perry, M.C. 1997. Rhenium catalyzed sulfurization of phosphorus(III) compounds with thiiranes: New reagents for phosphorothioate ester synthesis. Tetrahedron Lett. 38:7701-7704. Barone, A.D., Tang, J.-Y., and Caruthers, M.H. 1984. In situ activation of bis-dialkylaminophosphines—A new method for synthesizing deoxyoligonucleotides on polymer supports. Nucl. Acids Res. 12:4051-4061. Beaucage, S.L. 1984. A simple and efficient preparation of deoxynucleoside phosphoramidites in situ. Tetrahedron Lett. 25:375-378. Beaucage, S.L. and Caruthers, M.H. 1981. Deoxynucleoside phosphoramidites—A new class of key intermediates for deoxypolynucleotide synthesis. Tetrahedron Lett. 22:1859-1862. Beaucage, S.L. and Iyer, R.P. 1992. Advances in the synthesis of oligonucleotides by the phosphoramidite approach. Tetrahedron 48:22232311.

B

R'O

O

LITERATURE CITED

O O

O P OR B

O

Beaucage, S.L. and Iyer, R.P. 1993. The synthesis of modified oligonucleotides by the phosphoramidite approach and their applications. Tetrahedron 49:6123-6194.

O P OPhCl-2

O

O O

O P OR O

i-Pr2N

B

O

O

O P OPhCl-2 O

O P

B

OR

69 R = CH3 70 R = CH2CH2CN

O

i-Pr2N

O O P

B

Beier, M. and Pfleiderer, W. 1999. Pyridinium salts–An effective class of catalysts for oligonucleotide synthesis. Helv. Chim. Acta 82:879887.

OCH2CH2CN

71 R' = DMTr 72 R' = Fmoc

Fmoc, 9-fluorenylmethoxycarbonyl PhCl-2, o-chlorophenyl

Figure 3.3.22 Trinucleotide phosphoramidite blocks for the controlled, codon-by-codon, construction of combinatorial gene libraries.

Bentrude, W.G., Sopchik, A.E., and Gajda, T. 1989. Stereo- and regiochemistries of the oxidations of 2-methoxy-5-tert-butyl-1,3,2-dioxaphosphori nanes and the cyclic methyl 3′,5′-phosphite of thymidine by H2O/I2 and O2/AIBN to P-chiral phosphates. 17O NMR assignment of phosphorus configuration to the diastereomeric thymidine cyclic methyl 3′,5′-monophosphates. J. Am. Chem. Soc. 111:3981-3987.



Bergstrom, D.E. and Gerry, N. 1994. Precision sequence-specific cleavage of a nucleic acid by a minor-groove-directed metal-binding ligand linked through N-2 of deoxyguanosine. J. Am. Chem. Soc. 116:12067-12068. Berner, S., Mühlegger, K., and Seliger, H. 1989. Studies on the role of tetrazole in the activation of phosphoramidites. Nucl. Acids Res. 17:853864. Boal, J.H., Wilk, A., Harindranath, N., Max, E.E., Kempe, T., and Beaucage, S.L. 1996. Cleavage of oligodeoxyribonucleotides from controlledpore glass supports and their rapid deprotection by gaseous amines. Nucl. Acids Res. 24:31153117. Boudjebel, H., Gonçalves, H., and Mathis, F. 1975. Étude de la liaison P—N dans le motif S2P— NMe3 en résonance magnétique nucléaire et par la réaction d’échange avec le trifluoroacétate de méthyle. Bull. Chem. Soc. Chim. France 628634. Brill, W.K.-D., Nielsen, J., and Caruthers, M.H. 1991. Synthesis of deoxydinucleoside phosphorodithioates. J. Am. Chem. Soc. 113:39723980. Caruthers, M.H., Beaucage, S.L., Becker, C., Efcavitch, W., Fisher, E.F., Galluppi, G., Goldman, R., deHaseth, P., Martin, F., Matteucci, M., and Stabinsky, Y. 1982. New methods for synthesizing deoxyoligonucleotides. In Genetic Engineering: Principles and Methods, Vol. 4 (J.K. Setlow and A. Hollaender, eds.) pp. 1-17. Plenum, New York. Caruthers, M.H., Barone, A.D., Beaucage, S.L., Dodds, D.R., Fisher, E.F., McBride, L.J., Matteucci, M., Stabinsky, Z., and Tang, Y.-Y. 1987a. Chemical synthesis of deoxyoligonucleotides by the phosphoramidite method. In Methods and Enzymology; Vol. 154 (R. Wu and L. Grossman, eds.) pp. 287-313. Academic Press, San Diego; and references therein. Caruthers, M.H., Kierzek, R., and Tang, J.Y. 1987b. Synthesis of oligonucleotides using the phosphoramidite method. In Biophosphates and Their Analogues—Synthesis, Structure, Metabolism and Activity (K.S. Bruzik and W.J. Stec, eds.) pp. 3-21. Elsevier/North Holland, Amsterdam. Casale, R. and McLaughlin, L.W. 1990. Synthesis and properties of an oligodeoxynucleotide containing a polycyclic aromatic hydrocarbon site specifically bound to the N2 amino group of a 2′-deoxyguanosine residue. J. Am. Chem. Soc. 112:5264-5271.


Dahl, B.H., Nielsen, J., and Dahl, O. 1987. Mechanistic studies on the phosphoramidite coupling reaction in oligonucleotide synthesis. I. Evidence for nucleophilic catalysis by tetrazole and rate variations with the phosphorus substituents. Nucl. Acids Res. 15:1729-1743. Dahl, B.H., Bjergårde, K., Henriksen, L., and Dahl, O. 1990. A highly reactive, odourless substitute for thiophenol/triethylamine as a deprotection reagent in the synthesis of oligonucleotides and their analogues. Acta Chem. Scand. 44:639-641. Daub, G.W. and van Tamelen, E.E. 1977. Synthesis of oligoribonucleotides based on the facile cleavage of methyl phosphotriester intermediates. J. Am. Chem. Soc. 99:3526-3528. deBear, J.S., Hayes, J.A., Koleck, M.P., and Gough, G.R. 1987. A universal glass support for oligonucleotide synthesis. Nucleosides Nucleotides 6:821-830. Eadie, J.S. and Davidson, D.S. 1987. Guanine modification during chemical DNA synthesis. Nucl. Acids Res. 15:8333-8349. Eckstein, F. 1985. Nucleoside phosphorothioates. Annu. Rev. Biochem. 54:367-402. Efimov, V.A., Kalinkina, A.L., Chakhmakhcheva, O.G., Schmaltz Hill, T. and Jayaraman, K. 1995. New efficient sulfurizing reagents for the preparation of oligodeoxyribonucleotide phosphorothioate analogues. Nucl. Acids Res. 23:4029-4033. Farrance, I.K., Eadie, J.S., and Ivarie, R. 1989. Improved chemistry for oligodeoxyribonucleotide synthesis substantially improves restriction enzyme cleavage of a synthetic 35 mer. Nucl. Acids Res. 17:1231-1245. Fourrey, J.-L. and Varenne, J. 1984. Improved procedure for the preparation of deoxynucleoside phosphoramidites: Arylphosphoramidites as new convenient intermediates for oligodeoxynucleotide synthesis. Tetrahedron Lett. 25:45114514. Fourrey, J.-L. and Varenne, J. 1985. Introduction of a nonaqueous oxidation procedure in the phosphite triester route for oligonucleotide synthesis. Tetrahedron Lett. 26:1217-1220. Fourrey, J.-L., Varenne, J., Fontaine, C., Guittet, E., and Yang, Z.W. 1987. A new method for the synthesis of branched ribonucleotides. Tetrahedron Lett. 28:1769-1772. Froehler, B. and Matteucci, M.D. 1983. Substituted 5-phenyltetrazoles: Improved activators of deoxynucleoside phosphoramidites in deoxyoligonucleotide synthesis. Tetrahedron Lett. 24:3171-3174.

Claesen, C., Tesser, G.I., Dreef, C.E., Marugg, J.E., van der Marel, G.A., and van Boom, J.H. 1984. Use of 2-methylsulfonylethyl as a phosphorus protecting group in oligonucleotide synthesis via a phosphite triester approach. Tetrahedron Lett. 25:1307-1310.

Gaytán, P., Yañez, J., Sánchez, F., Mackie, H., and Soberón, X. 1998. Combination of DMTmononucleotide and Fmoc-trinucleotide phosphoramidites in oligonucleotide synthesis affords an automatable codon-level mutagenesis method. Chem. Biol. 5:519-527.

Crooke, S.T. and Bennett, C.F. 1996. Progress in antisense oligonucleotide therapeutics. Annu. Rev. Pharmacol. Toxicol. 36:107-129.

Gryaznov, S.M. and Letsinger, R.L. 1991. Synthesis of oligonucleotides via monomers with unprotected bases. J. Am. Chem. Soc. 113:5876-5877.


Gryaznov, S.M. and Letsinger, R.L. 1992. Selective O-phosphitilation with nucleoside phosphoramidite reagents. Nucl. Acids Res. 20:18791882. Guo, M., Yu, D., Iyer, R.P., and Agrawal, S. 1998. Solid-phase stereoselective synthesis of 2′-Omethyl oligoribonucleoside phosphorothioates using nucleoside oxazaphospholidines. Bioorg. Med. Chem. Lett. 8:2539-2544. Hayakawa, Y. and Kataoka, M. 1998. Facile synthesis of oligodeoxyribonucleotides via the phosphoramidite method without nucleoside base protection. J. Am. Chem. Soc. 120:12395-12401. Hayakawa, Y., Uchiyama, M., and Noyori, R. 1986. Nonaqueous oxidation of nucleoside phosphites to the phosphates. Tetrahedron Lett. 27:41914194. Hayakawa, Y., Kataoka, M., and Noyori, R. 1996. Benzimidazolium triflate as an efficient promoter for nucleotide synthesis via the phosphoramidite method. J. Org. Chem. 61:79967997. Hering, G., Stöcklein-Schneiderwind, R., Ugi, I., Pathak, T., Balgobin, N., and Chattopadhyaya, J. 1985. Preparation and properties of chloroN,N-dialkylamino-2,2,2-trichloroethoxy- and chloro-N,N-dialkylamino-2,2,2-trichloro-1,1dimethylethoxyphosphines and their deoxynucleoside phosphiteamidates. Nucleosides Nucleotides 4:169-171. Iyer, R.P., Phillips, L.R., Egan, W., Regan, J.B., and Beaucage, S.L. 1990. The automated synthesis of sulfur-containing oligodeoxyribonucleotides using 3H-1,2-benzodithiol-3-one 1,1-dioxide as a sulfur-transfer reagent. J. Org. Chem. 55:46934698. Iyer, R.P., Yu, D., Habus, I., Ho, N.-H., Johnson, S., Devlin, T., Jiang, Z., Zhou, W., Xie, J., and Agrawal, S. 1997. N-Pent-4-enoyl (PNT) group as a universal nucleobase protector: Applications in the rapid and facile synthesis of oligonucleotides, analogs, and conjugates. Tetrahedron 53:2731-2750. Iyer, R.P., Guo, M.-J., Yu, D., and Agrawal, S. 1998. Solid-phase stereoselective synthesis of oligonucleoside phosphorothioates: The nucleoside bicyclic oxazaphospholidines as novel synthons. Tetrahedron Lett. 39:2491-2494. Iyer, R.P., Roland, A., Zhou, W., and Ghosh, K. 1999. Modified oligonucleotides—Synthesis, properties and applications. Curr. Opin. Mol. Ther. 1:344-358. Jørgensen, P.N., Stein, P.C., and Wengel, J. 1994. Synthesis of 3′-C-(hydroxymethyl) thymidine: Introduction of a novel class of dexoynucleosides and oligodeoxynucleotides. J. Am. Chem. Soc. 116:2231-2232. Josephson, S., Lagerholm, E., and Palm, G. 1984. Automatic synthesis of oligodeoxynucleotides and mixed oligodeoxynucleotides using the phosphoramidite method. Acta Chem. Scand. B38:539-545.

Kamer, P.C.J., Roelen, H.C.P.F., van den Elst, H., van der Marel, G.A., and van Boom, J.H. 1989. An efficient approach toward the synthesis of phosphorothioate diesters via the Schönberg reaction. Tetrahedron Lett. 30:6757-6760. Kayushin, A.L., Korosteleva, M.D., Miroshnikov, A.I., Kosch, W., Zubov, D., and Piel, N. 1996. A convenient approach to the synthesis of trinucleotide phosphoramidites—Synthons for the generation of oligonucleotide/peptide libraries. Nucl. Acids Res. 24:3748-3755. Khorana, H.G. 1968. Nucleic acid synthesis. Pure Appl. Chem. 17:349-381. Kierzek, R., Rozek, M., and Markiewicz, W.T. 1987. Some steric aspects of synthesis of oligoribonucleotides by phosphoramidite approach on solid support. Nucl. Acids Res. Symp. Ser. No. 18:201204. Kofoed, T. and Caruthers, M.H. 1996. Synthesis of 5′-phosphonate linked thymidine deoxyoligonucleotides. Tetrahedron Lett. 37:6457-6460. Krotz, A.H., Klopchin, P.G., Walker, K.L., Srivatsa, G.S., Cole, D.L., and Ravikumar, V.T. 1997a. On the formation of longmers in phosphorothioate oligodeoxyribonucleotide synthesis. Tetrahedron Lett. 38:3875-3878. Krotz, A.H., Klopchin, P., Cole, D.L., and Ravikumar, V.T. 1997b. Improved purity profile of phosphorothioate oligonucleotides through the use of dimeric phosphoramidite synthons. Nucleosides Nucleotides 16:1637-1640. Kuijpers, W.H.A., Huskens, J., and van Boeckel, C.A.A. 1990. The 2-(acetoxymethyl)benzoyl (AMB) group as a new base-protecting group, designed for the protection of (phosphate) modified oligonucleotides. Tetrahedron Lett. 31:6729-6732. Kumar, G. and Poonian, M.S. 1984. Improvements in oligodeoxyribonucleotide synthesis: Methyl N,N-dialkylphosphoramidite dimer units for solid support phosphite methodology. J. Org. Chem. 49:4905-4912. Lee, H.-J. and Moon, S.-H. 1984. Bis-(N,N-dialkylamino)-alkoxyphosphines as a new class of phosphite coupling agent for the synthesis of oligonucleotides. Chem. Lett. 1229-1232. Letsinger, R.L. and Lunsford, W.B. 1976. Synthesis of thymidine oligonucleotides by phosphite triester intermediates. J. Am. Chem. Soc. 98:36553661. Letsinger, R.L. and Mahadevan, V. 1966. Stepwise synthesis of oligodeoxyribonucleotides on an insoluble polymer support. J. Am. Chem. Soc. 88:5319-5324. Letsinger, R.L. and Ogilvie, K.K. 1969. Synthesis of oligothymidylates via phosphotriester intermediates. J. Am. Chem. Soc. 91:3350-3355. Lyttle, M.H., Napolitano, E.W., Calio, B.L., and Kauvar, L.M. 1995. Mutagenesis using trinucleotide β-cyanoethyl phosphoramidites. BioTechniques 19:274-280.



Mathis, R., Lafaille, L., and Burgada, R. 1974. Fréquence d’absorption de la liaison P-N dans des composés du phosphore tricoordonné. Spectrochim. Acta, Part A 30:357-370. Matteucci, M.D. and Caruthers, M.H. 1981. Synthesis of deoxyoligonucleotides on a polymer support. J. Am. Chem. Soc. 103:3185-3191. McBride, L.J. and Caruthers, M.H. 1983. An investigation of several deoxynucleoside phosphoramidites useful for synthesizing deoxyoligonucleotides. Tetrahedron Lett. 24:245-248. McBride, L.J., Kierzek, R., Beaucage, S.L., and Caruthers, M.H. 1986. Amidine protecting groups for oligonucleotide synthesis. J. Am. Chem. Soc. 108:2040-2048. Montserrat, F.X., Cerandas, A., Eritja, R., and Pedroso, E. 1994. Criteria for the economic large scale solid-phase synthesis of oligonucleotides. Tetrahedron 50:2617-2622. Moore, M.F. and Beaucage, S.L. 1985. Conceptual basis of the selective activation of bis(dialkylamino)methoxyphosphines by weak acids and its application toward the preparation of deoxynucleoside phosphoramidites in situ. J. Org. Chem. 50:2019-2025. Mullah, B., Andrus, A., Zhao, H., and Jones, R.A. 1995. Oxidative conversion of N-dimethylformamidine nucleosides to N-cyano nucleosides. Tetrahedron Lett. 36:4373-4376. Natt, F. and Häner, R. 1997. Lipocap: A lipophilic phosphoramidite-based capping reagent. Tetrahedron 53:9629-9636. Neuner, P., Cortese, R., and Monaci, P. 1998. Codon-based mutagenesis using dimer-phosphoramidites. Nucl. Acids Res. 26:1223-1227. Nielsen, J. and Caruthers, M.H. 1988. Directed Arbuzov-type reactions of 2-cyano-1,1-dimethylethyl deoxynucleoside phosphites. J. Am. Chem. Soc. 110:6275-6276. Ono, A., Matsuda, A., Zhao, J., and Santi, D.V. 1995. The synthesis of blocked triplet-phosphoramidites and their use in mutagenesis. Nucl. Acids Res. 23:4677-4682. Polushin, N.N. 1996. Synthesis of functionally modified oligonucleotides from methoxyoxalamido precursors. Tetrahedron Lett. 37:32313234. Pon, R.T., Usman, N., Damha, M.J., and Ogilvie, K.K. 1986. Prevention of guanine modification and chain cleavage during the solid phase synthesis of oligonucleotides using phosphoramidite derivatives. Nucl. Acids Res. 14:6453-6470.


Pon, R.T. 1987. Enhanced coupling efficiency using 4-dimethylamino pyridine (DMAP) and either tetrazole, 5-(o-nitrophenyl) tetrazole or 5-(pnitrophenyl) tetrazole in the solid phase synthesis of oligoribonucleotides by the phosphoramidite procedure. Tetrahedron Lett. 28: 3643-3646.

Rao, M.V. and Macfarlane, K. 1994. Solid phase synthesis of phosphorothioate oligonucleotides using benzyltriethylammonium tetrathiomolybdate as a rapid sulfur transfer reagent. Tetrahedron Lett. 35:6741-6744. Rao, M.V., Reese, C.B., and Zhengyun, Z. 1992. Dibenzoyl tetrasulphide—A rapid sulphur transfer agent in the synthesis of phosphorothioate analogues of oligonucleotides. Tetrahedron Lett. 33:4839-4842. Reddy, M.P., Hanna, N.B., and Farooqui, F. 1994. Fast cleavage and deprotection of oligonucleotides. Tetrahedron Lett. 35:4311-4314. Reddy, M.P., Hanna, N.B., and Farooqui, F. 1997. Ultrafast cleavage and deprotection of oligonucleotides—Synthesis and use of CAc derivatives. Nucleosides Nucleotides 16:1589-1598. Regan, J.B., Phillips, L.R., and Beaucage, S.L. 1992. Large-scale preparation of the sulfurtransfer reagent 3H-1,2-benzodithiol-3-one 1,1dioxide. Org. Prep. Proc. Int. 24:488-492. Roelen, H.C.P.F., Kamer, P.C.J., van den Elst, H., van der Marel, G.A., and van Boom, J.H. 1991. A study on the use of phenylacetyl disulfide in the solid-phase synthesis of oligodeoxynucleoside phosphorothioates. Recl. Trav. Chim. Pays-Bas 110:325-331. Sanghvi, Y.S. and Cook, P.D. 1994. Carbohydrates: Synthetic methods and applications in antisense therapeutics. In ACS Symposium Series 580— Carbohydrate Modifications in Antisense Research (Y.S. Sanghvi and P.D. Cook, eds.) pp. 1-22. American Chemical Society, Washington, D.C. Schulhof, J.C., Molko, D., and Teoule, R. 1987. The final deprotection step in oligonucleotide synthesis is reduced to a mild and rapid ammonia treatment by using labile base-protecting groups. Nucl. Acids Res. 15:397-416. Schwartz, M.E., Breaker, R.R., Asteriadis, G.T., deBear, J.S., and Gough, G.R. 1992. Rapid synthesis of oligoribonucleotides using 2′-O-(o-nitrobenzyloxymethyl)-protected monomers. Bioorg. Med. Chem. Lett. 2:1019-1024. Sekine, M., Masuda, N., and Hata, T. 1986. Synthesis of oligodeoxyribonucleotides involving a rapid procedure for the removal of base-protecting groups by use of the 4,4′,4′′-tris(benzoyloxy)trityl (TBTr) group. Bull. Chem. Soc. Jpn. 59:1781-1789. Seliger, H. and Gupta, K.C. 1985. Three-phase synthesis of oligonucleotides. Angew. Chem. Int. Ed. Engl. 24:685-687. Sinha, N.D., Biernat, J., McManus, J., and Köster, H. 1984. Polymer support oligonucleotide synthesis XVIII: Use of β-cyanoethyl-N,N-dialkylamino-/N-morpholino phosphoramidite of deoxynucleosides for the synthesis of DNA fragments simplifying deprotection and isolation of the final product. Nucl. Acids Res. 12:45394557.


Sinha, N.D., Davis, P., Usman, N., Pérez, J., Hodge, R., Kremsky, J., and Casale, R. 1993. Labile exocyclic amine protection of nucleosides in DNA, RNA and oligonucleotide analog synthesis facilitating N-deacylation, minimizing depurination and chain degradation. Biochimie 75:1323. Sondek, J. and Shortle, D. 1992. A general strategy for random insertion and substitution mutagenesis: Substoichiometric coupling of trinucleotide phosphoramidites. Proc. Natl. Acad. Sci. U.S.A. 89:3581-3585. Stec, W.J. and Zon, G. 1984. Stereochemical studies of the formation of chiral internucleotide linkages by phosphoramidite coupling in the synthesis of oligodeoxyribonucleotides. Tetrahedron Lett. 25:5279-5282. Stec, W.J., Uznanski, B., Wilk, A., Hirschbein, B.L., Fearon, K.L., and Bergot, B.J. 1993. Bis(O,O-diisopropoxy phosphinothioyl) disulfide—A highly efficient sulfurizing reagent for cost-effective synthesis of oligo(nucleoside phosphorothioate)s. Tetrahedron Lett. 34:5317-5320. Tanaka, T. and Letsinger, R.L. 1982. Syringe method for the stepwise chemical synthesis of oligonucleotides. Nucl. Acids Res. 10:32493260. Tener, G.M. 1961. 2-Cyanoethyl phosphate and its use in the synthesis of phosphate esters. J. Am. Chem. Soc. 83:159-168. Ugi, I., Jacob, P., Landgraf, B., Rupp, C., Lemmen, P., and Verfürth, U. 1988. Phosphite oxidation and the preparation of five-membered cyclic phosphorylating reagents via the phosphites. Nucleosides Nucleotides 7:605-608. Usman, N., Pon, R.T., and Ogilvie, K.K. 1985. Preparation of ribonucleoside 3′-O-phosphoramidites and their application to the automated solid phase synthesis of oligonucleotides. Tetrahedron Lett. 26:4567-4570. Uznanski, B., Grajkowski, A., and Wilk, A. 1989. The isopropoxyacetic group for convenient base protection during solid-support synthesis of oligodeoxyribonucleotides and their triester analogs. Nucl. Acids Res. 17:4863-4871. Vargeese, C., Carter, J., Yegge, J., Krivjansky, S., Settle, A., Kropp, E., Peterson, K., and Pieken, W. 1998. Efficient activation of nucleoside phosphoramidites with 4,5-dicyanoimidazole during oligonucleotide synthesis. Nucl. Acids Res. 26:1046-1050. Virnekäs, B., Ge, L., Plückthun, A., Schneider, K.C., Wellnhofer, G., and Moroney, S.E. 1994. Trinucleotide phosphoramidites: ideal reagents for the synthesis of mixed oligonucleotides for random mutagenesis. Nucl. Acids Res. 22:5600-5607. Vu, H. and Hirschbein, B.L. 1991. Internucleotide phosphite sulfurization with tetraethylthiuram disulfide. Phosphorothioate oligonucleotide synthesis via phosphoramidite chemistry. Tetrahedron Lett. 32:3005-3008.

Vu, H., McCollum, C., Jacobson, K., Theisen, P., Vinayak, R., Spiess, E., and Andrus, A. 1990. Fast oligonucleotide deprotection phosphoramidite chemistry for DNA synthesis. Tetrahedron Lett. 31:7269-7272. Weber, H. and Khorana, H.G. 1972. CIV. Total synthesis of the structural gene for an alanine transfer ribonucleic acid from yeast. Chemical synthesis of an icosadeoxynucleotide corresponding to the nucleotide sequence 21 to 40. J. Mol. Biol. 72:219-249. Weiss, P. 1998. TOM-protecting group—A major improvement in RNA synthesis. The Glen Report 11:2-4. Wilk, A., Srinivasachar, K., and Beaucage, S.L. 1997. N-Trifluoroacetylamino alcohols as phosphodiester protecting groups in the synthesis of oligodeoxyribonucleotides. J. Org. Chem. 62:6712-6713. Wincott, F., DiRenzo, A., Shaffer, C., Grimm, S., Tracz, D., Workman, C., Sweedler, D., Gonzalez, C., Scaringe, S., and Usman, N. 1995. Synthesis, deprotection, analysis and purification of RNA and ribosymes. Nucl. Acids Res. 23:2677-2684. Wolter, A., Biernat, J., and Köster, H. 1986. Polymer support oligonucleotide synthesis XX: Synthesis of a henhectacosa deoxynucleotide by use of a dimeric phosphoramidite synthon. Nucleosides Nucleotides 5:65-77. Wu, X. and Pitsch, S. 1998. Synthesis and pairing properties of oligoribonucleotide analogues containing a metal-binding site attached to β-D-allofuranosyl cytosine. Nucl. Acids Res. 26:43154323. Xin, Z. and Just, G. 1996. Diastereoselective synthesis of phosphite triesters. Tetrahedron Lett. 37:969-972. Xu, Q., Barany, G., Hammer, R.P., Musier-Forsyth, K. 1996. Efficient introduction of phosphorothioates into RNA oligonucleotides by 3ethoxy-1,2,4-dithiazoline-5-one (EDITH). Nucl. Acids Res. 24:3643-3644. Yamana, K., Nishijima, Y., Oka, A., Nakano, H., Sangen, O., Ozaki, H., and Shimidzu, T. 1989. A simple preparation of 5′-O-dimethoxytrityl deoxyribonucleoside 3′-O-phosphor-bisdiethylamidites as useful intermediates in the synthesis of oligodeoxyribonucleotides and their phosphorodiethylamidate analogs on a solid support. Tetrahedron 45:4135-4140. Yu, D., Tang, J.-Y., Iyer, R.P., and Agrawal, S. 1994. Diethoxy N,N-diisopropyl phosphoramidite as an improved capping reagent in the synthesis of oligonucleotides using phosphoramidite chemistry. Tetrahedron Lett. 35:8565-8568. Zehl, A., Starke, A., Cech, D., Hartsch, T., Merkl, R., and Fritz, H.-J. 1996. Efficient and flexible access to fully protected trinucleotides suitable for DNA synthesis by automated phosphoramidite chemistry. Chem. Commun. 26772678.



Zemlicka, J. and Holy, A. 1967. Preparation of N-dimethylaminomethylene derivatives—A new method of a selective substitution of nucleoside amino groups. Coll. Czech. Chem. Commun. 32:3159-3168. Zhang, Z., Nichols, A., Alsbeti, M., Tang, J.X., and Tang, J.Y. 1998. Solid phase synthesis of oligonucleotide phosphorothioate analogues using bis(ethoxythiocarbonyl)tetrasulfide as a new sulfur-transfer reagent. Tetrahedron Lett. 39:24672470. Zhang, Z., Nichols, A., Tang, J.X., Han, Y., and Tang, J.Y. 1999. Solid phase synthesis of oligonucleotide phosphorothioate analogues using 3methyl-1,2,4-dithiazolin-5-one (MEDITH) as a new sulfur-transfer reagent. Tetrahedron Lett. 40:2095-2098.

Zon, G., Gallo, K.A., Samson, C.J., Shao, K., Summers, M.F., and Byrd, R.A. 1985. Analytical studies of ‘mixed sequence’ oligodeoxyribonucleotides synthesized by competitive coupling of either methyl- or β-cyanoethyl-N,N-diisopropylamino phosphoramidite reagents, including 2′deoxyinosine. Nucl. Acids Res. 13:8181-8196.

Contributed by Serge L. Beaucage Center for Biologics Evaluation and Research Food and Drug Administration Bethesda, Maryland Marvin H. Caruthers University of Colorado Boulder, Colorado

Zhao, Z. and Caruthers, M.H. 1996. Synthesis and preliminary biochemical studies with 5′-deoxy5′-methylidyne phosphonate linked thymidine oligonucleotides. Tetrahedron Lett. 37:62396242.



Synthesis of Oligodeoxyribo- and Oligoribonucleotides According to the H-Phosphonate Method This protocol outlines a general procedure for the preparation of oligodeoxyribo- and oligoribonucleotides using H-phosphonate monomers. It is followed by an in-depth discussion of the advantages of the H-phosphonate approach as well as its underlying chemistry (see Commentary). The preparation of the H-phosphonate monomers can be achieved by a variety of methods; these are presented in UNIT 2.6.

UNIT 3.4

BASIC PROTOCOL

Oligonucleotide synthesis employing H-phosphonates is considerably simpler than synthesis using the phosphotriester or phosphoramidite procedures. The elongation cycle includes only two chemical steps: deprotection of the terminal 5 -OH function of the support-bound oligonucleotide, and coupling of the 5 -OH with a nucleoside 3 -Hphosphonate in the presence of a condensing agent (Fig. 3.4.1). After completion of the desired number of elongation cycles (i.e., assembly of the oligomeric chain), a single oxidation cycle is performed to convert the internucleoside H-phosphonate functions to phosphodiesters (or some analog, such as phosphorothioates). Finally, the linkage between the oligomer and the support is cleaved under ammonolytic conditions, which is also the final deprotection step for oligodeoxyribonucleotide and oligoribonucleotide synthesis with 2 -O-2-chlorobenzoyl groups (also see UNIT 2.6). Purification by standard methods is then carried out to isolate the oligonucleotides. The synthetic protocol described below was optimized for oligoribonucleotide synthesis with 5 -O-MMTr and 2 -O-TBDMS protection on a modified Gene Assembler (Pharmacia) with a polystyrene support. It also gives good results with 2 -O-alkyl RNA and fairly good results for oligodeoxyribonucleotide synthesis. Controlled-pore glass (CPG) can be used, but has proven to be less reliable in the authors’ experience. The efficiency of each elongation step in solid-phase oligonucleotide synthesis is usually high, but technical aspects of the procedure may have to be adjusted for each particular machine. The most important of these are probably (1) the time of H-phosphonate preactivation before it reaches the solid support, (2) the concentration of the condensing agent, and (3) the proportion of pyridine in the solvent mixture. Some recently introduced condensing agents—e.g., bis(pentafluorophenyl) carbonate (Efimov et al., 1993) and carbonium- and phosphonium-based condensing agents (Wada et al., 1997)—seem to make the condensation less sensitive to these factors. The reaction conditions for the removal of the acid-labile 5 -O-MMTr or 5 -O-DMTr groups (usually 1% to 2% of various haloacetic acids in an anhydrous chlorinated solvent) have been shown not to affect the integrity of the H-phosphonate linkages within a relevant time (Stawinski et al., 1988). The most important factors affecting the condensation and oxidation steps are discussed later (see Commentary).

Materials 1,2-Dichloroethane (DCE; BDH) over 4-Å molecular sieves Trifluoroacetic acid (TFA; Fluka) Dichloroacetic acid (DCA; Lancaster, 99%), distilled Acetonitrile (MeCN; Lab-Scan) over 3-Å molecular sieves Synthesis of Unmodified Oligonucleotides Contributed by Roger Strömberg and Jacek Stawinski Current Protocols in Nucleic Acid Chemistry (2004) 3.4.1-3.4.15 C 2004 by John Wiley & Sons, Inc. Copyright

3.4.1 Supplement 19

Figure 3.4.1 Condensation of protected nucleoside H-phosphonate monoester with a nucleoside and conversion to the dinucleoside phosphate or backbone-modified analog.

Pyridine (Py; Lab-Scan, Anhydroscan) over 4-Å molecular sieves Protected nucleoside 3 -H-phosphonate building blocks (triethylammonium salts; see Commentary and UNIT 2.6) Pivaloyl chloride (Pv-Cl; Acros Organics, 99%), freshly distilled Polystyrene support (PE Applied Biosystems) loaded at 20 to 30 µmol/g with a 5 -O-(4,4 -dimethoxytrityl) (DMTr)– or 5 -O-(4-monomethoxytrityl) (MMTr)–protected nucleoside succinate (or equivalent nucleoside-loaded solid support) I2 Diethyl ether Concentrated (28% to 32%) aqueous ammonium hydroxide (NH4 OH) or 3:1 (v/v) concentrated NH4 OH/ethanol Triethylamine trihydrofluoride (for RNA synthesis with 2 -O-TBDMS protection) n-Butanol 20 mM sodium acetate buffer, pH 6.5, containing 30% and 10% MeCN LiClO4 0.1 M triethylammonium acetate (TEAA) buffer, pH 6.5 Automated oligonucleotide synthesizer (Gene Assembler, Pharmacia) 5-mL syringes with Luer lock 1.5-mL cryovials with screw caps Glass sintered funnel of coarse porosity Speedvac evaporator and a vacuum pump C18 cartridge (Waters Sep-Pac) Syringe filters (Millex-GV13 filter, 0.22-µm, 13 mm) and disposable syringes 4 × 250–mm Dionex NucleoPac PA-100 column Lyophilizer 4.6 × 150–mm Supelcosil LC-18 (3 µm) column Additional reagents and equipment for automated oligonucleotide synthesis (APPENDIX 3C) and purification by ion-exchange and reversed-phase HPLC (UNIT 10.5) Synthesis of Oligonucleotides According to the H-Phosphonate Method

3.4.2 Supplement 19


Table 3.4.1 Stepwise Oligonucleotide Synthesis Program

Step

Reagenta

Wash

DCE

Detritylation

Time (min)

Flow rate (mL/min)

2.0

2

1.0

2

2.0

2

3.5% DCA/DCE

2.5

2

DCE

2.0

2

MeCN

1.0

2

3:1 (v/v) MeCN/Py

1.0

2

50 mM phosphonate

0.1

1

225 mM Pv-Cl

0.1

1

50 mM phosphonate

0.1

1

225 mM Pv-Cl

0.1

1

50 mM phosphonate

0.1

1

Pump forward

0.4

1

Pump reverse

1.0

0.5

3:1 (v/v) MeCN/Py

1.0

2

MeCN

1.0

2

For 5 -O-MMTr-RNA or 2 -O-alkyl-RNA: 1% TFA/DCE

For 5 -O-DMTr-DNA: 3.5% DCA/DCE

For mixed 5 -O-MMTr-RNA/2 -O-alkyl-RNA and DNA: Wash

Coupling

Wash Total time per cycle

10.9–12.4

a Abbrevations: DCA, dichloroacetic acid; DCE, 1,2-dichloroethane; DMTr, 4,4 -dimethoxytrityl; MeCN, acetonitrile;

MMTr, 4-methoxytrityl; Pv-Cl, pivaloyl chloride; Py, pyridine; TFA, trifluoroacetic acid.

Prepare reagents 1. Charge the synthesizer with solvents and detritylation solution needed for the steps shown in Table 3.4.1. These can usually be kept on the synthesizer until they are consumed.

2. Dissolve the protected nucleoside 3 -H-phosphonates (triethylammonium salts) in pyridine and evaporate the solvent (two times) under reduced pressure in a rotary evaporator. Use 15 µmol/coupling plus 15 µmol extra for margins and priming of solutions. 3. Dissolve the residue in 3:1 (v/v) MeCN/pyridine to a concentration of 50 mM and transfer to appropriate vessels for attachment to the synthesizer. 4. Prepare a solution of 225 mM pivaloyl chloride in 3:1 (v/v) MeCN/pyridine directly in the vessel that will be used for synthesis. Prepare 0.2 mL/coupling plus 1 to 2 mL extra for margins. Attach the vessel to the synthesizer. 5. Prime the solutions and connect a column/cartridge filled with 0.2 to 1 µmol nucleoside-loaded support. 6. Immediately before the final oxidation step, prepare an iodine solution by dissolving 0.4 g I2 in 20 mL of 98:2 (v/v) pyridine/water.

Synthesis of Unmodified Oligonucleotides

The solution should be prepared no more than 30 min before use.


Supplement 19

Table 3.4.2 Final Oxidation Program

Step

Reagenta

Wash

Time (min)

Flow rate (mL/min)

DCE

2

2

Detritylation

1% TFA/DCE

1

2

Wash

DCE

2

2

MeCN

1

2

Oxidation

2% I2 in 98:2 (v/v) Py/H2 O

30

0.5

Wash

MeCN

10

2

Total time

46

a Abbrevations: DCE, 1,2-dichloroethane; MeCN, acetonitrile; Py, pyridine; TFA, trifluoroacetic acid.

Synthesize oligonucleotide 7. Perform automated oligonucleotide synthesis (APPENDIX 3C) using the synthesis cycle shown in Table 3.4.1, with a final oxidation step as shown in Table 3.4.2. In the coupling step, program the synthesis so that the H-phosphonate and pivaloyl chloride are taken up in alternating 100-µL portions. The volume of the tubing between the valve and the column (including the pump hose of the peristaltic pump, through which the reagents flow) is 0.4 mL, which means that the front of the condensation mixture reaches the column when the last segment is taken up from the reagent bottles. The segments are then passed through the column (0.4 min at 1 mL/min), and then the flow is lowered and the direction reversed (1 min at 0.5 mL/min), giving a total condensation cycle time of 1.9 min, with an effective condensation time of ∼1.5 min. The effective time of preactivation for the nucleoside H-phosphonate when it first reaches the column is ∼0.3 min, for a total of 1.8 min at the end of condensation.

Purify oligonucleotide 8. Remove the cartridge/column from the synthesizer and wash the support with ∼10 mL diethyl ether using a 5-mL syringe with a Luer fitting. 9. Air dry the support using the above 5-mL syringe with a Luer fitting by pushing air through the column a few times. 10. Remove the support from the cartridge and transfer it to a 1.5-mL cryovial (with screw cap). 11. Carry out ammonolysis by adding 3:1 (v/v) concentrated NH4 OH/ethanol (∼1 mL for up to a 1-µmol scale synthesis) and incubating 8 to 16 hr at 20◦ to 25◦ C (8-hr incubations for sequences up to 25-mers; 16-hr incubations for longer sequences, e.g., 50- to 60-mers). The above conditions are for RNA oligomers with N2 -phenoxyacetyl guanosine protection, N6 -butyryl adenosine protection, and N4 -propionyl cytidine protection, which are recommended to avoid cleavage of TBDMS groups (and subsequent cleavage of RNA) upon ammonolysis at elevated temperatures (Stawinski et al., 1988), which is required for more stable base protection. Other reagents and conditions may be needed for other protecting groups. For DNA or for 2 -O-alkyl-RNA, concentrated NH4 OH should be used alone for ∼20 hr at 55◦ to 60◦ C.

12. Filter off the support using a glass sintered funnel of medium porosity. Synthesis of Oligonucleotides According to the H-Phosphonate Method

13. Wash support with 1 mL of 3:1 (v/v) concentrated NH4 OH/ethanol (or concentrated NH4 OH alone) and combine the filtrates. 14. Concentrate the oligonucleotide using a Speedvac evaporator and a vacuum pump.

3.4.4 Supplement 19


15. For RNA synthesis with 2 -O-TBDMS protection, perform deprotection step with triethylamine trihydrofluoride (Westman and Strömberg, 1994) using the following steps: a. Dissolve the residue in 0.3 mL neat triethylamine trihydrofluoride and incubate 14 to 16 hr at room temperature. b. Add 30 µL water and 1 mL n-butanol, and incubate 1 hr at –20◦ C. c. Centrifuge and remove the liquid supernatant. Dissolve the pellet in HPLC buffer (see step 16). 16. Purify by ion-exchange and reversed-phase HPLC (UNIT 10.5). a. Dissolve the deprotected oligoribonucleotides in 0.5 mL of 20 mM sodium acetate buffer, pH 6.5, containing 30% MeCN and filter through a disposable C18 cartridge. b. Wash the cartridge with 1 mL of buffer, then combine fractions and filter through a disposable syringe attached to a 0.22-µm filter before HPLC purification. c. Purify by anion-exchange HPLC on a 4 × 250–mm Dionex NucleoPac PA-100 column using a linear gradient of LiClO4 in 20 mM sodium acetate buffer, pH 6.5, containing 10% MeCN at a flow rate of 1 mL/min. For analysis, inject 0.2 OD260 units; for purification, inject 15 to 30 OD260 units of crude oligoribonucleotides. d. Lyophilize the collected fractions, dissolve in 1 mL of 0.1 M triethylammonium acetate (TEAA) buffer, pH 6.5, and filter through a disposable syringe attached to a 0.22-µm filter. e. Further purify by reversed-phase HPLC on a 4.6 × 150–mm Supelcosil LC-18 column using a linear gradient of MeCN in 0.1 M TEAA buffer, pH 6.5, at a flow rate of 1 mL/min. f. Collect the fractions containing the product, lyophilize, dissolve in 1 mL water, and lyophilize again.

COMMENTARY Background Information General information and synthetic strategies Although the most common method today for synthesis of oligonucleotides and their analogs is the phosphoramidite approach (Beaucage and Iyer, 1993, see also UNIT 3.3), the newer H-phosphonate methodology can often be a preferred alternative (Garegg et al., 1985, 1986a,b,c; Froehler and Matteucci, 1986; Froehler et al., 1986). The use of H-phosphonates in nucleotide synthesis was pioneered by Sir Todd’s group in Cambridge, UK, who in 1952 demonstrated the formation of H-phosphonate diesters in a condensation reaction of H-phosphonate monoesters with a protected nucleoside, promoted by diphenyl phosphorochloridate (Corby et al., 1952; Hall et al., 1957). This chemical principle was, however, not explored further; it was rediscovered three decades later (Garegg et al., 1985, 1986c) and explored for oligonu-

cleotide synthesis (Froehler and Matteucci, 1986; Froehler et al., 1986; Garegg et al., 1986a,b,c, 1987a). The method consists of condensing a protected nucleoside H-phosphonate monoester (UNIT 2.6) with a nucleoside in the presence of a coupling agent to produce the corresponding dinucleoside H-phosphonate diester. This, under various experimental conditions, can be converted to the dinucleoside phosphate or to a variety of backbone-modified analogs, e.g., phosphorothioates, phosphoramidates, etc. (Fig. 3.4.1). The condensation step of the elongation cycles (i.e., the formation of an internucleoside H-phosphonate linkage between a nucleoside 3 -H-phosphonate monoester and the supportbound 5 -hydroxylic component) is usually carried out in pyridine-acetonitrile mixtures. Out of the various condensing agents initially tested, pivaloyl chloride (Pv-Cl) gave the best results in automated solid-support synthesis

Synthesis of Unmodified Oligonucleotides


Supplement 19

Synthesis of Oligonucleotides According to the H-Phosphonate Method

of oligonucleotides, and it is still the most frequently used reagent. The reaction in pyridine or acetonitrile-pyridine mixtures using 2 to 5 equiv of Pv-Cl is usually fast and goes to completion in 99%. Once the oligomer has been synthesized, deprotection of the methyl phosphate group is effected by disodium-2-carbamoyl-2-cyanoethylene-1,1-dithiolate (10 min;- Dahl et al., 1990); followed by treatment with aqueous 40% methylamine at 55°C for 10 min. The 2′-protected oligomer can then be analyzed; purified, if necessary; and then stored. To remove the modified 2′-O-orthoester, the oligoribonucleotide is heated to 55°C for 10 min in a pH 3 buffer, followed by incubation at pH 7.7 to 8.0 for 10 min at 55°C. This final step cleaves any remaining 2′-O-formyl groups that result from the orthoester deprotection. Syntheses of

Significant advances in RNA biology and biochemistry can be achieved only through concomitant advances in RNA chemistry. The current state of the art in ribozyme research would not have been possible without the recent improvements in RNA synthesis. The current technology, however, is still limiting. There is no report of routine syntheses of tRNAs or even hairpin ribozymes. Until RNA synthesis chemistry can provide oligoribonucleotides as readily as DNA, the search for new and better methods for the synthesis of RNA will continue. Currently, the 5′-O-DMTr/2′-O-TBDMS is the benchmark for the synthesis of oligoribonucleotides (Usman et al., 1987; Scaringe et al., 1990; Sproat et al., 1995; Wincott et al., 1995). The use of the TBDMS-protecting group (S.12) was first described in the 1970s. In the ensuing years, many other methods for the synthesis of RNA were developed, but none has gained the popularity of the TBDMS chemistry. Recent advances in the use of this silyl chemistry in terms of synthesis (Sproat et al., 1995; Wincott et al., 1995; Vargeese et al., 1998) and deprotection (Sproat et al., 1995; Wincott et al., 1995; Bellon, 1999) have made it an even more viable approach to the production of oligoribonucleotides. In the early 1990s the 5′-O-DMTr/2′O-Fpmp strategy to RNA synthesis showed great promise (Rao et al., 1993; Capaldi and Reese, 1994). Since that time, however, there have been very few reports of successful RNA syntheses using this protocol, although these monomers are commercially available. The results obtained with the 2′-O-(o-nitrobenzyloxymethyl) (S.21; Schwartz et al., 1992) and 2′-O-(p-nitrobenzyloxymethyl) (S.22; Gough et al., 1996) groups also appeared quite encouraging. Again, since the initial reports describing this chemistry, there have been few follow-ups,



and the use of these 2′-protecting groups does not appear to have gained an appreciable audience beyond its initial developers. Other 2′-protecting groups that, like TBDMS and Fpmp, are compatible with current DNA synthesis protocols are the convertible protecting groups S.18 (Rastogi and Usher, 1995) and S.19 (Pfleiderer et al., 1996). These 2′-O-acetal-derived groups look interesting, but there have been few reports since the initial publications. Of the synthetic methods that have been designed specifically for RNA synthesis, none is currently commercially available. Around 1990, there were reports citing the combination of 5′-O-Fmoc and either 2′-O-Mthp (Lehmann et al., 1989) or 2′-O-IPE (Ogawa et al., 1991) that provided good-quality oligoribonucleotides; however, the longest oligomer synthesized was a 21-mer. No further communication regarding oligoribonucleotide synthesis with these protecting groups have surfaced. The 5′-O-SIL/2′-O-ACE protocol, however, looks very attractive (Scaringe et al., 1998). The quality of the product is excellent, and oligomers of up to 36 residues have been synthesized. Currently, none of these amidites is commercially available, although efforts are under way to commercialize the 5′-O-SIL/2′-O-ACE method. It seems clear that TBDMS chemistry is the current choice for the synthesis of oligoribonucleotides. The amidites are commercially available, and quality products can be produced on a reasonable scale. RNA synthesis chemistry using the 2′-O-TBDMS group, however, has not yet reached the level achieved by DNA synthesis. As a result, the search for improved protocols or new approaches altogether persists.

LITERATURE CITED Andrus, A., Beaucage, S., Ohms, J., and Wert, K. 1986. American Chemical Society Meeting, New York, Organic Division, Abstract 333. Beaucage, S.L. and Iyer, R.P. 1992. Advances in the synthesis of oligonucleotides by the phosphoramidite approach. Tetrahedron 48:22232311. Beijer, B., Sulston, I., Sproat, B.S., Rider, P., Lamond, A.I., and Neuner, P. 1990. Synthesis and applications of oligoribonucleotides with selected 2′-O-methylation using the 2′-O-[1-(2fluorophenyl)-4-methoxypiperidin-4-yl] protecting group. Nucl. Acids Res. 18:5143-5151. Strategies for Oligoribonucleotide Synthesis According to the Phosphoramidite Method

Bellon, L. 2000. Oligoribonucleotides with 2′-O-(tbutyldimethylsilyl) groups. In Current Protocols in Nucleic Acid Chemistry (S.L. Beaucage, D.E. Bergstrom, G.D. Glick and R.A. Jones, eds.) in press. John Wiley & Sons, New York.

Bergmann, F. and Pfleiderer, W. 1994a. Solid-phase synthesis of oligoribonucleotides using the 2dansylethoxycarbonyl group for 5′-hydroxy protection. Helvetica Chim. Acta 77:481-500. Bergmann, F. and Pfleiderer, W. 1994b. The 2-dansylethoxycarbonyl group for the protection of the 5′-hydroxy function in oligoribonucleotide synthesis. Helvetica Chim. Acta 77:989-998. Capaldi, D.C. and Reese, C.B. 1994. Use of the 1-(2-fluorophenyl)-4-methoxypiperidin-4-yl (Fpmp) and related protecting groups in oligoribonucleotide synthesis: Stability of internucleotide linkages to aqueous acid. Nucl. Acids Res. 22:2209-2216. Cech, T. 1992. Ribozyme engineering. Curr. Opin. Struct. Biol. 2:605-609. Chaix, C., Duplaa, A.M., Molko, D., and Téoule, R. 1989. Solid phase synthesis of the 5′-half of the initiator t-RNA from B. subtilis. Nucl. Acids Res. 17:7381-7393. Christodoulou, C., Agrawal, S., and Gait, M.J. 1986. Incompatibility of acid-labile 2′ and 5′ protecting groups for solid-phase synthesis of oligoribonucleotides. Tetrahedron Lett. 27:1521-1522. Cook, K.S., Fisk, G.J., Hauber, J., Usman, N., Daly, T.J., and Rusche, J.R. 1991. Characterization of HIV-1 REV protein: Binding stoichiometry and minimal RNA substrate. Nucl. Acids Res. 19:1577-1583. Dahl, B.J., Bjergarde, K., Henriksen, L., and Dahl, O. 1990. Deoxyribonucleoside phosphorodithioates. Preparation of dinucleoside phosphorodithioates from nucleoside thiophosphoramidites. Acta Chem. Scand. 44:639-641. Damha, M.J., Giannaris, P.A., and Zabarylo, S.V. 1990. An improved procedure for derivitization of controlled-pore glass beads for solid-phase oligonucleotide synthesis. Nucl. Acids Res. 18:3813-3820. deBear, J.S., Hayes, J.A., Koleck, M.P., and Gough, G.R. 1987. A universal glass support for oligonucleotide synthesis. Nucleosides Nucleotides 6:821-830. Francklyn, C. and Schimmel, P.R. 1989. Aminoacylation of RNA minihelixes with alanine. Nature 337:478-481. Gait, M.J., Pritchard, C., and Slim, G. 1991. Oligoribonucleotide synthesis. In Oligonucleotides and Analogues, A Practical Approach (F. Eckstein, ed.) pp 25-48. Oxford University Press, Oxford. Gasparutto, D., Livache, T., Bazin, H., Duplaa, A.M., Guy, A., Khorlin, A., Molko, D., Roget, A., and Téoule, R. 1992. Chemical synthesis of a biologically active natural tRNA with its minor bases. Nucl. Acids Res. 20:5159-5166. Gold, L. 1988. Posttranscriptional regulatory mechanisms in Escherichia coli. Annu. Rev. Biochem. 57:199-233. Gough, G.R., Miller, T.J., and Mantick, N.A. 1996. p-Nitrobenzyloxymethyl: A new fluoride-removable protecting group for ribonucleoside 2′hydroxyls. Tetrahedron Lett. 37:981-982.


Hayakawa, Y., Kataoka, M., and Noyori, R. 1996. Benzimidazolium triflate as an efficient promoter for nucleotide synthesis via the phosphoramidite approach. J. Org. Chem. 61:79967997.

Ogawa, T., Hosaka, H., Makita, T., and Takaku, H. 1991. Solid-phase synthesis of oligoribonucleotides using 5′-9-fluorenylmethoxycarbonyl and 2′-1-(isopropoxyl)ethyl protection. Chem. Lett. 1169-1172.

Hayes, J.A., Brunden, M.J., Gilham, P.T. and Gough, G.R. 1985. High-yield synthesis of oligoribonucleotides using o-nitrobenzyl protection of 2′-hydroxyls. Tetrahedron Lett. 26:24072410.

Ohtsuka, E., Fujiyama, K., and Ikehara, M. 1981. Studies on transfer ribonucleic acids and related compounds. XL. Synthesis of an eicosaribonucleotide corresponding to residues 35-54 of tRNAfMet from E. coli. Nucl. Acids Res. 9:35033522.

Himmelsbach, F., Schulz, B.S., Trichtinger, T., Charubala, R., and Pfleiderer, W. 1984. The pnitrophenylethyl (Npe) group. Tetrahedron 40:59-72. Hogrefe, R.I., McCaffrey, A.P., Borozdina, L.U., McCampbell, E.S., and Vaghefi, M.M. 1994. Effect of excess water on the desilylation of oligoribonucleotides using tetrabutylammonium fluoride. Nucl. Acids Res. 21:4739-4741. Iwai, S. and Ohtsuka, E. 1988. 5′-Levulinyl and 2′-tetrahydrofuranyl protection for the synthesis of oligoribonucleotides by the phosphoramidite approach. Nucl. Acids Res. 16:9443-9456. Iwai, S., Yamada, E., Asaka, M., Hayase,Y., Inoue, H., and Ohtsuka, E. 1987. A new solid-phase synthesis of oligoribonucleotides by the phosphoro-p-anisidate method using tetrahydrofuranyl protection of 2′-hydroxyl groups. Nucl. Acids Res. 15:3761-3772. Johnson, K.A. and Benkovic, S.J. 1990. Analysis of protein function by mutagenesis. In The Enzymes, Vol. 19 (Sigman, D.S. and Boyer, P.D., eds) pp. 159-211. Academic Press, San Diego. Karaoglu, D. and Thurlow, D.L. 1991 A chemical interference study on the interaction of ribosomal protein L11 from in Escherichia coli with RNA molecules containing its binding site from 23S rRNA. Nucl. Acids Res. 19:5293-5300. Klosel, R., Konig, S., Lehnhoff, S., and Karl, R.M. 1996. The 1,1-dianisyl-2,2,2-trichloroethyl group as a 2′-hydroxyl protection of ribonucleotides. Tetrahedron 52:1493-1502. Lehmann, C., Xu, Y.Z., Christodoulou, C., Tan, Z.K., and Gait, M.J. 1989. Solid-phase synthesis of oligoribonucleotides using 9-fluorenylmethoxycarbonyl (Fmoc) for 5′-hydroxyl protection. Nucl. Acids Res. 17:2379-2390. Leiber, E. and Enkoju, T. 1961. Synthesis and properties of 5-(substituted) mercaptotetrazoles. J. Org. Chem. 26:4472-4479. Lyttle, M.H., Wright, P.B., Sinha, N.D., Bain, J.D., and Chamberlin, A.R. 1991. New nucleoside phosphoramidites and coupling protocols for solid-phase RNA synthesis. J. Org. Chem. 56:4608-4615. McCollum, C. and Andrus, A. 1991. An optimized polystyrene support for rapid, efficient oligonucleotide synthesis. Tetrahedron Lett. 32:40694072.

Pfister, M., Farkas, S., Charubala, R., and Pfleiderer, W. 1988. Recent progress in oligoribonucleotide synthesis. Nucleosides Nucleotides 7:595-600. Pfleiderer, W., Matysiak, S., Bergmann, F., and Schnell, R. 1996. Recent progress in oligonucleotide synthesis. Acta Biochim. Polonica 43:37-44. Pon, R.T., Usman, N., and Ogilvie, K.K. 1988. Derivitization of controlled pore glass beads for solid phase oligonucleotide synthesis. BioTechniques 6:768-774. Rao, M.V., Reese, C.B., Schehlmann, V., and Yu, P.S. 1993. Use of the 1-(2-fluorophenyl)-4methoxypiperidin-4-yl (Fpmp) protecting group in the solid-phase synthesis of oligo- and polyribonucleotides. J. Chem. Soc. Perkin Trans. I:43-55. Rastogi, J. and Usher, D. 1995. A new 2′-hydroxyl protecting group for the automated synthesis of oligoribonucleotides. Nucleic Acids Res. 23:4872-4877. Reddy, M.P., Hanna, N.B., and Farooqui, F. 1994. Fast cleavage and deprotection of oligonucleotides. Tetrahedron Lett. 35:4311-4314. Reddy, M.P., Hanna, N.B., and Farooqui, F. 1995. Methylamine deprotection provides increased yield of oligoribonucleotides. Tetrahedron Lett. 36:8929-8932. Reese, C.B. and Skone, P.A. 1985. Action of acid on oligoribonucleotide phosphotriester intermediates. Effect of released vicinal hydroxy functions. Nucl. Acids Res. 13:3501-. Reese, C.B., Serafinowska, H.T., and Zappia, G. 1986. An acetal group suitable for the protection of 2′-hydroxy functions in rapid oligoribonucleotide synthesis. Tetrahedron Lett. 27:22912294. Sakatsume, O., Yamaguchi, T., Ishikawa, M., Hirao, I., Miura, K., and Takaku, H. 1991a. Solid phase synthesis of oligoribonucleotides by the phosphoramidite approach using 2′-O-1-(2-chloroethox y)ethyl p ro tection. Tetrahedron 47:8717-8728. Sakatsume, O., Ogawa, T., Hosaka, H., Kawashima, M., Takaki, M., and Takaku, H. 1991b. Synthesis and properties of non-hammerhead RNA using 1-(2-chloroethoxy)ethyl group for the protection of 2′-hydroxyl function. Nucleosides Nucleotides 10:141-153. Synthesis of Unmodified Oligonucleotides


Scaringe, S.A., Francklyn, C., and Usman, N. 1990. Chemical synthesis of biologically active oligoribonucleotides using β-cyanoethyl protected ribonucleoside phosphoramidites. Nucl. Acids Res. 18:5433-5341. Scaringe, S.A., Wincott, F.E., and Caruthers, M.H. 1998. Novel RNA synthesis method using 5′-Osilyl-2′-O-orthoester protecting groups. J. Am. Chem Soc. 120:11820-11821. Schwartz, M.E., Breaker, R.R., Asteriadis, G.T., deBear, J.S., and Gough, G.R. 1992. Rapid synthesis of oligoribonucleotides using 2′-O-(o-nitrobenzyloxymethyl)-protected monomers. Bioorg. Med. Chem. Lett. 2:1019-1024. Sinha, N.D., Davis, P., Usman, N., Pérez, J., Hodge, R., Kremsky, J., and Casale, R. 1993. Labile exocyclic amine protection in DNA, RNA and oligonucleotide analog synthesis facilitating Ndeacylation, minimizing depurination and chain degradation. Biochimie 75:13-23. Sproat, B., Colonna, F., Mullah, B., Tsou, D., Andrus, A., Hampel, A., and Vinayak, R. 1995. An efficient method for the isolation and purification of oligoribonucleotides. Nucleosides Nucleotides 14:255-273. Tanaka, T., Tamatsukuri, S., and Ikehara, M. 1986. Solid phase synthesis of oligoribonucleotides using o-nitrobenzyl protection of 2′-hydroxyl via a phosphite triester approach. Nucl. Acids Res. 14:6265-6279. Theisen, P., McCollum, C., and Andrus, A. 1993. N-6-Dialkylformamidine-2′-deoxyadenosine phosphoramidites in oligodeoxynucleotide synthesis. Rapid deprotection of oligodeoxynucleotides. Nucleosides Nucleotides 12:10331046. Usman, N., Pon, R.T., and Ogilvie, K.K. 1985. Preparation of ribonucleoside 3′-O-phosphoramidites and their application to the automated solid phase synthesis of oligonucleotides. Tetrahedron Lett. 26:4567-4570.

Usman, N., Ogilvie, K.K., Jiang, M.-Y., and Cedergren, R.J. 1987. Automated chemical synthesis of long oligoribonucleotides using 2′-O-silylated ribonucleoside 3′-O-phosphoramidites on a controlled-pore glass support: Synthesis of a 43-nucleotide sequence similar to the 3′-half molecule of an in Escherichia coli formylmethionine tRNA. J. Am. Chem. Soc. 109:78457854. Vargeese, C., Carter, J., Yegge, J., Krivjansky, S., Settle, A., Kropp, E., Peterson, K., and Pieken, W. 1998. Efficient activation of nucleoside phosphoramidites with 4,5-dicyanoimidazole during oligonucleotide synthesis. Nucl. Acids Res. 26:1046-1050. Vinayak, R., Ratmeyer, L., Wright, P., Andrus, A., and Wilson, D. 1994. Chemical synthesis of biologically active RNA using labile protecting groups. In Innovations and Perspectives in Solid Phase Synthesis (R. Epton, ed.) pp 45-50. Mayflower Worldwide, Birmingham. Westman, E. and Strömberg, R. 1994. Removal of t-butyldimethylsilyl protection in RNA-synthesis. Triethylamine trihydrofluoride (TEA, 3HF) is a more reliable alternative to tetrabutylammonium fluoride (TBAF). Nucl. Acids Res. 22:2430-2431. Wincott, F., DiRenzo, A., Shaffer, C., Grimm, S., Tracz, D., Workman, C., Sweedler, D., Gonzalez, C., Scaringe, S., and Usman, N. 1995. Synthesis, deprotection, analysis and purification of RNA and ribozymes. Nucl. Acids Res. 23:2677-2684. Wu, T., Ogilvie, K.K., and Pon, R.T. 1988. N-Phenoxyacetylated guanosine and adenosine phosphoramidites in the solid phase synthesis of oligoribonucleotides: Synthesis of a ribozyme sequence. Tetrahedron Lett. 34:4249-4252.

Contributed by Francine E. Wincott Ribozyme Pharmaceuticals, Inc. Boulder, Colorado

Strategies for Oligoribonucleotide Synthesis According to the Phosphoramidite Method


Oligoribonucleotides with 2′-O-(tert-Butyldimethylsilyl) Groups

UNIT 3.6

This unit describes current methodologies to chemically synthesize oligoribonucleotides on solid support by means of automated DNA synthesizers. It also includes an updated collection of protocols describing the deprotection of base-labile phosphate, and nucleobase protecting groups, and fluoride-labile 2′-O-(tert-butyldimethylsilyl) protecting groups of crude synthetic oligoribonucleotides. Small-scale synthesis (i.e., 0.2 µmol scale) of oligoribonucleotide provides ∼400 to 600 µg of purified material, which is usually enough for a wide range of biochemical applications. Such synthesis can be readily achieved using the exocyclic amine-protected 5′-O-dimethoxytrityl-2′-O-tert-butyldimethylsilyl-3′-O-(2-cyanoethyl-N,N-diisopropyl) -ribonucleoside phosphoramidites in combination with the acidic 5-ethylthio-1H-tetrazole (SET, pKa = 4.28) as the activator (Wincott et al., 1995; Vinayak et al., 1995; see Basic Protocol 1). Alternatively, the use of less acidic activators such as 4,5-dicyanoimidazole (DCI, pKa = 5.2; Vargeese et al., 1998) or 1H-tetrazole (TET, pKa = 4.8; Usman et al., 1987; Scaringe et al., 1990; Usman and Cedergren, 1992) also allows for efficient oligoribonucleotide synthesis at comparable scales. DCI is better suited for larger-scale (i.e., >500 µmol) syntheses due to lesser acidity (Vargeese et al., 1998) that may limit the activator-induced detritylation of incoming phosphoramidite during extended coupling time (Krotz et al., 1997). Historically, crude oligoribonucleotides have been fully deprotected with a 3:1 mixture of concentrated ammonium hydroxide/ethanol followed by an n-tetrabutylammonium fluoride (TBAF) treatment (Usman et al., 1987; Stawinsky et al., 1988; see Basic Protocol 2). The use of aqueous methylamine followed by treatment with the triethylamine trihydrofluoride complex (see Alternate Protocol 1) constitutes a significant improvement in the deprotection process (Wincott et al., 1995; Vinayak et al., 1995), alleviating premature deprotection of the 2′-hydroxyl during the basic treatment, eliminating the well-known sensitivity of the TBAF to water (Hogrefe et al., 1993), and shortening the overall deprotection time considerably. Crude oligoribonucleotides can also be efficiently deprotected in a “one-pot” reaction using anhydrous methylamine and neat triethylamine trihydrofluoride (see Alternate Protocol 2). This alternate deprotection protocol eliminates the time-consuming evaporation step, thereby reducing the overall deprotection time to 45 min, which allows for a high-throughput production mode (Bellon, 2000). AUTOMATED OLIGORIBONUCLEOTIDE SYNTHESIS This protocol describes automated chemical synthesis of oligoribonucleotides by means of the phosphoramidite method (UNIT 3.5) according to the synthetic scheme pictured in Figure 3.6.1. The procedure described below was developed for the ABI 394 DNA/RNA synthesizer at the 0.2 µmol scale although it can be modified to utilize any standard synthesizer.

BASIC PROTOCOL 1

Materials Aminomethyl polystyrene (RNA primer solid support) derivatized with 5′-O-DMTr-2′-O-TBDMS-3′-O-succinyl ribonucleosides (Amersham Pharmacia Biotech) Synthesis of Unmodified Oligonucleotides Contributed by Laurent Bellon Current Protocols in Nucleic Acid Chemistry (2000) 3.6.1-3.6.13 Copyright © 2000 by John Wiley & Sons, Inc.

3.6.1 Supplement 1

Phosphoramidite Coupling Reaction DMTrO

B DMTrO

Cycle Entry

5'-Hydroxyl Deprotection

DMTrO

B

O

HO

O

i-Pr2N

B

O

OR

O

N H

O

O P

B

O

OR O

OCE

OR

P OCE

O O

N H

O

OR

EtS

O

O

H N

O

HO

+

B

O

N N N

O

N H

O

B

O O

N H

OR

OR

O

O

To Next Cycle Capping 5'-Unreacted Hydroxyl DMTrO

B

O

DMTrO O

AcO

OR

+

O P OCE O

O

B

O N H

O O

O

PhosphiteTriester Oxidation

O N H

O O

B

O

B

OR

OR

P OCE O

O

OR

O

O N H

O

AcO

+ B

OR

O

B

O N H

O

OR

O

O

Cycle End

DMTr, 4,4′-dimethoxytrityl; R, tert-butyldimethylsilyl; CE, 2-cyanoethyl; Ac, acetyl; B, uracil-1-yl (U); N 4-acetylcytosin-1-yl (CAc); N 6-phenoxyacetyladenin-9-yl (APAC); N2-[(4-isopropyl)phenoxy]acetylguanin-9-yl (GiPrPAC)

Figure 3.6.1 Synthesis of oligoribonucleotide with 2′-O-TBDMS groups on solid-support via the phosphoramidite approach.

RNA phosphoramidites (Amersham Pharmacia Biotech; diluted on the synthesizer to 0.1 M in acetonitrile, using automated protocols) 5′-O-DMTr-N6-(phenoxyacetyl)-2′-O-TBDMS-adenosine-3′-O-(β-cyanoethylN,N-diisopropyl) phosphoramidite 5′-O-DMTr-N2-(isopropylphenoxyacetyl)-2′-O-TBDMS-guanosine-3′-O-(βcyanoethyl-N,N-diisopropyl) phosphoramidite 5′-O-DMTr-N4-(acetyl)-2′-O-TBDMS-cytidine-3′-O-(β-cyanoethyl-N,Ndiisopropyl) phosphoramidite 5′-O-DMTr-2′-O-TBDMS-uridine-3′-O-(β-cyanoethyl-N,N-diisopropyl) phosphoramidite. 3% (v/v) TCA in methylene chloride (PE Biosystems) Cap A: 10% (v/v) acetic anhydride/10% (v/v) 2,6-lutidine in THF (PE Biosystems) Cap B: 16% (v/v) 1-methyl imidazole in THF (PE Biosystems) Iodine solution: 16.9 mM I2/49 mM pyridine/9% (v/v) water in THF (PE Biosystems) Synthesis grade acetonitrile (Burdick & Jackson) Activator (prepare in acetonitrile): 0.25 M 5-ethylthio-1H-tetrazole (SET), made from solid (American International Chemical) or 0.5 M 4,5-dicyanoimidazole solution (DCI), made from solid (Proligo) or 0.45 M 1H-tetrazole (TET; Glen Research). Synthesis columns for 0.2-µmol-scale syntheses (PE Biosystems) ABI 394 DNA/RNA synthesizer (PE Biosystems)

Oligoribonucleotides with 2′-O(tert-Butyldimethylsilyl) Groups

1. Load an empty synthesis column with ∼8 mg of the RNA primer solid-support (i.e., ∼25 µmol/g) corresponding to the first nucleotide at the 3′- end of the oligoribonucleotide.

3.6.2 Supplement 1


Table 3.6.1 Equivalents to the Synthesis Scale and Wait Time Required for Optimal Solid-Phase Synthesis

Reagentsa

Equivalents Wait time (sec)

Phosphoramidites 15 Activator: SET 39 DCI 80 TET 70 Acetic anhydride (Cap A) 655 1-Methylimidazole (Cap B) 1245 TCA 700 Iodine solution 21

465 465 465 465 5 5 10 15

aDescriptions of these reagents have been shortened; see Basic Protocol 1 materials list for complete information.

2. Perform synthesis on an ABI 394 synthesizer according to the cycle outlined in Tables 3.6.1 and 3.6.2. See materials list above as well as Tables 3.6.1 and 3.6.2 for synthesis materials.

3. At the end of the synthesis, perform a manual detritylation cycle on the synthesizer (optional) to remove the dimethoxytrityl group at the 5′-end of the oligonucleotide. 4. Remove the synthesis column from the synthesizer and dry it either under a stream of argon or in a vacuum dessicator for 10 to 15 min. 5. Deprotect the oligoribonucleotide (see Basic Protocol 2, Alternate Protocol 1, or Alternate Protocol 2). OLIGORIBONUCLEOTIDE DEPROTECTION WITH NH4OH/ETHANOL AND TBAF

BASIC PROTOCOL 2

This protocol describes a deprotection scheme using a 3:1 cocktail of concentrated ammonium hydroxide and ethanol to cleave the oligoribonucleotide from the solid support, perform the β-elimination of the cyanoethyl phosphodiester protecting group, and cleave the exocyclic N-acyl protecting groups. A subsequent treatment with ntetrabutylammonium fluoride effects cleavage of the tert-butyldimethylsilyl group protecting the 2′-hydroxyl functionality. (Fig. 3.6.2). Materials Oligoribonucleotide attached to solid support (see Basic Protocol 1) 3:1 (v/v) 29% ammonium hydroxide (Mallinckrodt Baker)/100% ethanol (prepare immediately before use) 3:1:1 (v/v/v) ethanol/acetonitrile/H2O 1.0 M n-tetrabutylammonium fluoride (TBAF) in THF (Aldrich) 50 mM and 2 M triethylammonium bicarbonate (TEAB), pH 7.8 (see recipe) Heating blocks 4-mL glass screw-top vial with Teflon lined lid (Wheaton) 14-mL centrifuge tubes (Falcon) Qiagen-tip 500 column (Qiagen) 1. Transfer the dried oligoribonucleotide on solid support from the synthesis column (see Basic Protocol 1) to a 4-mL glass screw top vial with Teflon-lined lid.



Supplement 1

Table 3.6.2 Summary of a 0.2 µmol–Scale Cycle for Oligoribonucleotide Synthesis on ABI 394 DNA/RNA Synthesizera

Step

Function

Wash steps 1 Acetonitrile to waste 2 Acetonitrile to column 3 Argon reverse flush 4 Argon block flush Chain extension steps 5 Activator to waste 6 Amidite + activator to column 7 Push to column 8 Wait 9 Push to column 10 Wait 11 Repeat steps 9-10 (6 times) 12 Argon flush to waste Wash steps 13 Acetonitrile to waste 14 Repeat steps 3 and 4 Capping steps 15 Cap A and B to column 16 Wait Wash steps 17 Repeat steps 13 and 14 Oxidation steps 18 Iodine to column 19 Wait Wash steps 20 Repeat steps 13 and 14 21 Acetonitrile to column 22 Argon flush to waste 23 Acetonitrile to column 24 Repeat steps 3 and 4 25 Repeat steps 21, 23 and 24 Detritylation steps 26 TCA/DCM to column 27 Wait 28 Argon trityl flush 29 Repeat steps 26-28 Wash steps 30 Acetonitrile to column 31 Argon trityl flush 32 Repeat steps 2, 3, and 4 33 End

Time (sec) 3 10 8 4 1.7 1.2 NA 150 0.1 45 0.1 4

4 5

4 15

10 4 10

6 5 5

10 5

aDelivery flow rate are ∼3.1 mL/min for phosphoramidites and activators and ∼3.6 mL/min for all other reagents.


3.6.4 Supplement 1


2. Add 2 mL of a 3:1 ammonium hydroxide/ethanol solution to the vial, screw cap on tightly, and place in a heat block for 4 hr at 65°C. 3. Remove the vial from the heat block, place it in a block kept at room temperature, and place the block in a –20°C freezer until cooled (i.e., ∼30 min). IMPORTANT NOTE: To avoid loss of contents, it is important to cool the sample vial in step 4 before opening the screw cap.

4. Decant the solution (containing the deprotected oligonucleotide) into a 14-mL Falcon tube. Add 1 mL of ethanol/acetonitrile/H2O solution, vortex well, and allow the support to settle. Decant wash and add to deprotected oligonucleotide solution. Repeat wash twice. Due to the presence of the hydrophobic 2′-O-TBDMS groups on the RNA, this organic wash helps increase the recovery yield.

5. Evaporate the combined supernatant from step 4 in the 14-mL tube on a Speedvac evaporator (i.e., ∼2.5 hr on medium heat). 6. Add 1 mL of 1.0 M TBAF to the 14-mL tube containing the dried RNA and allow to react at room temperature for 24 hr. 7. Quench the desilylation reaction by adding 9 mL of 50 mM TEAB, then refrigerate at 4°C until ready for desalting. Deprotected oligoribonucleotides are highly sensitive to nuclease degradation. Therefore, gloves should always be worn when manipulating deprotected synthetic RNA; sterile disposable containers, nuclease-free laboratory reagents, and Milli-Q water should always be used to limit potential exposure to nucleases.

8. Prewash the Qiagen-tip 500 cartridge with 10 mL of 50 mM TEAB. 9. Load the quenched reaction in TEAB onto the Qiagen-tip 500 anion-exchange cartridge. 10. Wash the loaded cartridge with 10 mL of 50 mM TEAB and discard the eluent. Elute the RNA with 10 mL of 2 M TEAB into a sterile tube, and dry to a white powder on a Speedvac evaporator. OLIGORIBONUCLEOTIDE DEPROTECTION WITH AQUEOUS METHYLAMINE AND TRIETHYLAMINE TRIHYDROFLUORIDE

ALTERNATE PROTOCOL 1

This protocol describes a deprotection scheme using aqueous methylamine and triethylamine trihydrofluoride as alternate reagents to effect nucleobase, 2′-hydroxyl and phosphodiester deprotection (Fig. 3.6.2.). Also see APPENDIX 3C, Basic Protocol 3, for general discussion of RNA oligonucleotide of deprotection. Additional Materials (also see Basic Protocol 2) 40% (w/v) aqueous methylamine (Aldrich) Triethylamine trihydrofluoride/NMP/TEA solution (see recipe) 3 M aqueous sodium acetate (e.g., Fluka) n-butanol 70% aqueous ethanol 1. Transfer the oligoribonucleotide attached to the solid-support from the synthesis column (see Basic Protocol 1) to a 4-mL glass screw top vial. 2. Add 1 mL of 40% aqueous methylamine to the vial, screw the cap on tightly, and place in a heat block at 65°C for 10 min.



Supplement 1

3. Remove the vial from the heat block, place it in block kept at room temperature, and place the block in −20°C freezer until cooled (e.g., ~20 min). 4. Decant the solution into a 14-mL tube. Add 1 mL of 3:1:1 ethanol/acetonitrile/H2O solution, vortex well, and allow the support to settle. Decant wash solution and add to deprotection solution. Repeat wash twice, for a total of three washes. 5. Evaporate the combined supernatants in the tube on a Speedvac evaporator (i.e., ∼2.5 hr on medium heat). 6. Add 0.3 mL of triethylamine trihydrofluoride/NMP/TEA solution to the tube containing the dried RNA, cap the tube, and place on a heat block for 90 min at 65°C. After incubation, bring to room temperature. 7. Precipitate the oligoribonucleotide directly from the desilylation reaction by adding 25 µL of 3 M aqueous sodium acetate, followed by 1 mL of n-butanol. If the oligoribonucleotide has been synthesized trityl-off, proceed to quenching and desalting steps (see Basic Protocol 2, steps 7 to 10).

8. Cool the mixture to −20°C for 2 hr to overnight and centrifuge 30 min at 4000 × g (3750 rpm in a Beckman GS-GR rotor), 4°C. 9. Decant the solution and wash the pellet with 70% ethanol. Centrifuge 10 min at 4000 × g (3750 rpm in a Beckman GS-GR rotor), 4°C. Decant the supernatant and dry the oligoribonucleotide pellet by using a Speedvac evaporator. This precipitation procedure cannot be applied to the TBAF procedure (see Basic Protocol 2) because of the high organic content of the desilylation reaction. Alternately, if a trityl-on deprotected oligoribonucleotide is sought, quench the desilylation reaction by adding 5 mL of 1.5 M ammonium bicarbonate, pH 7.5 (see recipe in Reagents and Solutions). ALTERNATE PROTOCOL 2

“ONE-POT” OLIGORIBONUCLEOTIDE DEPROTECTION WITH ANHYDROUS METHYLAMINE AND NEAT TRIETHYLAMINE TRIHYDROFLUORIDE This protocol describes an expedited deprotection scheme for oligoribonucleotides using anhydrous ethanolic methylamine and triethylamine trihydrofluoride to effect nucleobase, 2′-hydroxyl, and phosphodiester deprotection (Fig. 3.6.2).

HO

B

O O

R

HO

OTBDMS

Nucleobase Deprotection

O

B

NH4OH:EtOH (3:1) or 40% aq. CH3NH2 or 33% CH3NH2 in EtOH

R

O

n

2'-O-Deprotection

−

O

H

B

O O

OH

O P O O

O

OTBDMS

O

HO

OTBDMS

O O P O

O P OCE O

B

O

1.0 M TBAF in THF or TEA•3HF/NMP/TEA or TEA•3HF

B

−

O

O O

OTBDMS n

H

B

OH n

O NH


BR = U, CAc, APAC, GiPrPAC B = U, C, A, G TBDMS, tert-butyldimethylsilyl; TBAF, tetra-n-butylammonium fluoride; NMP, 1-methyl-2-pyrrolidinone; TEA, triethylamine

Figure 3.6.2 Deprotection of chemically synthesized oligoribonucleotides with 2′-O-TBDMS groups.

3.6.6 Supplement 1


Additional Materials (also see Basic Protocol 2) 1:1 (v/v) mixture of 33% ethanolic methylamine and anhydrous DMSO 1.5 M ammonium bicarbonate, pH 7.5 (see recipe) Acetonitrile 1:1:1 (v/v/v) acetonitrile/methanol/H2O RNase-free H2O/DEPC-treated C18 SepPak cartridges (Waters) 1. Transfer the dried solid support-bound oligoribonucleotide from the synthesis column to a 4-mL glass screw top vial. 2. Add 0.8 mL of a 1:1 (v/v) mixture of 33% methylamine/DMSO to the vial, screw the cap on tightly, and place on a heat block at 65°C for 15 min. DMSO is useful for solubilizing the partially deprotected oligoribonucleotide and helps prevent alkaline hydrolysis of the fully deprotected RNA.

3. Remove the vial from the heat block and place it in a block kept at room temperature. 4. Add 0.1 mL of neat triethylamine trihydrofluoride, vortex well, and place on a heating block for 15 min at 65°C. Cool the sample vial at room temperature and then at –20°C for 10 min. The solution usually gels after the addition of triethylamine trihydrofluoride.

5. Quench the reaction by adding 1 mL of 1.5 M ammonium bicarbonate, pH 7.5. Allow the support to settle and decant the supernatant. If sample is not cool enough, addition of ammonium bicarbonate solution may lead to significant effervescence.

6. Prewash the C18 Sep-pak cartridge successively with 10 mL of acetonitrile, 10 mL of 1:1:1 acetonitrile/methanol/H2O solution, and 20 mL of RNase free H2O. 7. Apply the quenched solution to the prewashed C18 Sep-pak cartridge and wash the loaded cartridge with 10 mL of RNase-free H 2O to remove salts. 8. Elute the product from the column by using 10 mL of 1:1:1 acetonitrile/methanol/H2O and evaporate the eluate to dryness on a Speedvac. This desalting step will detritylate a trityl-on deprotected oligoribonucleotide. The precipitation procedure of Alternate Protocol 1 (steps 7 to 9) cannot be applied to the “one-pot” deprotection protocol because of the high organic content.

REAGENTS AND SOLUTIONS Use deionized, distilled water in all recipes and protocol steps. For common stock solutions, see APPENDIX 2A; for suppliers, see SUPPLIERS APPENDIX.

Ammonium bicarbonate,1.5 M (pH 7.5) Weigh 118.59 g of solid NH4HCO3 and bring up to 1 L with Milli-Q water. Store the buffer solution up to 2 months at 4°C. Triethylamine trihydrofluoride/NMP/TEA solution, 1.5:0.75:1 (v/v/v) Combine, in the following order, 0.75 mL 1-methyl-2-pyrrolidinone (NMP), 1.0 mL triethylamine (TEA), and 1.5 mL triethylamine trihydrofluoride. If the reagent is not used immediately after preparation, store it in a capped container on a warm (55° to 65°C) heat block. The reagent will form an intractable gel if allowed to stand at room temperature.



Supplement 1

Triethylammonium bicarbonate (TEAB), 50 mM and 2 M (pH 7.8) Place a 4-L bottle containing 3 L of Super-Q water in an ethanol/ice bath. Bubble carbon dioxide through the water. After 15 min add 279 mL of neat triethylamine every 2 hr for 8 hr (total 4 additions). After 8 hr of CO 2 saturation, check the pH on an aliquot. Continue bubbling CO2 until the pH reaches 7.8 then bring to 4 L with water. Store the buffer solution up to 6 months at 4°C. To prepare 50 mM TEAB, dilute 2M TEAB solution 40-fold with water. COMMENTARY Background Information


Synthesis of oligoribonucleotides Similar to that of oligodeoxyribonucleotides, the chemical synthesis of oligoribonucleotides on solid support is routinely performed via the phosphoramidite method (Fig. 3.6.1; also see UNITS 3.3 & 3.5). However, the additional 2′-hydroxyl function of the ribofuranosyl sugar requires suitable protection during oligoribonucleotide synthesis. Among the various protecting groups available, trialkylsilyl ethers (Ogilvie et al., 1976, 1977; Wincott and Usman, 1994), and particularly the tertbutyldimethylsilyl ether (TBDMS; Usman et al., 1985), have been the most extensively studied, since fluoride-mediated silyl ether deprotection is quite orthogonal to the other acid- and base-labile protecting groups commonly used in oligonucleotide chemistry. Because of the steric bulk introduced by the TBDMS group, coupling 2′-O-TBDMS protected ribophosphoramidites to a growing oligonucleotide chain is notoriously more sluggish as compared to their 2′-deoxy analogs. The elongation cycle for automated oligoribonucleotide synthesis is similar to that of oligodeoxynucleotides and consists of the detritylation, coupling, capping, and oxidation steps. Each of these steps is critical for successful RNA synthesis. One important difference regarding the detritylation step is the possibility of using higher concentrations of haloacids due to the reduced sensitivity of ribofuranosylnucleotides to acid-mediated depurination. The oxidation and capping steps do not differ significantly from traditional oligodeoxyribonucleotide synthesis. Conversely, the coupling step has been the focus of much attention through the development of a number of phosphate protecting groups (Usman et al., 1987; Sinha et al., 1983; Schwarz et al., 1984; Hamamoto et al., 1986; Kayakawa et al., 1990; Wada and Sekine, 1994; Ravikumar and Cole, 1994) and different phosphoramidite dialkylamino functions (Lyttle et al., 1991; Sinha et al., 1984; Beaucage and Caruthers,

1981) aimed at providing faster coupling rates. Although some of these modifications have been commercialized, the phosphoramidites carrying the 2-cyanoethyl-N,N-diisopropyl combination (Sinha et al., 1983, 1984) are predominantly used. According to this chemistry, the coupling time for 2′-O-TBDMS ribophosphoramidites is considerably longer than that of the 2′-deoxy series for synthesis scales ranging from 0.2 to 2.5 µmol. Particularly important to commercially available ribophosphoramidites is the acidic activation step that allows rapid conversion of stable N,N-diisopropyl phophoramidite to azolide intermediates. These highly reactive transient species are then coupled to the free 5′-hydroxyl of any oligonucleotide bound to a support to form phosphite triester internucleotidic bonds which are further oxidized to the acid-stable pentavalent phosphotriester linkages. Dramatic improvements in oligoribonucleotide synthesis are achieved when activators more acidic than the standard 1-H-tetrazole (pKa = 4.8) are used. Such activators may include 5-(4-nitrophenyl)1H-tetrazole (pKa = 3.7) (Sproat et al., 1989) or 5-ethylthio-1H-tetrazole (pKa = 4.28) (Wincott et al., 1995). These tetrazole derivatives are supposedly more efficient at protonating the trivalent phosphorus. Once protonated, this electrophilic phosphorus center reacts with a tetrazole molecule to displace an N,N-diisopropylamino group. 5-Ethylthio-1H-tetrazole (SET) became the preferred activator for oligoribonucleotides, and has been used in a number of reports describing successful RNA synthesis. However, the acidic properties of 5ethylthio-1H-tetrazole can cripple larger-scale oligoribonucleotide syntheses because of concomitant acidic activator–mediated detritylation of nucleoside phosphoramidites, occurring as a result of an extended coupling reaction time required for satisfactory coupling efficiency (Krotz et al., 1997). This problem spearheaded the development of less acidic but more nucleophilic activators (Vargeese et al., 1998) like 4,5-dicyanoimidazole (pKa = 5.2) or benzimi-

3.6.8 Supplement 1


dazolium triflate (pKa = 4.5) (Hayakawa et al., 1996). Although no mechanistic studies have been presented, these activators are reported to speed up coupling reaction rates due to the increased nucleophilicity of their conjugated base while exhibiting sufficient acidity to protonate the tricoordinated phosphorus center. Automated oligoribonucleotide synthesis can be easily performed on long chain alkylamine controlled pore glass (CPG; Pon et al., 1988; UNITS 3.1 & 3.2). Non-swellable highly cross-linked polystyrene solid-supports have shown to be generally superior to CPG due to their increased mechanical and chemical resistance (McCollum and Andrus, 1991). Deprotection of oligoribonucleotides The deprotection of crude oligoribonucleotides traditionally requires a basic treatment, which allows for the transamination of the N-acyl exocyclic protecting groups, βelimination of the 2-cyanoethyl phosphate protecting group, and cleavage of the succinic ester bond linking the oligoribonucleotide to the solid support (Fig. 3.6.2). Of these three concomitant reactions, base deprotection is by far the most rate-limiting step. Once these reactions are accomplished, a fluoride treatment removes the substituted silyl ether function protecting the 2′-hydroxyl of the ribofuranose ring. Nucleobase deprotection In the early days of oligoribonucleotide chemistry, the heterocyclic amino function of each nucleobase was almost exclusively protected with the benzoyl (rA and rC) and isobutyryl (rG) groups. Although these protecting groups on synthetic DNA are efficiently removed by treatment with concentrated ammonium hydroxide, the presence of the 2′-OTBDMS group in synthesized RNA necessitated the use of ethanolic ammonia (Lyttle, 1993) or a solution of ammonium hydroxide in ethanol (3:1, v/v) for 12 to 16 hours at 55°C to minimize the cleavage of the silyl group (Stawinski, 1988; Wu et al., 1989). Premature deprotection of the 2′-hydroxyl under basic conditions results in extensive oligoribonucleotide chain cleavage from intramolecular nucleophilic attack of the 2′-hydroxyl group on the phophodiester function. To circumvent this unwanted side reaction and shorten the deprotection time, two strategies were investigated over the last decade. One of these strategies involves the use of various combinations of hydrazine, ethano-

lamine, and alcohol (Polushin et al., 1991) or more nucleophilic alkylamines (Wincott et al., 1995; Reddy et al., 1995) for oligoribonucleotide deprotection. For example, 40% aqueous methylamine, in place of or in addition to 30% ammonium hydroxide, cleaves the nucleobase N-acyl protecting groups in a few minutes at 65°C or ~1 hr at room temperature. However, this procedure requires the use of N4-acetyl cytidine phosphoramidite derivatives to avoid a well documented transamination side reaction (Reddy et al., 1994). The second strategy relates to the development of base-labile amino-protecting groups that would further shorten exposure of the 2′O-TBDMS group to basic conditions. These groups belong to the phenoxyacetyl (PAC) (Wu et al., 1988; Chaix et al., 1989) and amidinetype (McBride et al., 1986) protecting groups, mainly used for ribopurine phosphoramidites given that the N4-acetyl protection of cytidine appears quite optimal. Amidine protecting groups include acetamidines (McBride et al., 1986) and dialkylformamidines (Vinayak et al., 1992), which can be cleaved in 2 to 3 hours at 55°C using concentrated ammonium hydroxide/ethanol (3:1, v/v). The main advantage of the dimethylformamidine protecting group when used in conjunction with 2′-deoxyribonucleoside phosphoramidites (Vu et al., 1990) is that it confers an increased resistance towards depurination during the detritylation step. However, because oligoribonucleotides are inherently less sensitive to depurination, N-phenoxyacetyl protection has been preferred for the commercial manufacturing of these biomolecules. The phenoxyacetyl and 4-tert-butyl- or (4-isopropylphenoxy)acetyl protecting groups (Sinha et al., 1993) are considerably more labile than amidines under basic conditions. Typically, these can be quantitatively cleaved from the exocyclic amino function of nucleobases after a 15 min to 1 hr incubation in concentrated ammonium hydroxide/ethanol (3:1) at 65°C, or 2 to 4 hr at room temperature. The use of “fast deprotecting groups” of the PAC family does not preclude the concomitant use of more nucleophilic alkylamines. Indeed, combining these two strategies allows for expedited base-labile removal of protecting groups as exemplified in Alternate Protocol 1. 2′-O-tert-butyldimethylsilyl deprotection The fluoride-sensitive tert-butyldimethylsilyl group allows for an efficient orthogonal deprotection of the 2′-hydroxyl of synthetic RNA. The use of fluoride-based reagents is the



Supplement 1


preferred methodology for the removal of the 2′-O-TBDMS group, although Sekine and colleagues (Kawahara et al., 1996) recently reported the application of an acid-catalyzed desilylation scheme to oligoribonucleotides. After completion of nucleobase and phosphotriester deprotection, and subsequent evaporation of the basic solution, addition of n-tetrabutylammonium fluoride (TBAF) in THF to the partially deprotected RNA cleaves the 2′-O-TBDMS group at room temperature within 24 hr (Basic Protocol 2). However, limited solubility of the RNA in this apolar solvent hampers the efficient deprotection of longer oligoribonucleotides. To circumvent this problem, RNA dissolution in either dimethylsulfoxide (Gasparutto et al., 1992) or 50% ethanol (Scaringe, 1995) has been investigated. The notorious water sensitivity of TBAF (Hogrefe et al., 1993) and the desalting step required after quenching the desilylation reaction have led to the development of other fluoride-based reagents. Triethylamine trihydrofluoride (Alternate Protocol 1) has been shown to be a superior reagent either neat (Pirrung et al., 1994; Westman and Stromberg, 1994) or in combination with polar aprotic solvents such as dimethylformamide (Vinayak et al., 1995) or 1-methyl2-pyrrolidinone (Wincott et al., 1995). This reagent promotes 2′-O-TBDMS deprotection within 30 to 90 min at 65°C or 4 to 8 hr at room temperature. Using triethylamine trihydrofluoride, the time-consuming desalting step may be replaced by a sodium acetate/1-butanol precipitation procedure (Wincott et al., 1995) that is not compatible with the high organic content present in TBAF deprotection mixtures. Further application of neat triethylamine trihydrofluoride in a “one-pot deprotection” procedure (Alternate Protocol 2; Bellon, 2000) that employs a mixture containing anhydrous ethanolic methylamine allows for expedititious RNA deprotection. This procedure requires a quenching step with ammonium bicarbonate if one desires to retain the trityl group on the oligonucleotide, and a subsequent desalting/purification on a reverse-phase cartridge because the presence of ethanol and dimethylsulfoxide prevent RNA precipitation from butanol. HPLC analysis and purification of synthetic RNA is now well-documented (Wincott et al., 1995; Sproat et al., 1995; and Vinayak et al., 1995). In particular, the use of perchloratebased buffer (Na or Li form) in conjunction with anion-exchange Nucleo Pac columns, allows for easy purification of full-length RNA product from truncated sequences.

Critical Parameters Automated oligoribonucleotide synthesis according to the phosphoramidite method has not significantly evolved over the last decade. Therefore, all precautions mentioned in the major textbooks (Gait, 1984; Eckstein, 1991; Agrawal, 1993) still remain valid. It is particularly important to emphasize that phosphoramidite chemistry is highly water sensitive. Great care should therefore be taken to ensure that the phosphoramidites and activator are dissolved in strictly anhydrous acetonitrile. All ancillary reagents (i.e., acetonitrile, detrilylation, capping, and oxidation solutions) are commercially available, guaranteeing high performance reproducibility in the syntheses and relieving the chemist from time-consuming anhydrous distillations. Sterile, disposable pipet tips and plastic tubes should be used for storing and handling RNA. Troubleshooting an oligoribonucleotide synthesis is a relatively easy task when the trityl assay is used to spectrophotometrically monitor the synthesis (Gait, 1984; see also APPENDIX 3C, Basic Protocol 1, Support Protocol 1). Deprotection of oligoribonucleotides according to the three protocols presented in this unit is quite straightforward. However, Alternate Protocol 2 or the “one-pot” deprotection protocol should not be used in conjunction with controlled-pore-glass (CPG) synthesized RNA because of the inherent incompatibility between triethylamine trihydrofluoride and the silyl components of CPG. Because all the basic solutions used are composed of gaseous amines dissolved in water or ethanol, freshly opened bottles will ensure that the effective concentration of the amine (i.e., 29% NH4OH or 40% methylamine in water) is close to its stated nominal value. Typically, reagent bottles should be replaced every two weeks if opened on a regular basis. This may be especially true for the TBAF solution in THF (used in Basic Protocol 2) because a low water content is critical for efficient desilylation. The deprotection times at 65°C are suggested for 2 mL of basic solution, and need to be extended if larger amounts of reagents are used. Finally, deprotected oligoribonucleotides are highly sensitive to nuclease degradation. Therefore, gloves should always be worn when manipulating deprotected synthetic RNA; sterile disposable containers and Milli-Q water should also always be used to limit potential exposure to nucleases. DEPC treated water should be used when desalting RNA on C18 cartridges.

3.6.10 Supplement 1


Anticipated Results Because of the iterative nature of solidphase oligoribonucleotide synthesis, only modest chemical yields of oligoribonucleotide can be produced, especially for longer synthetic RNAs (i.e., >30 residues). When applying Basic Protocol 1 using 5-ethylthio-1H-tetrazole, a realistic averaged stepwise chemical yield (ASWY) of 97.5% can be routinely obtained for RNA synthesis. This ASWY is determined by the ratio (µmol FLR/µmol scale)1/n × 100 where µmol FLR is the amount of full length RNA in the crude mixture, µmol scale is the synthesis scale, and n is the number of synthesis cycles. This corresponds to an isolated yield of 41% for the all-RNA 36-mer pictured in Figure 3.6.3.A. Figure 3.6.3.B and C shows HPLC profiles of the same sequence synthesized according to Basic Protocol 1 using the alternate activators DCI and TET, respectively. Figure 3.6.3 indicates that 5-ethylthio-1H-tetrazole is the activator of choice for small scale RNA synthesis. At small scale (i.e., 50% of the respective label. A more common approach to synthesizing 3′-oligonucleotide conjugates takes advantage of the nucleophilicity of phosphorothioates. In one example, 3′-phosphorothioate-containing oligonucleotides were prepared on a solid phase synthesis support (Fig. 4.5.6, S.8; Asseline et al., 1992). Following standard oligonucleotide synthesis and ammoniacal deprotection/cleavage under reducing conditions, the fully deprotected oligonucleotides were conjugated to halogenated substrates (Fig. 4.5.12). Conjugation to derivatives of daunomycin, fluorescein, and 1,10phenanthroline (S.25) containing alkyl halide

−

O

−

SO3 Na

SH

2

O

+

−

SO3 Na

+

oligo

O P O

−

O

S 3

O N H

NH

24

O

I

5'-HO

+

N H

Figure 4.5.11

NH

Postsynthetic modification of oligonucleotide 3′-thiols.

Synthesis of Modified Oligonucleotides and Conjugates


Supplement 2

O oligo

5'-HO

O P OH S

−

N O

+

5'-HO

oligo

O

N

O

O P S

−

N H

N N

O I

N H

Figure 4.5.12

Attachment of Reporter and Conjugate Groups to the 3′ Termini of Oligonucleotides

25

Postsynthetic modification of oligonucleotide 3′-phosphorothioates.

functional groups are believed to have proceeded in essentially quantitative yields over 24 hours in either protic organic solvents containing crown ethers to enhance biopolymer solubility, or mixtures of dimethylformamide and water. This strategy was utilized recently in studies on truncated derivatives of the potent antitumor antibiotics, CC-1065 and the duocarmycins, which alkylate deoxyadenosine at N-3 (Lukhtanov et al., 1996). The active pharmacophore, the cyclopropapyrroloindole, was covalently linked via an α-bromoacetamido linkage to a 3′-phosphorothioate in 50% to 60% yield. Several reports of template-mediated oligonucleotide ligation involving 3′-phosphorothioate DNA have also appeared. Oxidative coupling of a 3′-phosphorothioate oligonucleotide to a 5′-phosphorothioated biopolymer is achieved rapidly (5 min) under mild conditions (0°C) using 1 µM K3Fe(CN)6 (Gryaznov and Letsinger, 1993a). Little or no coupling is detected under these conditions in the absence of a template. In contrast, ligation of a 5′-phosphorothioated oligonucleotide to a 3′-bromoacetamide-containing oligonucleotide proceeded in ∼80% yield after 48 hr at 0°C in the absence of a template (Gryaznov and Letsinger, 1993b). As expected, the presence of a template, which increases the effective concentration of the reaction partners, significantly accelerated the reaction. Essentially quantitative yields of ligated product were obtained in only 20 minutes when a stoichiometric template was present. A distinct advantage of this latter coupling method is that the reaction does not require any exogenous condensing agents. In a subsequent investigation, autoligation was effectively carried out using reaction partners of the opposite polarity, 3′-phosphorothioate and 5′-bromoacetamide (Fig. 4.5.13; Gryaznov et al., 1994).

Conjugation to 3′-Aldehydes Template-mediated synthesis has also been very useful for the conjugation of oligonucleotides containing 3′-aldehydes (Goodwin and Lynn, 1992; Zhan and Lynn, 1997). Aldehydes are attractive electrophiles with which to form oligonucleotide conjugates. Condensation with primary amines under reductive conditions provides secondary amines which are stable to acid and base. Solid supports that produce oligonucleotides containing 3′-aldehydes directly are unknown. Consequently, 3′aldehydes are typically produced via periodate oxidation of a vicinal diol. A common practice for preparing 3′-oligonucleotide conjugates via aldehyde condensation takes advantage of incorporating a ribonucleoside at the 3′ terminus (Fig. 4.5.14; Leonetti et al., 1988). Periodate oxidation of the oligonucleotide containing a 3′-terminal ribonucleoside produces a dialdehyde which under reductive amination conditions generates a morpholine upon reaction with a primary amine. An often cited application of this method concerns the synthesis of poly-L-lysine conjugates of oligonucleotides (Leonetti et al., 1988, 1990). Haralambidis et al. (1994) have utilized this ability to introduce amino acids at the 3′ termini of oligonucleotides to enable attachment of a substituted benzaldehyde, which is then conjugated to an enzyme via reductive amination. Condensation of a dialdehyde with a primary amine was also used recently in the segmental synthesis of a biologically active hammerhead ribozyme (Bellon et al., 1996). Linkage of the two segments was carried out in a loop region of the hammerhead ribozyme which had been shown to not be crucial for catalytic activity. One segment was synthesized so as to incorporate a 5′-alkylamine, while the other half contained a 3′-terminal uridine which served as the source of the dialdehyde. Follow-

4.5.8 Supplement 2


oligo

5'-HO

O

O − O P S OH

Br

oligo

N H

OH-3'

Template

5'-HO

oligo

O O P S O

Figure 4.5.13

O

−

oligo

N H

OH-3'

Template-mediated oligonucleotide coupling.

ing rapid and quantitative periodate oxidation, reductive condensation was carried out to 95% conversion over the course of seven days. In a more recent study, the electrophilic half of a ribozyme was synthesized on a glyceryl sup-

port (Fig. 4.5.6, S.9; Bellon et al., 1997; Urata and Akagi, 1993). In these experiments, conjugation of the two halves of a ribozyme within the loop II region of the unmodified hammerhead proceeded in as high as 81.2% yield

O 5'-DMTrO

prot. oligo

Prot.

O P O − O O

B

O

OTBDMS

O N H

O 1. deprotect 2. oxidize (NaIO4)

5'-HO

oligo

O O P O − O O

B

O H H

O

RNH2, NaCNBH3

5'-HO

oligo

O O P O − O

O

B

N R Prot., nucleobase protecting group TBDMS, tert-butyldimethylsilyl

Figure 4.5.14

Postsynthetic modification of oligonucleotides by reductive amination.



Supplement 2

(48 hr) when borane pyridine was employed as a reducing agent.

Solution Phase Conjugation of Protected Oligonucleotides

Attachment of Reporter and Conjugate Groups to the 3′ Termini of Oligonucleotides

The need for orthogonal linkers that enable the removal of fully protected oligonucleotides from the support for further elaboration in solution was recognized by at least one leader in the field of oligonucleotide synthesis a number of years ago (Zon and Geiser, 1991). The first reports of conjugating protected oligonucleotides in solution appeared in 1997 (McMinn et al., 1997; also see McMinn and Greenberg, 1998, 1999). This new method for synthesizing oligonucleotide conjugates was made possible by the development of a family of orthogonal solid phase synthesis supports (Fig. 4.5.8, S.14 to S.20). Using light or Pd(0), these supports enable one to release oligonucleotides containing 3′-alkylamines, 3′-alkyl carboxylic acids, or 3′-phosphate diesters which retain their exocyclic amine, phosphate diester, and 5′-hydroxyl protecting groups. One advantage of this method is that potential deleterious side reactions are eliminated by utilizing protected oligonucleotides. Nonspecific covalent modifications of unprotected oligonucleotides can be a significant problem when conjugating biopolymers (Bischoff et al., 1987; Erout et al., 1996; Ghosh and Musso, 1987; Lund et al., 1988). In addition, the rates at which conjugation reactions proceed are considerably faster than similar bond-forming reactions using unprotected oligonucleotides. An explanation for this acceleration is uncertain at this time, but may be related to the solvent conditions. Conjugation reactions of protected oligonucleotides are carried out in aprotic organic solvents, whereas unprotected oligonucleotides are often conjugated in aqueous solvents. Amide bond formation in aqueous solvents may be adversely affected by stronger solvation (hydrogen bonding) of the reactants, as well as a lower effective molarity of amines due to protonation. The original report of solution phase conjugation of protected oligonucleotides utilized a redox condensation or a Mukaiyama reaction to activate carboxylic acids (Fig. 4.5.15; McMinn et al., 1997; Mukaiyama, 1976). The oligonucleotides contained 3′-alkylamines. During the course of developing the conjugation chemistry, it was discovered that the “fast deprotecting” amides used to protect deoxyadenosine and deoxyguanosine underwent transamidation with the 3′-terminal alkyl-

amines (McMinn and Greenberg, 1997). Thus, phenoxyacetyl-protected phosphoramidites should not be used in conjunction with alkylamine modifiers. The observed transamidation is of general importance, because of the commercial availability of alkylamine modifiers for oligonucleotide synthesis. This problem was overcome by using isobutyryl exocyclic amine protecting groups for deoxyadenosine, deoxycytidine, and deoxyguanosine. Subsequently, a variety of biologically relevant reporter groups such as biotin and acridine were conjugated in excellent yields (88%) under mild reaction conditions (2 hr at 55°C) using only ten molar equivalents of carboxylic acids and activating reagents relative to oligonucleotide substrate. Only cholesterol coupled in 50 OD254 units of DNA, increase DMSO by 10 mL per additional 10 OD units.

20. Add this solution to a 1.5-mL plastic screw-cap tube containing the dried DNA, followed by 500 µL of 1 M sodium carbonate/1 M sodium bicarbonate solution. Close the tube and allow to stand 18 to 24 hr at room temperature. For >50 OD254 units of DNA, increase buffer by 100 mL per additional 10 OD units.



Supplement 3

21. Evaporate the liquid in a Speedvac evaporator, and dissolve the residue in 1.5 mL of 0.05 M aqueous ammonium acetate. 22. Mix 20 g Sephadex G-25 resin with 100 mL of 0.05 M aqueous ammonium acetate, and pour the slurry into a 1 × 30–cm glass tube with a frit and valve at the bottom. Add enough of the slurry so that the settled bed volume is 20 to 25 cm long. Allow the liquid to elute from the column until the liquid is level with the Sephadex slurry. Use fresh Sephadex for each purification.

23. Add the DNA solution (step 21) to the column, and allow the liquid to elute from the column until the liquid is level with the Sephadex column. 24. Elute the column with 0.05 M aqueous ammonium acetate at a flow rate of ∼2 mL/min. Collect the first colored band that elutes after the first 10 to 15 mL. Evaporate the collected fraction in 1.5-mL tubes in a Speedvac. For many applications, the fluorescein-DNA oligonucleotide conjugate is pure enough for good results. Analysis of purity can be performed by PAGE (UNIT 10.4) or HPLC (UNIT 10.5). If necessary, purification can be performed using a reversed-phase cartridge (see below).

Purify fluorescein-DNA oligonucleotide conjugate 25. Preequilibrate a reversed-phase DNA purification cartridge by eluting with 4 mL acetonitrile followed by 4 mL of 1 M TEAA at a flow rate of ∼1 mL/min using a 10-mL syringe. 26. Dissolve the sample obtained in step 24 in 1 mL of 14% aqueous ammonia. Apply the solution to the cartridge. Collect the effluent and reload it onto the cartridge two times. 27. Elute the cartridge with 4 mL of 2.8% aqueous ammonia followed by 4 mL water at ∼1 mL/min. Discard noncolored eluant. 28. Elute the cartridge with 3 mL of 20% acetonitrile in water at ∼1 mL/min, and collect the strongly green effluent. 29. Evaporate in a Speedvac and store the purified 3′-fluorescein DNA oligonucleotide conjugate at −20°C until needed. Samples retain fluorescence for 2 to 3 months if kept in the dark.

Analyze conjugate by HPLC 30. Dissolve the sample in 500 µL of 20% acetonitrile in water, and inject 2 to 20 µL, depending on the concentration, onto an anion-exchange column. Elute the conjugate with a linear gradient of 100% buffer A to 100% buffer B over 20 min at a flow rate of 1 mL/min. The fluorescein conjugate will elute 2 to 4 min later than the underivatized 3′-aminohexyl DNA oligonucleotide. ALTERNATE PROTOCOL 1

3′-Modified Oligonucleotides and their Conjugates

PREPARATION OF 3′-THIOALKYL-FUNCTIONALIZED DNA OLIGONUCLEOTIDES This procedure uses many of the same reagents and steps as the Basic Protocol. However, the oxygen of the sulfur-bearing spacer must be protected, or else mixtures of products will be obtained after DNA synthesis and cleavage. A synthesis of these O-DMTr-protected thioalkyl spacers has been previously reported (Gupta et al., 1991).

4.6.4 Supplement 3


Additional Materials (also see Basic Protocol) 6-Mercapto-1-O-DMTr hexanol (Biosearch Technologies) Dithiothreitol (DTT) Triethylamine Fluorescein-5-maleimide (Molecular Probes) Sodium phosphate/sodium chloride buffer: 50 mM sodium phosphate and 150 mM sodium chloride, adjust to pH 7.2 (if necessary) Synthesize 3′-thiohexyl DNA oligonucleotides on CPG 1. Prepare anhydride-functionalized CPG (see Basic Protocol, steps 1 to 7) and place 2 g in a 125-mL Erlenmeyer flask with stopper. 2. In a separate 125-mL flask, dissolve 1 g of 6-mercapto-1-O-DMTr hexanol in 10 mL DMF, and add 1 mL of triethylamine. 3. Add this solution to the Erlenmeyer flask containing CPG. Stopper the flask and swirl the CPG until there is a uniform slurry in the flask. Allow the flask containing the CPG to stand for 3 hr at room temperature. 4. Perform washing and drying as described (see Basic Protocol, steps 6 and 7). 5. Assay the loading of the spacer on the CPG (see Basic Protocol, steps 13 and 14). Loading should be ≥20 mmol/g.

6. Synthesize DNA oligonucleotides on the CPG, and deprotect and purify the product (see Basic Protocol, steps 15 to 18), but add 2 to 3 mg DTT to the 28% aqueous ammonia solution (step 17) before heating to prevent oxidative dimerization of the thioakyl oligonucleotides. Conjugate 3′-thiohexyl DNA oligonucleotides to fluorescein-5-maleimide 7. Weigh out 2 mg fluorescein-5-maleimide for every 10 OD254 units of DNA oligonucleotide to be conjugated in a 1.5-mL plastic screw-cap tube, and dissolve in 100 µl DMF. 8. Add this solution to a 1.5-mL plastic screw-cap tube containing the thiohexyl DNA. 9. Add 500 µL sodium phosphate/sodium chloride buffer, pH 7.2, and allow to stand 18 to 24 hr in the dark at room temperature. 10. Perform Sephadex G-25 size exclusion (see Basic Protocol, steps 21 to 24). Reversed-phase cartridge purification does not work well for thiomaleimido fluoresceinDNA oligonucleotide conjugates. Purification of the conjugates can be accomplished by either preparative HPLC (UNIT 10.5) or preparative PAGE (UNIT 10.4).

PREPARATION OF 3′-POLYETHYLENE-GLYCOL-FUNCTIONALIZED DNA OLIGONUCLEOTIDES


Dry solvents and equipment are essential for the key reaction in this sequence, which is the addition of a hydroxyl functionality to an anhydride. This procedure uses many of the same reagents and steps as previous protocols. Various polyethylene glycols are commercially available; selection depends on the experimental design. The example given below uses triethylene glycol. Additional Materials (also see Basic Protocol) Triethylene glycol (Aldrich) N-Methylimidazole (Aldrich)



Supplement 3

1. Prepare anhydride-functionalized CPG (see Basic Protocol, steps 1 to 4), and place 2 g in a 125-mL Erlenmeyer flask with stopper. 2. In a separate 125-mL flask, dissolve 1 g triethylene glycol in 10 mL DMF, and add 500 µL N-methylimidazole. 3. Add this solution to the Erlenmeyer flask containing CPG. Stopper the flask and swirl the CPG until there is a uniform slurry in the flask. Allow the flask containing the CPG to stand for 18 to 24 hr at room temperature. 4. Perform washing and drying (see Basic Protocol, steps 6 and 7). 5. Assay the loading of the triethylene glycol spacer on the CPG (see Basic Protocol, steps 11 through 14). Loading should be ≥20 mmol/g.

6. Synthesize DNA oligonucleotides on the PEGylated CPG, and deprotect and purify the product (see Basic Protocol, steps 15 to 18). 7. Analyze 3′-PEGylated DNA oligonucleotides by ion-exchange HPLC (see Basic Protocol, step 30). The desired product will elute slightly later than the corresponding unmodified DNA.

COMMENTARY Background Information

3′-Modified Oligonucleotides and their Conjugates

The basic strategy employed in these protocols (Lyttle et al., 1997) calls for the synthesis of an anhydride-functionalized solid support as a common intermediate for each functional group attachment. A bifunctional molecule (spacer) that has an OH group at one end and an SH, NH2, or OH group at the other end is then added. The most nucleophilic functional group reacts with the anhydride to form a thioester, amide, or ester bond, respectively, while the OH group at the other end of the spacer is available to react with nucleoside phosphoramidites (see Figure 4.6.1). There is then an optional step of adding a 4,4′-dimethoxytrityl (DMTr) group to this alcohol group to spectrophotometrically gauge the amount of addition of the first phosphoramidite during automated oligonucleotide synthesis. In the case of the thiol functionality, this DMTr alcohol protection is mandatory, or else a mixture of products (resulting from SH and OH addition to the anhydride) will be obtained. The synthesis of the required O-DMTr-protected hydroxylalkylthiols is described in the literature (Gupta et al., 1991). 6-O-DMTr-hydroxyhexylthiol is available from Biosearch Technologies. Once the support is made, automated oligonucleotide synthesis is performed to construct the desired sequence, then the usual aqueous ammonia treatment is employed for DNA

deprotection and solid-support cleavage to provide the desired 3′-terminal functionality upon basic hydrolysis of the thioester, amide, or ester bonds. The product DNA is then purified with reversed-phase cartridges, and can be characterized by PAGE, HPLC, and matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS). In the case of 3′-NH2- or SHmodified oligos, a procedure for the attachment of fluorescein is described, including an example of a successful reversed-phase cartridge purification of a 3′-amino-modified DNA-fluorescein conjugate.

Critical Parameters and Troubleshooting These protocols describe three techniques for the 3′-terminal labeling of DNA. Successful execution of each protocol depends on the skill and patience of the researcher, as well as the quality of reagents and solvents. The approximate expected loading is mentioned in each protocol; if loading is lower than expected, there are several points to check. For steps that require the immobilized anhydride, use a new bottle of trimellitic anhydride chloride for the coupling reaction. A negative ninhydrin test (Stewart and Young, 1984) will assure that the anhydride was quantitatively linked to the resin prepared according to the Basic Protocol. DMF must be dry (i.e., 18 hr. The deprotected pnODN can be stored in the ammonia solution at −20°C until purification for a maximum of 1 week, although it is best to remove the ammonia and purify the crude oligonucleotide as soon as possible after synthesis. BASIC PROTOCOL 2

Synthesis and Purification of Oligonucleotide N3′→P5′ Phosphoramidates

IEC PURIFICATION, ISOLATION, AND ANALYSIS The IEC purification method is used for pnODNs synthesized with a terminal 3′-amine or 3′-hydroxyl following removal of the trityl protecting group. Preparative IEC is able to separate failure sequences from the full-length product (n-mer). However, the resolution between the n-1 failure sequence and the product is not great; therefore, the main product peak must be fractioned and analyzed by analytical IEC or capillary gel electrophoresis (CGE) in order to decide which fractions to combine. Materials Deprotected oligonucleotide solution (see Basic Protocol 1), 4°C Buffer A: 0.01 M aqueous NaOH/0.01 M NaCl, pH 12 Buffer B: 0.01 M aqueous NaOH/1.5 M NaCl, pH 12 Concentrated aqueous ammonia, 4°C 0.5 M aqueous NaOH solution 100% ethanol UV/visible spectrometer

4.7.6 Supplement 3


Analytical IEC column (preferably a 4 × 250–mm Dionex PA-100 NucleoPac column) HPLC or FPLC system compatible with high pH buffer systems equipped with a UV detector, data collection system, and a 1- or 2-mL sample injection loop 3-mL disposable syringe with luer lock 0.45-µm filter that fits the end of a luer lock syringe Speedvac evaporator (Savant) Preparative IEC column (preferably a Pharmacia MonoQ 10/10 column) Sample holder with 1.5-mL centrifuge tubes or fraction collector Sephadex G-25 column (e.g., Pharmacia NAP-10; optional) Analyze quality of crude oligonucleotide 1. Dilute 10 µL cold, deprotected oligonucleotide solution in 990 µL water and scan from 200 to 400 nm using a UV spectrometer. Determine the absorbance at 260 nm and multiply by 100 (dilution factor) to determine the concentration of crude oligonucleotide in the 1 mL sample. To prevent loss of sample and to improve the accuracy of the dilution, make sure the ammonia solution is cold before opening the vial.

2. Dilute 0.5 OD260 units oligonucleotide into 0.5 mL water to serve as an analytical IEC sample. 3. Preequilibrate the analytical IEC column for ≥10 min with buffer A at a flow rate of 1 mL/min. 4. Program an HPLC or FPLC to run a gradient of 0% to 50% buffer B versus buffer A over 40 min at a flow rate of 1 mL/min and inject the 0.5-mL ODN sample. Monitor the run at 260 nm. Reequilibrate the column for ≥10 min with buffer A before performing a second run.

Prepare sample and IEC purify 5. Filter the CPG away from the remaining cold ammonia solution using a 3-mL disposable syringe with an attached 0.45-µm filter. 6. Wash the CPG twice with 0.5 mL cold, concentrated aqueous ammonia. 7. Add 10 µL of 0.5 M aqueous NaOH solution, and concentrate in a Speedvac evaporator to ∼0.5 mL. Do not completely dry the sample.

8. Filter the concentrated oligonucleotide again using a new 3-mL disposable syringe with an attached 0.45-µm filter and wash with 0.3 mL water. 9. Preequilibrate the preparative IEC column for ≥15 min with buffer A. Run a “blank” gradient if the column has not been used recently, and between purifications of samples with different sequences.

10. Program the HPLC to run a gradient ramping at 1%/min of buffer B versus buffer A at a flow rate of 1 mL/min. Use the analytical IEC (step 4) to determine the approximate percent of buffer B that will be needed to elute the sample. 11. Prepare a sample holder with at least ten 1.5-mL centrifuge tubes or use a fraction collector. This is necessary to obtain the highest level of purity; preparative IEC does not have the resolution of analytical IEC.



Supplement 3

12. Inject the entire crude ODN sample from the 1-µmol synthesis and monitor elution at 260 nm. Collect ∼0.5-mL fractions during the elution of the product peak. Store the fractions at 4°C until they have been analyzed and are ready for desalting. Analyze and desalt 13. Analyze a small amount (i.e., ∼0.1 OD260) of the fractions by analytical IEC. 14. Combine the fractions that are ≥85% pure and concentrate in a Speedvac evaporator to a volume of ~1 mL. Do not let the samples evaporate completely in the Speedvac evaporator; NaOH can potentially degrade pnODNs.

15. Precipitate the pnODN with 2.5 mL of 100% ethanol and cool for ≥30 min at −20°C. 16. Centrifuge for 2 min at 3000 × g, 4°C, and carefully remove the supernatant. 17. Dissolve the pellet in 1 mL of deionized water and repeat ethanol precipitation (steps 15 and 16) two more times to desalt the sample. Alternatively, desalt the sample on a Sephadex G-25 column using the manufacturer’s protocol.

18. Dissolve pellet in 1 mL of water and measure the OD260 as in step 1 to determine the yield of pnODN. 19. Inject a 0.2 OD260 sample on the analytical IEC column to determine the purity of the product. Alternatively, determine the purity using capillary gel electrophoresis (CGE) or polyacrylamide gel electrophoresis (PAGE; APPENDIX 3B).

20. Concentrate the pnODN to dryness and store up to 1 year at −20°C. ALTERNATE PROTOCOL 2

REVERSED-PHASE HPLC PURIFICATION, ISOLATION, AND ANALYSIS The RP-HPLC method developed for the pnODNs and their chimeras relies on the 3′-addition of a commercially available RNA monomer, 5′-O-(4,4′-dimethoxytrityl)-2′-Otert-butyldimethylsilyl-3′-O-[(N,N-diisopropylamino)(2-cyanoethoxy)]phosphinyl uridine, to the terminal 3′-OH via 1H-tetrazole activation, followed by oxidation to a 3′→3′ phosphodiester linkage (Figure 4.7.4). The DMTr group is retained at the end of the synthesis to enable hydrophobic purification. After the RP-HPLC purification, the 3′-terminal uridine phosphodiester is cleaved from the oligonucleotide product by treatment with fluoride and base. Additional Materials (also see Basic Protocol 2) Deprotected oligonucleotide solution (see Alternate Protocol 1), 4°C Buffer C: acetonitrile Buffer D (see recipe): 0.1 M TEAB/2% acetonitrile, pH 8 3:1 (v/v) concentrated aqueous ammonia/ethanol 1 M TEAB buffer, pH 8 (see recipe) Acetonitrile 1 M aqueous NaF (0.45-µm filtered)


Analytical RP-HPLC column (e.g., Polymer Laboratories 0.46 × 15–cm PLRP-S column) HPLC system compatible with reversed-phase buffers and solvents, equipped with a UV detector, data collection system, and a 2-mL sample injection loop

4.7.8 Supplement 3


1. DCA/CH2Cl2 2. DMTrO

O

OTBDMS

O

i-Pr2N

Ura

P OCH2CH2CN

DMTrO

Ura

O

DMTrO

Ura

O

1H-tetrazole/CH3CN 3'-DMTrOpnDNA-5'-CPG

1. Purify by RP-HPLC 2. 1 M aq. NaF/NH4OH, o 58 C, 16 hr

O P O

3. Desalt

−

+

−

O-3'-pnDNA-5'

Ura

O

OH

O O P O

O-3'-pnDNA-5'

DMTrO 3'-HOpnDNA-5'

+

OTBDMS

O 3. H2O2/H2O/C5H5N/THF o 4. NH3/EtOH, 58 C, 8-12 hr

DMTrO

Ura

O

+ O O P O

OH

−

OH

HO

O O P O

−

OH

Figure 4.7.4 Method facilitating the purification of oligonucleotide phosphoramidates by RPHPLC. DCA, dichloroacetic acid; Ura, uracil-1-yl; TBDMS, tert-butyldimethylsilyl; i-Pr, isopropyl.

Semipreparative RP-HPLC column (e.g., Polymer Labs 0.8 × 30–cm PLRP-S column) Heat block or oven set at 58°C Analyze quality of crude oligonucleotide 1. Optional: Follow the procedure for measurement of the crude OD260 and analysis by analytical IEC (see Basic Protocol 2, steps 1 to 4). 2. Dilute a second 0.5 OD260 units of ODN sample into 0.5 mL water for analysis by analytical RP-HPLC. 3. Preequilibrate the analytical RP-HPLC column for ≥10 min with 5% buffer C versus buffer D at a flow rate of 1 mL/min. 4. Program the HPLC to run a gradient of 5% to 40% buffer C versus buffer D over 40 min, followed by holding at 40% buffer C for 10 min at 1 mL/min. 5. Inject the 0.5-mL ODN sample and monitor the run at 260 nm. Reequilibrate the column for ≥10 min with 5% buffer C versus buffer D before performing a second run. Typical analytical IEC and RP-HPLC chromatograms of a crude pnODN containing a hydrophobic 3′-terminal uridine phosphodiester are shown in Figure 4.7.5. There are two product peaks in the RP-HPLC chromatogram because of partial loss of the tert-butyldimethylsilyl (TBDMS) group from uridine. Add the two peaks together to determine the amount of ODN product present. The byproduct peak generated from diphenylcarbamoyl (DPC) deprotection of G is observed near the product at 260 nm.

Prepare ODN sample and RP-HPLC purify 6. Filter the CPG away from the remaining cold ammonia solution using a 3-mL syringe attached to a 0.45-µm filter. 7. Wash the CPG twice with 0.5 mL of 3:1 concentrated aqueous ammonia/ethanol. 8. Concentrate in a Speedvac evaporator to ~0.5 mL. Do not completely dry the sample.



Supplement 3

Figure 4.7.5 (A) Analytical IEC chromatogram (40.1% pure) and (B) RP-HPLC chromatogram (42.9% pure) of the phosphoramidate oligonucleotide, 5′-CCCTCCTCCGGAGCCpUDMTr where p is a (3′,3′)-phosphodiester linkage. Two ODN product peaks are seen in the RP-HPLC because some of the TBDMS group on the 3′-terminal uridine is removed prematurely by ammonia treatment and/or subsequent workup. Peak 1, product containing uridine with 5′-O-DMTr but not 2′-O-TBDMS; peak 2, byproduct generated from diphenylcarbamoyl (DPC) deprotection of G; peak 3, product containing uridine with 5′-O-DMTr and 2′-O-TBDMS.

Do not add 0.5 M aqueous NaOH to this sample as the hydrophobic 3′-terminal uridine phosphodiester will cleave prematurely.

9. Filter the concentrated oligonucleotide again using a new 3-mL syringe attached to a 0.45-µm filter and wash twice with 0.3 mL of water. 10. Add 0.2 mL of 1 M TEAB buffer, pH 8, and 25 µL acetonitrile. The sample, once concentrated and buffered, should be purified within 12 hr. Occasionally the byproduct produced from DPC deprotection of G continues to precipitate out after filtration, especially if the sample is frozen; refilter the solution just prior to purification, if necessary, to prevent clogging of the column.

11. Preequilibrate a semipreparative RP-HPLC column for ≥15 min in 5% buffer C versus buffer D at a flow rate of 2 mL/min. Run a “blank” gradient if the column has not been used recently, and between purifications of samples with different sequences.


TEAB (buffer D) is used because, unlike the more commonly used triethylammonium acetate (TEAA), it remains basic during the post-RP-HPLC concentration and enables the isolation of pure pnODN without accompanying acid-mediated degradation.

12. Program the HPLC to run a gradient of 5% to 40% buffer C versus buffer D over 40 min, followed by a hold for 10 min at 40% buffer C at a flow rate of 2 mL/min.

4.7.10 Supplement 3


13. Inject 75 to 120 OD260 crude pnODN and collect both product peaks. Monitor the chromatography at 296 nm for preparative runs. The byproduct peak generated from DPC deprotection of G is not observed at 296 nm; the higher wavelength is used to attenuate the peak height. In general, there are some impurities just prior to the products, as well as a backside shoulder; both of these should be avoided during collection of the major fractions. It is usually best to collect only to approximately half the highest UV reading on the backside of the peak because this region contains more short-mer impurities. An example of a semipreparative RP-HPLC chromatogram and the fractionation of the peaks is shown in Figure 4.7.6.

14. Combine the two product peaks, concentrate in the Speedvac evaporator until the sample can be transferred to a 4-mL screw-cap vial, and then concentrate the sample to dryness. It is not necessary to add NaOH to the fractions; the TEAB buffer will stay basic during the concentration.

Remove 3′-terminal uridine phosphodiester, desalt, and analyze 15. Add 200 µL concentrated aqueous ammonia and 200 µL of 1 M aqueous NaF to the dry pnODN, vortex the mixture until the pnODN is dissolved, and heat the sample for 12 to 16 hr at 58°C. The fluoride removes the TBDMS group and the base causes intramolecular cleavage of the uridine phosphodiester function.

16. Cool the pnODN solution and check a small aliquot (0.2 OD260) by analytical IEC for complete cleavage. The retention time of the cleaved product is shorter than that of the uridine-containing product, and the two early-eluting uridine byproducts (see Fig. 4.7.4) are present.

17. Concentrate the solution to ∼200 µL in the Speedvac evaporator and precipitate the pnODN with 0.6 mL of 100% ethanol to remove the bulk of NaF. Freeze the sample for ≥30 min at –20°C. 18. Centrifuge for 2 min at 10,000 × g, room temperature, and carefully remove the supernatant.

Figure 4.7.6 Semipreparative RP-HPLC chromatogram of a pnODN with the sequence 5′CCCTCCTCCGGAGCCpUDMTr where p is a (3′,3′)-phosphodiester linkage.



Supplement 9

19. Dissolve the product in 1 mL of water and desalt the sample on a Sephadex G-25 column using the manufacturer’s protocol. The Sephadex G-25 column is necessary in this case because ethanol precipitation does not remove the uridine byproducts.

20. Measure the OD units at 260 nm to determine the yield of pnODN. Inject 0.2 OD260 units on an analytical IEC column to determine the purity of the product. Alternatively, determine ODN purity by CGE or PAGE (APPENDIX 3B).

21. Concentrate the pnODN to dryness in the Speedvac evaporator and store up to 1 year at −20°C. SUPPORT PROTOCOL 1

SYNTHESIS OF 3′-TRITYLAMINO-2′,3′-DEOXYTHYMIDINE The synthesis of 3′-tritylamino-2′,3′-deoxythymidine (S.5) from thymidine is shown in Figure 4.7.7. The 3′-azido-5′-O-(4-methoxybenzoyl)-2′,3′-deoxythymidine (S.3) is prepared as previously reported (Czernecki and Valéry, 1991). Materials Thymidine N,N-Dimethylformamide (DMF) Triphenylphosphine p-Anisic acid Diisopropylazodicarboxylate (DIAD) Diethyl ether, 5°C Lithium azide (LiN3) Ethyl acetate Saturated aqueous NaCl Sodium sulfate (Na2SO4, anhydrous) 8:2 (v/v) ethyl acetate/hexane 1:1 (v/v) ethanol/dichloromethane (CH2Cl2) Hydrogen 10% Pd/C catalyst (Aldrich) Pyridine (anhydrous) Triethylamine Trityl chloride Chromatography-grade silica gel, 70-230 mesh 60 Å (Aldrich) 0.5% to 5% (v/v) triethylamine in 2% (v/v) methanol/CH2Cl2 2% to 5% (v/v) methanol/CH2Cl2 5:95, 1:9, and 2:8 (v/v) methanol/CH2Cl2 57:43 (v/v) 1,4-dioxane/methanol 2 M aqueous NaOH (APPENDIX 2A) Dowex 50W-X8 cation-exchange resin (pyridinium H+ form; see recipe) Saturated aqueous NaHCO3 1:1 (v/v) ethyl ether/hexane


Heating mantle, variac, and temperature controller Mechanical overhead stirrer (Fisher) TLC plates (e.g., 0.2-mm-thick precoated Merck silica gel 60 F254 plates) Rotary evaporator Parr shaker-type hydrogenator able to hold pressures up to 60 psi

4.7.12 Supplement 9


O HO

O

Thy

2. DIAD/Ph3P/DMF 3. LiN3/DMF, 110 oC

OH

O

1. DIAD/Ph3P/4-MeOC6H4CO2H

O

Thy

MeO

1. H2,10% Pd/C, EtOH/CH2Cl2 2. TrCl/Et3N/C5H5N

N3 3 (98%)

T

CH3 OCH2CH2CN O

N O

O

MeO

Thy

2 M aq. NaOH

HO

O

Thy

1,4-Dioxan/MeOH NHTr 4 (83%)

NHTr

P Cl CH3 1t

DBU/CH2Cl2

5 (90%)

Figure 4.7.7 Synthetic steps in the preparation of 3′-tritylamino-2′,3′-dideoxythymidine (S.5) from thymidine. Thy, thymin-1-yl; DIAD, diisopropylazodicarboxylate; DBU, 1,8-diazabicyclo[5.4.0]undec7-ene.

Additional reagents and equipment for thin-layer chromatography (TLC, APPENDIX 3D) and column chromatography (APPENDIX 3E) Synthesize 3′-azido-5′-O-(4-methoxybenzoyl)-2′,3′-deoxythymidine (S.3) 1. Dissolve 141.0 g (582.2 mmol) thymidine in 1125 mL DMF in a 3-liter round-bottom flask. 2. Add 183.2 g (698.4 mmol) triphenylphosphine and 106.3 g (698.4 mmol) p-anisic acid. 3. Add 137.5 mL (698.4 mmol) diisopropylazodicarboxylate, diluted in 150 mL DMF, over 35 min using an additional funnel. 4. After 40 min, add another 183.2 g (698.4 mmol) triphenylphosphine all at once, and another 141.3 g (698.4 mmol) diisopropylazodicarboxylate, dissolved in 150 mL DMF, over a 40-min period using an additional funnel. 5. Stir the resultant mixture for an additional 65 min. 6. Quench the reaction with 10 mL of water and concentrate the solution to a volume of ∼800 mL using a rotary evaporator equipped with a vacuum pump. 7. Precipitate the product by pouring it into 6000 mL cold diethyl ether, 5°C, with rapid stirring. 8. Filter the solid using a Büchner funnel and house vacuum, wash with 2000 mL cold diethyl ether, and dry in vacuo to give semipure 2,3′-anhydro-5′-O-(4-methoxybenzoyl)-2′,3′-deoxythymidine. Crude yield = 110% (230.2 g).

9. Using a mechanical stirrer, dissolve 230.0 g (642.5 mmol) of 2,3′-anhydro-5′-O-(4methoxybenzoyl)-2′,3′-deoxythymidine in 1200 mL DMF, add 47.2 g (963.8 mmol) lithium azide, and heat the mixture for 48 hr at 100° to 110°C. 10. Concentrate the solution on the rotary evaporator using a vacuum pump, dissolve the residue in 3000 mL ethyl acetate, and wash two times with 1000 mL water and three times with 600 mL saturated aqueous NaCl.



Supplement 9

11. Dry the organic solution over anhydrous Na2SO4, filter, and concentrate on the rotary evaporator with a vacuum line to afford 3′-azido-5′-O-(4-methoxybenzoyl)-2′,3′-deoxythymidine (S.3) as an amber foam. 12. Perform TLC analysis (APPENDIX 3D) on 0.2-mm-thick precoated Merck silica gel 60 F254 plates to confirm the purity of the product. Elute with 8:2 (v/v) ethyl acetate/hexane. Crude yield = 97.9% (228.9 g, 570.0 mmol). Rf (8:2 ethyl acetate/hexane) = 0.51. 1H NMR (CDCl3 /TMS): δ 9.36 (1H, br s), 7.98 (2H, d, J = 8.7 Hz), 7.22 (1H, s), 6.95 (2H, d, J = 8.7 Hz), 6.18 (1H, t, J = 6.5 Hz), 4.65 (1H, dd, J = 12.3, 3.3 Hz), 4.53 (1H, dd, J = 12.3, 3.6 Hz), 4.35 (1H, m), 4.21 (1H, dt, J = 3.6, 2.5 Hz), 3.87 (3H, s), 2.54 (1H, m), 2.35 (1H, m), 1.71 (3H, s).

Synthesize 5′-O-(4-methoxybenzoyl)-3′-tritylamino-2′,3′-deoxythymidine (S.4) 13. Dissolve 10.0 g (24.9 mmol) of S.3 in 500 mL of 1:1 (v/v) ethanol/CH2Cl2 and reduce via hydrogenation (60 psi H2) in the presence of 1 g of 10% Pd/C catalyst for 16 hr. 14. Remove the catalyst by vacuum filtration and evaporate the solvent in vacuo. Yield = 92% (8.6 g, 22.9 mmol) of the corresponding 3′-amine, which is taken directly to the next reaction.

15. Dry 8.6 g (22.9 mmol) of 3′-amino-5′-O-(4-methoxybenzoyl)-2′,3′-deoxythymidine by azeotropic removal of water two times with 50 mL pyridine, evaporating to dryness on a rotary evaporator each time. Dissolve in 50 mL anhydrous pyridine. 16. Add 6.17 mL (48.1 mmol) triethylamine and 7.0 g (25.2 mmol) trityl chloride. Stir this mixture for 2 hr at room temperature. Perform TLC analysis (see step 12), eluting with 5:95 (v/v) methanol/CH2Cl2, to determine if the reaction is complete 17. If the reaction is complete, go to step 18, otherwise add an additional 1.9 g (6.9 mmol) trityl chloride, if necessary, and continue stirring for an additional 2 hr. 18. Remove the solvents using a rotary evaporator equipped with a vacuum pump and purify the crude product (S.4) by gravity on a silica gel column (APPENDIX 3E) preequilibrated with 0.5% triethylamine in 2% methanol/CH2Cl2. Obtain product by eluting with 2% to 5% methanol/CH2Cl2. For acid-sensitive (trityl-containing and phosphoramidite) compounds, the column should be packed and equilibrated with 0.5% triethylamine.

19. Perform TLC analysis (step 12) to confirm the purity of the product. Elute with 5:95 (v/v) methanol/CH2Cl2. Yield = 90% (12.7 g, 20.6 mmol). Rf (5:95 methanol/CH2Cl2) = 0.45. 1H NMR (CDCl3 /TMS): δ 8.26 (1H, br s, exchanges with D2O), 7.84, 6.90 (4H, AB, J = 8.86 Hz), 7.55 (6H, d, J = 7.45 Hz), 7.29 (6H, t, J = 7.56 Hz), 7.21 (3H, t, J = 7.27 Hz), 7.02 (1H, s), 6.08 (1H, t, J = 6.19 Hz), 4.59 (1H, dd, J = 12.35, 2.35 Hz), 4.29 (1H, dd, J = 12.43, 3.94 Hz), 3.98 (1H, m), 3.88 (3H, s), 3.41 (1H, m), 1.97 (1H, br, exchanges with D2O), 1.65-1.75 (1H, m), 1.58 (3H, s), 1.30-1.40 (1H, m). HRMS (FAB+): calcd for [M + Cs]+, 750.1580, observed 750.1559.

Synthesize 3′-tritylamino-2′,3′-deoxythymidine (S.5) 20. Remove the 5′-O-anisoyl protecting group by dissolving 30.1 g (48.7 mmol) of S.4 in 150 mL of 57:43 (v/v) 1,4-dioxane/methanol, and then adding 73.1 mL (146.2 mmol) of 2 M aqueous NaOH. Stir the solution for 1.5 hr at room temperature. Synthesis and Purification of Oligonucleotide N3′→P5′ Phosphoramidates

21. Neutralize with ~150 g of Dowex 50W-X8 cation-exchange resin (dry pyridinium H+ form, 1.6 meq/g).

4.7.14 Supplement 9


22. Once the pH is neutral (∼10 min), filter the resin, wash three times with 40 mL of 2:8 methanol/CH2Cl2, and concentrate the crude product on the rotary evaporator using a vacuum line. 23. Dissolve the residue in 500 mL ethyl acetate and extract two times with 250 mL saturated aqueous NaHCO3, once with 250 mL water, and once with 250 mL saturated aqueous NaCl. 24. Dry the organic phase over anhydrous Na2SO4 and filter. 25. Remove the solvents using a rotary evaporator with a vacuum line and dissolve the resulting foam in 300 mL of 5:95 methanol/CH2Cl2. 26. Add this solution slowly to 1250 mL of a rapidly stirring mixture of 1:1 (v/v) diethyl ether/hexane to precipitate the pure 3′-tritylamino-2′,3′-deoxythymidine (S.5). 27. Perform TLC analysis (step 12) to confirm the purity of the product. Elute with 1:9 (v/v) methanol/CH2Cl2. Yield = 90% (21.2 g, 43.8 mmol). Rf (1:9 methanol/CH2Cl2) = 0.50. 1H NMR (CDCl3 /TMS): δ 8.30 (1H, br s, exchanges with D2O), 7.52 (6H, d, J = 7.46 Hz), 7.29 (6H, t, J = 7.55 Hz), 7.21 (3H, t, J = 7.25 Hz), 7.16 (1H, s), 6.01 (1H, t, J = 6.38 Hz), 3.85 (1H, d, J = 11.71 Hz), 3.74 (1H, m), 3.65 (1H, dd, J = 11.99, 2.59), 3.34 (1H, q, J = 6.54 Hz), 1.80-2.00 (1H, br, exchanges with D2O), 1.83 (3H, s), 1.45-1.55 (1H, m), 1.30-1.40 (1H, m). HRMS (FAB+): calcd for [M + Cs]+, 616.1212, observed 616.1226.

SYNTHESIS OF N4-BENZOYL-3′-TRITYLAMINO-2′,3′-DIDEOXYCYTIDINE The synthesis of N4-benzoyl-3′-tritylamino-2′,3′-deoxycytidine (S.9) is shown in Figure 4.7.8. The C monomer is more readily and efficiently synthesized by the dU→dC route, rather than by lithium azide ring opening of a 2,3′-anhydro-2′-deoxycytidine derivative (Reese and Skone, 1984; Nelson et al., 1997).

SUPPORT PROTOCOL 2

Additional Materials (also see Support Protocol 1) 2′-Deoxyuridine 4-Dimethylaminopyridine tert-Butyldimethylsilyl (TBDMS) chloride 2:1 (v/v) ethanol/CH2Cl2 1,2,4-Triazole Phosphorus oxychloride (POCl3) 1,4-Dioxane Benzoyl chloride Concentrated aqueous ammonia (28%), 4°C Tetrahydrofuran (THF) 1 M tetra-n-butylammonium fluoride (TBAF) in THF 1% (v/v) triethylamine in 30% to 50% (v/v) ethyl acetate/hexane Synthesize 3′-azido-5′-O-(tert-butyldimethylsilyl)-2′,3′-dideoxyuridine (S.6) 1. Dry 11.4 g (50 mmol) of 2′-deoxyuridine two times thoroughly via co-evaporation with 100 mL anhydrous DMF in vacuo. 2. Add 100 mL DMF followed by 8.36 mL (60 mmol) triethylamine, 0.31 g (2.5 mmol) of 4-dimethylaminopyridine, and 8.29 g (55.0 mmol) tert-butyldimethylsilyl chloride. Stir the reaction mixture for 1 hr at room temperature. 3. Dilute with 600 mL CH2Cl2 and extract three times with 200 mL water and once with 200 mL saturated aqueous NaCl.



Supplement 9

O

O

NH

NH HO

O

N

O

1. TBDMS-Cl/cat. DMAP/Et3N/DMF

TBDMSO

2. DIAD/Ph3P/DMF 3. LiN3/ DMF/∆

OH

O

N

O

1. H2,10% Pd/C, EtOH/CH2Cl2 2. TrCl/Et3N/C5H5N

N3 6 (63%)

dU

NHBz

O

N

NH TBDMSO

O

N

O

1. POCl3/Et3N/1,2,4-triazole/MeCN 2. NH4OH/1,4-dioxan

TBDMSO

3. BzCl/C5H5N

NHTr

O

N

O TBAF/THF

NHTr

7 (85%)

8 (92%) NHBz N HO

O

N

O

CH3 OCH2CH2CN P N Cl CH3 1c DBU/CH2Cl2

NHTr 9 (88%)

Figure 4.7.8 Synthetic steps in the preparation of N4-benzoyl-3′-tritylamino-2′,3′-dideoxycytidine (S.9) from 2′-deoxyuridine. DMAP, 4-dimethylaminopyridine; Bz, benzoyl; TBAF, tetra-n-butylammonium fluoride.

4. Dry the organic layer over anhydrous Na2SO4, vacuum filter, and concentrate on a rotary evaporator using a vacuum pump. 5. Purify the resulting residue by gravity column chromatography (APPENDIX 3E) on silica gel with 2% to 10% methanol/CH2Cl2 to afford 5′-O-(tert-butyldimethylsilyl)-2′-deoxyuridine. Yield = 80% (13.7 g, 40.0 mmol).

6. Dissolve 13.7 g (40.0 mmol) of 5′-protected nucleoside and 16.8 g (64.0 mmol) triphenylphosphine in 100 mL DMF. While stirring, add 12.6 mL (64.0 mmol) diisopropylazodicarboxylate in 20 mL DMF. 7. Stir 2 hr at room temperature, concentrate the reaction mixture on a rotary evaporator using a vacuum pump to ∼30 mL, and pour into 1200 mL diethyl ether. The desired product precipitates out after ∼10 min of rapid stirring.

8. Place the resulting mixture overnight at 4°C. 9. Collect the precipitate by vacuum filtration, wash two times with 300 mL cold diethyl ether, and dry in vacuo to afford 2,3′-anhydro-5′-O-(tert-butyldimethylsilyl)-2′-deoxyuridine as a white solid. Yield = 90% (11.7 g, 36.0 mmol).

10. React 33.8 g (104.2 mmol) of 2,3′-anhydro-5′-O-(tert-butyldimethylsilyl)-2′-deoxyuridine with 7.65 g (156.3 mmol) LiN3 in 300 mL DMF for 48 hr at 95° to 100°C. Synthesis and Purification of Oligonucleotide N3′→P5′ Phosphoramidates

11. Cool the resulting brown homogeneous mixture to room temperature, and concentrate to an oil using a rotary evaporator and vacuum pump. 12. Dissolve the residue in 800 mL ethyl acetate and extract with 200 mL water.

4.7.16 Supplement 9


13. Extract the aqueous layer twice more with 75 mL ethyl acetate and combine the organics. Wash three times with 250 mL water and once with 250 mL saturated aqueous NaCl. 14. Dry the ethyl acetate solution over anhydrous Na2SO4, vacuum filter, and concentrate on a rotary evaporator with a vacuum line to afford 3′-azido-5′-O-(tert-butyldimethylsilyl)-2′,3′-dideoxyuridine (S.6) as a brownish foam. Proceed directly to hydrogenation. 15. Perform TLC analysis (APPENDIX 3D) on 0.2-mm-thick precoated Merck silica gel 60 F254 plates to confirm the purity of the product. Elute with 8:92 (v/v) methanol/CH2Cl2. Yield = 87% (33.2 g, 90.3 mmol). Rf (8:92 methanol/CH2Cl2) = 0.57. 1H NMR (CDCl3 /TMS): δ 8.87 (1H, br s, exchanges with D2O), 7.91 (1H, d, J = 8.10 Hz), 6.23 (1H, t, J = 5.88 Hz), 5.71 (1H, d, J = 8.18 Hz), 4.25 (1H, q, J = 5.91 Hz), 3.95-4.05 (2H, m), 3.83 (1H, dd, J = 11.40, 1.68 Hz), 2.45-2.55 (1H, m), 2.25-2.35 (1H, m), 0.95 (9H, s), 0.15 (3H, s), 0.14 (3H, s). HRMS (FAB+): calcd for [M + H]+, 368.1754, observed 368.1747.

Synthesize 5′-O-(tert-butyldimethylsilyl)-3′-tritylamino-2′,3′-dideoxyuridine (S.7) 16. Dissolve 33.2 g (90.3 mmol) of crude S.6 in 300 mL of 2:1 (v/v) ethanol/CH2Cl2 and reduce via hydrogenation (60 psi H2) in the presence of 3.0 g of 10% Pd/C catalyst for 18 hr. 17. Remove the catalyst by vacuum filtration and evaporate the solvent on a rotary evaporator using a vacuum line to afford the corresponding 3′-amine. Proceed directly to the next reaction. Yield = 99.4% (30.4 g, 89.8 mmol).

18. Azeotrope 30.4 g (89.8 mmol) of 3′-amino-5′-O-(tert-butyldimethylsilyl)-2′,3′ -dideoxyuridine two times with 300 mL pyridine and dissolve the solid in a mixture of 600 mL CH2Cl2 and 70 mL anhydrous pyridine. 19. Add 25.0 mL (179.6 mmol) triethylamine and 30.0 g (125.7 mmol) trityl chloride to this solution and stir for 2 hr at room temperature. 20. Purify (see Support Protocol 1, step 18) to afford 5′-O-(tert-butyldimethylsilyl)-3′tritylamino-2′,3′-dideoxyuridine (S.7). 21. Perform TLC analysis (step 15) to confirm the purity of the product. Elute with 8:2 (v/v) ethyl acetate/hexane. Yield = 85% (44.3 g, 75.9 mmol). Rf (8:2 ethyl acetate/hexane) = 0.58. 1H NMR (CDCl3/TMS): δ 8.24 (1H, br s, exchanges with D2O), 7.73 (1H, d, J = 8.25 Hz), 7.52 (6H, d, J = 7.78 Hz), 7.31 (6H, m), 7.23 (3H, t, J = 7.23 Hz), 6.21 (1H, t, J = 6.69 Hz), 5.60 (1H, d, J = 8.17 Hz), 3.84 (1H, m), 3.76 (1H, dd, J = 11.34, 2.00 Hz), 3.48 (1H, dd, J = 11.37, 2.27 Hz), 3.32 (1H, m), 2.07 (1H, br, exchanges with D2O), 1.60-1.70 (1H, m), 1.45-1.55 (1H, m), 0.84 (9H, s), 0.01 (3H, s), −0.05 (3H, s). HRMS (FAB+): calcd for [M + Na]+, 606.2764, observed 606.2751.

Synthesize N4-benzoyl-5′-O-(tert-butyldimethylsilyl)-3′-tritylamino-2′,3′-dideoxycytidine (S.8) 22. Add 22.5 mL (161.1 mmol) triethylamine dropwise over a period of 10 min to a stirring mixture of 11.1 g (161.1 mmol) of 1,2,4-triazole and 3.5 mL (37.1 mmol) phosphorus oxychloride in 125 mL anhydrous CH3CN at 0°C. 23. To this cold stirring mixture, add 9.4 g (16.1 mmol) of S.7 as a solution in 50 mL acetonitrile. Stir 2 hr at room temperature.



Supplement 3

24. Quench the reaction with 30 mL triethylamine and 10 mL water. 25. Remove the solvents using a rotary evaporator and vacuum line. 26. Dissolve the resulting brown solid in 250 mL CH2Cl2. Extract three times with 150 mL saturated aqueous NaHCO3 and once with 150 mL saturated aqueous NaCl. 27. Dry the organic solution over anhydrous Na2SO4, vacuum filter, and concentrate on a rotary evaporator using a vacuum line to afford 4-(1,2,4-triazol-1-yl)-5′-O-(tertbutyldimethylsilyl)-3′-tritylamino-2′,3′-dideoxyuridine as an orange solid. Crude yield = 100% (10.2 g, 16.1 mmol).

28. Dissolve this crude material in 200 mL of 1,4-dioxane and add 50 mL concentrated NH4OH, 4°C. 29. Stir the reaction mixture for 4 hr at room temperature and concentrate on a rotary evaporator using a vacuum pump to afford 5′-O-(tert-butyldimethylsilyl)-3′-tritylamino-2′,3′-dideoxycytidine as a beige solid. Crude yield = 100% (9.4 g, 16.1 mmol).

30. Azeotrope the crude material two times with 200 mL anhydrous pyridine. 31. Dissolve in 200 mL pyridine, cool externally in a 0°C ice bath, and add 2.2 mL (19.3 mmol) benzoyl chloride. 32. Allow the reaction to slowly warm to room temperature and stir an additional 16 hr at room temperature. 33. Cool the reaction mixture externally to 0°C and quench with 40 mL water. Stir 5 min. 34. Add 40 mL concentrated aqueous ammonia, 4°C, and stir the reaction mixture for an additional 15 min at 0°C. 35. Remove the solvents using a rotary evaporator and vacuum pump, dissolve the residue in 125 mL CH2Cl2, and extract three times with 75 mL saturated aqueous NaHCO3. 36. Dry the organic phase over Na2SO4, vacuum filter, and evaporate the solvents using a rotary evaporator and vacuum line. 37. Purify the crude material (see Support Protocol 1, step 18) to afford N4-benzoyl-5′O-(tert-butyldimethylsilyl)-3′-tritylamino-2′,3′-dideoxycytidine (S.8). 38. Perform TLC analysis (step 15) to confirm the purity of the product. Elute with 5:95 (v/v) methanol/CH2Cl2. Yield = 92% (10.2 g, 14.8 mmol). Rf (5:95 methanol/CH2Cl2) = 0.71. 1H NMR (CDCl3 / TMS): δ 8.70 (1H, d, J = 7.36 Hz, exchanges with D2O), 8.27 (1H, d, J = 7.36 Hz), 7.91 (2H, d, J = 7.46 Hz), 7.62 (1H, t, J = 7.20 Hz), 7.50-7.60 (8H, m; with 6H, d, J = 7.74 Hz at 7.52), 7.42 (1H, br d, J = 7.41 Hz), 7.30 (6H, t, J = 7.39 Hz), 7.22 (3H, t, J = 7.39 Hz), 6.26 (1H, t, J = 6.26 Hz), 3.80 (1H, br m), 3.77 (1H, br d, J = 11.39 Hz), 3.49 (1H, dd, J = 11.24, 2.33 Hz), 3.30 (1H, m), 1.90-2.10 (2H, br m; 1H exchanges in D2O), 1.52 (1H, dt, J = 13.57, 6.76 Hz), 0.86 (9H, s), 0.04 (3H, s), −0.01 (3H, s). HRMS (FAB+): calcd for [M + Cs]+, 819.2343, observed 819.2366.


4.7.18 Supplement 3


Synthesize N4-benzoyl-3′-tritylamino-2′,3′-dideoxycytidine (S.9) 39. Remove the 5′-TBDMS protecting group by dissolving 2.0 g (2.85 mmol) S.8 in 15 mL THF and reacting it with 15 mL of 1 M TBAF in THF for 16 hr. 40. Concentrate the reaction mixture to a syrup on the rotary evaporator using a vacuum line and dissolve the residue in 25 mL CH2Cl2. Extract four times with 25 mL water and once with 25 mL saturated aqueous NaCl. 41. Dry the organic layer over Na2SO4, vacuum filter, and remove the solvent using the rotary evaporator and vacuum line. 42. Purify the crude product on a silica gel column preequilibrated with 1% triethylamine in 30% ethyl acetate/hexane, and elute with 30% to 50% ethyl acetate/hexane to afford N4-benzoyl-3′-tritylamino-2′,3′-dideoxycytidine (S.9). 43. Perform TLC analysis (step 15) to confirm the purity of the product. Elute with 5:95 (v/v) methanol/CH2Cl2. Yield = 88% (1.4 g, 2.50 mmol). Rf (5:95 methanol/CH2Cl2) = 0.55. 1H NMR (CDCl3 / TMS): δ 8.65 (1H, br s, exchanges with D2O), 8.19 (1H, d, J = 7.36 Hz), 7.87 (2H, d, J = 7.57 Hz), 7.62 (1H, t, J = 7.37 Hz), 7.47-7.57 (9H, m), 7.30 (6H, t, J = 7.50 Hz), 7.23 (3H, t, J = 7.24 Hz), 6.07 (1H, dd, J = 6.66, 4.31), 3.91 (1H, d, J = 12.00 Hz), 3.79 (1H, m), 3.73 (1H, d, J = 12.10 Hz), 3.30 (1H, q, J = 6.38 Hz), 1.80-2.00 (2H, m, 1 br H exchanges in D2O), 1.40 (1H, ddd, J = 13.89, 7.03, 4.41 Hz). HRMS (FAB+): calcd for [M + Na]+, 595.2321, observed 595.2310.

SYNTHESIS OF N2-ISOBUTYRYL-O6-(N,N-DIPHENYLCARBAMOYL)3′-TRITYLAMINO-2′,3′-DIDEOXYGUANOSINE

SUPPORT PROTOCOL 3

The synthesis of N2-isobutyryl-O6-(N,N-diphenylcarbamoyl)-3′-tritylamino-2′,3′-dideoxyguanosine (S.13) is depicted in Figure 4.7.9. The 3′-O-benzoyl-N2-isobutyryl- 2′-deoxyxyloguanosine (S.10) is synthesized as previously reported (Nishino et al., 1986; Herdewijn and Van Aerschot, 1989). Additional Materials (also see Support Protocol 1) N2-Isobutyryl-2′-deoxyguanosine Benzoyl chloride Trifluoromethanesulfonic anhydride 4-Dimethylaminopyridine tert-Butyldimethylsilyl chloride 1:1 (v/v) methanol/1,4-dioxane 1 M aqueous HCl Diethylazodicarboxylate Argon N,N-Diisopropylethylamine N,N-Diphenylcarbamyl chloride Triethylamine trihydrofluoride Toluene 7:3 and 6:4 (v/v) ethyl acetate/hexane 2-liter large-mouth Erlenmeyer flask



Supplement 4

O N

O

NH N

HO

N

O

N

NHiBu

HO 2. (CF3SO2)2O/C5H5N/CH2Cl2, -15 oC 3. H2O

OH

NH

1. BzCl/C5H5N R O

dGiBu

N

N

NHiBu

10 R = OBz (35%) O N

1. TBDMS-Cl/DMAP/Et3N/DMF 2. 2 M NaOH, MeOH/dioxan (1:1 v/v), 5 oC

TBDMSO

3. LiN3/DEAD/Ph3P/DMF 4. H2,10% Pd/C, EtOH/CH2Cl2

O

N

NH N

NHiBu

TrCl/Et3N/C5H5N

NPh2

CH3 OCH2CH2CN P N Cl

NH2 11 (49%) O

O N R1O

O

N

O N

NH N

NHiBu 1. DPC-Cl/i-Pr NEt/C H N 2 5 5

HO

O

N

N N

NHiBu

CH3 1g

NHTr

2. Et3N•3HF/C5H5N/CH2Cl2

12 R1 = TBDMS (100%)

DBU/CH2Cl2

NHTr 13 (82%)

Figure 4.7.9 Synthetic steps in the preparation of N2-isobutyryl-O6-(N,N-diphenylcarbamoyl)-3′tritylamino-2′,3′-dideoxyguanosine (S.13) from N2-isobutyryl-2′-deoxyguanosine. DEAD, diethylazodicarboxylate; iBu, isobutyryl; DPC-Cl, N,N-diphenylcarbamyl chloride.

Synthesize 3′-O-benzoyl-N2-isobutyryl-2′-deoxyxyloguanosine (S.10) 1. Azeotrope 119.1 g (353.1 mmol) N2-isobutyryl-2′-deoxyguanosine two times with 600 mL pyridine and then add 1500 mL pyridine. The solid does not completely dissolve.

2. Add 45.1 mL (388.4 mmol) benzoyl chloride, dissolved in 250 mL pyridine, dropwise over 5 hr using an additional funnel, and then continue stirring for another 11 to 16 hr. 3. Quench the reaction with 20 mL water, then remove the solvent using a rotary evaporator and vacuum pump. 4. Dissolve the solid in 900 mL CH2Cl2, wash with 400 mL water, and then transfer the organic phase to a 2-liter large-mouth Erlenmeyer flask. 5. Add 400 mL water and stir vigorously to precipitate the white solid product. 6. Vacuum filter the solid and wash three times with 100 mL of water. 7. Dry the solid using a vacuum pump to obtain the 5′-O-benzoyl-N2-isobutyryl-2′-deoxyguanosine. Yield = 91.6% (142.7 g, 323.4 mmol). Rf (1:9 methanol/CH2Cl2) = 0.27.

8. Azeotrope 142.7 g (323.4 mmol) of 5′-O-benzoyl-N2-isobutyryl-2′-deoxyguanosine three times with 400 mL pyridine, and then add 240 mL pyridine and 2200 mL CH2Cl2. Synthesis and Purification of Oligonucleotide N3′→P5′ Phosphoramidates

9. Cool the mixture to −15°C and add, with stirring, 92.5 mL (549.7 mmol) trifluoromethanesulfonic anhydride dissolved in 500 mL CH2Cl2, dropwise over 1 hr, while maintaining the temperature between −10° and −15°C.

4.7.20 Supplement 4


10. Perform TLC analysis (APPENDIX 3D) on 0.2-mm-thick precoated Merck silica gel 60 F254 plates to confirm that the reaction is complete. Elute with 1:9 (v/v) methanol/CH2Cl2 (Rf = 0.6). 11. When the reaction is complete (after ∼1.5 hr), slowly quench the reaction with 150 mL water, maintaining the reaction temperature below 10°C. 12. Allow the solution to warm to room temperature and continue stirring for an additional 16 hr to complete the inversion. 13. Wash the organic layer two times with 1000 mL water and once with 1000 mL saturated aqueous NaCl. 14. Dry the organic solution over anhydrous Na2SO4, vacuum filter, and concentrate on the rotary evaporator and vacuum line to a brown solid. 15. Dissolve the solid in 550 mL CH2Cl2, stir for 1 hr, and then cool overnight at 4°C. 16. Vacuum filter the white solid, wash with 750 mL cold CH 2Cl2, and then dry using a vacuum pump to afford 3′-O-benzoyl-N2-isobutyryl-2′-deoxyxyloguanosine (S.10). 17. Concentrate the mother liquor using a rotary evaporator and vacuum line and repeat the recrystallization to obtain a second crop of pure product. 18. Perform TLC analysis (step 10) to confirm the purity of the product. Combined yield = 38.6% (55.1 g, 124.9 mmol). Rf (1:9 methanol/CH2Cl2) = 0.27. 1H NMR (DMSO-d6): δ 12.04 (1H, s), 11.70 (1H, s), 8.16 (1H, s), 7.84 (2H, d, J = 7.79 Hz), 7.67 (1H, t, J = 7.54 Hz), 7.52 (2H, t, J = 7.69 Hz), 6.25 (1H, dd, J = 7.55, 2.00 Hz), 5.69 (1H, t, J = 4.3 Hz), 4.96 (1H, t, J = 5.49 Hz), 4.32 (1H, dt, J = 7.79, 5.88 Hz), 3.77 (2H, m), 3.00 (1H, m), 2.76 (2H, m), 1.11 (6H, d, J = 6.76 Hz). The product (S.10) coelutes by TLC with the 5′-O-benzoyl-N2-isobutyryl-2′-deoxyguanosine but can be distinguished by 1H NMR spectroscopy.

Synthesize 3′-amino-5′-O-(tert-butyldimethylsilyl)-N2-isobutyryl-2′,3′-dideoxyguanosine (S.11) 19. To a stirring solution of 4.86 g (11.0 mmol) S.10 in 20 mL DMF, add 3.4 mL (24.2 mmol) triethylamine, 54 mg (0.44 mmol) 4-dimethylaminopyridine, and 3.31 g (22.0 mmol) tert-butyldimethylsilyl chloride. 20. Stir the reaction for 2 hr at room temperature. 21. Add 10 mL methanol and, after stirring an additional 5 min, concentrate the reaction mixture on a rotary evaporator using a high vacuum. 22. Dissolve the residue in 150 mL CH2Cl2 and wash three times with 40 mL water and once with 60 mL saturated aqueous NaCl. 23. Dry the organic layer over Na2SO4, vacuum filter, and concentrate with a rotary evaporator using a vacuum line to afford 6.40 g (>100% crude yield) of 3′-O-benzoyl-5′-O-(tert-butyldimethylsilyl)-N2-isobutyryl-2′-deoxyxyloguanosine. 24. Dissolve the crude material in 100 mL 1:1 methanol/1,4-dioxane and cool to 5°C. 25. Add 44.0 mL (87.9 mmol) prechilled (5°C) 2 M aqueous NaOH and stir the reaction mixture for 15 to 20 min in an ice bath. Monitor this reaction carefully and neutralize the hydroxide as soon as possible in order to avoid loss of the isobutyryl group.

26. Neutralize the reaction with 97 mL of 1 M aqueous HCl to pH 6 to 7.



Supplement 3

27. Remove the ice bath and concentrate the reaction mixture to ∼50 mL on the rotary evaporator with a high vacuum. 28. Extract three times with 75 mL CH2Cl2. 29. Wash the combined organics three times with 50 mL saturated aqueous NaHCO3 and two times with 50 mL saturated aqueous NaCl. 30. Dry the organic solution over Na2SO4, vacuum filter, and concentrate with a rotary evaporator using a vacuum line to afford 5′-O-(tert-butyldimethylsilyl)-N2-isobutyryl-2′-deoxyxyloguanosine. Proceed to the next reaction without further purification. Yield = 82% (4.1 g, 9.1 mmol).

31. To 47.3 g (104.7 mmol) crude 5′-O-(tert-butyldimethylsilyl)-N2-isobutyryl-2′-deoxyxyloguanosine dissolved in 1000 mL anhydrous DMF, add 15.4 g (314.1 mmol) LiN3 and 41.2 g (157.1 mmol) triphenylphosphine. 32. Add 24.7 mL (157.1 mmol) diethylazodicarboxylate and stir the reaction mixture for 5 hr at room temperature under argon. 33. Quench the reaction with 20 mL water and concentrate the reaction mixture on the rotary evaporator using a vacuum pump. 34. Dissolve the residue in 1500 mL ethyl acetate. 35. Wash three times with 1000 mL water and once with 1000 mL saturated aqueous NaCl. 36. Dry the organic solution over Na2SO4, vacuum filter, and concentrate using a rotary evaporator and vacuum line. Proceed directly to hydrogenation and purification of the 3′-amine. 37. Dissolve ≤104.7 mmol crude azide in 1600 mL of 1:1 (v/v) ethanol/CH2Cl2 and hydrogenate (60 psi H2) in the presence of 2.5 g of 10% Pd/C catalyst for 16 hr at room temperature. 38. Remove the catalyst by vacuum filtration and evaporate the solvent using a rotary evaporator and vacuum line to afford the crude 3′-amine. 39. Purify by gravity on a silica gel column (APPENDIX 3E) using 2% to 6% methanol/CH2Cl2 and then 1% triethylamine/6% methanol/CH2Cl2 to afford 3′-amino-5′O-(tert-butyldimethylsilyl)-N2-isobutyryl-2′,3′-dideoxyguanosine (S.11) as an off-white foam. 40. Perform TLC analysis (step 10) to confirm the purity of the product. Yield = 60% (28.2 g, 63.2 mmol). Rf (1:9 methanol/CH2Cl2) = 0.14. 1H NMR (CDCl3/TMS): δ 8.01 (1H, s), 6.17 (1H, dd, J = 6.77, 3.98 Hz), 3.80-3.90 (4H, mm), 2.83 (1H, septet, J = 6.80 Hz), 2.59 (1H, ddd, J = 13.26, 6.16, 4.03 Hz), 2.33 (1H, dt, J = 13.19, 6.79 Hz), 1.26 (6H, dd, J = 6.86, 2.79 Hz), 0.88 (9H, s), 0.07 (3H, s), 0.06 (3H, s). HRMS (FAB+): calcd for [M + H]+, 451.2489, observed 451.2480.


Synthesize 5′-O-(tert-butyldimethylsilyl)-N2-isobutyryl-3′-tritylamino-2′,3′-dideoxyguanosine (S.12) 41. Dissolve 28.5 g (63.2 mmol) S.11 in 500 mL pyridine, add 17.6 mL (126.4 mmol) triethylamine and 28.2 g (101.1 mmol) trityl chloride, and stir for 16 hr at room temperature.

4.7.22 Supplement 3


42. Concentrate the reaction product to a solid using a rotary evaporator and vacuum pump. 43. Purify crude product (see Support Protocol 1, step 18) to afford 5′-O-(tert-butyldimethylsilyl)-N2-isobutyryl-3′-tritylamino-2′,3′-dideoxyguanosine (S.12). 44. Perform TLC analysis (step 10) to confirm the purity of the product. Yield = 100% (43.8 g, 63.2 mmol). Rf (1:9 methanol/CH2Cl2) = 0.72. 1H NMR (CDCl3 / TMS): δ 11.90 (1H, br s, exchanges with D2O), 8.01 (1H, br s, exchanges with D2O), 7.58 (1H, s), 7.56 (6H, d, J = 7.39 Hz), 7.31 (6H, t, J = 7.58 Hz), 7.23 (3H, t, J = 7.28 Hz), 6.00 (1H, dd, J = 6.86, 4.63 Hz), 3.88 (1H, dt, J = 5.91, 3.01 Hz), 3.75 (2H, ABX, JAB = 11.25 Hz), 3.52 (1H, m), 2.57 (1H, septet, J = 6.91 Hz), 2.00- 2.10 (1H, br s, exchanges with D2O), 1.72 (1H, dt, J = 13.63, 6.89 Hz), 1.59 (1H, ddd, J = 13.73, 6.69, 4.87 Hz), 1.28 (6H, dd, J = 6.90, 3.22 Hz), 0.81 (9H, s), −0.03 (3H, s), −0.04 (3H, s). HRMS (FAB+): calcd for [M + Cs]+, 825.2561, observed 825.2540.

Synthesize O6-(N,N-diphenylcarbamoyl)-N2-isobutyryl-3′-tritylamino-2′,3′-dideoxyguanosine (S.13) 45. Dissolve 30.3 g (43.7 mmol) S.12 in 90 mL anhydrous pyridine. 46. Add 11.4 mL (65.6 mmol) N,N-diisopropylethylamine and 11.1 g (48.1 mmol) N,N-diphenylcarbamyl chloride under argon and stir for 1.5 hr at room temperature. 47. Concentrate the intensely red/purple reaction mixture in vacuo (see step 42). 48. Dissolve the residue in 600 mL CH2Cl2, extract two times with 400 mL water and once with 400 mL saturated aqueous NaCl. 49. Dry the CH2Cl2 solution over Na2SO4, vacuum filter, and concentrate in vacuo (see step 36) to afford impure 5′-O-(tert-butyldimethylsilyl)-O6-(N,N-diphenylcarbamoyl)-N2-isobutyryl-3′-tritylamino-2′,3′-dideoxyguanosine. This product is generally taken on directly to desilylation, although it can also be purified on silica.

50. Perform TLC analysis (step 10) using 6:4 (v/v) ethyl acetate/hexane. Crude yield > 100% (43.8 g). Rf (6:4 ethyl acetate/hexane) = 0.58. 1H NMR (CDCl3 /TMS): δ 8.03 (1H, s), 7.90 (1H, br s, exchanges with D2O), 7.55 (6H, d, J = 7.63 Hz), 7.24-7.50 (16H, mm), 7.21 (3H, t, J = 7.22 Hz), 6.29 (1H, t, J = 6.07 Hz), 3.89 (1H, m), 3.75 (2H, ABX, JAB = 11.25 Hz), 3.49 (1H, br m), 3.01 (1H, br m), 2.77 (1H, septet, J = 6.78 Hz), 2.00-2.10 (br s, exchanges with D2O), 1.65-1.75 (2H, m), 1.28 (6H, d, J = 6.64 Hz), 0.83 (9H, s), −0.01 (3H, s), −0.02 (3H, s). HRMS (FAB+): calcd for [M + Cs]+, 1020.3245, observed 1020.3281.

51. Dissolve ∼43.7 mmol crude 5′-O-(tert-butyldimethylsilyl)-O6-(N,N-diphenylcarbamoyl)-N2-isobutyryl-3′-tritylamino-2′,3′-dideoxyguanosine in 200 mL CH2Cl2 and 25 mL pyridine. 52. Add 49.8 mL (305.8 mmol) triethylamine trihydrofluoride, followed by a 25-mL CH2Cl2 rinse, and stir the reaction mixture for 20 hr under argon at room temperature. 53. Dilute the reaction mixture with 600 mL CH2Cl2 and extract two times with 400 mL water. 54. Back-extract the first aqueous layer with 50 mL CH2Cl2. 55. Dry the combined organics over Na2SO4, vacuum filter, and concentrate in vacuo (see step 36).



Supplement 3

56. Dissolve the residue in 100 mL CH2Cl2 and azeotrope three times with 50 mL toluene to remove traces of pyridine using a rotary evaporator and vacuum pump. 57. Purify on silica gel, using a column packed in 2% triethylamine in 7:3 (v/v) ethyl acetate/hexane and eluting with 7:3 ethyl acetate/hexane, to afford O6-(N,N-diphenylcarbamoyl)-N 2-isobutyryl-3′-tritylamino-2′,3′-dideoxyguanosine (S.13). 58. Confirm the purity of the product by TLC analysis (step 10), eluting with 6:4 (v/v) ethyl acetate/hexane. Yield = 82% (27.6 g, 35.7 mmol). Rf (6:4 ethyl acetate/hexane) = 0.20. 1H NMR (CDCl3 / TMS): δ 7.95 (1H, s), 7.84 (1H, br s, exchanges with D2O), 7.55 (6H, d, J = 7.85 Hz), 7.25-7.45 (16H, mm), 7.21 (3H, t, J = 7.25 Hz), 6.15 (1H, t, J = 6.31 Hz), 3.77-3.87 (2H, br m), 3.69 (1H, m), 3.62 (1H, m), 3.19 (1H, m), 2.80 (1H, septet, J = 6.86 Hz), 1.92-2.05 (2H, mm, 1H exchanges with D2O), 1.65 (1H, m), 1.24 (6H, d, J = 6.86 Hz). HRMS (FAB+): calcd for [M + Cs]+, 906.2380, observed 906.2350. Do not use 1 M tetra-n-butylammonium fluoride in THF to remove the TBDMS group because the O6-(N,N-diphenylcarbamoyl) group is not stable to this reagent. SUPPORT PROTOCOL 4

SYNTHESIS OF N6-BENZOYL-3′-TRITYLAMINO- 2′,3′-DIDEOXYADENOSINE The synthesis of N6-benzoyl-3′-tritylamino-2′,3′-dideoxyadenosine (S.18) from adenosine is illustrated in Figure 4.7.10. The 5′-O-(tert-butyldimethylsilyl)-2′-deoxyxyloadenosine is synthesized by slightly modified literature procedures (Wagner et al., 1974; Hansske and Robins, 1983), which is more efficient than the inversion route that is used for the pyroration of S.10. Additional Materials (also see Support Protocol 1) (−)-Adenosine Dibutyltin oxide p-Toluenesulfonyl chloride tert-Butyldimethylsilyl chloride 1 M lithium triethyl borohydride in THF Ammonium chloride Benzoyl chloride 7:10 (v/v) methanol/1,4-dioxane Pyridinium hydrochloride Diethylazodicarboxylate Argon 4:6 (v/v) ethyl acetate/hexane Tetrahydrofuran (THF) 1.0 M tetrabutylammonium fluoride (TBAF) in THF Synthesize 2′-O-(p-toluenesulfonyl)adenosine (S.14) 1. Gently reflux 25.0 g (93.5 mmol) (−)-adenosine and 23.3 g (93.5 mmol) dibutyltin oxide in 1000 mL methanol for 2 hr, until the cloudy suspension becomes clear. 2. Cool the solution to 4°C, then cautiously add 53.5 g (281 mmol) p-toluenesulfonyl chloride and 39.1 mL (281 mmol) triethylamine, keeping the reaction temperature at 4°C. 3. Stir the resulting cloudy suspension overnight at room temperature.


4. Vacuum filter the white solid, wash two times with 100 mL cold methanol, and dry using a vacuum pump to afford 2′-O-p-toluenesulfonyladenosine (S.14).

4.7.24 Supplement 3


NH2 N HO

N

N

O

NH2 N

N

1. Bu2SnO/MeOH/reflux

HO

N

N

O

1. TBDMS-Cl/C5H5N

N

2. 1M LiEt3BH/THF, 4 oC 3. NH4Cl

o

HO

2. Ts-Cl/Et3N, 4 C

OH

HO

OTs

14 (100% crude)

A NH2 N R1O

R O

NHBz N

N

N

1. BzCl/C5H5N

N

R1O

R O

N

N

N

o

2. 2 M NaOH, 4 C MeOH/dioxan (7:10 v/v) 16 R = OH (84%) R1 = TBDMS

15 R = OH (59%) R1 = TBDMS

NHBz

NHBz N TBDMSO

O

N

1. LiN3/DEAD/Ph3P/DMF 2. H2,10% Pd/C, EtOH/CH2Cl2 3. TrCl/Et3N/C5H5N

N

N HO

N 1 M TBAF/THF

NHTr 17 (67%)

O

N

N N

CH3 OCH2CH2CN P N Cl CH3 1a

NHTr

DBU/CH2Cl2

18 (94%)

Figure 4.7.10 Synthetic steps in the preparation of N6-benzoyl-3′-tritylamino-2′,3′-dideoxyadenosine (S.18) from adenosine. Bu, butyl; Ts, p-toluenesulfonyl.

Crude yield = 114% (45 g). 1H NMR (DMSO-d6): δ 8.19 (1H, s), 8.02 (1H, s), 7.37 (2H, s), 7.42 (2H, d), 7.03 (2H, d), 6.11 (1H, d), 6.03 (1H, d), 5.75 (1H, t), 5.49 (1H, dd), 4.39 (1H, ddd), 4.07 (1H, br d), 3.62 (2H, m), 2.25 (3 H, s).

Synthesize 5′-O-(tert-butyldimethylsilyl)-2′-deoxyxyloadenosine (S.15) 5. Dissolve 45 g (< 93.5 mmol) S.14 in 1000 mL pyridine, add 28.2 g (187 mmol) tert-butyldimethylsilyl chloride, and stir overnight at room temperature. 6. Quench the reaction with 100 mL methanol, concentrate on a rotary evaporator using a vacuum pump, and azeotrope three times with 25 mL toluene. 7. Dissolve the solid in 200 mL methanol, stir for 2 hr to desilylate the N6-amino group, and concentrate with a rotary evaporator and vacuum line. 8. Purify the reaction product by gravity on a silica gel column (APPENDIX 3E) using 3% to 4% methanol/CH2Cl2 to give 5′-O-(tert-butyldimethylsilyl)-2′-O-(p-toluenesulfonyl)adenosine. Yield = 75.6% (37.9 g, 70.7 mmol). TLC (1:9 methanol/CH2Cl2) Rf = 0.51.

9. Dissolve 37.9 g (70.7 mmol) 5′-O-(tert-butyldimethylsilyl)-2′-O-(p-toluenesulfonyl)adenosine in 100 mL anhydrous THF and cool to 4°C. 10. Add dropwise 283 mL (283 mmol) prechilled 1.0 M lithium triethyl borohydride in THF. 11. Stir the solution for 30 min at 4°C and then overnight at room temperature. 12. Cool the solution to 4°C, carefully quench with 11.3 g (212 mmol) NH4Cl, and concentrate on the rotary evaporator with a vacuum line. 13. Dissolve the residue first in 50 mL methanol and then 500 mL diethyl ether. 14. Wash with 100 mL saturated aqueous NaCl, and concentrate to a foam (see step 12).



Supplement 3

15. Purify on silica (2% to 4% methanol/CH2Cl2) to afford 5′-O-(tert-butyldimethylsilyl)-2′-deoxyxyloadenosine (S.15). Yield = 77.5% (20.1 g, 54.8 mmol). TLC (1:9 methanol/CH2Cl2) Rf = 0.43. 1H NMR (CDCl3 / TMS): δ 8.35 (1H, s), 7.98 (1H, s), 7.08 (1H, d, J = 9.34 Hz), 6.14 (1H, dd, J = 9.36, 2.68 Hz), 5.98 (2H, br s), 4.47 (1H, m), 4.10 (1H, m ), 3.97 (2H, m), 2.88 (1H, ddd, J = 15.47, 9.40, 6.32 Hz), 2.56 (1H, dd, J = 15.35, 2.76 Hz), 0.88 (9H, s), 0.063 (3H, s), 0.060 (3H, s). Make sure the reaction is fully quenched before concentrating the solution because lithium triethyl borohydride is pyrophoric. Other solvents, such as ethyl acetate or CH2Cl2, should not be used for the extraction as they lead to severe emulsions.

Synthesize N6-benzoyl-5′-O-(tert-butyldimethylsilyl)-2′-deoxyxyloadenosine (S.16) 16. Dissolve 5.0 g (13.6 mmol) S.15 in 25 mL pyridine, add 3.28 mL (27.4 mmol) benzoyl chloride, and stir for 2 hr at room temperature. 17. Quench the reaction with 1 mL water and concentrate on a rotary evaporator with vacuum pump. 18. Dissolve the residue in 80 mL of 7:10 (v/v) methanol/1,4-dioxane, cool to 4 °C. 19. Add 34 mL (68 mmol) prechilled 2.0 M aqueous NaOH and stir for 5 min to selectively remove the 3′-benzoyl group. Monitor the hydrolysis carefully and neutralize the hydroxide as soon as possible in order to avoid loss of the N6-benzoyl group.

20. Neutralize the solution to pH 7 with 4.0 g (35 mmol) pyridinium hydrochloride and concentrate on a rotary evaporator with vacuum pump. 21. Dissolve the residue in 100 mL CH2Cl2, extract two times with 50 mL saturated aqueous NaHCO3 and two times with 50 mL saturated aqueous NaCl, and concentrate in vacuo (see step 7). 22. Purify on silica (2% methanol/CH2Cl2) to give N6-benzoyl-5′-O-(tert-butyldimethylsilyl)-2′-deoxyxyloadenosine (S.16). Yield = 84.6% (5.4 g, 11.5 mmol). TLC (5:95 methanol/CH2Cl2) Rf = 0.38. 1H NMR (CDCl3 / TMS): δ 9.00 (1H, br s), 8.82 (1H, s), 8.32 (1H, s), 8.02 (2H, d, J = 7.26 Hz), 7.63 (1H, t, J = 7.41 Hz), 7.54 (2H, t, J = 7.56 Hz), 6.31 (1H, dd, J = 9.06, 2.44 Hz), 5.95 (1H, d, J = 7.63 Hz), 4.55 (1H, m), 4.11 (1H, pseudo q, J = 6.21 Hz), 4.02 (2H, dd, J = 8.08, 2.68 Hz), 2.90 (1H, ddd, J = 15.22, 9.12, 5.93 Hz), 2.58 (1H, dd, J = 15.33, 2.47 Hz), 0.88 (9H, s), 0.07 (3H, s), 0.06 (3H, s).

Synthesize N6-benzoyl-5′-O-(tert-butyldimethylsilyl)-3′-tritylamino-2′,3′-dideoxyadenosine (S.17) 23. Dissolve 17.6 g (37.5 mmol) S.16 in 375 mL anhydrous DMF and add 5.5 g (113.0 mmol) LiN3 and 14.8 g (56.3 mmol) triphenylphosphine. 24. Add 8.9 mL (56.3 mmol) diethylazodicarboxylate and stir the reaction mixture for 6 hr under argon at room temperature. 25. Quench the reaction with 10 mL water and concentrate in a rotary evaporator with a vacuum pump. Synthesis and Purification of Oligonucleotide N3′→P5′ Phosphoramidates

26. Dissolve the residue in 500 mL ethyl acetate, and wash three times with 300 mL water and once with 300 mL saturated aqueous NaCl.

4.7.26 Supplement 3


27. Dry the ethyl acetate solution over Na2SO4, filter, and concentrate with a rotary evaporator and vacuum line. This crude (triphenylphosphine oxide–contaminated) 3′-azido-N6-benzoyl-5′-O-(tertbutyldimethylsilyl)-2′,3′-dideoxyadenosine (18.8 g) is taken on directly to hydrogenation and purified as the 3′-amine.

28. Dissolve 18.8 g of the crude azide in 250 mL of 1:1 (v/v) ethanol/CH 2Cl2 and reduce by hydrogenation (60 psi H2) in the presence of 1.0 g of 10% Pd/C catalyst for 16 hr at room temperature. 29. Remove the catalyst by vacuum filtration, and evaporate the solvent in a rotary evaporator with vacuum line to afford the crude 3′-amine. 30. Purify on silica (preequilibrate with 2% methanol/CH2Cl2 and elute with 2% to 6% methanol/CH2Cl2 and then 1% triethylamine/6% methanol/CH2Cl2) to afford 3′amino-N6-benzoyl-5′-O-(tert-butyldimethylsilyl)-2′,3′-dideoxyadenosine as an offwhite foam. 31. Perform TLC analysis (APPENDIX 3D) on 0.2-mm-thick precoated Merck silica gel 60 F254 plates to confirm the purity of the product. Elute with 8:92 (v/v) methanol/CH2Cl2. Yield = 68% (12.0 g, 25.6 mmol). Rf (8:92 methanol/CH2Cl2) = 0.30. 1H NMR (CDCl3 / TMS): δ 8.95 (1H, br s, exchanges with D2O), 8.81 (1H, s), 8.40 (1H, s), 8.02 (2H, d, J = 7.23 Hz), 7.62 (1H, t, J = 7.43 Hz), 7.54 (2H, t, J = 7.48), 6.49 (1H, dd, J = 6.81, 3.68 Hz), 3.80-3.98 (4H, mm), 2.76 (1H, ddd, J = 13.27, 6.42, 3.69), 2.39 (1H, dt, J = 13.43, 6.93), 0.92 (9H, s), 0.11 (3H, s), 0.00 (3H, s). HRMS (FAB+): calcd for [M + Cs]+, 601.1360, observed 601.1373.

32. Protect 29.5 g (63.0 mmol) of 3′-amino-N6-benzoyl-5′-O-(tert-butyldimethylsilyl)2′,3′-dideoxyadenosine by reacting with 12.9 mL (94.5 mmol) triethylamine and 21.1 g (75.6 mmol) trityl chloride in 350 mL CH2Cl2 for 16 hr at room temperature. 33. Dilute the reaction mixture with an additional 150 mL CH2Cl2. 34. Extract once with 400 mL water, three times with 300 mL saturated aqueous NaHCO3, and two times with 300 mL saturated aqueous NaCl. 35. Concentrate to a glassy foam with a rotary evaporator and vacuum line. 36. Purify the crude product on a silica gel column preequilibrated with 1% triethylamine in 4:6 (v/v) ethyl acetate/hexane, and elute with 4:6 ethyl acetate/hexane to afford N6-benzoyl-5′-O-(tert-butyldimethylsilyl)-3′-tritylamino-2′,3′-dideoxyadenosine (S.17). 37. Perform TLC analysis (step 33) using 5:95 (v/v) methanol/CH2Cl2 as the eluent. Yield = 98% (44.1 g, 62.1 mmol). Rf (5:95 methanol/CH2Cl2) = 0.48. 1H NMR (CDCl3 / TMS): δ 8.99 (1H, br s, exchanges with D2O), 8.76 (1H, s), 8.12 (1H, s), 8.00 (2H, d, J = 7.29 Hz), 7.60 (1H, t, J = 7.41 Hz), 7.54 (6H, d, J = 7.44 Hz), 7.51 (2H, t, J = 7.28 Hz), 7.28 (6H, t, J = 7.56 Hz), 7.20 (2H, t, J = 7.22 Hz), 6.36 (1H, t, J = 5.96 Hz), 3.90 (1H, m), 3.82 (1H, dd, J = 11.28, 2.74 Hz), 3.67 (1H, dd, J = 11.26, 3.10 Hz), 3.50 (1H, br m), 2.00-2.10 (1H, br s, exchanges with D2O), 1.68-1.83 (2H, mm), 0.82 (9H, s), −0.02 (3H, s), −0.03 (3H, s). HRMS (FAB+): calcd for [M + Cs]+, 843.2455, observed 843.2477.

Synthesize N6-benzoyl-3′-tritylamino-2′,3′-dideoxyadenosine (S.18) 38. Remove the 5′-TBDMS protecting group by dissolving 43.7 g (61.5 mmol) S.17 in 123.0 mL THF and reacting with 123.0 mL (123.0 mmol) of 1.0 M TBAF in THF for 24 hr, room temperature.



Supplement 3

39. Concentrate the reaction mixture with vacuum line and dissolve the residue in 400 mL ethyl acetate. 40. Extract three times with 250 mL water and two times with 250 mL saturated aqueous NaCl. 41. Dry the organic layer over Na2SO4, vacuum filter, and remove the solvent with a rotary evaporator and vacuum line. 42. Purify the crude product on a silica gel column preequilibrated with 2% triethylamine in 8:2 (v/v) ethyl acetate/hexane, and elute with 8:2 ethyl acetate/hexane to 100% ethyl acetate to afford N6-benzoyl-3′-tritylamino-2′,3′-dideoxyadenosine (S.18). 43. Perform TLC analysis (step 33) using 5:95 (v/v) methanol/CH2Cl2 as the eluent. Yield = 94% (34.5 g, 57.9 mmol). Rf (5:95 methanol/CH2Cl2) = 0.40. 1H NMR (CDCl3 / TMS): δ 9.06 (1H, br s, exchanges with D2O), 8.68 (1H, s), 8.02 (1H, s), 8.01 (2H, d, J = 7.33 Hz), 7.60 (1H, t, J = 7.47 Hz), 7.53 (6H, d, J = 7.38 Hz), 7.51 (2H, t, J = 7.20 Hz), 7.29 (6H, t, J = 7.58 Hz), 7.21 (3H, t, J = 7.25 Hz), 6.24 (1H, dd, J = 7.56, 6.30 Hz), 4.85 (1H, dd, J = 9.87, 3.19 Hz, exchanges with D2O), 3 65-3.82 (3H, mm), 3.37 (1H, t, J = 10.16 Hz), 2.38 (1H, dt, J = 13.46, 7.02 Hz), 2.00-2.20 (1H, br s, exchanges with D2O), 1.75 (1H, ddd, J = 13.28, 5.96, 2.97 Hz). HRMS (FAB+): calcd for [M + Na]+, 619.2434, observed 619.2421. SUPPORT PROTOCOL 5

SYNTHESIS OF 3′-O-(4,4′-DIMETHOXYTRITYL)-PROTECTED DEOXYRIBONUCLEOSIDES An approach to the synthesis of 3′-O-(4,4′-dimethoxytrityl)-protected deoxyribonucleosides (S.20) is presented in Figure 4.7.11. These nucleosides are necessary for the synthesis of the phosphodiester or phosphorothioate portion of chimeric oligonucleotides. Additional Materials (also see Support Protocol 1) N6-Benzoyl-2′-deoxyadenosine N4-Benzoyl-2′-deoxycytidine N2-Isobutyryl-2′-deoxyguanosine Thymidine 4-Dimethylaminopyridine tert-Butyldimethylsilyl chloride Tetrahydrofuran (THF) 1 M tetrabutylammonium fluoride (TBAF) in THF N,N-Diisopropylethylamine

HO

O

1. TBDMS-Cl/DMAP/Et3N/C5H5N 2. DMTr-Cl/C5H5N

B

3. DPC-Cl/i-Pr2NEt/C5H5N iBu (only when B = G )

OH Bz

TBDMSO

Bz

B

DMTrO

iBu

B=A ,C ,G

O

, or T

19

CH3 OCH2CH2CN P N Cl HO

TBAF/THF (when B = ABz, CBz, T)

O

iBu,DPC

Et3N•3HF/C5H5N/CH2Cl2 (when B = G


)

B

CH3 2 DBU/CH2Cl2

DMTrO 20

Figure 4.7.11 General approach to the synthesis of 3′-O-(4,4′-dimethoxytrityl)-2′-deoxyribonucleosides (S.20) from N-protected 2′-deoxyribonucleosides.

4.7.28 Supplement 3


N,N-Diphenylcarbamyl chloride 9:1 (v/v) CH2Cl2/pyridine Triethylamine trihydrofluoride 50% to 70% ethyl acetate/hexane Prepare 5′-O-(tert-butyldimethylsilyl)-2′-deoxyribonucleosides (S.19) 1. Azeotrope the N-protected 2′-deoxyribonucleoside (dABz, dCBz, dGi-Bu, or T) two times from 10 mL/g pyridine and suspend in pyridine at 10 mL/g. 2. To this stirring mixture, add sequentially 0.1 eq 4-dimethylaminopyridine, 1.2 eq triethylamine, and 1.05 to 1.2 eq tert-butyldimethylsilyl chloride. Stir for 8 to 24 hr at room temperature. 3. Remove the pyridine using a rotary evaporator and vacuum pump. 4. Dissolve the residue in 15 mL/g CH2Cl2 and extract two times with 10 mL/g water and one time with 10 mL/g saturated aqueous NaCl. 5. Dry the organic solution over anhydrous Na2SO4, vacuum filter, and concentrate under reduced pressure to a solid that is used in the next reaction without further purification. Prepare 5′-O-(tert-butyldimethylsilyl)-3′-O-(4,4′-dimethoxytrityl)-2′-deoxyribonucleosides 6. Azeotrope the 5′-O-(tert-butyldimethylsilyl)-protected 2′-deoxyribonucleoside two times from 10 mL/g pyridine and dissolve in pyridine at 10 mL/g. 7. While stirring this solution, add 1.2 to 1.3 eq 4,4′-dimethoxytrityl chloride. Stir the solution for 16 to 24 hr at room temperature. 8. Concentrate on a rotary evaporator using a vacuum pump. 9. Dissolve the residue in 15 mL/g CH2Cl2 and extract once each with 10 mL/g water, 10 mL/g saturated aqueous NaHCO3, and 10 mL/g saturated aqueous NaCl. 10. Dry the CH2Cl2 solution over Na2SO4, vacuum filter, and concentrate under reduced pressure to a foam. The product can be used directly in the next reaction (T) or purified on silica (dABz, dCBz, and dGi-Bu) using a gradient of 1% to 5% methanol in CH2Cl2.

Prepare 3′-O-(4,4′-dimethoxytrityl)-2′-deoxyribonucleosides (S.20) For dABz, dCBz, and T: 11a. Remove the 5′-TBDMS protecting group by dissolving the 5′-(tert-butyldimethylsilyl)-3′-O-(4,4′-dimethoxytrityl)-2′-deoxyribonucleoside in THF at 3 mL/g and reacting it with 1 M (2.0 eq) TBAF in THF for 16 to 24 hr, room temperature. 12a. Concentrate the solution under reduced pressure. 13a. Dissolve the residue in 15 mL/g CH2Cl2, and extract two times with 10 mL/g water and one time with 10 mL/g saturated aqueous NaCl. 14a. Dry the organic layer over Na2SO4, vacuum filter, and evaporate on a rotary evaporator with pump. 15a. Purify the crude 3′-O-(4,4′-dimethoxytrityl)-2′-deoxyribonucleosides (S.20) (see Support Protocol 1, step 18).



Supplement 3

16a. Perform TLC analysis (APPENDIX 3D) on 0.2-mm-thick precoated Merck silica gel 60 F254 plates to confirm the purity of the product. Elute with 1:9 (v/v) methanol/CH2Cl2. N6-Benzoyl-3′-O-(4,4′-dimethoxytrityl)-2′-deoxyadenosine (S.20). Yield from dABz = 56.7% (55.4 g). Rf (1:9 methanol/CH2Cl2) = 0.68. 1H NMR (CDCl3/TMS): δ 9.14 (1H, br s, exchanges with D2O), 8.72 (1H, s), 8.06 (1H, s), 8.02 (2H, d, J = 7.43 Hz), 7.62 (1H, t, J = 7.33 Hz), 7.53 (2H, d, J = 7.74 Hz), 7.50 (2H, d, J = 7.54 Hz), 7.39 (4H, d, J = 8.81 Hz), 7.34 (2H, t, J = 7.54 Hz), 7.26 (2H, t, J = 7.95 Hz), 6.87 (4H, dd, J = 8.91, 2.43 Hz), 6.37 (1H, dd, J = 9.95, 5.26 Hz), 5.79 (1H, br d, J = 10.38 Hz, exchanges with D2O), 4.66 (1H, d, J = 5.32 Hz), 4.08 (1H, s), 3.81 (6H, s), 3.76 (1H, d, J = 12.78 Hz), 3.35 (1H, t, J = 11.86 Hz), 2.73 (1H, ddd, J = 13.21, 10.10, 7.99 Hz), 1.76 (1H, dd, J = 13.31, 5.30 Hz). HRMS (FAB+): calcd for [M + H] +, 658.2666; found, 658.2666. N4-Benzoyl-3′-O-(4,4′-dimethoxytrityl)-2′-deoxycytidine (S.20). Yield from dCBz = 74.7% (70.0 g) including additional mixed fractions that were purified further by precipitation from CH2Cl2 into a 20× volume of 3:1 hexane/diethyl ether over 1.5 hr. Rf (1:9 methanol/CH2Cl2) = 0.66. 1H NMR (CDCl3/TMS): δ 8.66 (1H, br s, exchanges with D2O), 8.09 (1H, d, J = 7.35 Hz), 7.87 (2H, d, J = 7.43 Hz), 7.62 (1H, t, J = 7.36 Hz), 7.53 (2H, d, J = 7.80 Hz), 7.48 (2H, d, J = 7.63 Hz), 7.37 (4H, d, J = 8.86 Hz), 7.32 (2H, t, J = 7.53 Hz), 7.25 (1H, t, J = 7.17 Hz), 6.85 (4H, d, J = 8.76 Hz), 6.25 (1H, dd, J = 7.63, 6.12 Hz), 4.36-4.43 (1H, br m), 3.94 (1H, d, J = 2.19 Hz), 3.81 (6H, s), 3.66 (1H, br d, J = 11.86 Hz), 3.26 (1H, br d, J = 11.90 Hz), 2.48 (1H, br s, exchanges with D2O), 2.22 (1H, dd, J = 13.13, 5.20 Hz), 2.08 (1H, quintet, J = 6.94 Hz). HRMS (FAB+): calcd for [M +Na]+, 656.2373; found, 656.2383. 3′-O-(4,4′-Dimethoxytrityl)-thymidine (S.20). Yield from T = 81.8% (45.2 g). Rf (1:9 methanol/CH2Cl2) = 0.56. 1H NMR (CDCl3 /TMS): δ 8.61 (1H, br s, exchanges with D2O), 7.46 (2H, d, J = 7.47 Hz), 7.36 (4H, d, J = 8.83 Hz), 7.32 (2H, t, J = 7.94 Hz), 7.25 (1H, t, J = 7.43 Hz), 6.86 (4H, d, J = 7.39 Hz), 6.15 (1H, dd, J = 8.87, 5.76 Hz), 4.38 (1H, d, J = 6.20 Hz), 3.99 (1H, d, J = 2.13 Hz), 3.81 (6H, s), 3.68 (1H, br d, J = 11.79 Hz), 3.30-3.37 (1H, br m), 2.47-2.55 (1H, br m, exchanges with D2O), 1.95 (1H, ddd, J = 13.98, 8.42, 6.00 Hz), 1.87 (3H, s), 1.67-1.74 (1H, m). HRMS (FAB+): calcd for [M + Na] +, 567.2107; found, 567.2111.

For dGi-Bu,DPC: 11b. Prepare a stirring solution of 101.9 g (135.2 mmol) 5′-O-(tert-butyldimethylsilyl)3′-O-(4,4′-dimethoxytrityl)-N2-isobutyryl-2′-deoxyguanosine in 300 mL pyridine, add 26.6 g (206.1 mmol) N,N-diisopropylethylamine and 41.4 g (178.7 mmol) N,N-diphenylcarbamyl chloride. 12b. Stir the dark solution for 2 hr and then concentrate on a rotary evaporator with a vacuum pump. 13b. Dissolve the residue in 600 mL CH2Cl2 and extract two times with 250 mL water and once with 250 mL saturated aqueous NaCl. 14b. Dry the organic solution over Na2SO4, vacuum filter, and concentrate to a purplecolored foam with a vacuum line on a rotary evaporator. 15b. Dissolve the crude nucleoside in 800 mL of a 9:1 (v/v) CH2Cl2/pyridine solution, then add 155.0 g (961.5 mmol) triethylamine trihydrofluoride and react for 16 hr at room temperature.


16b. Remove the solvents with a rotary evaporator and vacuum pump, dissolve the residue in 600 mL CH2Cl2, and wash two times with 250 mL water and 250 mL saturated aqueous NaCl. 17b. Dry the organic solution over Na2SO4, vacuum filter, and concentrate to a dark red foam with a vacuum line on a rotary evaporator.

4.7.30 Supplement 3


18b. Purify the crude product by gravity on a silica gel column (APPENDIX 3E) preequilibrated with 1% Et3N/50% ethyl acetate/hexane and elute with 50% to 70% ethyl acetate/hexane to afford the protected 2′-deoxyguanosine (S.20). 19b. Perform TLC analysis (step 16a), eluting with 75:25 (v/v) ethyl acetate/hexane. Yield from dGi-Bu = 35.6% (43.3 g). Rf (75:25 ethyl acetate/hexane) = 0.38. 1H NMR (CDCl3/TMS): δ 8.00 (1H, s), 7.91 (1H, br s, exchanges with D2O), 7.49 (2H, d, J = 7.75 Hz), 7.20-7.45 (19H, mm with 4H, d, J = 8.93 Hz at 7.38), 6.86 (4H, dd, J = 8.82, 2.09 Hz), 6.26 (1H, dd, J = 9.80, 5.12 Hz), 4.65 (1H, d, J = 5.25 Hz), 4.35 (1H, dd, J = 10.35, 3.26 Hz, exchanges with D2O), 4.04 (1H, s), 3.80 (6H, s), 3.73 (1H, br d, J = 11.48 Hz), 3.39 (1H, br t, J = 11.54 Hz), 2.64-2.80 (2H, m), 1.68 (1H, dd, J = 13.18, 5.16 Hz), 1.24 (6H, d, J = 6.93 Hz). HRMS (FAB+): calcd for [M + Na]+, 857.3275; found, 857.3270. Do not use 1 M tetra-n-butylammonium fluoride in THF to remove the TBDMS group because the O6-(N,N-diphenylcarbamoyl) group is not stable to this reagent.

SYNTHESIS OF 3′-AMINONUCLEOSIDE-CONTAINING SOLID SUPPORT

SUPPORT PROTOCOL 6

Additional Materials (also see Support Protocol 1) 3′-Tritylamino-2′,3′-dideoxynucleosides (S.5, S.9, S.13, or S.18; see Support Protocols 1 to 4) 4-Dimethylaminopyridine Succinic anhydride 10% (v/v) aqueous citric acid, cold 1-Hydroxybenzotriazole 1:1 (v/v) 1-methyl-2-pyrrolidinone (anhydrous)/dimethyl sulfoxide (anhydrous) N,N-Diisopropylethylamine 2-(1H-Benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate Aminopropyl-conjugated controlled-pore glass (aminopropyl-CPG; Sigma) 1:1:8 (v/v/v) acetic anhydride/2,6-lutidine/THF 16.5% (v/v) N-methylimidazole in THF (see recipe) Mechanical shaker Synthesize 3′-tritylamino-2′,3′-dideoxynucleoside-5′-O-hemisuccinates 1. To a solution of 1.5 mmol of 3′-tritylamino-2′,3′-dideoxynucleoside (S.5, S.9, S.13, or S.18;) in 5 mL CH2Cl2, add 0.22 g (1.8 mmol) 4-dimethylaminopyridine and 0.18 g (1.8 mmol) succinic anhydride. Stir for 1 hr at room temperature. 2. Quench the reaction with 0.6 mL methanol and dilute with 30 mL CH2Cl2. 3. Extract once each with 20 mL cold 10% aqueous citric acid, 20 mL water, and 20 mL saturated aqueous NaCl. 4. Dry the organic layer over Na2SO4, vacuum filter, and concentrate the product (S.21; Figure 4.7.12) to a foam on a rotary evaporator using a vacuum line. N6-Benzoyl-3′-tritylamino-2′,3′-dideoxyadenosine-5′-O-hemisuccinate: 100% yield (1.15 g). N4-Benzoyl-3′-tritylamino-2′,3′-dideoxycytidine-5′-O-hemisuccinate: 76% yield (0.77 g). 3′-Tritylamino-2′,3′-deoxythymidine-5′-O-hemisuccinate: 94% yield (0.82 g). O6-(N,N-Diphenylcarbamoyl)-N2-isobutyryl-3′-tritylamino-2′,3′-dideoxyguanosine5′-O-hemisuccinate: 78% yield (1.02 g).



Supplement 9

O O

HO

O

B

O NHTr 21a 21c 21g 21t

Bz

B=A Bz B=C iBu,DPC B=G B= T

Figure 4.7.12 Structures of N-protected-3′-tritylamino-2′,3′-dideoxyribonucleoside-5′-O-hemisuccinates.

Conjugate 3′-tritylamino-2′,3′-dideoxynucleoside-5′-O-hemisuccinates to CPG 5. To a solution of 1 mmol 3′-tritylamino-2′,3′-dideoxynucleoside-5′-O-hemisuccinate and 0.13 g (0.95 mmol) of 1-hydroxybenzotriazole in 10 mL of 1:1 (v/v) 1-methyl2-pyrrolidinone/DMSO, add 0.35 mL (2.0 mmol) N,N-diisopropylethylamine and 0.36 g (0.95 mmol) 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate. 6. Stir the solution 5 min, add 10.0 g aminopropyl-CPG, and put on a shaker for 6 hr, room temperature. 7. Vacuum filter the CPG and wash successively with 20 mL each of DMF, methanol, and ethyl ether. 8. Prepare a 1:1 (v/v) mixture of 1:1:8 acetic anhydride/2,6-lutidine/THF and 16.5% N-methylimidazole/THF, add this mixture to the support on the funnel, and let stand for 30 min to cap any unreacted amino groups on the CPG. 9. Vacuum filter the CPG and wash successively with 20 mL each of acetonitrile, methanol, and diethyl ether. 10. Dry the CPG using a vacuum pump. The nucleoside loadings, determined by trityl assay at 432 nm in 20% TFA/CHCl3 using a molar extinction coefficient of 40.7 ìmol−1 cm−1, were 38.6 ìmol/g for A, 33.6 ìmol/g for C, 29.0 ìmol/g for T, and 39.0 ìmol/g for G. For larger scales, use an overhead stirrer instead of a shaker. Do not use a magnetic stir bar because the CPG will be crushed. SUPPORT PROTOCOL 7


PHOSPHORAMIDITE SYNTHESIS The preparation of nucleoside 5′-O-cis-(2,6-dimethylpiperidinyl)-2-cyanoethylphosphoramidite monomers is described below. While either cis-2,6-dimethylpiperidino phosphoramidites or N,N-diisopropylamino phosphoramidites can be prepared and used for the preparation of pnODNs, the former are preferred because they allow the use of significantly lower equivalents per coupling (Fearon et al., 1998). Additional Materials (also see Support Protocols 1 to 5) 3′-Tritylamino-2′,3′-dideoxynucleosides (see Support Protocols 1 to 4) or 3′-O-(4,4′-dimethoxytrityl)-2′-deoxyribonucleosides (see Support Protocol 5) Phosphorus trichloride 3-Hydroxypropionitrile 10% (w/v) aqueous KOH 1:4 (v/v) toluene/hexane

4.7.32 Supplement 9


cis-2,6-Dimethylpiperidine 1,8-Diazabicyclo[5.4.0]undec-7-ene (DBU), distilled from CaH2 before use Synthesize 2-cyanoethylphosphorodichloridite 1. To a solution of 500 mL (5.73 mol) phosphorus trichloride in 250 mL acetonitrile, add dropwise at room temperature, with stirring and bubbling argon, a solution of 47 mL (0.69 mol) 3-hydroxypropionitrile in 250 mL acetonitrile. 2. Stir the solution 15 min at room temperature, with absorption of evolving HCl into a solution of 10% aqueous KOH. 3. Concentrate on a rotary evaporator using a pump and vacuum filter, under argon, into a distillation flask. 4. Distill the product under reduced pressure. The 2-cyanoethylphosphorodichloridite distills as a colorless liquid at 78° to 80°C at 1.0 mmHg. Yield = 75.7% (88.5 g). 31P NMR (CDCl3): δ 180.3.

Synthesize 2-cyanoethyl-cis-(2,6-dimethylpiperidinyl)chlorophosphoramidite 5. To a solution of 35.0 g (203.6 mmol) 2-cyanoethylphosphorodichloridite in 300 mL of 1:4 toluene/hexane, add 55 mL (408.1 mmol) cis-2,6-dimethylpiperidine dropwise, with stirring at 4°C. 6. Stir the reaction for 2 hr at room temperature. 7. Vacuum filter and wash the solid with 40 mL of 1:4 toluene/hexane under argon. 8. Concentrate the filtrate using a rotary evaporator and vacuum pump. 9. To the resultant oil, add 5 mL CH2Cl2 and 300 mL hexane, and crystallize the product overnight at 4°C. 10. Filter the 2-cyanoethyl-cis-(2,6-dimethylpiperidinyl)chlorophosphoramidite under argon. 11. Crush with a spatula, wash with 100 mL of 100:3 (v/v) hexane/CH2Cl2, and dry using a rotary evaporator and vacuum pump. 12. Concentrate the mother liquor (step 3) and recrystallize again (step 9) to obtain a second crop of pale yellow 2-cyanoethyl-cis-(2,6-dimethylpiperidinyl)chlorophosphoramidite. Combined yield = 76.5% (38.8 g). 31P NMR (CDCl3): δ 172.7.

Synthesize nucleoside 5′-O-[cis-(2,6-dimethylpiperidinyl)(2-cyanoethyl)]phosphoramidite monomers (S.1 and S.2) 13. Azeotrope 10.0 mmol of 3′-tritylamino-2′,3′-dideoxynucleoside (S.5, S.9, S.13, or S.18) or 3′-O-(4,4′-dimethoxytrityl)-2′-deoxyribonucleoside (S.20) two times from 50 mL CH3CN. 14. Dissolve azeotroped nucleoside in 30 mL CH2Cl2 and then add 3.0 mL (20.0 mmol) DBU. 15. With stirring, add a solution of 3.0 g (12.0 mmol) 2-cyanoethyl-cis-(2,6-dimethylpiperidinyl)chlorophosphoramidite in 8 mL CH2Cl2, under an argon atmosphere. Stir the reaction mixture for 15 min at ambient temperature. 16. Check the reaction by TLC.



Supplement 3

In order to obtain accurate TLC of the product, pre-wet the TLC plate by immersing it in 10% triethylamine/CH2Cl2, quickly let it dry, then immediately spot the sample, and elute with 5:70:25 Et3N/ethyl acetate/hexane.

17. To avoid decomposition, desalt the crude reaction by loading the mixture directly onto a silica gel column (APPENDIX 3E) preequilibrated in 5% triethylamine/CH2Cl2 and quickly elute it in the same solvent system. These phosphoramidites are not stable to an aqueous workup after the reaction.

18. Remove solvents under reduced pressure and purify the crude product on a silica gel column preequilibrated with 0.5% to 5% triethylamine in 2% methanol/CH2Cl2. Elute as indicated below. N6-Benzoyl-3′-tritylamino-5′-O-[(cis-2,6-dimethylpiperidino)(2-cyanoethoxy)]phosphinyl-2′,3′-dideoxyadenosine (S.1). Purify on silica (60% to 70% ethyl acetate/hexane containing 3% triethylamine). Yield = 83.1% (6.72 g). 31P NMR (CD3CN): δ 148.82, 149.16. Rf = 0.45. N4-Benzoyl-3′-tritylamino-5′-O-[(cis-2,6-dimethylpiperidino)(2-cyanoethoxy)]phosphinyl2′,3′-dideoxycytidine (S.1). Purify on silica (70% ethyl acetate/hexane containing 3% triethylamine). Yield = 82.7% (6.50 g). 31P NMR (CD3CN): δ 149.31, 149.68. Rf = 0.43. O6-(N,N-Diphenylcarbamoyl)-N2-isobutyryl-3′-tritylamino-5′-O-[(cis-2,6-dimethylpiperidino)(2-cyanoethoxy)]phosphinyl-2′,3′-dideoxyguanosine (S.1). Purify compound on silica (60% ethyl acetate/hexane containing 3% triethylamine). Yield = 76.9% (7.58 g). 31P NMR (CD3CN): δ 148.93, 149.50. Rf = 0.57. 3′-Tritylamino-5′-O-[(cis-2,6-dimethylpiperidino)(2-cyanoethoxy)]phosphinyl-2′,3′-dideoxythymidine (S.1). Purify compound on silica (50% ethyl acetate/hexane containing 3% triethylamine). Yield = 79.4% (5.52 g). 31P NMR (CD3CN): δ 149.13, 149.49. Rf = 0.60. N6-Benzoyl-3′-O-(4,4′-dimethoxytrityl)-5′-O-[(cis-2,6-dimethylpiperidino)(2-cyanoethoxy)]phosphinyl-2′-deoxyadenosine (S.2). Purify compound on silica (60% to 70% ethyl acetate/hexane containing 3% triethylamine). Yield= 76.7% (6.67 g). 31P NMR (CD3CN): δ 149.26, 149.39. Rf = 0.47. N4-Benzoyl-3′-O-(4,4′-dimethoxytrityl)-5′-O-[(cis-2,6-dimethylpiperidino)(2-cyanoethoxy)] phosphinyl-2′-deoxycytidine (S.2). Purify compound on silica (60% to 75% ethyl acetate/hexane containing 3% triethylamine). Yield = 74.4% (6.29 g). 31P NMR (CD3CN): δ 149.37, 149.76. Rf = 0.43. O6-(N,N-diphenylcarbamoyl)-N2-isobutyryl-3′-O-(4,4′-Dimethoxytrityl)-5′-O-[(cis2,6-dimethylpiperidino)(2-cyanoethoxy)]phosphinyl-2′-deoxyguanosine (S.2). Purify compound on silica (50% ethyl acetate/hexane containing 3% triethylamine). Yield = 71.0% (7.43 g). 31P NMR (CD3CN): δ 149.32, 149.51. Rf = 0.60. 3′-O-(4,4′-Dimethoxytrityl)-5′-O-[(cis-2,6-dimethylpiperidino)(2-cyanoethoxy)]phosphinyl2′-deoxythymidine (S.2). Purify compound on silica (60% ethyl acetate/hexane containing 3% triethylamine). Yield = 74.1% (5.61 g). 31P NMR (CD3CN): δ 149.24, 149.65. Rf = 0.60. Synthesis and Purification of Oligonucleotide N3′→P5′ Phosphoramidates

4.7.34 Supplement 3


REAGENTS AND SOLUTIONS Use deionized, distilled water in all recipes and protocol steps. The acetonitrile used for all formulations must contain ≤0.001% water. Oven-dry all bottles and syringes. For common stock solutions, see APPENDIX 2A; for suppliers, see SUPPLIERS APPENDIX.

Buffer D: 0.1 M TEAB/2% acetonitrile, pH 8 Dilute 100 mL of 1 M TEAB buffer, pH 8 (see recipe), in 880 mL water and add 20 mL acetonitrile. Check that the pH is 8.0 and correct with triethylamine or dry ice, if necessary. Filter through a 0.2-µm filter before use on the HPLC. Store up to 6 months at 4°C. Dichloroacetic acid in CH2Cl2, 3% (v/v) Dissolve 12 mL of dichloroacetic acid in 388 mL of CH2Cl2. Store up to 1 week at room temperature. This solution should only be kept for about 1 week due to its potential to generate HCl, which is extremely detrimental to the pnODN synthesis. Also, only use bottles of DCA 40 couplings.

5. Dissolve AEEA linker in 2.4 mL NMP diluent (final 209 mM). Perform PNA synthesis 6. Load the above reagents as well as base solution, amino acids, deblocking solution, and capping solution onto an Expedite 8909 synthesizer with an FMOC-XAL-PEGPS column. Monomers, activator, and linker should not remain on the machine for >2 weeks. Amino acids should not be left on the machine for >2 days due to their tendency to crystallize. Amino acids are the least expensive reagent and it is cost effective to replace these solutions frequently rather than risk failed syntheses or damage to the instrument.

7. Perform synthesis according to manufacturer’s programs and specifications, stopping before removal of the final FMOC group. It is critical that the final FMOC be left on the PNA until a decision has been made to cleave the PNA from the resin or add another group. If synthesis is complete and the PNA is to be cleaved from the resin, proceed to step 9.

8. Optional: Perform any desired manual additions (e.g., see Support Protocols 1 and 2). 9. Implement the final deblock option in the prime menu of the Expedite PNA software. Cycle the final deblock procedure one time to complete the double deblock procedure. Wash and prepare PNA-bound resin 10. Remove the column from the synthesizer and wash four times with 10 mL DMF, reversing the direction of flow through the column each time. Synthesis and Purification of Peptide Nucleic Acids

11. Wash the column four times with 10 mL isopropyl alcohol, reversing the direction of flow through the column each time to facilitate drying of the resin.

4.11.4 Supplement 8


12. Dry the resin in vacuo for a minimum of 30 min or by blowing filtered house air across the column for 3 to 5 min, reversing the ends frequently. Resin is sufficiently dry when the resin plug slides easily from end to end as the air input is reversed.

Cleave PNA from solid-phase resin 13. Transfer the dried, resin-bound, deprotected PNA to a 1.5-mL, 0.2-µm PTFE or regenerated cellulose spin column. 14. Add 250 µL cleavage cocktail and incubate 90 min at room temperature. Although cleavage of PNAs from the FMOC-XAL-PEG-PS resin occurs in 5 min, cleavage of the protecting groups on the PNA side-chains requires at least 90 min. Complete cleavage of the BHOC side chains can be verified by MALDI-TOF mass spectrometry. Incomplete cleavage will result in a mass that exceeds the expected value by increments of the BHOC mass (100.1). Also, in the authors’ experience, products that retain one or more protecting groups will have longer retention times by RP-HPLC (the protecting groups are hydrophobic). Uniform cleavage is dependent on the freshness of the TFA, while the m-cresol serves as a molecular scavenger. The quality of the TFA can be checked by observing the color of the solution. It should be clear to slightly yellow. If the color progresses to a more orangebrown color, it should not be used for cleavage, but may still be used in preparing RP-HPLC buffer A (see Basic Protocol 3). In addition to solution color, TFA should release a small amount of fume (smoke) when the container is opened. If this is not observed, it should not be used for any procedure involving PNAs. CAUTION: TFA is caustic and should be dispensed using only glass pipets or pipet tips containing a charcoal filter (Intermountain Scientific). The charcoal filter serves as a barrier that protects the pipettor seals from the TFA fumes.

15. Centrifuge 2 min at 1300 × g for a PTFE filter or 2 min at 8400 × g for a regenerated cellulose filter. 16. Repeat steps 14 and 15, but reduce cleavage time to 5 min. 17. Collect the cleavage filtrate, remove the filter unit, and precipitate the PNA by adding 1 mL cold (–20°C) diethyl ether. Invert the tube several times to ensure complete precipitation. If the synthesis has been successful, the precipitated PNA should be obvious.

18. Centrifuge precipitated PNA 2 min at 1300 × g. Discard the supernatant. 19. Wash the pellet three times with 1 mL diethyl ether, vortexing to suspend the pellet. 20. Centrifuge 2 min at 8400 × g to repack the pellet. Centrifugation speeds of more than 8400 × g should be avoided because tight packing of PNA makes it difficult to dissolve the pellet.

21. Remove as much of the supernatant as possible by aspiration and then air dry the pellet for 5 to 10 min in a chemical fume hood. 22. Hydrate the pellet with 200 µL sterile water (for a 2-µmol synthesis) and allow the tube to remain undisturbed for 10 to 15 min at 65°C. Only slight vortexing should be required to complete PNA solubilization.

23. Purify and analyze PNA (see Basic Protocol 3).



Supplement 8

SUPPORT PROTOCOL 1

ADDING PEPTIDES TO PNAS The addition of peptide sequences to PNAs is a convenient method for obtaining conjugates in which the peptide domain enhances hybridization (Zhang et al., 2000) or cell uptake (Simmons et al., 1997) of the attached PNA . Peptides can be added to a PNA in a number of ways. If the peptide contains three or fewer different amino acids, it can be conveniently added immediately before or after automated synthesis using the three open ports on the Expedite 8909 synthesizer in addition to the four dedicated to PNA monomers. All amino acids should be double coupled since amino acid coupling is sometimes inefficient. Also, since amino acids are inexpensive relative to PNA monomers, generous use of amino acids during coupling is a cost-effective strategy for optimizing synthesis yields. If more than three different amino acids need to be added, it is often more convenient to contract a dedicated peptide synthesis facility to add the completed peptide. In this case, it is recommended that the first amino acid of the peptide be coupled to the newly synthesized PNA prior to shipping the resin. The authors have found that this procedure reduces the likelihood of the N terminus becoming blocked during shipping and storage. The synthesis of the PNA should be coordinated with the facility that will add the peptide, so that delays between syntheses are avoided. Delay between syntheses can result in the spontaneous loss of FMOC groups, exposing the N terminus to modification and preventing its extension. The facility adding the peptide should be instructed to double or triple couple the first amino acid that is added at their facility. Prior to shipping, the column should be washed with DMF and dried in vacuo overnight. The thiol group of cysteine provides a convenient reactive group for PNA modification. Additional care is required to adequately cleave PNAs or PNA conjugates containing sulfhydryl groups, as the authors and others (Goodwin et al., 1998) have noted that the sulfhydryl can be modified during deprotection and purification. To avoid modification of cysteine, the cleavage cocktail should be supplemented with 7.5 mg pure crystalline phenol (Fisher) and 250 µL ethanedithiol (Sigma-Aldrich) per 1 mL of cleavage cocktail. Cysteine-containing PNAs should be neutralized immediately after purification to avoid the formation of TFA adducts. Neutralization can be achieved by adding 0.5 mL of an aqueous solution of 0.1 M ammonium acetate to the RP-HPLC fraction collection tube prior to collecting the PNA fraction. Incomplete neutralization will result in a product that is 97 mass units higher than expected, corresponding to an adduct with TFA.

SUPPORT PROTOCOL 2

ADDITION OF BIOTIN Once a PNA is synthesized, fluorescent groups, biotin, or other labels can be added to the free N terminus prior to deprotection. The labeling of PNAs with biotin is described as an example. Additional Materials (also see Basic Protocol 1) Biotin (Sigma) 42°C water bath 1-mL syringe 1. Perform a normal automated PNA synthesis (see Basic Protocol 1, steps 1 to 8). 2. Dissolve 20 mg biotin, 6.1 mg HATU, and 1.4 mg HOAt in 800 µl DMF. Warm the solution to 42°C and vortex intermittently until completely dissolved.

Synthesis and Purification of Peptide Nucleic Acids

Biotin is not readily soluble and will require warming and vortexing. Due to its poor solubility, biotin solutions should never be put on the synthesizer, as it will clog the lines.

3. Add 200 µL PNA base solution and mix well. Allow components to activate for 5 min at 42°C.

4.11.6 Supplement 8


4. Remove N-terminal FMOC group by performing the final two deblock cycles to remove the N-terminal cap of the resin-bound PNA (see Basic Protocol 1, step 9). 5. Wash the column two times from each end with 10 mL DMF. 6. Draw up the biotin solution into a 1-mL syringe and place into one end of the column. 7. Place a second syringe into the other end of the column and push the solution back and forth for 30 min. 8. Repeat steps 5 to 7 with another 1-mL preparation of biotin solution. Addition of biotin is repeated because of the inherently poor coupling efficiency of biotin.

9. Proceed to the procedure for cleavage of PNAs from solid-phase resin (see Basic Protocol 1, steps 10 to 23). MANUAL SYNTHESIS OF PEPTIDE NUCLEIC ACIDS The manual synthesis of PNAs is illustrated in Figure 4.11.3. Manual synthesis of PNAs is advantageous because PNAs can be obtained in larger amounts (>2 µmol) than on the Expedite synthesizer. By changing the amount of resin used, one can prepare as much or as little PNA as needed for each experiment. Manual synthesis also avoids the need for a dedicated automated synthesizer. As with automated synthesis, it is important to keep reagents and materials as anhydrous as possible. However, the coupling reactions are more exposed to atmospheric water. The authors have found that syntheses that use tert-butyloxycarbonyl (BOC) monomers generally produce better yields than syntheses using FMOC monomers.

BASIC PROTOCOL 2

Manual PNA synthesis is often so efficient that a capping step can be dispensed with. It is sometimes advisable, however, to include a capping step after coupling of the monomer to simplify the HPLC purification. The individual experimenter will have to determine the necessity for capping. It is typically necessary for long syntheses or for syntheses that have failed in the past. Many different types of apparatus can be used for manual peptide synthesis, and it is likely that these can be adapted for PNA synthesis. The apparatus described here (Fig. 4.11.4; Norton et al., 1995) uses common laboratory glassware and offers robust performance. While the details of manual synthesis will vary with apparatus, the outline of the procedure and the precautions that need to be taken will remain the same. Because a single missed step can ruin a labor-intensive synthesis, the authors follow a detailed checklist for each step. The checklist provides a written record that all steps were performed. A sample spreadsheet detailing the amount of reagent needed for synthesis is shown in Figure 4.11.5. Materials Nitrogen source N,N-Dimethylformamide (DMF; OptiDry; Fisher) 4-Hydroxymethylphenylamidomethyl (PAM) resin protected with tert-butyloxycarbonyl (BOC; Applied Biosystems) Carrier resin: PAM resin capped with an acetyl group (see Support Protocol 3) BOC-PNA monomers (Fig. 4.11.3; Applied Biosystems): tert-butyloxycarbonyl-protected peptide nucleic acid monomers (A, C, G, and T), base protected with benzyloxycarbonyl 2-(1H-Benzotriazol-1-yl)-1,1,3,3-tetramethyl uronium hexafluorophosphate (HBTU) and 1-hydroxybenzotriazole (HOBt) activators (Applied Biosystems) Fresh dichloromethane (DCM; Fisher) m-Cresol Trifluoroacetic acid (TFA; Burdick Jackson)



Supplement 9

coupling step base*

NHBOC

P

O

deprotection 5% m -cresol in TFA P

base*

N

O N

HO

NH2

NHBOC

O

HBTU/HOBT/DIPEA in DMF

N H

P

n=n+1

n

base* O

P

N H

O

O

N H

n

base*

coupling step

N

O

N H

P

NHBOC

O

N

NH2

O

1. deprotection 2. cleavage from the support TFA/m -cresol/thioanisole/TFMSA

when n = n max

base

n max

base O

O

N

HO

deprotection

base*

N

NHBOC

O

N

N H

O

NH2

O

PNA monomers NHCO2Bn

O

NH

N

N

N

N

O

O

N

O

O

NHBOC

HO

N

N

N

O

NHBOC

HO

NH

N

NHCO2Bn

O N

N

O

N

O

O N

HO

NHCO2Bn

O

N

NHBOC

HO

O

NHBOC

Figure 4.11.3 Manual PNA synthesis as described in Basic Protocol 2. Although a capping step can be added, it is often not required and is not shown here. Abbreviations: base*, N-protected nucleobase (see PNA monomers); BOC, tert-butyloxycarbonyl; Bn, benzyl; DIPEA, diisopropylethylamine; DMF, dimethylformamide; HBTU, 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyl uronium hexafluorophospate; HOBt, 1hydroxybenzotriazole; P, PAM resin; TFA, trifluoroacetic acid; TFMSA, trifluoromethanesulfonic acid.

Pyridine Diisopropylethylamine (DIPEA) Methanol Thioanisole Trifluoromethanesulfonic acid (TFMSA; Aldrich) Diethyl ether, ice cold


250°C oven 125-mL vacuum filtration side-arm flasks 24/40 rubber septa 15-mL medium (C) fritted Pyrex funnel Vacuum tubing 3-way valves 250-mL Wheaton bottles with caps

4.11.8 Supplement 9


fritted filter

side-arm flask

coupling flask

deprotection flask

3-way valve

vacuum vacuum release valve N2

Figure 4.11.4 Apparatus for manual synthesis. Reprinted from Norton et al. (1995) with permission from Elsevier Science.

C-Terminus 1 2 3 4 5 6 7 8 9 10 11 12 13 14 N-Terminus

Monomer terminal amide lysine A A C A G A T T G G G A T

Molecular Weight g/mol 17.00 128.00 275.32 275.32 251.30 275.32 291.32 275.32 266.28 266.28 291.32 291.32 291.32 275.32 266.28

PNA mass after each coupling g/mol – 145.00 420.32 695.64 946.94 1222.26 1513.58 1788.90 2055.18 2321.46 2612.78 2904.10 3195.42 3470.74 3737.02 Total Mass

A Monomer G Monomer C Monomer T Monomer HBTU HOBt DIPEA DMF

Amount of Resin Used Substitution Number for Resin Total Active Sites PNA Mass

10 mg 0.66 mmol/g 0.0066 mmol 3737.02 mg/mmol

Amount of PNA to be Synthesized

24.6 mg

Equivalents 5.0 5.0 5.0 5.0 4.5 5.0 10.0 –

Amount Per Coupling 17.41 mg 17.94 mg 16.62 mg 12.68 mg 11.26 mg 4.46 mg 8.54 µL Up to 1 mL

3737.02

Figure 4.11.5 Example of a spreadsheet for the manual synthesis of a 14-base PNA including lysine at the C terminus. The resin used produces a theoretical yield of 24.6 mg of PNA upon cleavage.



Supplement 8

Lyophilizer 10-mL flask with a ground glass joint Desiccator Additional reagents and equipment for purification and analysis of PNAs (see Basic Protocol 3) NOTE: All powdered reagents such as monomers and activator should be unpacked on arrival and stored at –20°C in a sealed container containing Drierite desiccant. Monomers should be inspected upon arrival. Clumps of reagent may indicate that water has been introduced during shipping. DMF should have a low amine content to reduce the likelihood of side reactions. Dry reagents and supplies 1. Dry all monomers and coupling reagents overnight in vacuo before beginning synthesis. To prevent accumulation of moisture, BOC-PNA monomers that have been refrigerated should be warmed to room temperature before opening.

2. Dry all glassware, pipet tips, and 1.5-mL microcentrifuge tubes in a 250°C oven overnight prior to use. Flush pipet tips and microcentrifuge tubes with nitrogen to ensure they are dry prior to synthesis. Set up apparatus 3. Clear a work area in a chemical hood. 4. Fit rubber septa over the tops of two 125-mL side-arm flasks and make a single hole in the middle of each septum to allow a 15-mL fritted funnel to be inserted without too much force. 5. Assemble the manual synthesis apparatus as in Figure 4.11.4 using vacuum tubing and 3-way valves such that the vacuum and nitrogen bubbling can be easily manipulated on and off. 6. Ensure that the 15-mL fritted funnel is clean and unclogged. Pull 5 mL DMF through the funnel to make sure it drains quickly. Prepare reagents 7. In a 1.5-mL microcentrifuge tube, weigh out the appropriate amount of BOC-protected PAM resin and add enough carrier resin to give ∼50 mg total resin weight. The purpose of the carrier resin is to facilitate handling of the resin by allowing researchers to handle larger volumes. The amount of PNA to be synthesized is dependent on the amount of resin used and the number of active sites on the resin. The number of active sites is based on the substitution number (Fig. 4.11.5), which should be listed on the resin bottle when purchased. The total number of active sites is based only on the amount the BOC-protected (noncarrier) PAM resin, as the carrier resin has been capped and contains no active sites. PAM resin can also be purchased with the first amino acid already attached (Advanced Chemtech).

8. Load 1.5-mL microcentrifuge tubes with the proper amount of dry monomer for each coupling reaction. Use a 5-fold excess of monomer over the resin active sites. Synthesis and Purification of Peptide Nucleic Acids

It is easiest to set up a spreadsheet (Fig. 4.11.5) with quantities of monomers, HBTU, HOBt, and DIPEA used for each activation/coupling reaction, checking them off during synthesis of the PNA.



9. For each coupling reaction, measure out a 5-fold excess of HOBt (relative to resin active sites), a 4.5-fold excess of HBTU, 10 eq DIPEA, and 1 mL DMF. Draw DMF from the bottle using a syringe under nitrogen or argon gas to keep it as anhydrous as possible. HOBt, HBTU, DIPEA, and DMF will be added to the monomer in each microcentrifuge tube just minutes before starting the coupling reaction (step 17). This method is the most efficient way of activating the monomers, and thus preventing deletions in the final product. Aldrich sells DMF in Sure/Seal bottles, which have a rubber opening on a crimped lid so that a syringe can be inserted without uncapping the bottle and exposing the contents to air. A nitrogen-filled balloon may also be inserted through the rubber opening to continually flush the bottle with nitrogen during the synthesis.

10. Prepare 200 mL fresh 50:50 (v/v) DMF/DCM for washing. Also prepare 100 mL of 5% (v/v) m-cresol/TFA and 100 mL pyridine in separate clean, dry 250-mL Wheaton bottles with caps. The pyridine is used to wash the resin with the base prior to coupling. CAUTION: TFA is caustic and should be dispensed using only glass pipets or pipet tips containing a charcoal filter (Intermountain Scientific). The charcoal filter serves as a barrier that protects the pipettor seals from the TFA fumes.

11. Assemble disposable glass pipets and bulbs, one for each reagent, and place near or on each reagent bottle. For convenience and to prevent contamination, the authors usually tape a test tube onto each bottle as a pipet holder.

12. Place the 15-mL fritted funnel into the deblocking flask. Wet the rubber septum and the funnel with a little methanol to help it slide in. 13. Add the dry resin to the 15-mL fritted funnel without solubilizing it in DMF. It is easier to get it all in this way.

Perform synthesis 14. Swell resin in 1 mL DMF for 1 hr with nitrogen bubbling. 15. Close and bleed vacuum line. 16. Add 1 mL of 5% m-cresol/TFA to begin deprotection. Stir 3 min with nitrogen bubbling. 17. While the BOC group is being removed, activate (esterify) the first monomer to be added by adding HOBt, HBTU, DMF, and DIPEA (from step 9) to the monomer tube (step 8). Vortex until everything has gone into solution. Typically, these solutions turn a tan color after vortexing. Failure to change color can sometimes be an indication that the activation step is not proceeding properly. As the synthesis progresses, liquid will accumulate in the flasks. It is important to watch the level of liquids in the coupling and deprotection flasks and empty them when the volumes approach the fritted funnel. The deprotection flask can be emptied during a coupling step and vice versa.

18. Vacuum off deblocking solution from the resin. 19. Wash resin two times with 1 mL of 50:50 DMF/DCM. 20. Close and bleed vacuum line.



Supplement 8

21. Remove funnel from the deprotection flask and insert into the rubber septum over the coupling flask. 22. Add activated monomer to resin and mix 20 min, with nitrogen bubbling, to couple the monomer to the resin (or growing PNA). 23. Vacuum off solution. 24. Wash six times with 1 mL of 50:50 DMF/DCM. 25. Wash with 1 mL pyridine then with 1 mL dry DMF. If capping is necessary, add 1 mL of 1:1 (v/v) acetic anhydride/DMF to the resin and mix for 20 min. Wash well with DMF and proceed to the next step.

26. Vacuum off solution. 27. Close and bleed vacuum line. 28. Remove funnel from the coupling flask and insert into the septum over the deblocking flask. 29. Repeat steps 16 to 28 for remaining monomer additions. 30. Remove the final protecting group as in steps 15 and 16. 31. Vacuum off deblocking solution. 32. Wash six times with 1 mL of 50:50 DMF/DCM. 33. Wash six times with 1 mL methanol. 34. Lyophilize overnight. When the synthesis is complete, it is customary to wash the resin with methanol and to dry in vacuo overnight before the cleavage is carried out. If the synthesis takes >1 day, it will be necessary to store the resin. Stop the synthesis just prior to deblocking the PNA, leaving the protecting group on. Wash the resin several times with HPLC-grade methanol. Remove the funnel with the resin from the coupling flask, put a Kimwipe over the top, and secure it with a rubber band. Label the funnel and put it in a desiccator. Resume the synthesis the following day by starting with swelling of the resin in NMP and deblocking the PNA (steps 14 to 16).

Cleave PNA from resin 35. Connect a clean 125-mL vacuum flask to the synthesis apparatus in place of the coupling flask. 36. Attach a rubber septum over the top of the flask and insert the funnel that contains the resin. 37. Prepare 4 mL cleavage cocktail in a clean, dry 10-mL flask that has a ground glass joint. Mix 1 part m-cresol, 1 part thioanisole, and 6 parts TFA. Cap the flask, put on gloves and a shield, crack open a vial of TFMSA, and add 2 parts TFMSA. Typically, the cleavage cocktail turns brown at this point. CAUTION: TFMSA is corrosive and extremely destructive to mucous membranes, the upper respiratory tract, eyes, and skin. Avoid skin contact and inhalation. Always use suitable protection.

38. Using a pipettor, add 1 mL cleavage cocktail to the resin. Using a very small flow of nitrogen into the flask, allow solution to barely bubble for ∼1 hr. Synthesis and Purification of Peptide Nucleic Acids



39. Turn off the nitrogen and apply a vacuum to pull the cleavage solution into the clean flask. Remove as much of the cleavage solution as possible before proceeding to the precipitation. Excess TFA will make it more difficult to precipitate the PNA. Excess TFA can be removed from the solution by blowing a steady stream of nitrogen into the flask until most of the solution is gone, or by applying a vacuum to the flask for several minutes.

40. With most of the cleavage solution gone, add ≥45 mL ice-cold diethyl ether to the flask and set it on ice. 41. Take a spatula and scrape the sides of the flask to loosen all of the PNA into solution. There should be a white precipitate in the solution.

42. Pour the PNA/diethyl ether solution into a 50-mL conical centrifuge tube, cap tightly, and centrifuge 3 min at 1040 × g (e.g., 2500 rpm in a Beckman S4180 swinging bucket rotor), 5°C, pelleting the PNA to the bottom of the tube. 43. Carefully decant or aspirate off the diethyl ether. 44. Wash pellet three times by adding 50 mL ice-cold diethyl ether to the tube, vortexing, centrifuging, and removing the supernatant. 45. After decanting the last time, place a Kimwipe over the tube containing the wellwashed pellet and place it in a desiccator. Attach the house vacuum and allow PNA to dry overnight. 46. Purify and analyze PNA (see Basic Protocol 3). The PNA can be stored at −20°C as a lyophilized product or as an aqueous stock solution (in deionized water). Traces of TFA will slightly acidify the solution and readily dissolve the PNA.

PREPARATION OF CARRIER RESIN FOR MANUAL PNA SYNTHESIS Carrier resin is prepared by deprotecting a simple BOC-protected PAM resin and performing a capping step using acetic anhydride.

SUPPORT PROTOCOL 3

Additional Materials (also see Basic Protocol 2) tert-Butyloxycarbonyl-protected 4-hydroxymethylphenylamidomethyl resin (e.g., BOC-Ala-PAM, BOC-Val-PAM, BOC-Ile-PAM; Applied Biosystems) Acetic anhydride HPLC-grade dichloromethane HPLC-grade methanol 1. Set up the manual synthesis apparatus (Fig. 4.11.4) with a single flask. 2. Place ∼2 g BOC-protected PAM resin in the fritted funnel and swell with 1 mL DMF for 1 hr. 3. Vacuum off DMF. 4. Add a large excess (e.g., 1 mL) of 5% (v/v) m-cresol/TFA and deprotect the resin for 10 min with bubbling. 5. Wash two times with 1 mL DMF and repeat deprotection two more times. The resin should be fully deprotected at this point.

6. Wash five times with 1 mL DMF and once with 1 mL pyridine. 7. Add 1 mL of 1:1: (v/v) acetic anhydride/DMF and bubble for 20 min. 8. Wash once with 1 mL DMF and repeat step 7.



Supplement 8

9. Wash once each with 1 mL DMF, 1 mL dichloromethane, and 1 mL methanol. 10. Dry overnight in a desiccator. The resin can be prepared in bulk and stored at 4°C in the desiccator. BASIC PROTOCOL 3

PURIFICATION AND ANALYSIS OF PEPTIDE NUCLEIC ACIDS PNAs can be purified by reversed-phase HPLC (RP-HPLC) followed by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) (Fig. 4.11.6). PNAs are not, however, purified or analyzed by standard procedures used for oligonucleotides. PNA purification and analysis is more similar to that for peptides. Materials PNA sample solution (see Basic Protocols 1 and 2) RP-HPLC buffer A: 0.1% (v/v) trifluoroacetic acid (TFA; Burdick Jackson) in water, passed through a 47-mm, 0.4-µm nylon membrane (Whatman) RP-HPLC buffer B: 0.1% (v/v) TFA in acetonitrile (Optima grade; Fisher), filtered through an Anodisc 47 filter (0.22-µm; Whatman) α-Cyano-4-hydroxycinnamic acid (Sigma) Isopropanol High-performance liquid chromatograph (HPLC) with C18 reversed-phase column (300-Å Microsorb-MV column; Varian Analytical Instruments) Matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometer (Voyager-DE workstation; Applied Biosystems) Lyophilizer UV spectrophotometer Additional reagents and equipment for HPLC and MALDI-TOF-MS

A 12.358

B m.w. 5144.58

%B 100

60 40 20

11.153

0

5

10.073

9.261 9.910

4.898

3.171

10.181 10.418

80

Retention time (min)


Mass (amu)

Figure 4.11.6 Typical (A) HPLC and (B) MALDI-TOF mass spectrometry data.



Purify by HPLC 1. Centrifuge a PNA solution 3 min at 12,000 × g, room temperature, to remove particulate. 2. Heat a C18 reversed-phase HPLC column to 55°C. PNAs tend to form internal structure or higher order aggregates; sharper peaks will often be obtained if the column is maintained at 55°C using a heated water jacket.

3. Set up a gradient of 0% to 5% (v/v) RP-HPLC buffer B in buffer A for 6 min followed by 5% to 100% buffer B in buffer A for 24 min. 4. Inject sample. 5. Collect fractions corresponding to major peaks. Analyze by MALDI-TOF-MS 6. Spot a 1-µL aliquot of each HPLC fraction on the laser target of a MALDI-TOF mass spectrometer. 7. Overlay the sample with 1 µL matrix consisting of 10 mg/mL α-cyano-4-hydroxycinnamic acid in 25:75 (v/v) RP-HPLC buffers A/B. 8. Activate the laser and collect data. Resolution is best when lower laser energies are used (∼1400 mV) but higher energies are sometimes necessary, especially with long PNAs and PNA-peptide conjugates. Analytical scale analysis by HPLC or mass spectral analysis of the crude product often reveals that only one product has been formed. If this is the case, the PNA can be conveniently purified by Sep-Pak Vac 6-mL (1 g) C18 cartridge (Waters Chromatography). Alternatively, PNAs can be purified by preparative C18 HPLC (Baker Bond, J.T. Baker). PNAs typically elute between 32% and 37% RP-HPLC buffer B and are routinely >80% full-length material.

Solubilize and store PNA 9. Pool collected fractions containing the PNA with the appropriate mass. Freeze in an isopropanol/dry ice bath and lyophilize. PNAs that are not to be used immediately can be stored indefinitely in a lyophilized form at –20°C in a sealed storage box containing desiccant.

10. Dissolve the resulting pellet in 200 µL sterile water and allow the tube to sit 5 to 10 min undisturbed at room temperature. PNAs are readily soluble in aqueous solutions at pH 5.0 to 6.0 at high concentrations (millimolar) but are less soluble at higher pH’s. Solubility at neutral pH can be enhanced by the incorporation of charged amino acid residues at the termini of PNAs during synthesis. Heating solutions containing PNAs can also enhance solubility, and the authors recommend always heating PNAs to >50°C immediately prior to use to ensure that aggregation is minimized. The authors have repeatedly frozen and thawed solubilized PNAs without observing diminished hybridization, although for long-term storage it is best to store PNAs in lyophilized form.

Analyze by UV spectrophotometry 11. Heat PNA stock solutions 5 min at 65°C to reduce aggregation prior to dilution and measurement. PNAs readily aggregate because they are relatively hydrophobic. Heating is a convenient way to ensure that the measured concentration reflects the total concentration of PNA in the sample. Repeated heating of PNA samples does not affect their activity.

12. Observe the optical density at 260 nm (typically a 1:500 dilution).



Supplement 8

13. Determine the concentration using the following equation: c (mM) = (A260 × 30 ng/µL × 500)/mol. wt. where 30 ng/µL is the extinction coefficient, 500 is the dilution factor, and mol. wt. is the molecular weight of the PNA. Alternatively, the concentration can be determined as a cumulative function of the extinction coefficients of the PNA monomers. While more time-consuming, the latter method is more accurate since it is based on the extinction coefficient for a specific PNA sequence. The authors have found that it yields concentrations that vary by as much as 50% to 60% from those derived from use of the standard conversion factor. To determine the concentration using extinction coefficients, add up the extinction coefficients for the PNA sequence [A =13.7, C = 6.06, G = 11.7, T= 8.6, mL/ìmol(cm)]. Calculate the OD260/mL of crude PNA and divide by the total extinction coefficient to obtain the millimolar concentration. Determination of melting temperature provides a useful functional analysis for PNAs. A clean melting curve ensures that the PNA is present in solution at the anticipated concentration and can hybridize to its target sequences. The Tm value also provides important information for optimizing annealing conditions and interpreting results.



Peptide nucleic acids (PNAs; Nielsen et al., 1991) are DNA analogs that have a neutral amide backbone (Fig. 4.11.1) and possess physical properties that differ from those possessed by nucleic acids with traditional phosphodiester or phosphorothioate backbones. Although they hybridize with high affinity to DNA and RNA according to normal WatsonCrick base-pairing rules (Egholm et al., 1993), the neutral backbone eliminates the electrostatic repulsion that characterizes the hybridization of DNA and RNA strands. PNA hybridization to single-stranded DNA or RNA occurs with high affinity, and hybridization to duplex DNA is characterized by an outstanding potential for strand invasion (Smulevitch et al., 1996; Lohse et al., 1999; Nielsen, 2001). The absence of a negatively charged backbone also reduces the likelihood that PNAs will associate with cellular proteins (Hamilton et al., 1996) and generate misleading phenotypes. Another difference relative to DNA or RNA is that the strength of PNA hybridization is independent of salt concentration. Given the many nucleic acid derivatives available, why should researchers consider using PNAs? PNAs possess distinctive chemical properties that confer numerous favorable properties, including high-affinity binding, rapid rates of hybridization, efficient strand invasion, resistance to digestion by nucleases and proteases, and low propensity to bind to proteins. PNAs do not spontaneously enter cultured cells, but can be introduced through simple transfection protocols (Hamilton et al.,

1999; Herbert et al., 1999; Braasch and Corey, 2001; Doyle et al., 2001). PNA synthesis is efficient and versatile and PNAs can be readily purified by HPLC and characterized by mass spectral analysis. Most PNAs will be less soluble than DNA or RNA, but lower solubility can be overcome by adjusting the pH of stock solutions or by heating PNA solutions prior to use. Learning how to obtain and work with PNAs is not trivial, but the power of PNA recognition amply justifies the effort for many applications. Advantages and disadvantages of PNAs are summarized in Table 4.11.1. PNAs can also be purchased directly from Applied Biosystems. Other vendors throughout the world have been licensed to sell PNAs. Currently, international vendors include Nippon Flour Mill, Omgen, Sawady Technologies, and OSWEL-University of Southhampton. Applied Biosystems can supply information for the vendor most convenient to a particular laboratory. Obtaining PNAs from commercial sources will probably be a less expensive option for laboratories that require a limited number of PNAs on a 2-µmol scale, especially if the laboratories do not have experience making peptides.

Critical Parameters The solid-phase synthesis of PNAs uses protocols similar to those developed for peptide synthesis and the physical properties of PNAs are more similar to hydrophobic peptides than to DNA or RNA. PNA solubility at neutral pH is relatively low, and PNAs tend to aggregate upon storage. These properties can pose prob-



Table 4.11.1 Advantages and Disadvantages of Peptide Nucleic Acids

Advantages

Disadvantages

Synthesis by standard peptide synthesis protocols

Less soluble than DNA or RNA

Easy to derivatize at N or C terminus High-affinity hybridization

Tendency for some sequences to aggregate Requires carrier DNA and lipid for delivery into mammalian cells or attachment to import peptides

High rates of hybridization Exceptional potential for strand invasion Low potential for binding to proteins that bind negatively charged polymers

lems for researchers unfamiliar with PNAs, and methods for obtaining useful concentrations of soluble PNA will be discussed below. The authors have not found poor solubility to be a major barrier to experiments with PNAs since the solubility of PNAs is a pH-dependent property. PNAs are soluble to high concentrations at pH ≤5.0, and usually remain soluble at neutral pH when diluted to lower concentrations or when pH is slowly adjusted upward. A substantial advantage is that PNAs can be readily derivatized by peptides or other groups that can form a covalent linkage with the free amino terminus. For automated synthesis, proper maintenance of the instrument is essential. One common problem is clogged lines that can lead to reduced flow rate. Line-filters should be replaced monthly on all internal bottles to ensure that the flow of reagents is consistent for all syntheses. When bottles are removed or placed on the instrument, the open lip of each bottle should be wiped with a clean Kimwipe to prevent cross-contamination of bottles and reduce wear on the O-ring seals. The O-ring seals should be replaced every 3 months regardless of any appearance of physical stress.

Troubleshooting A PNA that has been correctly synthesized will usually give rise to predominantly one peak by HPLC purification. However, PNAs that possess a high likelihood for forming internal structure may give rise to multiple peaks. The potential for multiple peaks can be lessened by heating the HPLC column at 55°C. If mass spectral analysis shows that the major peaks share a single predominant product, multiple peaks may not be a cause for concern. The most

common problem with PNA synthesis is formation of truncated products. One cause of failed syntheses is use of reagents that are contaminated with water, leading to a reduction in coupling efficiency. PNA monomer and activator should be routinely dried in vacuo prior to synthesis. The authors also use fresh bottles of solvent, ≤14 days old. Alternatively, failed sequences could arise due to problems with automated synthesis or human error during manual synthesis. If truncated products are formed, the HPLC will show a series of peaks with shorter retention times than would be expected for full-length product. Another problem is failure to fully cleave protecting groups after completion of synthesis. Since the protecting groups are hydrophobic, a failure to cleave them results in products with longer retention times after RP-HPLC. PNAs that have this problem will appear to have at least two peaks upon HPLC analysis and can be salvaged by a second treatment with TFA and m-cresol. Retention of FMOC groups can also be confirmed by observation of a mass of 222 atomic mass units (amu) greater than expected. To avoid this problem, the authors use TFA within 1 month of first opening bottles. The solubility properties of PNA are discussed above. PNA oligomers have different solubility properties than analogous DNA oligonucleotides and will often be less soluble. Dissolving PNA stocks at pH ≤5.0 and heating solutions prior to use should alleviated this. The authors routinely check the concentration of PNA stocks by monitoring their absorbance at 260 nm to ensure that the expected concentration is being maintained.



Supplement 8

Anticipated Results The authors’ laboratory has made several hundred PNAs. When the Expedite synthesizer is working properly and when all reagents are fresh and dry, an excellent yield of the desired product is almost invariably achieved. Crude material before purification contains ≥50% desired product, while HPLC-purified product contains >90%. PNAs and PNA-peptide conjugates that require as many as 40 coupling steps can be synthesized in good yield. It is possible to synthesize PNAs with long runs of purines or PNAs that are self-complementary as long as critical coupling steps are repeated. PNA synthesis is actually easier than peptide synthesis because the “personalities” of only four, rather than twenty, monomers are involved in determining coupling efficiency. With care, almost any PNA can be obtained.

Time Considerations Preparing the reagents for automated or manual PNA synthesis requires 1 to 2 hr. Automated PNA synthesis can be completed within hours, depending on the length of the PNA. Manual synthesis should take no more than 1 hr per cycle depending on whether a capping step is performed. Deprotection and purification can be accomplished in 1 day. HPLC analysis and purification should take 2 to 3 hr combined. MALDI-TOF-MS should require no more than 30 min.

Literature Cited Braasch, D.A. and Corey, D.R. 2001. Synthesis, analysis, purification, and intracellular delivery of peptide nucleic acids. Methods 23:97-107. Doyle, D.F., Braasch, D.A, Simmons, C.G., Janowski, B.A., and Corey, D.R. 2001. Intracellular delivery and inhibition of gene expression by peptide nucleic acids. Biochemistry 40:53-64.


Hamilton, S.E., Simmons, C.G., Kathriya, I., and Corey, D.R. 1999. Cellular delivery of peptide nucleic acids and inhibition of human telomerase. Chem. Biol. 6:343-351. Herbert, B.-S., Pitts, A.E., Baker, S.I., Hamilton, S.E., Wright, W.E., Shay, J.W., and Corey, D.R. 1999. Inhibition of telomerase in immortal human cells leads to progressive telomere shortening and cell death. Proc. Nat. Acad. Sci. U.S.A. 96:14726-14281. Lohse, J., Dahl, O., and Nielsen, P.E. 1999. Doubleduplex invasion by peptide nucleic acid: A general principle for sequence-specific targeting of double-stranded DNA. Proc. Natl. Acad. Sci. U.S.A. 96:11804-11808. Mayfield, L.D. and Corey, D.R. 1999. Automated synthesis of peptide nucleic acids (PNAs) and peptide nucleic acid-peptide conjugates. Anal. Biochem. 268:401-404. Nielsen, P.E. 2001. Targeting double-stranded DNA with PNA. Curr. Med. Chem. 8:545-550. Nielsen, P.E., Egholm, M., Berg, R.H., and Buchardt, O. 1991. Sequence-selective recognition of double stranded DNA by a thymine-substituted polyamide. Science 254:1497-1500. Norton, J.C., Waggenspack, J.J., Varnum, E., and Corey, D.R. 1995. Targeting peptide nucleic acid protein conjugates to structural features within duplex DNA. Bioorg. Med. Chem. 3:437-445. Simmons, C.G., Pitts, A.E., Mayfield, L.D., Shay, J.W., and Corey, D.R. 1997. Synthesis and membrane permeability of PNA-peptide conjugates. Bioorg. Med. Chem. Lett. 7:3001-3007. Smulevitch, S.V., Simmons, C.G., Norton, J.C., Wise, T.W., and Corey, D.R. 1996. Enhanced strand invasion by oligonucleotides through manipulation of backbone charge. Nature Biotech. 14:1700-1704. Zhang, X., Ishihara, T., and Corey, D.R. 2000. Strand invasion by mixed base PNAs and PNApeptide chimera. Nucl. Acids Res. 28:3332-3338.

Internet Resources http://www.appliedbiosystems.com/ds/pna.taf

Egholm, M., Buchardt, O., Christensen, L., Behrens, C., Freier, S.M., Driver, D.A., Berg, R.H., Kim, S.K., Norden, B., and Nielsen, P.E. 1993. PNA hybridizes to complementary oligonucleotides obeying the Watson-Crick hydrogenbonding rules. Nature 365:566-568.

Ordering information and bibliography.

Goodwin, T.E., Holland, R.D., Lay, J.O., and Raney, K.D. 1998. A simple procedure for solid-phase synthesis of peptide nucleic acids with N-terminal cysteine. Bioorg. Med. Chem. Lett. 8:22312234.

PNA synthesis provider in The Netherlands.

Hamilton, S.E., Iyer., M., Norton, J.C., and Corey, D.R. 1996. Specific and nonspecific inhibition of RNA synthesis by DNA, PNA and phosphorothioate promoter analog duplexes. Bioorg. Med. Chem. Lett. 6:2897-2900.

http://www.horizonpress.com/gateway/pna.html Links to PNA-related sites. http://www.isogen.nl/pna.html

http://www.bostonprobes.com Supplier of PNA diagnostic probes.

Contributed by Dwaine A. Braasch, Christopher J. Nulf, and David R. Corey University of Texas Southwestern Medical Center at Dallas Dallas, Texas



Locked Nucleic Acids: Synthesis and Characterization of LNA-T Diol

UNIT 4.12

This unit describes the eleven-step convergent synthesis of the thymidine analog of locked nucleic acids (LNAs) starting from a commercially available sugar. The first five protocols describe the synthesis of a glycosyl donor suitable for synthesis of several LNA monomers; the overall procedure is illustrated in Figure 4.12.1. First, the starting sugar, 1,2:5,6-di-O-isopropylidene-α-D-allofuranose (S.1), is benzylated (S.2; see Basic Protocol 1 and Alternate Protocol). In the next set of procedures (see Basic Protocol 2), the 5,6-O-isopropylidene group is then selectively removed, the product (S.3) is treated with sodium periodate to give the aldehyde derivative (S.4), and a 4-C-hydroxymethyl group is introduced (S.5). The product is then mesylated (S.6; see Basic Protocol 3) and acetylated (see Basic Protocol 4), yielding the glycosyl donor, 1,2-di-O-acetyl-3-O-benzyl-5-O-methanesulfonyl-4-C-methanesulfonyloxymethyl-D-erythropentofuranose (S.7). Finally, the last protocol describes the synthesis of LNA-T diol from this sugar precursor (see Basic Protocol 5; Fig. 4.12.2). All procedures are experimentally simple and use readily available standard reagents. The synthesis of the other LNA monomers follows the same general pathway. Glycosylation reactions of adenine and guanine are known to give isomeric mixtures. Thus, chromatographic purification is to some extent necessary to obtain pure compounds. Furthermore, these nucleosides contain exocyclic amino groups that need protection. CAUTION: All the reactions should be carried out in a fume hood and contact with chemicals should be avoided. SYNTHESIS OF 3-O-BENZYL-1,2:5,6-DI-O-ISOPROPYLIDENE-α-D-ALLOFURANOSE USING BENZYL BROMIDE IN TETRAHYDROFURAN This protocol describes the benzylation of 1,2:5,6-di-O-isopropylidene-α-D-allofuranose (S.1) by treatment with sodium hydride and benzyl bromide resulting in 3-O-benzyl1,2:5,6-di-O-isopropylidene-α-D-allofuranose (S.2; Fig. 4.12.1). Purification is performed by a combination of extraction and crystallization.

BASIC PROTOCOL 1

CAUTION: Hydrogen gas, which can be explosive, is evolved in steps 5 and 9. Perform steps 2 through 9 under a nitrogen stream and use extreme care. Materials 60% (v/v) sodium hydride in mineral oil Hexane (stored over 3A molecular sieves) Nitrogen (N2) stream (or argon stream) Tetrahydrofuran (THF; stored over 3A molecular sieves) Dimethylformamide (DMF; stored over 3A molecular sieves) 1,2:5,6-Di-O-isopropylidene-α-D-allofuranose (Pfanstiehl Laboratories) Benzyl bromide Brine (saturated aqueous NaCl) MgSO4 500-mL three-neck round-bottom flask 250-mL dropping funnel Nitrogen inlet 20-mL syringe Sintered glass funnel, pore size 3 Contributed by Henrik M. Pfundheller and Christian Lomholt Current Protocols in Nucleic Acid Chemistry (2002) 4.12.1-4.12.16 Copyright © 2002 by John Wiley & Sons, Inc.


4.12.1 Supplement 8

Rotary evaporator connected to vacuum pump Generate 3-alkoxyde 1. Equip a 500-mL three-neck round-bottom flask with a 250-mL dropping funnel, nitrogen inlet, and a magnetic stir bar. 2. Place 11.2 g of 60% sodium hydride in the flask, add 30 mL hexane, and stir for a few minutes under a stream of nitrogen. 3. Stop stirring and allow sodium hydride to settle. Remove the excess hexane carefully with a 20-mL syringe and discard. 4. Add 25 mL THF and 5 mL DMF to the sodium hydride and cool the suspension in an ice bath. 5. Place 52 g (0.20 mol) of 1,2:5,6-di-O-isopropylidene-α-D-allofuranose dissolved in 100 mL THF in the 250-mL dropping funnel and add the solution dropwise over a 30-min period to the sodium hydride suspension. 6. Remove from the ice bath and stir for 1.5 hr at room temperature. Benzylate 3-alkoxyde 7. Over a 20-min period, add 25 mL benzyl bromide (in small portions) to the solution and then stir the mixture for 16 hr. 8. Cool the mixture in an ice bath for 15 min. 9. Destroy excess sodium hydride by careful addition of 100 mL water.

O O

O O O

O

1. NaH O

OH

HO HO O

80% AcOH

2. BnBr

O

O

OBn O

O

1

OBn O

2

3

O

HO

O

NalO4

H2CO, NaOH HO

O OBn O

MsO

Ac2O, AcOH, conc. H2SO4 O

OBn O

6


5

O

MsO

O OBn O

4

MsCl, Pyridine

O

MsO

O

OAc

MsO OBn OAc

7

Figure 4.12.1 Synthesis of the universal glycosyl donor. Bn, benzyl; Ms, methanesulfonyl.

4.12.2 Supplement 8


Purify and isolate product 10. Separate phases and extract the aqueous phase (lower phase) with 75 mL THF. 11. Combine the two organic phases and wash two times with 100 mL brine. 12. Dry over MgSO4, remove the MgSO4 by vacuum filtration through a sintered glass funnel, and wash the solid with 50 mL THF. 13. Concentrate the filtrate in a rotary evaporator attached to a vacuum pump. 14. Dissolve the residue in 100 mL boiling hexane and place solution overnight at 5°C. 15. Isolate the formed crystals of 3-O-benzyl-1,2:5,6-di-O-isopropylidene-α-D-allofuranose (S.2) by filtration and dry under vacuum. Yield: 64 g (91%). An additional 4 g of product can be obtained from the mother liquor.

SYNTHESIS OF 3-O-BENZYL-1,2:5,6-DI-O-ISOPROPYLIDENE-α-D-ALLOFURANOSE USING BENZYL BROMIDE IN DIMETHYLFORMAMIDE

ALTERNATE PROTOCOL

This procedure describes the benzylation of 1,2:5,6-di-O-isopropylidene-α-D-allofuranose (S.1) using DMF as solvent. Using DMF instead of THF makes the reaction mixture less flammable, although DMF is toxic. Purification of the reaction mixture is obtained through precipitation of the product from the reaction mixture followed by crystallization. All reagents and materials are listed above (see Basic Protocol 1). CAUTION: Hydrogen gas, which can be explosive, is evolved in steps 5 and 9. Perform steps 2 through 9 under a nitrogen stream and use extreme care. Generate 3-alkoxyde 1. Equip a 500-mL three-neck round-bottom flask with a 250-mL dropping funnel, nitrogen inlet, and a magnetic stir bar. 2. Place 4.8 g of 60% sodium hydride in the flask, add 20 mL hexane, and stir for a few minutes under a stream of nitrogen. 3. Stop stirring and let the sodium hydride settle, then remove the excess hexane carefully with a 20-mL syringe and discard. 4. Add 20 mL DMF to the sodium hydride and cool the suspension in an ice bath. 5. Place 26 g (0.10 mol) of 1,2:5,6-di-O-isopropylidene-α-D-allofuranose dissolved in 30 mL DMF in the 250-mL dropping funnel and add solution dropwise over a 30-min period to the sodium hydride suspension. 6. Remove from the ice bath and stir for 1.5 hr at room temperature. Benzylate 3-alkoxyde 7. Over a 15-min period, add 14.3 mL benzyl bromide (in small portions) to the solution and then stir mixture for 1 hr at room temperature. 8. Cool the mixture in an ice bath. 9. Destroy excess sodium hydride by careful addition of 5 mL water. Purify and isolate product 10. Pour the reaction mixture into 175 mL ice water with stirring.



Supplement 8

11. When the ice has melted, filter off the precipitate by vacuum filtration through a sintered glass funnel, and wash the solid three times with 100 mL ice-cold water. 12. Dry the solid using a rotary evaporator connected to a vacuum pump. 13. Add 150 mL boiling hexane and keep overnight at 5°C to allow crystal formation. 14. Isolate the formed crystals of 3-O-benzyl-1,2:5,6-di-O-isopropylidene-α-D-allofuranose (S.2) by filtration and dry under vacuum. Yield: 31 g (89%). BASIC PROTOCOL 2

SYNTHESIS OF 3-O-BENZYL-4-C-HYDROXYMETHYL-1,2-O-ISOPROPYLIDENE-α-DERYTHROPENTOFURANOSE In this protocol, selective removal of the 5,6-O-isopropylidene group is performed by treatment of S.2 with 80% acetic acid (steps 1 to 5) to give 3-O-benzyl-1,2-O-isopropylidene-α-D-allofuranose (S.3). The formed 5,6-diol is then oxidatively cleaved with sodium periodate (steps 6 to 11) to give the corresponding 5-aldehyde derivative (S.4). Treatment of S.4 with formaldehyde and aqueous sodium hydroxide in a mixed aldol condensation/Cannizzaro reaction (steps 12 to 19) introduces a 4-C-hydroxymethyl group, giving 3-O-benzyl-4-C-hydroxymethyl-1,2-O-isopropylidene-α-D-erythropentofuranose (S.5). Materials 3-O-Benzyl-1,2:5,6-di-O-isopropylidene-α-D-allofuranose (S.2; see Basic Protocol 1 or Alternate Protocol) Acetic acid Dichloromethane Ethyl acetate Toluene Tetrahydrofuran (THF) NaIO4 Brine (saturated aqueous NaCl) 1,4-Dioxane 37% (w/v) aqueous formaldehyde 4 M NaOH MgSO4 Hexane 1-L round-bottom flask Rotary evaporator connected to vacuum pump Sintered glass funnel, pore size 3 Additional reagents and equipment for thin-layer chromatography (TLC; APPENDIX 3D) Deprotect 5,6-O-isopropylidene group 1. Place 63.1 g (0.18 mol) of 3-O-benzyl-1,2:5,6-di-O-isopropylidene-α-D-allofuranose (S.2) into a 1-L round-bottom flask containing a magnetic stir bar.


2. Add 70 mL water and 290 mL acetic acid and stir at room temperature until no starting material can be observed (∼25 to 35 hr) as monitored by TLC (APPENDIX 3D) using 9:1 (v/v) dichloromethane/ethyl acetate.

4.12.4 Supplement 8


3. Remove the magnetic stir bar and concentrate the solution in a rotary evaporator connected to a vacuum pump. Avoid temperatures that are >45°C in steps 3 and 4, since this will result in removal of the 1,2-O-isopropylidene group.

4. Co-evaporate the residue (S.3) under vacuum two times with 100 mL toluene. Perform periodate oxidation of vicinal 5,6-diol 5. Add 100 mL THF, 100 mL water, and a large magnetic stir bar to the residue. 6. Over a 30-min period, add 43 g NaIO4 in small portions under vigorous stirring. 7. Stir the mixture for 3 hr. 8. Filter off the solid by vacuum filtration through a sintered glass funnel, and wash two times with 50 mL ethyl acetate. 9. Separate the phases and extract the aqueous phase (lower layer) two times with 50 mL ethyl acetate. 10. Combine the two organic phases, wash with 50 mL brine, and concentrate the organic phase (containing S.4) under vacuum in the rotary evaporator. Perform mixed aldol condensation/Cannizzaro reaction 11. Dissolve the residue in 120 mL of 1,4-dioxane. Add a magnetic stir bar and 40 mL aqueous 37% formaldehyde. 12. Over a 40-min period, add dropwise 90 mL of 4 M NaOH and then stir the mixture overnight. 13. Separate the phases and extract the aqueous phase (lower layer) one time with 100 mL ethyl acetate and then two times with 50 mL ethyl acetate. 14. Combine the two organic phases, wash with 100 mL brine, and extract the brine with 50 mL ethyl acetate. 15. Dry the organic phase over MgSO4 and remove the MgSO4 by filtration. 16. Wash the solid two times with 50 mL ethyl acetate and concentrate the combined filtrates in the rotary evaporator. 17. Dissolve the residue in 50 mL dichloromethane and add this solution to 350 mL hexane under vigorous stirring. 18. Isolate the precipitated 3-O-benzyl-4-C-hydroxymethyl-1,2-O-isopropylidene-α-Derythropentofuranose (S.5) by filtration and dry under vacuum. Yield: 45.7 g (82%) of a white solid.



Supplement 8

BASIC PROTOCOL 3

SYNTHESIS OF 3-O-BENZYL-1,2-O-ISOPROPYLIDENE-5-O-METHANESULFONYL-4-CMETHANESULFONYLOXYMETHYL-α-D-ERYTHROPENTOFURANOSE This protocol describes the mesylation of the two hydroxyl groups in 3-O-benzyl-4-Chydroxymethyl-1,2-O-isopropylidene-α-D-erythropentofuranose (S.5) with methanesulonylchloride, giving 3-O-benzyl-1,2-O-isopropylidene-5-O-methanesulfonyl4-C-methanesulfonyloxymethyl-α-D-erythropentofuranose (S.6). The resulting mesyl groups are used in later steps for both protection and functionalization. Materials 3-O-Benzyl-4-C-hydroxymethyl-1,2-O-isopropylidene-α-D-erythropentofuranose (S.5; see Basic Protocol 2) Dichloromethane (stored over 3A molecular sieves) Pyridine (stored over 3A molecular sieves) Methanesulfonylchloride (e.g., Aldrich) Brine (saturated aqueous NaCl) 1 M HCl Saturated aqueous NaHCO3 MgSO4 Methanol 500-mL round-bottom flask 250-mL dropping funnel Guard tube/nitrogen inlet Sintered glass funnel, pore size 3 Rotary evaporator connected to vacuum pump Perform mesylation 1. Place 43.5 g (0.14 mol) of 3-O-benzyl-4-C-hydroxymethyl-1,2-O-isopropylidene-αD-erythropentofuranose (S.5) in a 500-mL round-bottom flask equipped with a magnetic stir bar, 250-mL dropping funnel, and a guard tube/nitrogen inlet. This reaction must be performed under a dry atmosphere.

2. Add 75 mL dichloromethane and 55 mL pyridine and cool mixture in an ice bath. 3. Place 24 mL (0.31 mol) methanesulfonylchloride in the dropping funnel and add dropwise over a 45-min period. 4. Remove from the ice bath and stir 3 hr at room temperature. 5. Add 100 mL water and 100 mL brine and stir 45 min at room temperature. Purify and isolate product 6. Separate the phases and extract the aqueous phase (top layer) three times with 75 mL dichloromethane. 7. Combine the two organic phases and wash three times with 150 mL of 1 M HCl, one time with 150 mL of saturated aqueous NaHCO3, and one time with 150 mL brine. 8. Dry the organic phase over MgSO4 and remove the MgSO4 by vacuum filtration through a sintered glass funnel. Locked Nucleic Acids: Synthesis and Characterization of LNA-T Diol

9. Wash the solid two times with 25 mL dichloromethane and concentrate the combined filtrates in a rotary evaporator with a vacuum pump.

4.12.6 Supplement 8


10. Dissolve the residue in 50 mL boiling methanol. Cool to room temperature and then let sit overnight at 5°C. 11. Isolate the formed crystals of 3-O-benzyl-1,2-O-isopropylidene-5-O-methanesulfonyl-4-C-methanesulfonyloxymethyl-α-D-erythropentofuranose (S.6) by filtration and dry under vacuum. Yield: 56.7 g (90%) of colorless crystals.

SYNTHESIS OF 1,2-DI-O-ACETYL-3-O-BENZYL-5-O-METHANESULFONYL-4-CMETHANESULFONYLOXYMETHYL-D-ERYTHROPENTOFURANOSE

BASIC PROTOCOL 4

This protocol describes the deprotection of the 1,2-O-isopropylidene group in 3-Obenzyl-1,2-O-isopropylidene-5-O-methanesulfonyl-4-C-methanesulfonyloxymethyl-αD-erythropentofuranose (S.6) followed by acetylation, giving the universal glycosyl donor 1,2-di-O-acetyl-3-O-benzyl-5-O-methanesulfonyl-4-C-methanesulfonyloxymethylD-erythropentofuranose (S.7) as a mixture of anomers. Deprotection and acetylation are performed in a one-pot procedure by treatment of S.6 with a mixture of acetic acid, acetic anhydride, and catalytic amounts of concentrated sulfuric acid. Materials 3-O-Benzyl-1,2-O-isopropylidene-5-O-methanesulfonyl-4-C-methanesulfonyloxymethyl-α-D-erythropentofuranose (S.6; see Basic Protocol 3) Acetic acid Acetic anhydride Concentrated H2SO4 Dichloromethane Saturated aqueous Na2CO3 Saturated aqueous NaHCO3 MgSO4, anhydrous 1-L round-bottom flask Guard tube/nitrogen inlet Sintered glass funnel, pore size 3 Rotary evaporator connected to vacuum pump Perform 1,2-O-isopropylidene acetolysis and acetylation 1. Place 46.6 g (0.10 mol) 3-O-benzyl-1,2-O-isopropylidene-5-O-methanesulfonyl-4C-methanesulfonyloxymethyl-α-D-erythropentofuranose (S.6) in a 1-L round-bottom flask equipped with a guard tube/nitrogen inlet and a magnetic stir bar. 2. Add 150 mL acetic acid and 30 mL acetic anhydride, and cool mixture in an ice bath. 3. Add 0.2 mL concentrated H2SO4 and let sit 5 min. 4. Remove from the ice bath and stir mixture for 18 hr at room temperature. Purify and isolate product 5. Add 150 mL dichloromethane to the reaction mixture followed by slow addition of 500 mL saturated aqueous Na2CO3. CAUTION: A large amount of CO2 gas is evolved in this step.

6. Stir the mixture for 3 hr at room temperature.



Supplement 9

7. Separate phases and extract the aqueous phase (top layer) two times with 150 mL dichloromethane. 8. Combine the two organic phases, add 300 mL saturated aqueous NaHCO3, and stir 2 hr at room temperature. 9. Separate phases and extract the aqueous phase (top layer) two times with 75 mL dichloromethane. 10. Dry the combined organic phases over MgSO4 and remove the MgSO4 by vacuum filtration through a sintered glass funnel. 11. Wash the solid two times with 25 mL dichloromethane. 12. Concentrate the combined filtrates to an oil in a rotary evaporator connected to a vacuum pump. The product, 1,2-di-O-acetyl-3-O-benzyl-5-O-methanesulfonyl-4-C-methanesulfonyloxymethyl-D-erythropentofuranose (S.7; mixture of anomers), is obtained as a yellow oil. Yield: 51.0 g (100%). BASIC PROTOCOL 5

SYNTHESIS OF (1S,3R,4R,7S)-7-HYDROXY-1-HYDROXYMETHYL-3-(THYMIN-1-YL)-2,5DIOXABICYCLO[2.2.1]HEPTANE (LNA-T DIOL) This protocol describes the synthesis of LNA-T diol from the universal glycosyl donor (S.7); the procedure is illustrated in Figure 4.12.2. The coupling of S.7 with silylated thymine (steps 1 to 9) gives 1-(2-O-acetyl-3-O-benzyl-5-O-methanesulfonyl-4-Cmethanesulfonyloxymethyl-β-D-erythropentofuranosyl)thymine (S.8). Basic hydrolysis of the 2′-O-acetyl group in S.8 (steps 10 to 16) liberates the free 2′-OH group, which instantly attacks the 4′-C-methanesulfonyloxy group, thereby causing ring formation between 2′-O and 4′-C. The resulting (1S,3R,4R,7S)-7-benzyloxy-1-methanesulfonyloxymethyl-3-(thymin-1-yl)-2,5-dioxabicyclo[2.2.1]heptane (S.9) constitutes the bicyclic LNA skeleton. Substitution of the methanesulfonyloxy group with benzoate (steps 17 to 22) and basic hydrolysis of the benzoate (steps 23 to 31) result in (1S,3R,4R,7S)-7-benzyloxy-1-hydroxymethyl-3-(thymin-1-yl)-2,5-dioxabicyclo[2.2.1]heptane (S.11) . Finally, reductive removal of the benzyl group in S.11 using Pd(OH)2/C-ammonium formate yields LNA-T diol (S.12; steps 32 to 38).


Materials 1,2-Di-O-acetyl-3-O-benzyl-5-O-methanesulfonyl-4-C-methanesulfonyloxymethylD-erythropentofuranose (S.7; see Basic Protocol 4) Acetonitrile (stored over 3A molecular sieves) Thymine N,O-Bis(trimethylsilyl)acetamide (Fluka) Trimethylsilyl trifluoromethanesulfonate (Fluka) Saturated aqueous NaHCO3 Dichloromethane Brine (saturated aqueous NaCl) Tetrahydrofuran (THF) LiOH⋅H2O Acetic acid Ethyl acetate MgSO4, anhydrous Anhydrous dimethylformamide (DMF; stored over 3A molecular sieves) Sodium benzoate

4.12.8 Supplement 9


Hexane Methanol 20% Pd(OH)2/C (palladium hydroxide catalyst on carbon; Fluka) Ammonium formate 2-cm-thick Celite pad 250-mL, 500-mL, and 1-L round-bottom flasks Rotary evaporator connected to vacuum pump Condensers Guard tubes/nitrogen inlets Sintered glass funnel, pore size 3 Glycosylate thymine 1. Place 25.5 g (50 mmol) of 1,2-di-O-acetyl-3-O-benzyl-5-O-methanesulfonyl-4-Cmethanesulfonyloxymethyl-D-erythropentofuranose (S.7) in a 500-mL round-bottom flask. Add 100 mL acetonitrile and concentrate the mixture using a rotary evaporator connected to a vacuum pump. It has been observed that skipping this evaporation step, which is performed to remove traces of water, can lead to lower yields from the reaction.

O

NH MsO

1. Thymine/BSA

O

MsO

OAc 2. TMS-OTf

MsO OBn OAc

N

O

O

MsO OBn OAc

7

8 O

O

NH LiOH MsO

N

NH

O

BzONa BzO

O

OBn O

N

OBn O

10

9 O

O

NH LiOH HO

N O

OBn O

11

O

O

NH

O

Pd(OH)2 /C NH4HCO2

HO

N

O

O

OH

O

12

Figure 4.12.2 Synthesis of LNA-T diol. Bn, benzyl; BSA, bis(trimethylsilyl)acetamide; Bz, benzoyl; Ms, methanesulfonyl; TMS-OTf, trimethylsilyl trifluoromethanesulfonate.



Supplement 8

2. Dissolve in 100 mL acetonitrile, and add 7.0 g (55 mmol) thymine, 24.5 mL (10 mmol) N,O-bis(trimethylsilyl)acetamide, and a magnetic stir bar. 3. Equip the 500-mL round-bottom flask with a condenser and a guard tube/nitrogen inlet, and heat the mixture to reflux (∼90°C) for 45 min. 4. Cool the reaction mixture to room temperature and add 10 mL trimethylsilyl trifluoromethanesulfonate. 5. Heat the mixture to reflux (∼90°C) overnight. 6. Cool the mixture to room temperature and pour it into 200 mL saturated aqueous NaHCO3. Stir the mixture for 5 min. 7. Extract three times with 200 mL dichloromethane. 8. Combine the organic phases and wash with 200 mL saturated aqueous NaHCO3 and then 100 mL brine. 9. Concentrate the organic phase (containing S.8) in a 1-L round-bottom flask in the rotary evaporator. Form 2′-O, 4′-C ring 10. Dissolve the residue in 100 mL THF and add 200 mL water and 10 g LiOH⋅H2O. Stir the mixture for 1.5 hr at room temperature. 11. Neutralize to pH ∼8 with acetic acid (∼10 mL). 12. Extract the mixture three times with 200 mL ethyl acetate. 13. Combine the organic phases and wash with 100 mL brine. 14. Dry the organic phase over MgSO4 and remove the MgSO4 by vacuum filtration through a sintered glass funnel. 15. Wash the solid two times with 25 mL ethyl acetate and concentrate in the rotary evaporator. 16. Dissolve the residue (S.9) in 200 mL acetonitrile and concentrate two times in a 500-mL round-bottom flask. Substitute methanesulfonyloxy group with benzoyloxy group 17. Dissolve the residue in 250 mL anhydrous DMF and add 20 g sodium benzoate. 18. Equip the flask with a guard tube/nitrogen inlet and heat the mixture for 16 hr at 90°C. The reaction results in a gummy mixture. If effective stirring cannot be achieved, an additional 250 mL anhydrous DMF should be added.

19. Cool the reaction mixture to room temperature. 20. Add water until the formed precipitate dissolves. 21. Pour the reaction mixture into 800 mL of stirring ice water (1600 mL of ice water if 500 mL of DMF has been used as solvent). 22. Isolate the precipitated product (S.10) by filtration and wash the product two times with 100 mL water. Locked Nucleic Acids: Synthesis and Characterization of LNA-T Diol



Hydrolyze benzoyl ester 23. Suspend the solid in 200 mL THF and 300 mL water in a 1-L round-bottom flask containing a magnetic stir bar. 24. Add 10 g LiOH⋅H2O and stir the mixture for 5 hr at room temperature. 25. Neutralize to pH ∼8 with acetic acid (∼10 mL) and concentrate to half volume in the rotary evaporator. 26. Extract three times with 200 mL dichloromethane 27. Combine the organic phases and wash with 200 mL saturated aqueous NaHCO3 followed by 200 mL brine. 28. Dry over MgSO4 and remove MgSO4 by filtration. 29. Wash the solid two times with 25 mL dichloromethane and concentrate the combined filtrates in the rotary evaporator. 30. Dissolve the residue in a minimum of ethyl acetate (∼50 mL) and pour this solution slowly into 600 mL hexane under vigorous stirring. 31. Collect the precipitated product (S.11) by filtration. Yield: 14.2 g (79%) of a white solid.

Remove 3′-O-benzyl group 32. Dissolve 14.2 g of S.11 in 100 mL methanol in a 250-mL round-bottom flask equipped with magnetic stir bar and condenser. 33. Add 1.5 g of 20% Pd(OH)2/C and 8.0 g ammonium formate. 34. Heat the mixture for 1 hr at 60°C. 35. Vacuum filter the hot solution through a 2-cm-thick Celite pad. 36. Wash the Celite with 200 mL methanol. Combine the filtrates and concentrate in the rotary evaporator. 37. Dissolve the residue in 50 mL boiling methanol and add 200 mL ethyl acetate followed by 600 mL hexane. 38. Isolate the precipitated product (1S,3R,4R,7S)-7-hydroxy-1-hydroxymethyl-3(thymin-1-yl)-2,5-dioxabicyclo[2.2.1]heptane (LNA-T diol; S.12) by filtration and dry under vacuum. Yield: 10.2 g (95%) as white precipitate.

COMMENTARY Background Information LNA oligonucleotides were introduced several years ago and were shown to have very high binding affinity and selectivity towards complementary nucleic acids (Koshkin et al., 1998b; Obika et al., 1998; Singh et al., 1998). These interesting properties have been evaluated in different diagnostic and therapeutic settings. LNA oligonucleotides have been used in diagnostic genotyping for the detection of the prothrombic mutations factor V Leiden (Ørum

et al., 1999) and have also been successfully evaluated in living rats for control of gene expression (Wahlestedt et al., 2000). Today, much effort is being put into the development of simple and reliable microarrays for multiplex genotyping in combination with multiplex target amplification using LNA oligonucleotides (Choleva et al., 2001). Furthermore, LNA oligonucleotides are successfully being investigated with respect to recognition of double-stranded DNA, as shown by inhibition of



Supplement 8

the NF-κB transcription factor p50 (Obika et al., 2001). The key steps in the synthesis of LNA nucleosides are the introduction of the 4-C-hydroxymethyl functionality followed by selective derivatization of the resulting two diastereoto pic h yd ro xy methy l groups. 4′-C-Hydroxymethyl nucleosides have been synthesized using both linear (Youssefyeh et al., 1977; Jones et al., 1979) and convergent (Leland and Kotick, 1974; Youssefyeh et al., 1977, 1979) strategies based on aldol condensation of 5(′)-aldehydes with formaldehyde followed by a Cannizzaro reaction or sodium borohydride reduction. LNA nucleosides were previously synthesized following linear (Obika et al., 1997; Koshkin et al., 1998a) as well as convergent (Singh et al., 1998; Koshkin et al., 1998a,b) strategies. Due to the reported problems (i.e., low yields, workup difficulties) for th e p reparation of 5′-aldehydo/4′-C-hydroxymethyl nucleosides from natural ribonucleosides, the authors decided to focus on the convergent strategy for the large-scale synthesis of LNA nucleosides further stimulated by the possibilities of introducing a variety of different nucleobases. This unit describes the authors’ efforts to develop a simplified and efficient synthesis of a sugar intermediate (S.7; Figure 4.12.1; see Basic Protocols 1 to 4), which can be used as a glycosyl donor in the coupling reactions with different nucleobases. This is exemplified by the synthesis of LNA-T diol (S.12; Figure 4.12.2; see Basic Protocol 5). These protocols give high yields of the desired products (>79% for each step), are experimentally simple, and avoid the use of time-consuming column chromatography. The synthesis of the adenine, cytosine, and guanine derivatives of LNA follow the same strategy and are described elsewhere (Koshkin et al., 2001).


Synthesis of the universal glycosyl donor S.7 Standard benzylation (see Basic Protocol 1) of the commercially available 1,2:5,6-di-O-isopropylidene-α-D-allofuranose with THF as solvent results in the corresponding 3-O-benzylated derivative S.2 (Horton and Tindall, 1970) in almost quantitative yield after crystallization from hexanes. In order to simplify the workup, the mineral oil from the sodium hydride suspension is initially removed with a syringe after addition of hexanes. As an alternative to THF, DMF can be used as the solvent (see Alternate Protocol) as reported earlier (Brimacombe and Ching, 1968). The starting sugar is commer-

cially available but can also be synthesized from the gluco-epimer via oxidation and selective reduction (Sowa and Thomas, 1966). Using 80% acetic acid, the 5,6-O-isopropylidene protecting group is selectively removed (see Basic Protocol 2) to give S.3 as reported (Horton and Tindall, 1970). It is important to keep the reaction at a low temperature to avoid removal of the 1,2-O-isopropylidene protecting group. After evaporation of the reaction mixture, the 5,6-glycol is oxidatively cleaved by periodate, giving the 5-aldehydo derivative after a simple work-up procedure. Finally, the aldehyde is condensed with formaldehyde followed by in situ crossed Cannizzaro reaction with excess formaldehyde, giving the desired 4-C-hydroxymethyl derivative S.5 (Youssefyeh et al., 1977, 1979) in 82% yield (3 steps) after crystallization from hexanes. In the first publications describing the synthesis of LNA nucleosides using the convergent strategy (Koshkin et al., 1998b; Singh et al., 1998), 3-O-benzyl-4-C-hydroxymethyl-1,2O-isopropylidene-α-D-ribofuranose S.5 was regioselectively 5-O-benzylated (Waga et al., 1993) in 71% yield, taking advantage of the different positioning of the two diastereotopic hydroxymethyl groups on each face of the bicyclo[3.3.0]octane system. However, chromatographic purification of the reaction mixture was necessary to separate the two isomers. Acetylation (Koshkin et al., 1998b; Singh et al., 1998) or tosylation (Koshkin et al., 1998a) of the 4-C-hydroxymethyl group, acetolysis, and subsequent 1,2-di-O-acetylation was followed by nucleobase-coupling and ring-closing reactions to give the desired LNA nucleoside derivative. Based on these results, the authors decided to synthesize 1,2-di-O-acetyl-3-Obenzyl-4-C-methanesulfonoxymethyl-5-Omethanesulfonyl-D-erythropentofuranose S.7 to overcome the early chromatographic step and increase the overall yield of the synthesis. The diol S.5 is permesylated using standard conditions (see Basic Protocol 3) to give the desired compound in 90% yield after crystallization from methanol. Subsequent standard acetolysis and basic diacetylation (see Basic Protocol 4) quantitatively yield S.7 as an anomeric mixture (∼1:5) that can be used as a glycosyl donor in coupling reactions with different nucleobases. The overall yield from S.1 is 67%. Synthesis of LNA-T diol S.12 Following the method of Vorbrüggen (Vorbrüggen et al., 1981; Vorbrüggen and Höfle,



Table 4.12.1

Selected Rf Values and Melting Points

Compound number

Rf (90:10 CH2Cl2/ethyl acetate)a

1 2 3 4 5 6 7 8 9 10 11 12

0.30 0.70

0.55 0.40 + 0.35 0.05

Rf (95:5 CH2Cl2/methanol)a 0.78 0.14 0.40 0.16 0.70 0.60 + 0.55 0.25 0.22 0.45 0.13 0.05

Melting point (°C)b

65c

99d 109

169 196e

aSolvent ratios are given in v/v. TLC plates: Merck silica gel 60 F . 254 bUncorrected, measured on a Büchi melting point B-540. c64°C to 65°C (Brimacombe, 1968). d101°C to 102°C (Youssefyeh, 1979). e204°C to 205°C (Obika, 2001).

1981), S.7 is stereoselectively coupled (see Basic Protocol 5) with silylated thymine followed by a one-pot deacetylation and intramolecular ring-closing reaction upon treatment with LiOH, giving the LNA derivative S.9. This two-step one-pot procedure is fast and smooth compared to the two-step procedure originally described (Koshkin et al., 1998a,b). The methanesulfonyloxy group is substituted with benzoyl upon reaction of S.9 with sodium benzoate in hot DMF (Codington et al., 1960), followed by hydrolysis of the benzoate ester using LiOH to give S.11 in 79% yield (four steps) after crystallization from hexanes. Direct basic hydrolysis of the 5′-O-methanesulfonyl group was unsuccessful (Koshkin et al., 2001). Debenzylation with Pd(OH)2/C and ammonium formate as the H donor is fast and efficient, giving the desired LNA-T diol S.12 in 95% yield (50% from S.1) after a simple workup procedure and crystallization from ethyl acetate/hexanes. The corresponding phosphoramidite for use in automated oligonucleotide synthesis (Caruthers, 1991) can be synthesized via 5′-O-dimethoxytritylation and 3′-O-phosphitylation (Sinha et al., 1983; Koshkin et al., 1998b).

Compound Characterization All isolated compounds appear as one spot on a TLC plate except S.7, which exists as a mixture of anomers (∼1:5). Rf values and ap-

propriate solvent systems are given in Table 4.12.1. For bicyclic LNA structures, 1H-NMR shows three singlet signals for H1′, H2′, and H3′. This can be taken as verification of the formation of the rigid LNA bicyclic structure, as small coupling constants are indicative of H1′-C1′-C2′-H2′ dihedral angles close to 90°, which are characteristic of the N conformation (Altona and Sundaralingam, 1973; Obika et al., 1997; Koshkin et al., 1998b). Data for selected compounds S.2: 1H-NMR (CDCl3) δ 7.39-7.26 (m, 5H), 5.75 (d, J = 3.6 Hz, 1H), 4.78 (d, J = 12.2 Hz, 2H), 4.61-4.56 (m, 2H), 4.36 (m, 1H), 4.164.12 (m, 1H), 4.01-3.95 (m, 2H), 3.91-2.85 (m, 1H), 1.60 (s, 3H), 1.40 + 1.38 + 1.36 (3 × s, 9H). 13C-NMR (CDCl3) δ 137.5, 128.5, 128.3, 128.0, 112.9, 109.7, 103.9, 78.0, 77.8, 77.4, 74.7, 72.1, 64.9, 26.7, 26.4, 26.0, 25.0. MALDI-MS m/z: 351.2 [M+H]+. S.5: 1H-NMR (CDCl3) δ 7.37-7.26 (m, 5H), 5.76 (d, J = 3.9 Hz, 1H), 4.80 (d, J = 11.7 Hz, 1H), 4.63 (t, J = 4.0 Hz, 1H), 4.56 (d, J = 11.7 Hz, 1H), 4.20 (d, J = 5.2 Hz, 1H), 3.93 (d, J = 12.0 Hz, 1H), 3.85 (d, J = 12.0 Hz, 1H), 3.78 (d, J = 11.9 Hz, 1H), 3.54 (d, J = 12.0 Hz, 1H), 2.50 (br, 1H), 2.32 (br, 1H), 1.63 (s, 3H), 1.33 (s, 3H). 13C-NMR (CDCl3) δ 137.2, 128.4, 128.1, 127.7, 113.4, 104.3, 86.3, 78.3, 78.2,



Supplement 8

72.6, 64.0, 63.0, 26.5, 25.8. MALDI-MS m/z: 311.1 [M+H]+. S.6: 1H-NMR (CDCl3) δ 7.38-7.26 (m, 5H), 5.73 (d, J = 3.8 Hz, 1H), 4.85 (d, J = 12.0 Hz, 1H), 4.75 (d, J = 11.6 Hz, 1H), 4.64 (t, J = 3.8 Hz, 1H), 4.55 (d, J = 11.6 Hz, 1H), 4.39 (d, J = 12.0 Hz, 1H), 4.30 (d, J = 11.0 Hz, 1H), 4.17 (d, J = 5.2 Hz, 1H), 4.12 (d, J = 11.0 Hz, 1H), 3.05 (s, 3H), 2.95 (s, 3H), 1.66 (s, 3H), 1.32 (s, 3H). 13C-NMR (CDCl3) δ 136.5, 128.4, 128.2, 127.9, 113.8, 104.3, 83.0, 78.2, 77.6, 72.6, 69.3, 68.5, 37.8, 37.2, 26.0, 25.4. MALDI-MS m/z: 467.1 [M+H]+. S.7 (main isomer): 1H-NMR (CDCl3) δ 7.37-7.28 (m, 5H), 6.17 (s, 1H), 5.37 (d, J = 4.8 Hz, 1H), 4.62 (d, J = 11.1 Hz, 1H), 4.52 (d, J = 11.1 Hz, 1H), 4.50 (d, J = 11.6 Hz, 1H), 4.42 (d, J = 4.8 Hz, 1H), 4.37 (d, J = 11.6 Hz, 1H), 4.30 (d, J = 10.6 Hz, 1H), 4.19 (d, J = 10.6 Hz, 1H), 3.01 (s, 6H), 2.14 (s, 3H), 2.10 (s, 3H). 13C-NMR (CDCl ) δ 169.1, 168.7, 136.3, 3 128.5, 128.3, 128.1, 97.2, 82.7, 78.6, 73.9, 73.1, 68.8, 68.4, 37.5, 37.3, 20.9, 20.5. MALDI-MS m/z: 511.1 [M+H]+. S.11: 1H-NMR (CDCl3) δ 9.28 (br, 1H), 7.45(d, J = 1.1 Hz, 1H), 7.38-7.22 (m, 5H), 5.66 (s, 1H), 4.67 (d, J = 11.6 Hz, 1H), 4.56 (d, J = 11.7 Hz, 1H), 4.54 (s, 1H), 4.05 (d, J = 7.9 Hz, 1H), 4.01 (d, J = 12.5 Hz, 1H), 3.96 (s, 1H), 3.95 (d, J = 12.6 Hz, 1H), 3.83 (d, J = 7.9 Hz, 1H), 1.88 (s, 3H). 13C-NMR (CDCl3) δ 163.9, 149.8, 137.0, 134.7, 128.5, 128.2, 127.8, 110.3, 88.2, 87.3, 76.9, 75.9, 72.3, 72.0, 57.6, 12.7. MALDI-MS m/z: 360.1 [M+H]+. S.12: 1H-NMR (DMSO-d6) δ 11.33 (br, 1H), 7.60 (d, J = 1.1 Hz, 1H), 5.68 (d, J = 4.1 Hz, 1H), 5.38 (s, 1H), 5.20 (br t, J = 5.6 Hz, 1H), 4.09 (s, 1H), 3.89 (d, J = 4.0 Hz, 1H), 3.80 (s, J = 7.8 Hz, 1H), 3.74 (d, J = 5.5 Hz, 2H), 3.61 (d, J = 7.8 Hz, 1H), 1.75 (d, J = 1.1 Hz, 3H). 13C-NMR (DMSO-d6) δ 164.1, 150.1, 135.1, 108.6, 89.0, 86.5, 79.1, 71.2, 68.9, 56.2, 12.6. Anal. calcd. for C11H14N2O6⋅2⁄3H2O: C 46.81, H 5.48, N 9.92; found: C 46.64, H 5.22, N 10.05. MALDI-MS m/z: 270.9 [M+H]+.

Critical Parameters and Troubleshooting


In general, the presented synthesis of LNAT has been developed to be tolerant to small changes in reaction conditions. Nonetheless, in some steps it is necessary to take precautions to ensure anhydrous conditions. When necessary, it was found sufficient to use solvents that were stored over molecular sieves. It is advisable to check all reactions by TLC (APPENDIX 3D; see Table 4.12.1 for Rf values)

before workup to ensure that complete conversion of starting material has been obtained. If the reaction has not gone to completion, a prolonged reaction time and/or additional reagent (typically 10% excess) will typically bring the reaction to completion. In Basic Protocol 1, the development of hydrogen gas demands a steady flow of inert atmosphere through the reaction flask. If such precaution is not taken, build up of hydrogen gas could potentially result in explosion of the reaction mixture. In Basic Protocol 2, steps 3 and 4, it is important to avoid temperatures higher than 45°C during the evaporation of acetic acid. If the temperature gets too high, the 1,2-O-isopropylidene group will be lost. A small amount of product that has lost the 1,2-O-isopropylidene group is acceptable in the following steps, but will lower the overall yield of the reaction. In steps 5 to 7, all the starting 5,6-diol S.4 must be consumed since unreacted starting material will be unaffected in subsequent reactions and is difficult to separate from the product during isolation of S.6. In Basic Protocols 3 and 4, the use of anhydrous reaction conditions is necessary to obtain a complete reaction. The use of anhydrous reaction conditions is essential during the glycosylation of S.7 with silylated nucleobases (see Basic Protocol 5, steps 1 to 5), since water will h yd ro lyze th e tr imethy lsilyl trifluoromethanesulfonate. It is also crucial to have anhydrous conditions during the substitution of the methanesulfonyloxy group with sodium benzoate (Basic Protocol 5, steps 17 to 18). It has been observed that even small amounts of water in the reaction mixture can not only slow down the reaction, but result in hydrolysis of the formed benzoate giving S.11, which will not be recovered during the precipitation step in the work-up procedure. If hydrolysis has occurred during the reaction, it is possible to extract S.11 from the aqueous phase by extraction with ethyl acetate. In Basic Protocol 5, step 34, it is advisable to keep the reaction at a temperature just below the boiling point. It has been observed that when the reaction mixture reaches the boiling point, large amounts of salt precipitate in the condenser.

Anticipated Results The overall yield for synthesis of LNA-T diol is 50% (eleven steps). Expected yields for isolated intermediates are S.2, 90%; S.5, 82%; S.6, 90%; S.7, 100%; S.11, 79%; and S.12,



95%. The scale of the synthesis can be increased to 1 mol for Basic Protocols 1 to 3 without any major problems. The synthesis was developed to avoid chromatographic purification. The yields can be increased slightly by purifying the mother liquors (e.g., by column chromatography); however, this is time consuming.

Time Considerations The time needed to complete the synthesis of LNA-T diol is estimated to be 10 to 14 days. Most of the reactions are not affected by a prolonged reaction time. Thus, one can leave the reaction mixtures overnight at room temperature. An exception is the removal of the 5,6-O-isopropylidene group from S.2 (see Basic Protocol 2, step 2), where extended exposure to acetic acid will result in loss of the 1,2-O-isopropylidene group. All isolated compounds are stable and can be stored without special precautions.

Literature Cited Altona, C. and Sundaralingam, M. 1973. Conformational analysis of the sugar ring nucleosides and nucleotides. Improved method for the interpretation of proton magnetic resonance coupling constants. J. Am. Chem. Soc. 95:2333-2344. Brimacombe, J.S. and Ching, O.A. 1968. Nucleophilic displacement reactions in carbohydrates. Carbohydr. Res. 8:82-88. Caruthers, M.H. 1991. Chemical synthesis of DNA and DNA analogues. Acc. Chem. Res. 24:278284. Choleva, Y., Nørholm, M., Pedersen, S., Mouritzen, P., Høiby, P.E., Nielsen, A.T., Møller, S., Jakobsen, M.H., and Kongsbak, L. 2001. Multiplex SNP genotyping using locked nucleic acids and microfluidics. J. Assoc. Lab. Automation 6:9297. Codington, J.F., Fecher, R., and Fox, J.J. 1960. Pyrimidine nucleosides. VII. Reactions of 2′,3′,5′-trimesyloxyuridine. J. Am. Chem. Soc. 82:2794-2803. Horton, D. and Tindall, C.G. Jr. 1970. Methyleneinsertion reactions with unsaturated sugars, synthesis of 4-C-cyclopropyl-D-ribo-tetrafuranose derivatives. Carbohydr. Res. 15:215-232. Jones, G.H., Taniguchi, M., Tegg, D., and Moffat, J.G. 1979. 4′-Substituted nucleosides. 5. Hydroxymethylation of nucleoside 5′-aldehydes. J. Org. Chem. 44:1309-1317. Koshkin, A.A., Rajwanshi, V.K., and Wengel, J. 1998a. Novel convenient syntheses of LNA [2.2.1]bicyclo nucleosides. Tetrahedron Lett. 39:4381-4384.

Koshkin, A.A., Singh, S.K., Nielsen, P., Rajwanshi, V.K., Kumar, R., Meldgaard, M., Olsen, C.E., and Wengel, J. 1998b. LNA (locked nucleic acids): Synthesis of the adenine, cytosine, guanine, 5-methylcytosine, thymine and uracil bicyclonucleoside monomers, oligomerisation, and unprecedented nucleic acid recognition. Tetrahedron 54:3607-3630. Koshkin, A.A., Fensholdt, J., Pfundheller, H.M., and Lomholt, C. 2001. A simplified and efficient route to 2′-O, 4′-C-methylene-linked bicyclic ribonucleosides (LNA). J. Org. Chem. 66:85048512. Leland, D.L. and Kotick, M.P. 1974. Studies on 4-C-(hydroxymethyl)pentofuranoses. Synthesis of 9-[4-C-(hydroxymethyl)-α-L-threo-pentofuranosyl]adenine. Carbohydr. Res. 38:C9-C11. Obika, S., Nanbu, D., Hari, Y., Morio, K., In, Y., Ishida, T., and Imanishi, T. 1997. Synthesis of 2′- O, 4′-C-methyleneuridine and -cytidine. Novel bicyclic nucleosides having a fixed C3′endo sugar puckering. Tetrahedron Lett. 38:8735-8738. Obika, S., Nanbu, D., Hari, Y., Andoh, J., Morio, K., Doi, T., and Imanishi, T. 1998. Stability and structural features of the duplexes containing nucleoside analogues with a fixed N-type conf o rm ation, 2′-O,4′-C-methyleneribonucleosides. Tetrahedron Lett. 39:5401-5404. Obika, S., Uneda, T., Sugimoto, T., Nanbu, D., Minami, T., Doi, T., and Imanishi, T. 2001. 2′-O,4′C-Methylene bridged nucleic acid (2′,4′-BNA): Synthesis and triplex-forming properties. Bioorg. Med. Chem. 9:1001-1011. Ørum, H., Jakobsen, M.H., Koch, T., Vuust, J., and Borre, M.B. 1999. Detection of the factor V Leiden mutation by direct allele-specific hybridization of PCR amplicons to photoimmobilized locked nucleic acids. Clin. Chem. 45:1898-1905. Singh, S.K., Nielsen, P., Koshkin, A.A., and Wengel, J. 1998. LNA (locked nucleic acids): Synthesis and high-affinity nucleic acid recognition. Chem. Commun. (1998):455-456. Sinha, N.D., Biernat, J., and Köster, H. 1983. β-Cyanoethyl N,N-dialkylamino/N-morpholinomonochloro phosphoramidites, new phosphitylating agents facilitating ease of deprotection and work-up of synthesized oligonucleotides. Tetrahedron Lett. 24:5843-5846. Sowa, W. and Thomas, G.H.S. 1966. The oxidation of 1,2:5,6-di-O-isopropylidene-D-glucose by dimethylsulfoxide-acetic anhydride. Can. J. Chem. 44:836-838. Vorbrüggen, H. and Höfle, G. 1981. On the mechanism of nucleoside synthesis. Chem. Ber. 114:1256-1268. Vorbrüggen, H., Krolikiewicz, K., and Bennua, B. 1981. Nucleoside synthesis with trimethylsilyl triflate and perchlorate as catalysts. Chem. Ber. 114:1234-1255.



Supplement 8

Waga, T., Nishizaki, T., Miyakawa, I., Ohrui, H., and Meguro, H. 1993. Synthesis of 4′-C-methylnucleosides. Biosci. Biotech. Biochem. 57:14331438. Wahlestedt, C., Salmi, P., Good, L., Kela, J., Johnsson, T., Høkfelt, T., Broberger, C., Porreca, F., Lai, J., Ren, K., Ossipov, M., Koshkin, A., Jakobsen, N., Skouv, J., Oerum, H., Jacobsen, M.H., and Wengel, J. 2000. Potent and nontoxic antisense oligonucleotides containing locked nucleic acids. Proc. Natl. Acad. Sci. U.S.A. 97:5633-5638. Youssefyeh, R., Tegg, D., Verheyden, J.P.H., Jones, G.H., and Moffat, J.G. 1977. Synthetic routes to 4′-hydroxymethylnucleosides. Tetrahedron Lett. 5:435-438. Youssefyeh, R.D, Verheyden, J.P.H., and Moffat, J.G. 1979. 4′-Substituted nucleosides. 4. Synthesis of some 4′-hydroxymethyl nucleosides. J. Org. Chem. 44:1301-1309.

Key References Koshkin et al., 1998a,b. See above. This unit’s protocols were developed based on the synthetic results described in these papers. Wengel, J. 1999. Development of locked nucleic acid. Acc. Chem. Res. 32:301-310. This article describes the development of LNA.

Contributed by Henrik M. Pfundheller and Christian Lomholt Exiqon A/S Vedbaek, Denmark




Cellular Delivery of Locked Nucleic Acids (LNAs)

UNIT 4.13

This unit describes the introduction of locked nucleic acid (LNA) oligomers (Fig. 4.13.1; Koshkin et al., 1998; Obika et al., 1998; Wang et al., 1999; reviewed in Braasch and Corey, 2001) into cells. It is intended to extend the discussion of the synthesis and characterization of LNA that is found in UNIT 4.12. INTRODUCTION OF LNA OLIGOMERS INTO CELLS As is the case for most other types of oligonucleotide, there is little reason to believe that LNA oligomers will be able to spontaneously enter most types of cultured cells and locate a cellular target. Since LNAs possess a negatively charged backbone, one simple method for promoting uptake is the use of cationic lipid. A step-wise sample procedure for a 48-well plate format is given below and a schematic summary is provided in Figure 4.13.2.

BASIC PROTOCOL

NOTE: The conditions required for successful transfections will vary from one cell line to the next. It is impossible to predict which combination of lipid and oligonucleotide will be most effective; this must be determined empirically for each cell line. Materials Cells grown to confluence in 75-cm2 tissue culture flasks Complete growth medium (see recipe) 100 µM LNA stock solution (see Support Protocol 1) Opti-MEM I (Invitrogen Life Technologies; reduced-serum medium, containing L-glutamine and no phenol red) LipofectAMINE (Invitrogen Life Technologies) 48-well tissue culture plate (Costar) Repeating pipettor (e.g., Eppendorf) 12.5-mL Combitips (Eppendorf) 65° and 37°C water baths or a thermal cycler 12 × 75–mm round-bottom tubes 37°C, 5% CO2 incubator

base

RO O

O O

P

O O– base

O O

Figure 4.13.1 Structure of LNA.


Contributed by Dwaine A. Braasch and David R. Corey

4.13.1

R′O

Current Protocols in Nucleic Acid Chemistry (2002) 4.13.1-4.13.9 Copyright © 2002 by John Wiley & Sons, Inc.

O

Supplement 9

Additional reagents and equipment for trypsinizing and counting cells (e.g., CPMB APPENDIX 3F) NOTE: LipofectAMINE and Opti-MEM I are important to the success of the experiment and should not be substituted. Prepare cells 1. Beginning with a 75-cm2 tissue culture flask of cells grown to confluence, trypsinize the cells according to standard procedures (e.g., CPMB APPENDIX 3F). The total number of cells (and thus the number of flasks) will depend on the number of LNAs being transfected. For transfection of a single LNA, 33,000 to 39,000 cells are needed (11,000 to 13,000 cells per well at three different concentrations of LNA-LipofectAMINE complex).

2. Suspend the cells in fresh complete growth medium.

inhibitor (100 µM LNA) 6.4 µL

Opti-MEM 143.6 µL

step 9

1.9 µL

Opti-MEM 148.1 µL

step 10 step 11

LipofectAMINE

total volume =

300 µL

15 min

200 nM, V t = 3200 µL

add 2900 µL Opti-MEM

step 12 complex formation

step 14

100 nM, V t = 3000 µL 1500 µL 200 nM solution step 15 1500 µL Opti-MEM 25 nM, V t = 2000 µL 500 µL 100 nM solution 1500 µL Opti-MEM

dispense 200 µL/well

step 15

step 16

step 17 transfection overnight


Figure 4.13.2 Preparation of LipofectAMINE-LNA complexes for cellular transfection, including subsequent dilutions of stock complexes. Vt = total volume. Volumes are sufficient for dispensing six replicates and have been optimized for COS-7 cells.

4.13.2 Supplement 9


3. Draw up the entire volume of cell suspension into a 10-mL pipet and dispense back into the flask with the tip of the pipet pressed lightly against the bottom of the flask. Aggregates of cells will be disrupted, yielding a single-cell suspension that gives more accurate cell counts.

4. Perform a cell count in triplicate using a Coulter counter or hemacytometer (e.g., CPMB APPENDIX 3F) and average the results. 5. Prepare cell suspension in complete growth medium at a density of 44,000 to 52,000 cells/mL. 6. Plate cells 4 to 6 hr prior to transfection. For each LNA to be transfected, plate 250 µL cell suspension (11,000 to 13,000 cells) in three wells of a 48-well tissue culture plate. For accuracy with larger numbers of LNAs, use a repeating pipettor and a 12.5-mL Combitip.

7. Replace cover on plate and disperse cells evenly within the wells by sliding the plate back and forth, gently bumping it against the front lip of a laminar flow hood work surface. Keep cells 4 to 6 hr at 37°C in a 5% CO2 incubator for cells to attach to the surface. Prepare LNA-LipofectAMINE complexes 8. Warm 100 µM LNA stock solution 5 min at 65°C for a 15-mer or 95°C for a 25-mer to disrupt aggregates, and then maintain at 37°C until transfection. Also warm Opti-MEM I to 37°C. 9. In a 12 × 75–mm round-bottom tube, dilute 6.4 µL of 100 µM LNA with 143.6 µL Opti-MEM I. 10. In a separate 12 × 75–mm round-bottom tube, dilute 1.9 µL LipofectAMINE with 148.1 µL Opti-MEM I. This solution can be scaled up in a single tube depending on how many unique conditions are being tested. If, for example, five inhibitors (LNAs) are to be tested, then it would be advantageous to prepare sufficient lipid mix for seven conditions.

11. Add 150 µL LipofectAMINE from step 10 to each LNA in step 9 (total 300 µL) and tap the tube briskly 15 times to mix the reagents and initiate the formation of LNA-LipofectAMINE complexes. 12. Allow the tube to sit 15 min at room temperature in the dark. Opti-MEM I is light sensitive.

Transfect cells 13. While waiting for complexes to form, aspirate off the complete growth medium in each well of the tissue culture plate and replace with 250 µL Opti-MEM I per well. Also set up two tubes for a dilution series of LNA-LipofectAMINE complexes and add 1.5 mL Opti-MEM I to each. 14. When the incubation (step 12) is complete, add 2.9 mL Opti-MEM I to the LNALipofectAMINE complexes and mix well. This gives 3.2 mL at 200 nM LNA for the starting concentration for the serial dilution. The authors typically use 200 nM or 500 nM as the starting concentration.



Supplement 8

15. Transfer 1.5 mL stock to one of the tubes in step 13 (final 3 mL at 100 nM) and mix well. Transfer 500 µL of this solution to the second tube (final 2 mL at 25 mM) and mix well. If 500 nM is used as the starting concentration, a 200 nM dilution should be included.

16. Aspirate the Opti-MEM I wash from the cells and immediately dispense 200 µL of each LNA-LipofectAMINE dilution to the appropriate wells, working backward through the dilution scheme for a given LNA. 17. Allow cells to incubate overnight at 37°C. 18. Aspirate off transfection solution and replace with 250 µL complete growth medium. 19. Incubate >24 hr at 37°C prior to conducting an assay for the effects of the LNA. When developing a protocol for delivering LNAs into cells, it is useful to obtain a fluorophore-labeled LNA. Delivery of the LNA can be visualized by microscopy, facilitating the evaluation and subsequent optimization of delivery conditions. SUPPORT PROTOCOL 1

PREPARATION OF LNA OLIGOMER STOCK SOLUTIONS LNA oligomers arrive lyophilized and should be handled like DNA or RNA oligomers. Materials Locked nucleic acid oligomers (LNAs; Proligo) DNase/RNase-free water (Life Technologies) Spectrophotometer 1. If LNAs have been refrigerated, allow them to equilibrate to room temperature. 2. Centrifuge the samples 2 min at 14,000 × g, room temperature, to collect LNA at the bottom of the tube. 3. Add DNase/RNase-free water to give a stock solution of ∼1 mM. Allow the oligomer to sit undisturbed for 10 to 15 min at room temperature. The estimated concentration is based on the volume, on the mass reported by the manufacturer, and on the molecular weight of a given LNA.

4. Vortex in 5-sec bursts several times. Heat to 65°C for up to 15 min and cool to room temperature. 5. Allow tubes to sit undisturbed for 5 min, room temperature. 6. Centrifuge 2 min at 14,000 × g, room temperature, to pellet any remaining undissolved material. Observe the tube contents carefully at this point, as occasionally there are insoluble materials that can interfere with cellular assays. It is best not to proceed with cellular assays if an LNA exhibits this behavior unless one has significant experience in desalting and purifying oligomers. Consult the manufacturer if solubility properties are not satisfactory.

7. Remove a 1-µL aliquot and dilute it with 144 µL distilled water. Ascertain the absorbance at 260 nm. 8. Calculate the concentration of the LNA using the following equation, where 33 ng/µL is the extinction coefficient of the LNA and 145 is the dilution factor: Cellular Delivery of Locked Nucleic Acids (LNAs)

c (mM) = (A260 × 33 ng/µL × 145)/mol. wt. of LNA

4.13.4 Supplement 8


The value 33 ng/ìL assumes an average extinction coefficient based on an equal population of all bases. If the LNA contains a preponderance of one or two bases, the equation can be modified (e.g., see CPMB APPENDIX 3D) using the appropriate extinction coefficients. The extinction coefficients of the LNA nucleotide analogs are not available to the authors, but it may be reasonable to assume that they are the same as for DNA nucleotides (Proligo, pers. comm.).

9. Adjust concentration to 100 µM and store the stock solution for up to 1 year at 4°C. DETERMINATION OF Tm FOR LNA OLIGOMERS To understand the potential for LNA oligonucleotides to recognize intracellular targets, it is useful to determine Tm values for LNAs with complementary RNA or DNA oligomers.

SUPPORT PROTOCOL 2

Materials LNA oligonucleotides DNA or RNA oligomers 10× Ca2+- and Mg2+-free phosphate-buffered saline (CMF-PBS; Invitrogen Life Technologies or see recipe) 0.1 M Na2HPO4 buffer, pH 7.5 (Fisher) Mineral oil (Sigma) Stoppered cuvette (1-cm pathlength and 1.5-cm Z dimension; Spectrosil Far UV Quartz, Uvonic Instruments) Spectrophotometer with temperature-controlled cuvette holder 1. Calculate the concentration of each of the single-stranded components (LNA and either DNA or RNA). The concentration of LNAs can be determined from the A260 value of a diluted aliquot (see Support Protocol 1, step 8). It is sometimes useful to heat the LNAs to 65°C or higher, depending on the oligomer length and the number of LNA bases, to break up intra- and intermolecular hydrogen bonding and aggregation. When preparing oligonucleotide pairs for Tm analysis or tranfection, it is prudent to heat the samples and ensure that the concentrations are accurate.

2. Prepare a small volume (∼20 to 30 µL) of 100 µM heteroduplex nucleic acid in a solution containing 2.5× CMF-PBS final concentration. 3. Dilute a 5-µL aliquot of the 100 µM heteroduplex mixture with 145 µL of 0.1 M Na2HPO4 buffer, pH 7.5, in a 1-cm-pathlength stoppered cuvette. 4. Overlay this solution with 145 µL mineral oil to minimize evaporation. 5. Monitor the change in the absorbance at 260 nm every 5°C as the temperature is ramped from 100° to 12°C and also back up to 100°C. 6. Fit the data collected from these analyses using van’t Hoff thermal denaturation/renaturation curve analysis to determine the Tm values from the denaturation and renaturation curves (UNIT 7.3). It is not uncommon for entirely LNA oligomers with >11 bases to possess Tm values >95°C.



Supplement 8


CMF-PBS (Ca2+- and Mg2+-free phosphate-buffered saline), 10× 10.4 mM KH2PO4 1551.7 mM NaCl 29.6 mM Na2HPO4⋅7H2O, pH 7.4 Store up to 1 year at 4°C Complete growth medium Dulbecco’s modified Eagle’s medium (DMEM) high-glucose without L-glutamine (e.g., Mediatech Cellgro, Fisher) containing: 20 mM HEPES buffer, pH 7.4 (cell-culture grade, Sigma) 10% (w/v) FBS (Atlanta Biologicals) 1× PSF (see recipe) 0.7 mg/mL tylosin (Sigma) 2 mM L-glutamine (Invitrogen Life Technologies) Store up to 4 months at 4°C PSF (penicillin/streptomycin/Fungizone), 100× 10,000 U penicillin 10 mg streptomycin 25 µg/mL amphotericin B (e.g., Fungizone, Invitrogen Life Technologies) Store up to 1 year at −20°C COMMENTARY Background Information LNAs offer several advantages for nucleic acid recognition (Table 4.13.1), which should encourage investigators to consider their use. Currently, oligonucleotides that contain LNA bases can be obtained commercially from Proligo (http://www.proligo.com). Synthesis is based on phosphoramidite chemistry and employs LNA monomers of A, T, G, and 5-methylC. The most striking advantage conferred by use of LNA bases is a dramatic increase in the

Table 4.13.1


affinity of binding to complementary sequences (Table 4.13.2). A single LNA base can increase the melting temperature (Tm) of binding by 10°C, and oligomers that contain several strategically positioned LNA bases can bind with even higher affinity than analogous peptide nucleic acid (PNA) oligomers (Braasch and Elayadi, unpub. observ.). LNA bases are introduced into oligonucleotides by standard synthesis methods (UNIT 4.12), allowing LNA bases to be interspersed among DNA or RNA bases. As a result, important properties such as

Advantages of Locked Nucleic Acids

High affinity hybridization

Tight binding by short LNAs or at high temperatures

Synthesized like DNA/RNA

Enables ready adaptation of existing synthesizers. Simple to intersperse DNA or RNA bases to modulate Tm values or RNase H sensitivity.

Negatively charged backbone

Good solubility; investigators who work with DNA or RNA will find LNA easy to work with. Ability to incorporate phosphorothioate linkages to improve stability or pharmacokinetic properties.

4.13.6 Supplement 8


Table 4.13.2 Melting Temperature (Tm) Values for LNAs and LNA-DNA Hybrids

LNA or LNA-DNAa

Tm

Tm(ref)b ∆Tm

∆Tm/LNA Reference base

GTGTTTTGC GTGTCCGAGACGTTG

52 72

28 59

24 13

5 1.5

Kumar et al. (1998) Wahlstedt et al. (2000)

GTGTCCGAGACGTTG GTGTCCGAGACGTTG

83 >90

59 59

24 >31

3 2

Wahlstedt et al. (2000) Wahlstedt et al. (2000)

CACTATACG CTGATATGC

40 36.8

29 27.2

11 9.6

3.3 9.6

Koshkin et al. (1998) Bondensgaard et al. (2000)

CTGATATGC AGGGTCGCTmeCGGTGT

51.6 >96

27.2 53

24.4 43

8.1 3

Bondensgaard et al. (2000) Braasch (unpub. observ.)

AGGGTCGCTmeCAATGT meCAGTTAGGGTTAG

83 81

NPc 50

— 31

— 3.1

Braasch (unpub. observ.) Braasch (unpub. observ.)

meCAGTTAGAATTAG

TAGGGT

65 56

NPc NDc

— —

— —


TAGGGTTA AGGATmeCTAGGTGAA

74 >96

22 53

52 25

6.5 2.9


AGGATmeCTAGG AGGATmeCTAGGTGAA

73 59

39 53

34 6

3.4 0.6


aUnderlined bases are LNA. meC, 5-methylcytosine. All oligomers are shown from 5′ to 3′. bReference T values are for analogous DNA oligonucleotides. m cND, melting temperature not detected; NP, analysis not performed. ∆T and ∆T /LNA base could thus not be calculated. m m

Tm or RNase H activation can be tailored to meet the specifications of individual applications. This combination of high-affinity hybridization with standard synthesis protocols is powerful because it encourages adapting LNAs to existing protocols that use oligonucleotides.

Critical Parameters When designing oligonucleotides that are intended to function inside cells, the ability of the oligonucleotides to form duplexes that act as substrates for RNase H is a primary consideration (Crooke, 1999). RNase H degrades RNA-DNA hybrids, allowing oligonucleotides that contain DNA to promote the cleavage of mRNA. If antisense inhibition of gene expression is desired, this can be an advantage since the target mRNA is permanently inactivated, and the antisense oligomer can then move on to inactivate additional mRNA molecules. Oligonucleotides that cannot recruit RNase H (i.e., that do not contain DNA portions) can block the binding of the translation apparatus when targeted to the 5′ terminus of the untranslated region (Baker et al., 1997; Doyle et al., 2001). Oligomers that do not activate RNase H can also redirect splicing when targeted to splice

sites (Kang et al., 1998), but when they are targeted to other mRNA sequences, they are likely to be displaced by the ribosome. Oligomers that contain only LNA bases activate RNase H poorly (Wahlestedt et al., 2000). However, because LNAs are made using protocols similar to those used for DNA synthesis, it is straightforward to incorporate DNA bases into LNA-DNA chimera. This provides the experimenter with the choice of whether or not to incorporate RNase H sensitivity into oligonucleotide design by including a contiguous run of at least six DNA bases. If antisense gene inhibition is desired, this strategy may allow a wider range of sequences to be targeted. If simple steric blocking of the RNA target is required, the potential to direct RNase H cleavage is unnecessary and even counterproductive because it might lead to unintentional destruction of nontargeted RNA substrates. LNA bases confer some increase in the stability of oligomers to degradation by nucleases. However, to achieve maximal stability in animal studies, it is likely that one or two phosphorothioate (PS) linkages will need to be substituted at both the 3′- and 5′-terminal linkages. As noted above, LNA synthesis is similar to the



Supplement 8

synthesis of DNA or RNA, allowing PS linkages to be added routinely (Kumar et al., 1998). Complete substitution of phosphodiester linkages with PS linkages has also been noted to improve the pharmacokinetic properties of antisense oligonucleotides (Geary et al., 2001). It is reasonable to believe that LNA-containing oligomers will also need to be modified with PS linkages to achieve in vivo efficacy, though animal studies will be necessary to establish if this truly is the case for LNA-containing chimeric oligonucleotides. The choice of whether or not to exploit the potential for RNase H activation influences how antisense activity will be examined on a case-by-case basis. LNA oligomers that contain DNA segments that can activate RNase H will cause RNA to be degraded, allowing efficacy to be judged by northern analysis. LNA oligomers that cannot activate RNase H can be evaluated by examining the expression of the protein target or by measuring its activity. As with any antisense experiment, use of control oligonucleotides that contain mismatched bases is necessary to support the belief that an effect is specific, i.e., due to binding to the intended mRNA target.

Troubleshooting


When oligonucleotides are introduced into cells, they may prove to be toxic. This toxicity could be due to successful inhibition of the target gene function. However, toxicity could also be due to (1) binding to one or more nontarget proteins, (2) hybridization to one or more nontarget nucleic acid sequences, or (3) poisoning of the cells by small molecule impurities or endotoxins. The primary consideration is that the observed effects should be consistent with the biology of the system being examined. If cell death is observed, LNAs can be purified by desalting to remove small molecule contaminants. Alternatively, a fresh synthesis of LNA can be performed to determine if newly made material behaves similarly. Toxicity could also be caused by improper choice of transfection conditions, reagent concentrations, cell line, or lipids. The window between the conditions that produce optimal LNA delivery and those that cause cells to die is likely to be small. Use of fluorophore-labeled LNAs provides a convenient method for evaluating the success of a given protocol. The authors have also observed that some syntheses of LNAs are relatively insoluble. If this occurs, the experimenter should consult with the manufacturer.

Anticipated Results There are many effective antisense oligonucleotides that do not contain LNA bases. Why use LNA? Why not stay with standard oligonucleotide designs? The use of LNAs is only three years old, and there are no definitive answers to these questions. However, it is reasonable to speculate that the ability of LNA bases to dramatically improve the affinity of binding might increase the potency, specificity, and predictability of antisense action. The resulting increase in efficacy would allow “knock down” phenotypes to be generated more easily, and the fact that LNA bases can be easily incorporated into oligonucleotides allows this hypothesis to be readily tested. Microscopic examination of the delivery of a fluorophore-labeled LNA is a useful demonstration that transfection conditions are promoting LNA uptake by cells. In the authors’ experience, micrographs show that LNA is distributed throughout the cytoplasm and nucleus, with some punctate staining indicating areas of high concentration.

Time Considerations LNAs normally can be obtained within 2 weeks of placing an order with Proligo. LNAs should completely dissolve in water within 20 min. Quantification of LNA concentration by UV spectrophotometry and determination of a Tm value should take ∼2 hr. Transfection of LNA into cells should require an additional 2 to 4 hr. The time required for observation of a phenotype will vary from hours to days, depending on the gene being targeted.

LITERATURE CITED Baker, B.F., Lot, S.S., Condon, T.P., Cheng-Flournoy, S., Lesnik, E.A., Sasmor, H.M., and Bennett, C.F. 1997. 2′-O-(2-Methoxy)ethyl-modified anti-intercellular adhesion molecule 1 (ICAM-1) oligonucleotides selectively increase the ICAM-1 mRNA level and inhibit formation of the ICAM-1 translation initiation complex in human umbilical vein endothelial cells. J. Biol. Chem. 272:11994-12000. Bondensgaard, K., Petersen, M., Singh, S.K., Rajwanshi, V.K., Kumar, R., Wengel, J., and Jacobsen, J.P. 2000. Structural studies of LNA:RNA duplexes by NMR: Conformations and RNase H activity. Chem. Eur. J. 6:2687-2695. Braasch, D.A. and Corey, D.R. 2001. Locked nucleic acids: Fine-tuning nucleic acid recognition. Chem. Biol. 8:1-7. Crooke, S.T. 1999. Molecular mechanisms of antisense drugs: Human RNase H. Antisense Nucl. Acid Drug Devel. 9:377-379.

4.13.8 Supplement 8


Doyle, D.F., Braasch, D.A., Simmons, C.G., Janowski, B.A., and Corey, D.R. 2001. Inhibition of gene expression inside cells by peptide nucleic acids: Effect of mRNA target sequence, mismatched bases, and PNA length. Biochemistry 40:53-64.

Obika, S., Nanbu, D., Hari, Y., Andoh, J., Morio, K., Doi, T., and Imanishi, T. 1998. Stability and structural features of the duplexes containing the nucleoside analogues with a fixed N-type conf o rmation, 2′-O,4′-C-methyleneribonucleosides. Tetrahedron Lett. 39:5401-5404.

Geary, R.S., Yu, R.Z., and Levin, A.A. 2001. Pharmacokinetics of phosphorothioate antisense oligonucleotides. Curr. Opin. Investigational New Drugs 2:562-573.

Wahlestedt, C., Salmi, P., Good, L., Kela, J., Johnsson, T., Hokfelt, T., Broberger, C., Porreca, F., Lai, J., Ren, K., Ossipov, M., Koshkin, A., Jakobsen, N., Skouv, J., Oerum, H., Havsteen Jacobsen, M., and Wengel, J. 2000. Potent and nontoxic antisense oligonucleotides containing locked nucleic acids. Proc. Natl. Acad. Sci. U.S.A. 97:5633-5638.

Kang, S.H., Cho, M.J., and Kole, R. 1998. Up-regulation of luciferase gene expression with antisense oligonucleotides—Implications and applications in functional assay developments. Biochemistry 37:6235-6239. Koshkin, A.A., Singh, S.K., Nielsen, P., Rajwanshi, V.K., Kumar, R., Meldgaard, M., Olsen, C.E., and Wengel, J. 1998. LNA (locked nucleic acids): Synthesis of the adenine, cytosine, guanine, 5-methylcytosine, thymine, and uracil bicyclonucleoside monomers, oligomerisation and unprecedented nucleic acid recognition. Tetrahedon 54:3607-3630. Kumar, R., Singh, S., Koshkin, A.A., Rajwanshi, V.K., Meldgaard, M. and Wengel, J. 1998. The first analogues of LNA (locked nucleic acids): Pho sphorothioate-LNA and 2′-thio-LNA. Bioorg. Med. Chem. Lett. 8:2219-2222.

Wang, G., Gunic, E., Girardet, J-L., and Stoisavljevic, V. 1999. Conformationally locked nucleosides. Synthesis and hybridization properties of oligodeoxynucleotides containing 2′4′-Cbridged 2′-deoxynucleosides. Bioorg. Med. Chem. Lett. 9:1147-1150.

Contributed by Dwaine A. Braasch and David R. Corey University of Texas Southwestern Medical Center at Dallas Dallas, Texas



Supplement 8

Solid-Phase Synthesis of Branched Oligonucleotides

UNIT 4.14

This unit describes the synthesis of nucleic acids containing vicinal 2′,5′- and 3′,5′-phosphodiester bonds. These molecules occur in the cell nucleus, and are formed during the splicing of precursor messenger RNA (pre-mRNA). As such they have many potential applications in nucleic acid biochemistry, particularly as tools for probing the substrate specificity of lariat debranching enzymes, and as tools for studying pre-mRNA splicing (e.g., Nam et al., 1994; Carriero et al., 2001). The assembly of these branched nucleic acids (bNAs) on a solid support can be achieved by following two strategies (Damha and Zabarylo, 1989; Braich and Damha, 1997). The first, referred to as the convergent strategy, is based on well-established automated phosphoramidite chemistry (UNITS 3.3 & 3.5). This method uses a ribonucleoside bisphosphoramidite as the branch-introduction synthon (see Basic Protocol 1). The branching reagent serves to couple together solid-support-bound chains, thus forming a branch juncture with the desired vicinal 2′,5′- and 3′,5′-phosphodiester bonds (see Basic Protocol 2). For efficient branching to occur, CPG supports with high nucleoside loadings are used (see Support Protocol 1). With this approach, Y- and V-shaped molecules having identical 2′ and 3′ chains are readily assembled. The second method is a divergent approach that permits the regiospecific synthesis of bNAs using readily available phosphoramidite reagents (see Basic Protocol 3). An important feature of this method is the assembly of a linear DNA:RNA chimera containing a single 2′-O-silylribonucleoside residue in the middle of the chain. Subsequent removal of the 2-cyanoethyl and silyl protecting groups without detaching the nascent oligonucleotide from the solid support is another salient feature of this approach. This releases an internal 2′-OH group from which orthogonal synthesis of a branch can be carried out. This unit also describes methods used in the authors’ laboratory for the deprotection (see Support Protocol 2), purification, and characterization of branched oligonucleotides. Preferred methods for purification of bNAs are anion-exchange HPLC (see Support Protocol 3) and polyacrylamide gel electrophoresis (see Support Protocol 4). The branched nature of the molecule is confirmed by enzymatic hydrolysis of the bNA to its constituent nucleosides using nuclease P1 (see Support Protocol 3). Further characterization may be conducted via nucleoside composition analysis using snake venom phosphodiesterase (UNIT 10.6) or MALDI-TOF-MS (UNIT 10.1). SYNTHESIS AND CHARACTERIZATION OF THE ADENOSINE BRANCHING SYNTHON N6-BENZOYL-5′-O-(4,4′-DIMETHOXYTRITYL)ADENOSINE-2′,3′-BIS-O-(2-CYANOETHYL-N,N-DIISOPROPYL) PHOSPHORAMIDITE The authors’ group has been predominantly interested in the synthesis of branched RNA fragments related to the lariat intermediates formed during pre-mRNA splicing. Such intermediates contain almost exclusively adenosine at the branch point; therefore, the protocol given below describes the synthesis of the adenosine branching phosphoramidite synthon (BIS-A; S.3; Fig. 4.14.1) used for the synthesis of symmetrical, branched DNA and RNA oligonucleotides (Damha and Ogilvie, 1988). The same protocol may be adapted to the synthesis of the corresponding U, C, and G bisphosphoramidites. The starting protected nucleoside, N6-benzoyl-5′-O-(4,4′-dimethoxytrityl)adenosine (S.1), is commercially available (ChemGenes). Alternatively, it may be synthesized from adenosine using the transient benzoylation procedure of Ti et al. (1982), followed by dimethoxytritylation of the 5′-hydroxyl group (Hakimelahi et al., 1982; Wu et al., 1989). Contributed by Sandra Carriero and Masad J. Damha Current Protocols in Nucleic Acid Chemistry (2002) 4.14.1-4.14.32 Copyright © 2002 by John Wiley & Sons, Inc.

BASIC PROTOCOL 1


4.14.1 Supplement 9

O

[ABz] [DMTr]

Ph

HN O

N

N

OMe Ph

HN N N O

O

N N

N

i -Pr2N

DMTrO

N

O

P OCH2CH2CN

Cl 2

NCCH2CH2O

O

O

P

P

i -Pr2N

i -Pr2N

OCH2CH2CN

DMAP, DIPEA, THF OH

HO OMe

1

3 (4 diastereomers)

Figure 4.14.1 Reaction scheme demonstrating the synthesis of N6-benzoyl-5′-O-(4,4′-dimethoxytrityl)adenosine-2′,3′-Obis-(2-cyanoethyl-N,N-diisopropyl)phosphoramidite (S.3) from the starting protected nucleoside N6-benzoyl-5′-O-(4,4′-dimethoxytrityl)adenosine (S.1). The structure of the chlorophosphoramidite (S.2) is also shown. Abbreviations: Bz, benzoyl; DIPEA, N,N-diisopropylethylamine; DMAP, 4-dimethylaminopyridine; DMTr, 4,4′-dimethoxytrityl.

The synthesis of the branching synthon involves the phosphitylation of the 2′ and 3′ secondary hydroxyls of the ribose sugar (Fig. 4.14.1) using an excess of 2-cyanoethylN,N-diisopropylchlorophosphoramidite (S.2). Reaction conditions, workup, chromatographic purification, and characterization of the product N6-benzoyl-5′-O-(4,4′-dimethoxytrityl)adenosine-2′,3′-bis-O-(2-cyanoethyl-N,N-diisopropyl)phosphoramidite (S.3) are described. Materials N6-Benzoyl-5′-O-(4,4′-dimethoxytrityl)adenosine (S.1; ChemGenes) 4-Dimethylaminopyridine (DMAP; 99%; Aldrich) Nitrogen or argon gas, dry Anhydrous THF (see recipe) in a septum-sealed distillation collection bulb Anhydrous DIPEA (see recipe) 2-Cyanoethyl-N,N-diisopropylchlorophosphoramidite (S.2; ChemGenes) 1:1 (v/v) dichloromethane/diethyl ether 20% (v/v) sulfuric acid (optional) Ethyl acetate prewashed with 5% (w/v) NaHCO3 NaCl solution, saturated Sodium sulfate (Na2SO4), anhydrous 50:47:3 (v/v/v) CH2Cl2/hexanes/triethylamine Silica gel (230- to 400-mesh) in 50:47:3 CH2Cl2/hexanes/triethylamine 95% (v/v) ethanol Diethyl ether


50-mL oven- or flame-dried round-bottom flask with rubber septum Glass syringe and needle, oven dried 2 × 5 cm silica-coated thin-layer chromatography (TLC) plate with fluorescent indicator (e.g., Kieselgel 60 F254 aluminum sheets) 254-nm UV light source 500-mL separatory funnel Gravity filtration device and filter paper

4.14.2 Supplement 9


250- and 500-mL round-bottom flasks Rotary evaporator with a water aspirator 5 × 25–cm glass chromatography column with solvent reservoir bulb Additional reagents and materials for thin-layer chromatography (TLC; APPENDIX 31 3D), column chromatography (APPENDIX 3E), P-NMR (UNIT 7.2), and mass spectrometry (UNITS 10.1 & 10.2) Phosphitylate 2′- and 3′-OH 1. Place 1.3 g (2.0 mmol) N6-benzoyl-5′-O-(4,4′-dimethoxytrityl)adenosine (S.1) and 84 mg (0.7 mmol) DMAP into a 50-mL oven- or flame-dried round-bottom flask. 2. Cap the flask with a rubber septum and purge the flask with dry nitrogen or argon. Care must be taken to avoid the presence of moisture throughout the entire reaction.

3. Withdraw anhydrous THF from a septum-sealed distillation collection bulb using an oven-dried glass syringe and needle. Add 6.0 mL THF to the purged flask with stirring until the starting material is completely dissolved. 4. Add 3.6 mL (21 mmol) anhydrous DIPEA and stir the mixture. 5. Slowly add 1.8 mL (8.3 mmol) 2-cyanoethyl-N,N-diisopropylchlorophosphoramidite (S.2). If the THF solution is sufficiently dry, a white precipitate should form after ∼1 min. This is the diisopropylethylammonium hydrochloride salt that forms during the reaction.

6. Stir the mixture 1 hr at room temperature or until the reaction is complete. Reaction of the secondary hydroxyls of the starting material with S.2 is fast. If TLC analysis (see below) reveals that the reaction is not complete after 1 hr, a further 0.5 mmol S.2 should be added dropwise and stirring continued until all of the starting material is consumed.

Monitor reaction by TLC 7. Spot the reaction mixture onto a precut 2 × 5–cm silica-coated TLC plate with fluorescent indicator and develop using 1:1 (v/v) dichloromethane/diethyl ether (APPENDIX 3D). Addition of 1% to 3% trietylamine may help prevent detritylation.

8. Visualize heterocyclic bases under a 254-nm UV light source. If desired, spray the plate with 20% sulfuric acid in order to visualize the dimethoxytrityl-bearing species. CAUTION: Wear protective eyewear. TLC analysis should indicate complete conversion to products, which exhibit larger Rf values than the starting protected nucleoside. Since the phosphitylation reaction gives rise to two new chiral centers (2′- and 3′-P), the product (S.3) consists of a mixture of four diastereomers. The products appear as two spots (Rf = 0.51 and 0.40) or as one dumbbellshaped spot, because the solvent system partially resolves the four diastereomeric products. A minor side product forms, which migrates between the product (S.3) and starting material (S.1). This is likely the nucleoside-2′,3′-H-bis-phosphonate (Rf = 0.25) that forms via hydrolysis of S.3.

Work up reaction 9. Transfer the reaction mixture to a 500-mL separatory funnel and add 100 mL prewashed ethyl acetate. The ethyl acetate is prewashed with 5% NaHCO3 in order to prevent detritylation and/or activation of the phosphoramidite moiety.

10. Wash the ethyl acetate layer five times each with 100 mL saturated NaCl solution. The diisopropylammonium hydrochloride salt dissolves.



Supplement 9

11. Dry the organic layer over anhydrous Na2SO4. Add more Na2SO4 if the salt crystals clump together upon swirling. When the solution is dry, the nonhydrated crystals will float in solution upon swirling.

12. Gravity filter the resulting solution through filter paper into a 250-mL round-bottom flask. Rinse the Na2SO4 crystals with 10 to 20 mL ethyl acetate. 13. Remove the solvent under reduced pressure (i.e., in a rotary evaporator with a water aspirator) to yield the crude product as a yellow oil. Isolate and characterize product 14. Prepare a 5 × 25–cm glass chromatography column by adding a slurry of 40 g silica gel in 50:47:3 CH2Cl2/hexanes/triethylamine. Precondition with the same solvent. 15. Dissolve the crude material in a minimum amount of 50:47:3 CH2Cl2/hexanes/triethylamine and load on column. Perform chromatography at a rate of ∼1 in. solvent/min using a small amount of air pressure (APPENDIX 3E). Collect product in 10-ml fractions in small test tubes. 16. Combine product-containing fractions into a 500-mL round-bottom flask and concentrate to an oil on a rotary evaporator. 17. Remove residual triethylamine by co-evaporating the oil first with 50 mL of 95% ethanol followed by 50 mL diethyl ether, to provide the pure product as a pale yellow foam. 18. Store bisphosphoramidite at –20°C under an inert atmosphere protected from light. Phosphoramidites are particularly sensitive to UV light; therefore, it is best to store the bisphosphoramidite in a dark bottle (or a bottle covered with aluminum foil) in a –20°C freezer. Under these conditions, the phosphoramidite may be stored for an indefinite period of time. Prior to use, its purity may be verified via TLC analysis. If partial decomposition has occurred, or the coupling reactions with S.3 are poor, the compound should be subjected to chromatography again as described above.

19. Characterize by TLC (APPENDIX 3D), 31P-NMR (UNIT 7.2), and mass spectrometry (UNITS 10.1 & 10.2). 31

P-NMR spectra (400 MHz) of S.3 were measured on a Varian XL-400 spectrometer using CD3CN as the solvent (Fig. 4.14.2). Chemical shifts are reported in parts per million (ppm) and are downfield (positive value) from 85% H3PO4 (external standard). Fast atom bombardment mass spectrometry (FAB-MS) analysis was conducted on a Kratos MS25RFA high-resolution mass spectrometer using a p-nitrobenzyl alcohol (NBA) matrix. N6-Benzoyl-5′-O-(4,4′-dimethoxytrityl)adenosine-2′,3′-bis-O-(2-cyanoethyl-N,N-diisopropylphosphoramidite) (S.3): yield 52% (560 mg); Rf (1:1 CH2Cl2/diethyl ether): 0.51 and 0.40; 31P-NMR (400 MHz, CD3CN): diastereomer 1, 152.3 and 150.7 ppm (5JP-P = 10.1 Hz); diastereomer 2, 151.9 and 150.5 ppm (5JP-P = 6.6 Hz); diastereomer 3, 151.8 and 151.0 ppm (5JP-P = 4.6 Hz); diastereomer 4, 151.3 and 151.2 ppm (5JP-P = 8.9 Hz); FAB-MS anal. calc’d.: 1074.17; observed: 1074.57; [M]+. BASIC PROTOCOL 2


CONVERGENT SYNTHESIS OF SYMMETRICAL BRANCHED NUCLEIC ACIDS The procedure described below for the synthesis of bNAs is carried out on a 1-µmol scale using an ABI 381A DNA synthesizer (Damha and Zabarylo, 1989; Damha et al., 1992). The condensation of two adjacent linear oligonucleotides (prepared from standard RNA or DNA phosphoramidites; Fig. 4.14.3) with the adenosine bisphosphoramidite synthon (S.3; Fig. 4.14.1) produces bNAs that contain identical branches connected via vicinal 2′,5′- and 3′,5′-phosphodiester linkages (Fig. 4.14.4). The same protocol may be utilized for the synthesis of bNAs containing D-xylose or D-arabinose instead of D-ribose at the

4.14.4 Supplement 9


branchpoint, using the appropriate bisphosphoramidite synthons (Damba and Ogilvie, 1988; Noronha, Carriero, Agha, and Damha, unpub. observ.). Branched nucleic acid synthesis works very well on commercially available solid supports (i.e., LCAA-CPG) containing 20 to 40 µmol nucleoside per gram support; however, even better yields are attainable on CPG supports with higher loadings (e.g., 90 µmol/g; Fig. 4.14.5). Such supports can be prepared using HATU/DMAP as the coupling reagents (see Support Protocol 1), and are ideal for the synthesis of short-length bNAs (e.g., trimers) since, in this case, high loadings ensure proper distance between the neighboring CPGbound nucleosides (Damha and Zabarylo, 1989). Materials 5′-O-(4,4′-Dimethoxytrityl)-N-protected-2′-deoxyribonucleoside- or -ribonucleoside-derivatized succinyl-LCAA-CPG (ChemGenes; also see Support Protocol 1) Cap A and B capping reagents (see recipes) DNA and/or RNA 3′-phosphoramidites (S.4a-d and S.5a-d; Fig. 4.14.3)

2 2 1 1

3

3 4

152.5

152.0

151.5

4

151.0

150.5 PPM

Figure 4.14.2 31P-NMR of N6-benzoyl-5′-O-DMTr-adenosine-2′,3′-bis-O-(2-cyanoethyl-N,N-diisopropyl)phosphoramidite (S.3). Due to the two chiral phosphorus centers, the compound exists as a mixture of four diastereomers (22 = 4; Fig. 4.14.1). Each diastereomer displays two sets of phosphorus signals (4 isomers × 2 31P signals = 8 signals). The doubling or splitting of each signal (8 × 2 = 16) is due to long-range coupling between the two chiral 31P atoms (nuclear spin = 1⁄2) of the bisphosphoramidite. The spectrum was recorded on a Varian XL-400 spectrometer using CD3CN as the solvent. The chemical shift and coupling constants for the four diastereomers (shown as numbers near peaks) are as follows: diastereomer 1, 152.3 and 150.7 ppm (5JP-P = 10.1 Hz); diastereomer 2, 151.9 and 150.5 ppm (5JP-P = 6.6 Hz); diastereomer 3, 151.8 and 151.0 ppm (5JP-P = 4.6 Hz); diastereomer 4, 151.3 and 151.2 ppm (5JP-P = 8.9 Hz).



Supplement 9

DNA 3′-phosphoramidites B

DMTrO

RNA 3′-phosphoramidites B

DMTrO

O O

i -Pr2N P

OCH2CH2CN 4a-d

DNA 5′-phosphoramidites OCH2CH2CN

O NCCH2CH2O

O

i -Pr2N

P

B

O

O

O Si

P

DMTrO

i -Pr2N

6a-d

5a-d

4-6a B = N 6-benzoyladenin-9-yl 4-6b B = N 2-isobutyrylguanin-9-yl 4-6c B = N 4-benzoylcytosin-1-yl 4d, 6d B = thymin-1-yl 5d B = uracil

Figure 4.14.3 Chemical structures of DNA and RNA phosphoramidites used for bNA synthesis: 5′-O-DMTr-N-protected-2′-deoxyribonucleoside-3′-O-(2-cyanoethyl-N,N-diisopropylamino)phosphoramidites (S.4a-d), 5′-O-DMTr-N-protected-2′-O-TBDMS-ribonucleoside-3′-O-(2-cyanoethylN,N-diisopropylamino)phosphoramidites (S.5a-d), and the inverted phosphoramidites 3′-O-DMTrN-protected-2′-deoxyribonucleoside-5′-O-(2-cyanoethyl-N,N-diisopropylamino)phosphoramidites (S.6a-d).

N6-Benzoyl-5′-O-(4,4′-dimethoxytrityl)adenosine-2′,3′-bis-O-(2-cyanoethyl-N,Ndiisopropyl)phosphoramidite (BIS-A; S.3; see Basic Protocol 1) Anhydrous acetonitrile (see recipe) Activator solution: 0.5 M 1H-tetrazole (sublimed) in anhydrous acetonitrile Oxidant solution (see recipe) Detritylation solution (see recipe) Nitrogen or argon gas (optional) Synthesis columns for 1-µmol scale synthesis, with seals and filters (PE Biosystems) and 13-mm aluminum seals (Chromatographic Specialties) ABI 381A automated DNA synthesizer (PE Biosystems) Synthesizer bottles for phosphoramidites, oven dried Fraction collector and 15-mL test tubes 50-mL buret Quartz cuvettes UV-Vis spectrophotometer Additional reagents and equipment for oligonucleotide synthesis (APPENDIX 3C), cleaving and deprotecting oligonucleotides (see Support Protocol 2), and anion-exchange HPLC (see Support Protocol 3) or denaturing PAGE (see Support Protocol 4) NOTE: 1H-Tetrazole is no longer commercially available in crystalline form. Solutions of 0.45 M 1H-tetrazole in anhydrous acetonitrile may be purchased from ChemGenes. For a list of supplementary phosphoramidite activating reagents, see UNIT 3.5. Solid-Phase Synthesis of Branched Oligonucleotides

CAUTION: All solutions for the DNA/RNA synthesizer should be prepared in a well-ventilated fume hood.

4.14.6 Supplement 9


Prepare columns for synthesis 1. Transfer an accurately weighed amount of 5′-O-(4,4′-dimethoxytrityl)-N-protected2′-deoxyribonucleoside- or -ribonucleoside-derivatized succinyl-LCAA-CPG for a 1-µmol synthesis to an assembled synthesis column. The support-bound nucleoside represents the first nucleotide at the 3′-end of the oligonucleotide to be synthesized. See Support Protocol 1 for high-loading CPG supports.

NNNNNN 7 1. NH4OH/EtOH 2. TREAT-HF (RNA only)

DMTrON

DMTrON

DMTrON

DNA/RNA synthesis (3′→5′)

DMTrON

HON

HON

HON

HON

N N N N N

N N N N N

N N N N N

N N N N N

trityl off succinyl-LCAA-CPG 3

failure sequences N N N N N N

N N N N N N

HOABz 2′ 3′

HOABz 3′ 2′

N N N N N N

N N N N N N

N N N N N N

trityl off

failure sequences HOABz 3′ 2′

P*

AcON

DNA/RNA synthesis (3′→5′)

N N N N N

trityl off

HOABz 3′ 2′

N N N N N N

N N N N N N

N N N N N N

P*

N N N N N

1. NH4OH/EtOH 2. TREAT-HF (RNA only)

2′

NNNNNN

3′

1. NH4OH/EtOH 2. TREAT-HF (RNA only)

NNNNNN

A

2′

Y-shaped

NNNNNN

A 3′

NNNNNN

V-shaped NNNNNN

10 2′

NNNNNN

A 3′

8

P*

2′

+

AcON

NNNNNN

3′

NNNNNN

11a

11b

NNNNNN

A

2′

P*

2′

+

A P*

3′

NNNNNN

A 3′

NNNNNN 9a

P* 9b

NNNNNN

NNNNNN

7

7

Figure 4.14.4 Schematic representation demonstrating the convergent synthesis of V- and Yshaped bDNA or bRNA oligonucleotides on a solid support. The method can also be used for the synthesis of short branched sequences—e.g., tetranucleoside triphosphates, NpA(2′pN)3′pN— particularly when the nucleoside loading on the solid support approaches 90 µmol/g. Abbreviations: A, bisphosphoramidite; Bz, benzoyl; DMTr, 4,4′-dimethoxytrityl; N, any nucleotide (RNA or DNA); TREAT-HF, triethylammonium trihydrofluoride; S.7, linear oligonucleotide; S.8 and S.10, full-length bNAs (V- and Y-shaped); S.9a-b and S.11a-b, unbranched linear failure sequences. P* indicates the presence of a 2′- or 3′-linked phosphate to the adenosine branch point, which is formed by the hydrolysis and oxidation of the residual phosphoramidite during solid-phase synthesis.



Supplement 9

(i) T10

A

T10

A

XC +

(ii) T10 (iii) T10

BPB

A

T10 T10 P* T10 T10 P*

Relative % oligonucleotide

CPG loading (µmol/g)

80 70 60 50 40 30 20 10 0 20

30

40

50

60

70

80

CPG loading (µmol/g) i: Branched DNA

ii: Extended isomeric DNA

Figure 4.14.5 Effect of CPG loading on yield of the Y-shaped bDNA, T10A2′,5′-T10 3′,5′-T10 synthesized via the convergent strategy (see Basic Protocol 2). (A) PAGE analysis of the amount of Y-shaped product (i; S.10) and failure sequences (ii; S.11a-b) formed as a function of CPG loading on a 20% denaturing polyacrylamide gel. P* indicates the presence of a 2′or 3′-linked phosphate. (B) Chart demonstrating the increase in the amount of S.10 (i) with increasing nucleoside-CPG loading. The chart also demonstrates the inverse relationship between the amount of extended isomeric linear failures (ii; S.11a-b) and nucleoside-CPG loading. The percentage oligonucleotide was determined by integration of the HPLC peak areas of the compounds in question (see Support Protocol 3) using gradient 3 (see Table 4.14.4).

Table 4.14.1 Automated Cycle on an ABI 381A DNA Synthesizer for Capping of Free CPG-Bound Amino and Hydroxyl Groupsa

Synthesis step

Function

Time (sec)

Column washing steps:


1 2 3 4

Acetonitrile to waste Acetonitrile to column Argon reverse flush Argon block flush

Column capping steps: 5 6 7

Cap A + cap B to column Wait Repeat steps 5 and 6

Column washing steps: 8 9 10 11 12

Argon reverse flush Acetonitrile to column Repeat steps 8 and 9 Argon reverse flush Argon block flush

5 60 5 5 15 300

5 30 5 5

aAlternatively, CPG may be capped manually using acetic anhydride (see Support Protocol 1).

4.14.8 Supplement 9


2. Acetylate any underivatized amino and hydroxyl groups on the solid support using an ABI 381A automated DNA synthesizer and the capping cycle given in Table 4.14.1 (APPENDIX 3C). This step also removes traces of water from the solid support. Alternatively, see Support Protocol 1 for the manual capping procedure on nucleosideloaded CPG.

Synthesize branched oligonucleotides 3. Weigh out the appropriate amount of DNA (S.4a-d) and/or RNA (S.5a-d) 3′-phosphoramidites (Fig. 4.14.3) and dilute to the appropriate concentration with anhydrous acetonitrile as indicated in Table 4.14.2. 4. Transfer 100 mg BIS-A (S.3) to an oven-dried synthesizer bottle and dilute to 0.03 M with anhydrous acetonitrile. Low concentrations of BIS-A should be employed in the branching reaction, as high concentrations minimize the yield of fully branched product (S.8 and S.10) and favor the extended isomeric side products (S.9a-b and S.11a-b). It is important to prepare the BIS-A stock solution using ≥100 mg material to ensure that (unavoidable) traces of moisture do not consume significant amounts of BIS-A during the coupling (branching) step and reduce the overall yield of bNA synthesis (Fig. 4.14.6E). When the ABI 381A DNA synthesizer is used, this stock solution can be used for as many as 18 branching reactions (170 ìL/addition). Once bNA synthesis is complete, the bottle containing the BIS-A reagent can be removed from the synthesizer, sealed, purged under an inert atmosphere, and left in a freezer (−20°C) for ∼2 weeks. Alternatively, the stock solution can be evaporated under vacuum, and the solid bisamidite recovered for future use.

5. Place all synthesizer reagents (i.e., activator, capping, oxidant, and detritylation solutions, and acetonitrile) and diluted phosphoramidites (step 3) on the appropriate ports of the synthesizer. 6. Place the BIS-A phosphoramidite bottle on the spare phosphoramidite port (the “X” port on the 381A synthesizer).

Table 4.14.2 Concentrations and Optimal Coupling Times of Phosphoramidites in Synthesis of bNAs on an ABI 381A DNA Synthesizer

Phosphoramiditea

Mol. wt. (g/mol)

Concentration (g/mL)

Coupling time (sec)

DNA phosphoramidites (0.1 M in CH3CN b): 857.7 S.4a, S.6a

86

90

S.4b, S.6b S.4c, S.6c S.4d, S.6d

839.7 833.7 744.6

84 83 74

120 90 90

RNA phosphoramidites (0.15 M in CH3CN): S.5a 988.2 S.5b 970.2 S.5c 964.2 S.5d 861.0

148 146 145 129

600 900 600 600

Bisphosphoramidite (0.03 M in CH3CN): S.3 1074.2

32.2

1800

aPhosphoramidite structures are shown in Figure 4.14.3. bFinal concentration of the first inverted DNA phosphoramidite coupled to the 2′-OH of the rA in divergent and

regiospecific synthesis is 0.3 M (see 5′-pD′′ in Fig. 4.14.8).



Supplement 9

Table 4.14.3 Automated 1-µmol Synthesis Cycle for bDNA and bRNA on an ABI 381A DNA Synthesizer

Synthesis step

Function

Time (sec)

Detritylation of support-bound nucleoside: 1 Acetonitrile to waste 2 Acetonitrile to column 3 Argon reverse flush 4 Argon block flush 5 Advance fraction collector 6 3% TCA to waste 7 3% TCA to column 8 Acetonitrile to column 9 3% TCA to column 10 Argon block flush Column washing steps: 11 Acetonitrile to waste 12 Acetonitrile to column 13 Argon reverse flush 14 Argon block flush 15 Acetonitrile to waste 16 Acetonitrile to column 17 Argon reverse flush 18 Argon block flush Phosphoramidite coupling steps: 19 Phosphoramidite preparation 20 Activator to column 21 Phosphoramidite + activator to column 22 Repeat steps 20 and 21 two times 23 Activator to column 24 Waita 25 Argon reverse flush 26 Argon block flush Column capping steps: 27 Cap A + cap B to column 28 Wait 29 Repeat steps 27 and 28 30 Acetonitrile to waste 31 Argon block flush 32 Acetonitrile to waste 33 Argon reverse flush 34 Argon block flush Oxidation steps: 35 Oxidant to waste 36 Oxidant to column 37 Acetonitrile to waste 38 Argon block flush 39 Wait Column washing steps: 40 Acetonitrile to waste 41 Argon reverse flush 42 Argon block flush 43 Acetonitrile to waste 44 Acetonitrile to column 45 Argon reverse flush 46 Repeat steps 44 and 45 six times 47 Argon block flush Solid-Phase Synthesis of Branched Oligonucleotides

5 45 5 5 1 10 140 30 80 10 5 120 5 5 5 60 5 5 3 5 5 3 5 5 17 45 5 5 5 5 5 5 20 5 5 20 5 10 5 5 18 5 5

aSee Table 4.14.2 for coupling times of various phosphoramidites.



A

B (i) T10 A

GT9 GT9

i

ii

AU T10

(iii) linear failure sequences

0

10

+

(ii) T10

20

A

30

A

P* GT9 GT9

iii

P*

40 Time (min)

50

60

70

80

D

C (i) CCCUACUAA

(iii) linear failure sequences

GUAUGCCC GUAUGCCC

P* CCCUACUAA GUAUGCCC + (ii) GUAUGCCC CCCUACUAA P*

AU

0

10

20

30

40 Time (min)

E i

i

ii

ii

iii

50

60

70

iii

80

Figure 4.14.6 Analysis of bDNA and bRNA molecules synthesized using the convergent strategy (see Basic Protocol 2). (A-B) Analysis of a successful synthesis of the Y-DNA 5′-T10A2′,5′-GT9 3′,5′-GT9 by (A) anion-exchange HPLC (see Support Protocol 3) using gradient 3 (Table 4.14.4) and (B) 20% denaturing PAGE (see Support Protocol 4). (C-E) Analysis of a successful (C-D) and unsuccessful (E) synthesis of the mixed base Y-RNA 5′-CCCUACUAA2′,5′-GUAUGCCC3′,5′-GUAUGCCC by (C) anion-exchange HPLC (see A for conditions) and (D-E) 20% denaturing PAGE. The regioisomeric extended failure sequences (ii) are resolved into two peaks by HPLC (A and C), but appear as one band by gel analysis (B, D, and E). In panel E, the major product is the unbranched 8-mer 5′-GUAUGCCC-3′, which accumulates due to the unsuccessful branching of the bisphosphoramidite. P* indicates the presence of a 2′- or 3′-linked phosphate.

7. Enter the sequence to be synthesized in the 5′-to-3′ direction, where the 3′-nucleotide corresponds to the nucleoside bound to the CPG. For example, to synthesize the hypothetical V-shaped branched oligonucleotide (S.8) shown in Figure 4.14.4, enter the sequence 5′-XNNNNNN-3′, where X corresponds to the bisphosphoramidite and N is any base (phosphoramidite) of choice. If the hypothetical Y-shaped branched oligonucleotide (S.10) is desired, enter the sequence 5′NNNNNNXNNNNNN-3′.

8. Perform synthesis in the trityl-off mode according to the synthesis cycle outlined in Table 4.14.3 and utilizing the coupling times recommended in Table 4.14.2. Collect dimethoxytrityl solutions in 15-mL test tubes using an external fraction collector. Turning the trityl mode off ensures that the last nucleotide at the 5′ end has a free hydroxyl group, which is desirable for purification using anion-exchange HPLC (see Support Protocol 3).



Supplement 9

9. Upon completion of the synthesis, dry the CPG by manually conducting an argon reverse flush operation on the synthesizer for 10 min. Alternatively, dry the CPG under a stream of nitrogen or argon, or in a vacuum desiccator for 30 min. 10. Cleave the oligonucleotides from the support and deprotect the exocyclic amino, phosphate, and 2′-silyl (RNA only) protecting groups (see Support Protocol 2). 11. Purify the bNAs from failure sequences by anion-exchange HPLC (see Support Protocol 3) or denaturing PAGE (see Support Protocol 4). Typical HPLC and PAGE profiles for the synthesis of branched V-shaped and Y-shaped DNA and RNA molecules are demonstrated in Figures 4.14.6 and 4.14.7.

Measure branching efficiency by trityl color analysis 12. Dilute the dimethoxytrityl solutions collected after each successive coupling (step 8) with 10 mL detritylation solution using a 50-mL buret. 13. Aliquot 100 µL into a quartz cuvette and dilute with 2 mL detritylation solution. Use only quartz cuvettes, as the 1,2-dichloroethane will dissolve disposable polystyrene cuvettes.

14. Measure the absorbance of the solution on a UV-Vis spectrophotometer between 450 and 550 nm, and record the absorbance peak at ∼505 nm for the dimethoxytrityl cation.

XC

d c a

a′ b

b

BPB

1


2

3

4

Figure 4.14.7 PAGE analysis of bDNA molecules synthesized via the convergent strategy (see Basic Protocol 2), demonstrating the mobilities of bNAs through a cross-linked 20% polyacrylamide gel. Lane 1, crude linear DNA 12-mer 5′-TACTAAGTATGT-3′ (a); lane 2, crude V-DNA 13-mer 5′-A2′,5′-GTATGT3′,5′-GTATGT (c); lane 3, crude Y-DNA 18-mer TACTAA2′,5′-GTATGT3′,5′-GTATGT (d); lane 4, running dyes (xylene cyanol and bromphenol blue). The fastest-migrating sequence in lanes 2 and 3 is the 6-mer 5′-GTATGT-3′ (b), which is the immediate precursor of the V-shaped molecule (c). Notice the large gap between the failure sequence (b) and product V-DNA (c) in lane 2, owing to more than doubling the molecular weight upon branching. In lane 3, extension of the failure sequence yields a separate band (a′).



15. Determine the efficiency of each coupling step using the equation (Ax/Ax−1) × 100%, where Ax is the absorbance of the trityl cation released at any given step and Ax−1 is the release in the previous step. For efficient coupling, these values should be close to 100% for linear portions of the oligonucleotide (also see UNIT 10.3). For efficient branching, the value for the coupling of S.3 should be 50%, since one tritylated bisphosphoramidite is coupled to two nucleotide chains. For this step, trityl yields that are significantly greater than or less than 50% indicate little branch formation (e.g., 80% indicates the formation of mainly the extended isomeric compounds S.9a-d and S.11a-d).

REGIOSPECIFIC SYNTHESIS OF BRANCHED NUCLEIC ACIDS The protocol described below outlines the regiospecific and divergent synthesis of bNAs via phosphoramidite chemistry according to the method of Braich and Damha (1997; Fig. 4.14.8). This protocol allows for the synthesis of bNA molecules with different DNA sequences surrounding the branchpoint nucleotide. The methodology requires the use of standard DNA and RNA 3′-phosphoramidites (S.4a-d and S.5a-d) for the synthesis of a linear DNA strand incorporating a single ribonucleotide unit. Once this sequence is assembled, all of the 2-cyanoethyl phosphate-protecting groups are selectively removed by treatment with triethylamine. This step is necessary as phosphotriesters are susceptible to cleavage/modification by the ensuing fluoride treatment. The CPG-bound oligomer is then treated with fluoride ions to cleave the tert-butyldimethylsilyl group at the 2′ position of the ribose unit, from which another chain (2′-branch) can be synthesized. This is accomplished using “inverted” DNA phosphoramidites (deoxyribonucleoside 5′-phosphoramidites; S.6a-d), allowing branch synthesis to occur in the opposite (5′-to-3′) direction. With the exception of the decyanoethylation and desilylation steps, the entire process is conducted using an ABI 381A DNA synthesizer.

BASIC PROTOCOL 3

Materials 5′-O-(4,4′-Dimethoxytrityl)-N-protected-2′-deoxyribonucleoside-derivatized succinyl-LCAA-CPG (ChemGenes; also see Support Protocol 1) Cap A and B capping reagents (see recipes) DNA 3′-phosphoramidites (S.4a-d; Chem Genes; Fig. 4.14.3) Anhydrous acetonitrile (see recipe) RNA 3′-phosphoramidite (S.5a-d; ChemGenes; Fig. 4.14.3) Activator solution: 0.5 M 1H-tetrazole (sublimed) in anhydrous acetonitrile Oxidant solution (see recipe) Detritylation solution (see recipe) 4:6 (v/v) triethylamine/acetonitrile (see recipe) Anhydrous THF (see recipe) 1 M tetra-n-butylammonium fluoride (TBAF; Aldrich) in THF, fresh Inverted DNA 5′-phosphoramidites (S.6a-d; ChemGenes; Fig. 4.14.3) Argon or nitrogen gas (optional) Synthesis columns for 1 µmol scale synthesis, with seals and filters (PE Biosystems) and 13-mm aluminum seals (Chromatographic Specialties) ABI 381A automated DNA synthesizer (PE Biosystems) External fraction collector and 15-mL test tubes Empty DNA synthesizer bottles, oven-dried 10- and 1-mL disposable syringes 25-mL glass syringe Additional reagents and equipment for cleaving and deprotecting the oligonucleotide (see Support Protocol 2), anion-exchange HPLC (see Support



Supplement 9

B

DMTrO

O OTBDMS

O

O P OCH2CH2CN O HOD

D D D D D D

D D D D D

DMTrOD

DNA synthesis (3′→5′)

3′-pR amidites

3′-pD amidites

succinylLCAA-CPG

1. DNA synthesis (3′→5′) 2. cap with Ac2O

AcO

AcOD

AcO

D D D D D D O

D D D D D D O

O P O− O

O

O P OCH2CH2CN

O P O−

extend from 2′-OH in 5′→3′ direction with 5′-pD′ amidites

B

D D D D D O B

O

O

O O

O P

O P O− O D D D D D D

OCH2CH2CN

O

OH

O P O−

O D″ D′ D′ D′ D′ D′

D D D D D D

1. cleave from support and deprotect 2. purify (3′-5′)

R

1. remove all 2cyanoethyl groups 2. remove 2′-OTBDMS group

O

B O O

OTBDMS

O P OCH2CH2CN

O

O

D D D D D D

D D D D D D

12

(2′-5′)

D″ D′ D′ D′ D′ D′

D D D D D D 13


Figure 4.14.8 Schematic representation for the divergent and regiospecific synthesis of branched DNA. The branching synthon is a standard RNA phosphoramidite (S.5a-d; Fig. 4.14.3). Abbreviations: 3′-pD, DNA 3′-phosphoramidites (S.4a-d); 5′-pD′, inverted DNA 5′-phosphoramidites (S.6ad); 5′-pD′′, higher concentration (0.3M) of inverted DNA 5′−phosphoramidite as first nucleotide coupled to the ribose branch point; 3′-pR, RNA 3′-phosphoramidite (S.5a-d); DMTr, 4,4′-dimethoxytrityl; TBDMS, tert-butyldimethysilyl.



Protocol 3) or denaturing PAGE (see Support Protocol 4), and measuring coupling efficiency by trityl color analysis (see Basic Protocol 2) CAUTION: All solutions required for bNA solid-phase synthesis should be prepared in a well-ventilated fume hood. Synthesize linear DNA (S.12) 1. Prepare synthesis column (1 µmol) with the appropriate 5′-O-dimethoxytrityl-2′-deoxyribonucleoside-derivatized succinyl-LCAA-CPG (see Basic Protocol 2, steps 1 and 2). 2. Weigh out the proper amount of DNA 3′-phosphoramidites (S.4a-d; Fig. 4.14.3) and dilute to 0.1 M with anhydrous acetonitrile (see Table 4.14.2). For synthesis on the ABI 381A DNA synthesizer, the volume of each phosphoramidite addition to the column is 170 ìL.

3. Weigh out the appropriate amount of 3′-RNA phosphoramidite (S.5a-d; Fig. 4.14.3) and dilute to 0.15 M with anhydrous acetonitrile (see Table 4.14.2). The RNA 3′-phosphoramidite is the branching synthon. Any of the four standard RNA 3′-phosphoramidites (A, G, C, or U) may be used depending on the specific branch point to be introduced.

4. Install all reagents (i.e., activator, capping, oxidant, and detritylation solutions, and acetronitrile) and phosphoramidite solutions on the synthesizer, placing the RNA 3′-phosphoramidite on the spare port (the “X” port on the 381A synthesizer). 5. Enter the base sequence of the linear oligonucleotide to be synthesized in the 5′-to-3′ direction, where the last entry (3′ nucleotide) corresponds to the nucleoside bound to the CPG. For example, to synthesize the hypothetical linear oligonucleotide (S.12) shown in Figure 4.14.8, enter the sequence 5′-NNNNNNXNNNNNN-3′, where N is any deoxyribonucleoside phosphoramidite and X represents the branch point of the RNA.

6. Perform synthesis in the trityl off mode according to the synthesis cycle outlined in Table 4.14.3 and utilizing the coupling times shown in Table 4.14.2. Collect the dimethoxytrityl solutions in 15-mL test tubes in an external fraction collector. Turning the trityl mode off ensures that the last nucleotide at the 5′ end has a free hydroxyl group, which is desirable for purification using anion-exchange HPLC (see Support Protocol 3).

7. Acetylate the hydroxyl group at the 5′ terminus by running the automated capping cycle (Table 4.14.1). Capping the free 5′-OH is necessary as it ensures that extension from this functional group will not occur during the synthesis of the “’orthogonal” 2′-branch.

Cleave 2-cyanoethyl protecting group 8. Dry the CPG by manually conducting an argon reverse flush operation on the synthesizer for 10 min. 9. Remove the synthesis column from the synthesizer and connect it to a 10-mL disposable syringe filled with 4:6 (v/v) triethylamine/acetonitrile. Slowly push the deprotection solution through the column over a 90-min period. Deprotection of the 2-cyanoethyl phosphate-protecting group converts the phosphotriester to the more stable phosphodiester, which withstands the conditions required for desilylation in the ensuing step. To ensure complete decyanoethylation, push the solution slowly through the column and then pull in on the syringe slightly in order to displace the CPG beads from the base of the column.



Supplement 9

10. Wash the CPG beads extensively with 30 mL acetonitrile followed by 30 mL THF using a 25-ml glass syringe attached to the column via the syringe adapter. Cleave 2′-O-TBDMS group 11. Push 1 mL of 1 M TBAF in THF through the column over a period of 10 min using a 1-mL disposable syringe. It is essential to use fresh TBAF. In order to ensure complete desilylation, push the solution slowly through the column and then pull in on the syringe slightly in order to displace the CPG beads from the base of the column. Prolonged treatment with TBAF results in cleavage of the oligonucleotide chain from the solid support (Braich and Damha, 1997). Incomplete desilylation results in the accumulation of silylated linear DNA (S.12; Fig. 4.15.8), which does not allow branch extension from the 2′ position of the branch point (see Fig. 4.14.9B).

12. Wash CPG beads with 50 mL THF followed by 50 mL acetonitrile. 13. Reinstall the column on the synthesizer. Synthesize 2′,5′-linked branch (S.13) 14. Modify the DNA synthesis cycle such that synthesis step 15 becomes the first step in the cycle. Steps 1 to 14 (TCA treatment) may be disregarded since the assembled chain lacks a dimethoxytrityl group.

15. Weigh out the appropriate amounts of inverted DNA 5′-phosphoramidites (S.6a-d; Fig. 4.14.3) into the amidite bottles. Add acetonitrile to the first DNA phophoramidite to prepare a 0.3 M solution. Add acetonitrile to make 0.1 M solutions of each of the remaining monomers. The first inverted DNA 5′-phosphoramidite is added as a 0.3 M solution due to steric hindrance around the ribose 2′-hydroxyl group.

16. Install all the inverted phosphoramidites on the synthesizer and place the first inverted phosphoramidite (0.3 M concentration) on the spare port (the “X” port on the 381A). 17. Enter the linear sequence to be synthesized in the 5′-to-3′ direction, where the last entry corresponds to the first phosphoramidite to be coupled to the 2′-hydroxyl of the branch point. For example, to synthesize the hypothetical branched DNA oligonucleotide (S.13) shown in Figure 4.14.8, enter the sequence 5′-NNNNNN-3′, where N is the first DNA 5′-phosphoramidite to be coupled to the 2′-hydroxyl group of ribose.

18. Synthesize the 2′-branch in the trityl off mode using the modified synthesis cycle starting from step 15 of Table 4.14.3, and utilizing the coupling times shown in Table 4.14.2, except for the first phophoramidite (0.3 M), which should have a coupling time of 30 min. 19. Upon completion of the synthesis, dry the CPG by reverse flushing the column with argon for 10 min. Alternatively, dry the CPG under a stream of nitrogen or argon, or in a vacuum desiccator for 30 min. 20. Cleave the oligonucleotides from the support and deprotect the amino- and phosphate-protecting groups (see Support Protocol 2). Solid-Phase Synthesis of Branched Oligonucleotides



21. Purify the bNAs from failure sequences by anion-exchange HPLC (see Support Protocol 3) or denaturing PAGE (see Support Protocol 4) and measure coupling efficiency by trityl color assay (see Basic Protocol 2, steps 12 to 15). Typical PAGE profiles for the successful and unsuccessful regiospecific synthesis of a branched Y-shaped DNA molecule are demonstrated in Figure 4.14.9.

PREPARATION OF LCAA-CPG SUPPORTS WITH HIGH NUCLEOSIDE LOADINGS

SUPPORT PROTOCOL 1

The method described allows for the rapid derivatization of LCAA-CPG having nucleoside loadings up to ∼90 µmol/g, which is 3 to 4 times the loading found in commercially available solid supports. While commercial samples provide more than adequate yields of bNAs (Damha et al., 1992), those with higher loadings (50 to 90 µmol/g) provide the best results (Fig. 4.14.5). For example, the synthesis of small bNAs (i.e., trimers and tetramers) requires that the CPG be densely loaded so that efficient branching may occur. The protocol below, adapted from the work of Pon et al. (1999) and Damha et al. (1990), allows for the rapid esterification of 5′-O-protected ribonucleosides and deoxyribonucleosides to succinyl-LCAA-CPG. The key condensing reagent is a mixture of HATU and 4-DMAP. Materials 5′-O-(4,4′-Dimethoxytrityl)-N-protected-ribonucleoside or -2′-deoxyribonucleoside (ChemGenes) O-(7-Azabenzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate (HATU; PE Biosystems) 4-Dimethylaminopyridine (DMAP; 99%, Aldrich)

A (i) A

B

T10 T10

XC

XC (iii) TACTAA (iv) TACTAA

(ii) A

GTATGT CAAGTT Si CAAGTT

Si T10 BPB BPB 1

2

3

4

Figure 4.14.9 20% denaturing PAGE analysis of a successful (A) and a less successful (B) synthesis of bNAs. (A) Lane 1, pure bNA (i) prepared by the divergent synthesis method (see Basic Protocol 3); lane 2, pure bNA (i) prepared by the convergent synthesis method (see Basic Protocol 2); lane 3, crude bNA (i) prepared by the divergent synthesis method; lane 4, running dyes (xylene cyanol and bromphenol blue). (B) Preparative PAGE purification of bNA (iii) prepared via the divergent (regiospecific) approach. The diminutive amount of fully branched product (iii) is due to the incomplete desilylation of the 2′-O-TBDMS group (Si) on the RNA branching synthon, as seen by the significant amount of (iv) that remains.



Supplement 9

Succinylated long-chain-aminoalkyl controlled-pore glass (succinyl-LCAA-CPG; UNIT 3.2) Anhydrous acetonitrile (see recipe) Dichloromethane, reagent grade (Fisher) Methanol, reagent grade (Fisher) Cap A and B capping reagents (see recipes) 15-mL glass bottle with septum 1-mL syringe and needle Wrist-action shaker Sintered glass funnel or Buchner funnel with filter paper Additional reagents and equipment for quantitation of released trityl groups (UNIT 3.2) and acetylation of support (UNIT 3.2; optional) 1. In a 15-mL glass bottle, combine the following: 0.1 mmol 5′-O-(4,4′-dimethoxytrityl)-N-protected-ribonucleoside or -2′-deoxyribonucleoside 0.1 mmol HATU 12 mg DMAP 250 mg succinyl-LCAA-CPG. 2. Cap the bottle with a septum and add 1 mL anhydrous acetonitrile via a 1-mL syringe and needle. 3. Shake 2 to 4 hr on a wrist-action shaker at room temperature. Do not stir the slurry with a magnetic stir bar as this will break up the glass beads into fine particles that may clog the frit on the DNA synthesizer.

4. Vacuum filter succinyl-LCAA-CPG into a side-arm filter flask through a sintered glass funnel or Buchner funnel with filter paper. 5. Wash the CPG sequentially with 25 mL reagent-grade dichloromethane, 25 mL methanol, and 25 mL dichloromethane. It has been reported that the free carboxylic groups are inconsequential and do not react during phosphoramidite synthesis (Lyttle et al., 1997).

6. Transfer the CPG to a glass vial and dry under vacuum using in a desiccator attached to a vacuum pump. The derivatized CPG may be stored indefinitely at room temperature, preferably in a vacuum desiccator.

7. Determine nucleoside loading through the quantitation of the released trityl groups from the support-bound nucleoside (UNIT 3.2). 8. Acetylate the solid support with cap A and B capping reagents on a DNA synthesizer using the capping cycle outlined in Table 4.14.1. Alternatively, perform capping as described in UNIT 3.2.




COMPLETE DEPROTECTION OF BRANCHED OLIGONUCLEOTIDES (DNA AND RNA)

SUPPORT PROTOCOL 2

This protocol describes the steps necessary for cleaving the bNA from the solid support and removing the protecting groups from the heterocyclic bases and sugar-phosphate backbone. The first step is treatment of the solid support with concentrated aqueous ammonia to concomitantly cleave both the bNA from the support and the N-acyl and 2-cyanoethyl phosphate-protecting groups. A subsequent deprotection step with triethylammonium trihydrofluoride (TREAT-HF) cleaves the 2′-O-tert-butyldimethylsilyl (TBDMS) protecting groups from branched oligoribonucleotides (Gasparutto et al., 1992). The desilylated material is then precipitated directly using 1-butanol (Sproat et al., 1995). A procedure for the quantitation of oligonucleotides is also described. Materials Branched oligonucleotide attached to CPG (bNA-CPG; see Basic Protocols 2 and 3) 29% ammonium hydroxide, 4°C (store up to 1 month at 4°C) 70% and 100% (v/v) ethanol, former at 4°C DEPC-treated water (optional; APPENDIX 2A) Autoclaved water Triethylammonium trihydrofluoride (TREAT-HF; 98%; Aldrich) 3 M sodium acetate, pH 5.5 (APPENDIX 2A) 1-Butanol, analytical grade, 4°C 1.5-mL screw-cap microcentrifuge tubes with O-ring seals (preferred) Wrist-action shaker Speedvac evaporator (Savant) Double-beam UV spectrophotometer, calibrated Cleave and deprotect oligonucleotide 1. Transfer bNA-CPG to a 1.5-mL screw-cap microcentrifuge tube, preferably with an O-ring seal. 2. Add 750 µL cold 29% ammonium hydroxide and 250 µL of 100% ethanol, screw the cap on tightly, and incubate 24 to 48 hr at room temperature on a wrist-action shaker. If the branched oligonucleotide sequence contains isobutyryl-protected guanosine nucleotides, then deprotection must proceed for ≥48 hr, room temperature. The ammonium hydroxide should be relatively fresh (10 nt). Smaller bNAs do not precipitate out efficiently and must be further purified by size-exclusion chromatography (or reversed-phase Sep-Pak cartridges) subsequent to HPLC separation. As an alternative to HPLC purification, the bNAs may be purified by denaturing PAGE (see Support Protocol 4). Characterization of the bNAs is conveniently done via MALDI-TOF-MS as described in UNIT 10.1. The matrix and co-matrix typically used are 6-aza-2-thiothymine (ATT) and dibasic ammonium citrate, respectively (Lecchi et al., 1995). The branched nature of the molecules may also be confirmed via the yeast debranching enzyme (yDBR), a phosphodiesterase specific to hydrolysis of the 2′,5′-phosphodiester bond of oligonucleotides that contain vicinal 2′,5′- and 3′,5′-phosphodiester linkages (Nam et al., 1994; Ooi et al., 2001). Nucleoside composition analysis of bNA is carried out using snake venom phosphodiesterase (SVPD) according to the method of Eadie et al. (1987; UNIT 10.6). This enzyme cleaves bDNA or bRNA from the 3′ termini yielding 5′-monophosphates, which can then be converted to their constituent nucleosides by in situ treatment with alkaline phosphatase (AP). The resulting nucleoside mixture is analyzed by reversed-phase HPLC as described in UNIT 10.5. Alternatively, bNA can be digested with nuclease P1 from Penicillium citrinum, an endonuclease that cleaves bNA to produce the constituent nucleoside 5′-monophosphates and its branch core trinucleoside diphosphate—i.e., A(2′p5′N)3′p5′N (Damha et al., 1992). The released branched trinucleoside diphosphate structure can be readily synthesized (see Basic Protocol 2) and used as a standard during HPLC analysis of the enzyme digest (UNIT 10.6). NOTE: For branced RNAs, use DEPC-treated water throughout (APPENDIX 2A).

UACUAA

GUAUGU GUAUGU

95% purity AU

0.00

10.00

20.00

30.00

40.00 Time (min)

50.00

60.00

70.00

80.00

Figure 4.14.10 Analysis of the purity of a bNA synthesized via the convergent strategy. Purification was conducted using anion-exchange HPLC (see Support Protocol 3) with linear gradient 2 (Table 4.14.4). The chromatogram was obtained using the same conditions.



Supplement 9

Table 4.14.4

Conditions for Separation of bNAs by Anion-Exchange HPLCa

Length of bNA (nt) 15

Gradient

% Buffer A (H2O)

% Buffer B (1 M LiClO4)

Run time (min)

1

100-90

0-10

60

2 3

100-80 90-80

0-20 10-20

60 60

aPerformed on a Waters 7.5 × 75–mm Protein Pak DEAE-5PW column.

Materials Deprotected branched oligonucleotide (see Support Protocol 2) Autoclaved water 1 M LiClO4 (see recipe) Reagent-grade 1-propanol, 4°C (>5-mers; Fisher) Sephadex G-25 columns (Amersham Pharmacia Biotech) or Sep-Pak cartridges (Waters Chromatography) for 60 A260 units may overload the column and compromise the separation of the bNA from the linear failure sequences.

7. Run the sample as described (steps 4 and 5), but set the detector to 290 nm and collect 1-ml fractions from the peaks of interest in sterile 1.5-mL microcentrifuge tubes. The detector wavelength is set to 290 nm in order to avoid saturation of the detector signal. If the HPLC is equipped with a detector capable of monitoring dual wavelengths, monitor both the 260- and 290-nm profiles. The anticipated retention time should be very similar to that obtained during routine HPLC analysis.

8. Pool peak fractions and dry them in a Speedvac evaporator. Add 250 µL autoclaved water. For large bRNAs (≥5-mers): 9a. Precipitate from perchlorate salts (in sample) by adding 4 vol (1 mL) reagent-grade cold 1-propanol and cooling 4 to 6 hr at −20°C. Lithium perchlorate (LiClO4) is much more soluble in organic solvents than other perchlorate salts, making precipitation easy and efficient, and thus preventing a final desalting step. The DNA or RNA isolated is in its lithium salt form.

10a. Microcentrifuge 10 min at maximum speed, room temperature, carefully remove the supernatant, and wash the white pellet twice with 500 µL cold 1-propanol. Disrupting the pellet during washing steps can result in loss of sample.

11a. Dry in a Speedvac evaporator, resuspend in 1 mL autoclaved water, and proceed to step 12. For small bNAs (10 nucleotides in length are very well resolved on a 20% denaturing polyacrylamide gel. If shorter sequences must be purified, better resolution is achieved with a 24% denaturing polyacrylamide gel. Setup and polymerization of the gel along with electrophoretic separation conditions are described. The resultant bands may be visualized by UV shadowing and photographed. A technique for the rapid extraction of oligonucleotides from the gel matrix is also described (Chen and Ruffner, 1996). Materials 20% (w/v) denaturing acrylamide gel solution (see recipe) Deprotected branched oligonucleotide sample, dry (see Support Protocol 2) Formamide loading buffer (see recipe) Gel extraction buffer (see recipe) Running dye (see recipe)


Gel electrophoresis equipment (APPENDIX 3B) with: 16 × 18–cm glass plates Spacers: 0.75 mm (analysis) or 1.5 mm (purification) Gel combs: 0.75 mm thick with 12 to 20 wells (analysis) or 1.5 mm thick with one to three large wells (purification) Sonicator or ∼50°C water bath (optional) 20 × 20–cm silica-coated thin-layer chromatography (TLC) plate with fluorescent indicator (e.g., Kieselgel 60 F254 aluminum sheets) Handheld UV lamp (254 nm) Camera equipped with UV filter (optional) 90°C water bath or heating block (optional) UV shadow box Wrist-action shaker (optional) Speedvac evaporator (Savant) Sephadex G-25 (Amersham Pharmacia Biotech) or reversed-phase cartridges (e.g., Sep-Pak cartridges; Waters Chromatography)



Additional reagents and equipment for denaturing PAGE (UNIT 10.4 and APPENDIX 3B), reversed-phase chromatography using Sep-Pak cartridges (UNIT 10.1), UV spectrophotometry (see Support Protocol 2), MALDI-TOF-MS (UNIT 10.1), and enzymatic digestion (see Support Protocol 3) CAUTION: Acrylamide and N,N′-methylenebisacrylamide are neurotoxins. Prepare all solutions containing these two reagents in a fume hood. Minimize exposure and contact to both crystalline forms and solutions by conducting all handling (including weighing) in a well-ventilated area and wearing disposable gloves at all times. Analyze bNAs by denaturing PAGE 1. Assemble a gel sandwich using 16 × 18–cm glass plates separated by 0.75-mm spacers. See APPENDIX 3B and UNIT 10.4 for a thorough description of denaturing PAGE.

2. Transfer 30 mL of 20% denaturing acrylamide gel solution to an Erlenmeyer flask and add 200 µL fresh 10% (w/v) APS immediately followed by 20 µL TEMED. Swirl and degas the solution by attaching the Erlenmeyer flask to a house vacuum line. Maintain rapid stirring to avoid bumping. The gel solution may also be degassed prior to APS and TEMED addition by placing the 20% acrylamide gel solution on a sonicator for 5 to 10 min. If using larger glass plates, adjust the volumes of acrylamide, APS, and TEMED accordingly.

3. Pour the solution between the plates, insert a 0.75-mm-thick comb, and allow gel to polymerize (30 to 45 min). 4. Place gel in electrophoresis apparatus, removing the comb and bottom spacer, rinse the wells, and prerun the gel 30 min at 500 V, room temperature. 5. Dissolve 0.6 A260 units deprotected bNA sample in 10 µL formamide loading buffer. If the dissolution process is not immediate, place the sample in a sonicator bath 1 to 2 min or heat the samples briefly at ∼50°C. 6. Load the samples into the wells. Load an equal amount of running dye in the first and last well as an external reference marker. Run the gel at 500 V until the bromphenol blue dye is 3⁄4 of the way down the gel. 7. Disassemble the glass plates and wrap the gel in plastic wrap. Place the wrapped gel over a 20 × 20–cm silica-coated TLC plate with fluorescent indicator and visualize the bands by shining a handheld UV lamp over the gel. Take a picture of the gel using a camera equipped with a UV filter. CAUTION: Wear safety glasses to avoid eye burn. A typical crude bNA reaction mixture will consist of at least three bands as shown in Figures 4.14.6 and 4.14.7. The fastest-migrating band is the linear precursor sequence (S.7; Fig. 4.14.4), followed by the extended isomeric failure sequences (S.9a-b for the synthesis of V-shaped bNAs and S.11a-b for the synthesis of Y-shaped bNAs), and finally the products (S.8 and S.10). If the coupling efficiencies between successive nucleotides is less than optimal, a ladder of failure sequences will be evident below the extended branch failure and linear sequences. Note that the extended isomeric failure sequences are a mixture of regioisomers (S.9a-b and S.11a-b) that are sometimes resolved into two close-moving bands (this is also evident by HPLC analyses). The slowest of the predominant bands is the bNA of interest. If linear markers are run alongside the purified bNA, one observes that bNA has retarded mobility relative to a linear oligonucleotide of identical composition (length and sequence composition). This is due to the increased frictional effects of the bNA relative to the linear sequences as they move through the highly cross-linked gel environment.



Supplement 9

Purify bNAs by denaturing PAGE 8. Assemble and run a preparative gel as for the analytical gel (steps 1 to 6) with the following modifications: a. Increase gel thickness (spacers and comb) to 1.5 cm and scale up volume of gel to 50 mL gel solution, 350 µL APS, and 35 µL TEMED. b. Use up to 100 A260 units crude bNA dissolved in 100 µL formamide loading buffer. c. Use a comb with one large well to purify 60 to 100 A260 units, one well of a two-well comb for 30 to 60 units, and one well of a three-well comb for 95% (Fig. 4.14.10). In the case of divergent bNA synthesis, typical crude yields are in the range of 50 to 100 A260 units. A standard gel analysis of bNA prepared by the divergent method is shown in Fig. 4.14.9. After HPLC or gel purification, 5 to 20 A260 units of bNA are typically recovered from divergent syntheses (slightly higher yields are obtained with HPLC). As for convergent synthesis, purity is >95%.

Time Considerations Provided that all reagents and materials required for each step are available, most of the procedures are simple and rapid. The synthesis of S.3 from 5′-DMTr-N6-benzoyl-riboadenosine requires 2 to 4 hr to complete, including the workup. The reaction should not be allowed to proceed overnight, as decomposition may occur. Column chromatography requires ∼1 to 2 hr including setup. Ideally, column purification should be conducted immediately following the phosphitylation reaction. When the branching synthon S.3 and all the other phosphoramidite derivatives are ready to



use, the time required to prepare, purify, and isolate a Y-shaped RNA or DNA via the convergent approach is 3 days and 4 to 6 days, respectively. The preparation of bNA via the divergent approach requires 3 days or more.

Literature Cited Braich, R.S. and Damha, M.J. 1997. Regiospecific solid-phase synthesis of branched oligonucleotides. Effect of vicinal 2′,5′- (or 2′,3′-) and 3′,5′-phosphosdiester linkages on the formation of hairpin DNA. Bioconjugate Chem. 8:370377. Carriero, S., Braich, R.S., Hudson, R.H.E., Anglin, D., Friesen, J.D., and Damha, M.J. 2001. Inhibition of in vitro pre-mRNA splicing in S. cerevisiae by branched oligonucleotides. Nucleosides, Nucleotides, and Nucl. Acids. 20:873-877. Chapman, K.B. and Boeke, J.D. 1991 Isolation and characterization of the gene encoding yeast debranching enzyme. Cell 65:483-492. Chen, Z. and Ruffner, D.E. 1996. Modified crushand-soak method for recovering oligodeoxynucleotides from polyacrylamide gel. BioTechniques 21:820-822. Damha, M.J., and Braich, R.S. 1998. Synthesis of branched DNA/RNA chimera similar to the msDNA molecule of Myxoccus xanthus. Tetrahedron lett. 39:3907-3910. Damha, M.J. and Ogilvie, K.K. 1988. Synthesis and spectroscopic analysis of branched RNA fragments: Messenger RNA splicing intermediates. J. Org. Chem. 53:3710-3722. Damha, M.J. and Zabarylo, S.V. 1989. Automated solid-phase synthesis of branched oligonucleotides. Tetrahedron Lett. 30:6295-6298. Damha, M.J., Pon, R.T., and Ogilvie, K.K. 1985. Chemical synthesis of branched RNA: Novel trinucleoside diphosphates containing vicinal 2′5′ and 3′-5′ phosphodiester linkages. Tetrahedron Lett. 26:4839-4842. Damha, M.J., Giannaris, P.A., and Zabarylo, S.V. 1990. An improved procedure for derivatization of controlled-pore glass beads for solid-phase oligonucleotide synthesis. Nucl. Acids Res. 18:3813-3821. Damha, M.J., Ganeshan, K., Hudson, R.H.E., and Zabarylo, S.V. 1992. Solid-phase synthesis of branched oligoribonucleotides related to messenger RNA splicing intermediates. Nucl. Acids Res. 20:6565-6573. Eadie, J.S., McBride, L.J., Efcavitch, J.W., Hoff, L.B., and Cathcart, R. 1987. High-performance liquid chromatographic analysis of oligodeoxyribonucleotide base composition. Anal. Biochem. 165:442-447. Ganeshan, K., Tadey, T., Nam, K., Braich, R., Purdy, W.C., Boeke, J.D., and Damha, M.J. 1995. Novel approaches to the synthesis and analysis of branched RNA. Nucleosides & Nucleotides 14:1009-1013.

Gasparutto, D., Livache, T., Bazin, H., Duplaa, A.M., Guy, A., Khorlin, A., Molko, D., Roget, A., and Teoule, R. 1992. Chemical synthesis of a biologically active natural tRNA with its minor bases. Nucl. Acids Res. 20:5159-5166. Hakimelahi, G.H., Proba, Z.A., and Ogilvie, K.K. 1982. New catalysts and procedures for the dimethoxytritylation and selective silylation of ribonucleosides. Can. J. Chem. 60:1106-1113. Hudson, R.H.E. and Damha, M.J. 1993. Nucleic acid dendrimers: Novel biopolymer structures. J. Am. Chem. Soc. 115:2119-2124. Hudson, R.H.E., Uddin, A.H., and Damha, M.J. 1995. Association of branched nucleic acids: Structural and physicochemical analysis of antiparallel TAT triple-helical DNA. J. Am. Chem. Soc. 117:12470-12477. Hudson, R.H.E., Robidoux, S., and Damha, M.J. 1998. Divergent synthesis of nucleic acid dendrimers. Tetrahedron Lett. 32:1299-1302. Inouye, S., Furuichi, T., Dhundle, A., and Inouye, M. 1987. Molecular Biology of RNA: New Perspectives (M. Inouye and B. S. Dudock, eds.) pp. 271. Academic Press, San Diego. Kierzek, R., Kopp, D.W., Edmonds, M., and Caruthers, M.H. 1986. Chemical synthesis of branched RNA. Nucl. Acids Res. 14:4751-4764. Lecchi, P., Le, H.M.T., and Pannell, L.K. 1995. 6-Aza-2-thiothymine: A matrix for MALDI spectra of oligonucleotides. Nucl. Acids Res. 23:1276-1277. Lyttle, M.H., Adams, H., Hudson, D., and Cook, R.M. 1997. Versatile linker chemistry for synthesis of 3′-modified DNA. Bioconjugate Chem. 8:193-198. Nam, K., Hudson, R.H.E., Chapman, K.B., Ganeshan, K., Damha, M.J., and Boeke, J.D. 1994. Yeast lariat debranching enzyme: Substrate and sequence specificity. J. Biol. Chem. 269:20613-20621. Ooi, S.L., Dann, C. III, Nam, K., Leahy, D., Damha, M.J., and Boeke, J.D. 2001. Ribonucleases part A: Functional roles and mechanisms. Methods Enzymol. 342:233-250. Pon, R.T., Yu, S., and Sanghvi, Y.S. 1999. Rapid esterification of nucleosides to solid-phase supports for oligonucleotide synthesis using uronium and phosphonium coupling reagents. Bioconjugate Chem. 10:1051-1057. Robidoux, S., Klinck, R., Gehring, K., and Damha, M.J. 1997. Association of branched oligonucleotides into the i-motif. J. Biomol. Struct. Dyn. 15:517-527. Rousse, B., Puri, N., Viswanadham, G., Agback, P., Glemarec, C., Sandstroem, A., Sund, C., and Chattopadhyaya, J. 1994. Solution conformation of hexameric and heptameric lariat-RNAs and their self-cleavage reactions which give products mimicking those from some catalytic RNAs (ribozymes). Tetrahedron 50:1777-1810. Ruskin, B. and Green, M. 1985. An RNA processing activity that debranches RNA lariats. Science 229:135-140.



Supplement 9

Ruskin, B., Krainer, A.R., Maniatis, T., and Green, M.R. 1984. Excision of an intact intron as a novel lariat structure during pre-mRNA spicing in vitro. Cell 38:317-331. Sproat, B.S., Beijer, B., Grotli, M., Ryder, U., Morand, K.L., and Lamond, A.I. 1994. Novel solidphase synthesis of branched oligoribonucleotides including a substrate for the RNA debranching enzyme. J. Am. Chem. Soc. Perkin. Trans. 1:419-431. Sproat, B., Colonna, F., Mullah, B., Tsou, D., Andrus, A., Hampel, A., and Vinayak, R. 1995. An efficient method for the isolation and purification of oligoribonucleotides. Nucleosides & Nucleotides 14:255-273. Still, W.C., Kahn, M., and Mitra, A. 1978. Rapid chromatographic technique for preparative separations with moderate resolution. J. Org. Chem. 43:2923-2925. Ti, G.S., Gaffney, B.L., and Jones, R.A. 1982. Transient protection: Efficient one-flask syntheses of protected deoxynucleosides. J. Am. Chem. Soc. 104:1316-1319. Uddin, A.H., Piunno, P.A.E., Hudson, R.H.E., Damha, M.J., and Krull, U.J. 1997. A fiber optic biosensor for fluorimetric detection of triplehelical DNA. Nucl. Acids Res. 25:4139-4146. Urdea, M.S., Horn, T., Fultz, T.J., Anderson, M., Running, J.A., Hamren, S., Ahle, D., and Chang, C.A. 1991. Branched DNA amplification multimers for the sensitive, direct detection of human hepatitis viruses. Nucl. Acids Symp. Series. 24:197-200. Wallace, J.C. and Edmonds, M. 1983. Polyadenylated nuclear RNA contains branches. Proc. Natl. Acad. Sci. U.S.A. 80:950-954. Wincott, F., Di Renzo, A., Shaffer, C., Grimm, S., Tracz, D., Workman, C., Sweedler, D., Gonzalez, C., Scaringe, S., and Usman, N. 1995. Synthesis, deprotection, analysis and purification of RNA and ribozymes. Nucl. Acids. Res. 23:2677-2684. Wu, T., Ogilvie, K.K., and Pon, R.T. 1989. Prevention of chain cleavage in the chemical synthesis of 2′-silylated oligoribonucleotides. Nucl. Acids Res. 17:3501-3517. Yee, T., Furuichi, T., Inouye, S., and Inouye, M. 1984. Multicopy single-stranded DNA isolated from a Gram-negative bacterium, Myxococcus xanthus. Cell 38:203-209.

Key References Damha and Zabarylo, 1989. See above.

Reports on the first general procedure for the convergent solid-phase synthesis of branched oligonucleotides via an adenosine bisphosphoramidite. Damha et al., 1992. See above. Reports on the convergent synthesis of branched RNA oligonucleotides using the standard silyl-phosphoramidite RNA synthesis methodology. The bNAs synthesized are related to the splicing intermediates derived from S. cerevisiae. Nam et al., 1994. See above. Reports on the substrate and sequence specificity of the yeast lariat debranching enzyme (yDBR), a unique 2′,5′-phosphodiesterase. The enzyme accepts a variety of substrates including group II intron lariats, msDNA, and synthetic bNAs. Padgett, R.A., Konarska, M.M., Grabowski, P.J., Hardy, S.F., and Sharp, P.A. 1984. Lariat RNA’s as intermediates and products in the splicing of messenger precursors. Science 225:898-903. This report provides evidence that the branched lariat structure is an intermediate of splicing of an adenovirus ML2 RNA transcript. Specifically demonstrated is that the excised intron contains an unusual nuclease-resistant core consisting of a branched trinucleotide structure with vicinal 2′,5′and 3′,5′-phosphodiester linkages. Sharp, P.A. 1994. Split genes and RNA splicing. Cell 77:805-815. A Nobel lecture. A paramount review describing the splicing of introns from nascent RNA, the evolutionary significance of introns, and the plethora of factors involved in post-transcriptional processing. Wallace and Edmonds, 1983. See above. A first account demonstrating the occurrence of a branched nuclear polyadenylated RNA containing vicinal 2′,5′- and 3′,5′-phosphodiester bonds. Such molecules were absent from cytoplasmic polyadenylated RNA, implicating these structures as intermediates during mRNA processing.

Internet Resources http://paris.chem.yale.edu/extinct.html A useful site for the calculation of molecular weights of oligonucleotides and peptides as well as the determination of extinction coefficients (ε).

Contributed by Sandra Carriero and Masad J. Damha McGill University Montreal, Canada




Solid-Phase Synthesis of 2′-Deoxy-2′-fluoroβ-D-Oligoarabinonucleotides (2′F-ANA) and Their Phosphorothioate Derivatives

UNIT 4.15

This unit describes the chemical synthesis of 2′-deoxy-2′-fluoro-β-D-oligoarabinonucleotides (2′F-ANA), both with phosphodiester and phosphorothioate linkages. The protocols described herein include araF phosphoramidite preparation (see Basic Protocol 1), assembly on DNA synthesizers (see Basic Protocol 2), and final deprotection and purification of oligonucleotides (see Basic Protocol 3). The preparation of araF phosphoramidite building blocks is carried out by introducing a 2-cyanoethyl-N,N-diisopropylaminophosphinyl group at the 3′-O-position of conveniently protected araF nucleosides. The preparation of araF-protected nucleosides is described in UNIT 1.7. Assembly of 2′F-ANA sequences can be carried out under similar conditions as for DNA sequences, but longer coupling times for ara F phosphoramidite monomers are required. For phosphorothioate analogs, longer sulfurization times are also required, as compared to S-DNA synthesis. Assembly of the oligonucleotides is carried out by the stepwise addition of phosphoramidite building blocks to nucleoside or nucleotide hydroxyl termini preimmobilized on a solid support until the desired sequence is obtained (APPENDIX 3C). Each addition of new building block requires five steps: detritylation, coupling, capping, oxidation or sulfurization, and capping. The last capping reaction also serves to dry the column prior to the next coupling cycle. Cleavage of the sequence from the solid support and the removal of the nucleobase- and phosphodiester-protecting groups is carried out using an aqueous ammonia/ethanol mixture. Crude oligonucleotides obtained in this way can be purified by ion-exchange high-performance liquid chromatography (HPLC; UNIT 10.5) or denaturing polyacrylamide gel electrophoresis (PAGE; UNIT 10.4). NOTE: All glassware should be oven dried.

MMTrO

O F OH

MMTrO

B

B

i -Pr2NP(Cl)OCH2CH2CN O

i -Pr2NEt, THF i -Pr2N

1a b c d

O F

B = N 6-benzoyladenin-9-yl B = N 2-isobutyrylguanin-9-yl B = N 4-benzoylcytosin-1-yl B = thymin-1-yl

P

2a b c d

OCH2CH2CN (75%) (70%) (80%) (95%)

Figure 4.15.1 Synthesis of the four araF-protected nucleoside phosphoramidites (S.2a-d). MMTr, 4-monomethoxytrityl; i-Pr2NEt, N-ethyl-N,N-diisopropylamine; i-Pr2-NP(Cl)OCH2CH2CN, 2-cyanoethyl-N,N-diisopropylchlorophosphoramidite.


Contributed by Ekaterina Viazovkina, Maria M. Mangos, Mohamed I. Elzagheid, and Masad J. Damha

4.15.1


Supplement 10

BASIC PROTOCOL 1

PREPARATION OF araF PHOSPHORAMIDITES Conversion of protected araF nucleosides into the corresponding phosphoramidite building blocks is shown in Figure 4.15.1 (Wilds and Damha, 2000). In order to obtain phosphoramidites in sufficiently high yields, the starting materials and reaction solvents should be as dry as possible and all glassware should be oven dried. Likewise, to simplify purification, the reaction should be completed using a minimal amount of phosphitylation reagent. Materials Protected araF nucleosides (UNIT 1.7): N6-Benzoyl-9-[2-deoxy-2-fluoro-5-O-(4-methoxytrityl)-β-D-arabinofuranosyl] adenine (S.1a) N2-Isobutyryl-9-[2-deoxy-2-fluoro-5-O-(4-methoxytrityl)-β-D-arabinofuranosyl] guanine (S.1b) 4 N -Benzoyl-1-[2-deoxy-2-fluoro-5-O-(4-methoxytrityl)-β-D-arabinofuranosyl] cytosine (S.1c) 1-[2-Deoxy-2-fluoro-5-O- (4-methoxytrityl)-β-D-arabinofuranosyl]thymine (S.1d) THF, anhydrous (see recipe) N-Ethyl-N,N-diisopropylamine (DIPEA; Aldrich), double distilled 2-Cyanoethyl-N,N-diisopropylchlorophosphoramidite (Chem Genes) Dichloromethane 1:9 and 1:19 (v/v) methanol/dichloromethane Saturated sodium bicarbonate solution Magnesium sulfate, anhydrous Solvent system: 1:99 (v/v) triethylamine/chloroform (for araF-T) 1:9:10 (v/v) triethylamine/dichloromethane/hexanes (for araF-A and araF-C) 1:99 (v/v) triethylamine/dichloromethane (for araF-G) Silica gel (230 to 400 mesh) Diethyl ether 50-mL round-bottom flask equipped with stir bar and rubber septum Syringe TLC Merck silica plates (Kieselgel 60 F-254) 254-nm UV lamp 500-mL separatory funnel Vacuum evaporator (e.g., Savant Speedvac) 3-cm-diameter chromatography column Additional reagents and equipment for TLC (APPENDIX 3D) and column chromatography (APPENDIX 3E) Prepare araF phosphoramidite monomers 1. In a 50-mL round bottom flask equipped with a stir bar and rubber septum, dissolve 1 mmol of each protected araF nucleoside (S.1a-d) in 3 to 5 mL of anhydrous THF. 2. While stirring, add 0.63 mL (3.6 mmol) double distilled N-ethyl-N,N-diisopropylamine (DIPEA) followed by the slow addition of 0.24 mL (1.1 mmol) 2-cyanoethylN,N-diisopropylchlorophosphoramidite.

Synthesis of 2′-Deoxy-2′-fluoroβ-D-oligoarabinonucleotides (2′F-ANA)

A white precipitate of diisopropylethylammonium hydrochloride salts forms after 10 min and is indicative of sufficiently anhydrous conditions for a successful reaction.



3. Stir the reaction mixture for an additional 2 hr or until complete consumption of starting material is observed. For TLC analysis (APPENDIX 3D), remove 10 to 20 µL reaction mixture by syringe and dilute with 100 µL dichloromethane in a small tube. Analyze on a TLC Merck silica plate using 1:19 (v/v) methanol/dichloromethane for araF-T, -C, and -A, and 1:9 (v/v) methanol/dichloromethane for araF-G. Co-spot the starting material for comparison. Visualize by exposure with a 254-nm UV lamp. TLC solvents in this step do not include TEA. However, TLC solvents listed under step 11, can be substituted for those listed here. The phosphoramidites migrate faster than the starting material. In this reaction, two stereoisomeric products are formed, but they are not separable in the methanol/dichloromethane system.

4. If, as demonstrated by TLC, the reaction is not complete (i.e., more than ∼5% of starting material remaining), add an additional portion (∼0.2 eq.) of 2-cyanoethylN,N-diisopropylchlorophosphoramidite, stir for 2 hr more, and repeat TLC analysis (step 3). If the reaction looks complete by TLC analysis, start workup of the phosphoramidite. It is important to minimize the amount of phosphitylating reagent used, otherwise the excess is hydrolyzed during the workup to form H-phosphonate impurities (31P NMR: singlet, ∼14 ppm), which can cause difficulties during subsequent purification of the amidites.

5. Dilute the reaction mixture with 150 mL dichloromethane, transfer it to a 500-mL separatory funnel, and wash with 150 mL saturated sodium bicarbonate solution. 6. Collect the organic layer, dry it by adding solid anhydrous magnesium sulfate until it no longer clumps, and filter and concentrate under reduced pressure to obtain a crude product. Chromatographic purification of the crude amidite is recommended, but if the reaction mixture looks pure by TLC and evaporates easily to give a stable foam, this step may be omitted. In case of a second day purification, it is strongly recommended to evaporate the crude compound to a foam, dry it under high vacuum, and keep the product in a vacuum desiccator. If the product persists as an oil after these steps are taken, co-evaporate it with diethyl ether. A foam can usually be obtained after co-evaporation.

Purify phosphoramidite monomers by flash column chromatography 7. Redissolve the crude material in a minimal amount of one of the following solvent systems: 1:99 (v/v) triethylamine/chloroform for araF-T 1:9:10 (v/v) triethylamine/dichloromethane/hexanes for araF-A and araF-C 1:99 (v/v) triethylamine/dichloromethane for araF-G. 8. Apply sample to a 3-cm-diameter chromatography column containing 50 g 230- to 400-mesh silica gel, preconditioned with the appropriate chromatography solvent (step 7). Column chromatography is performed according to the method of Still et al. (1978) using a small amount of air pressure at a rate of ∼1 inch of solvent per minute (also see APPENDIX 3E).



Supplement 10

9. Apply the sample to the column and begin eluting with the preconditioning solvent. a. For araF-C and araF-A: Elute in 1:9:10 (v/v) triethylamine/dichloromethane/hexanes. b. For araF-T: Follow a stepwise gradient of triethylamine/chloroform until a 1:31 ratio is reached. For 1 mmol crude sample, collect 100 mL eluate at 1% TEA, 100 mL at 2% TEA, and so on. c. For araF-G: Follow a step gradient of methanol/triethylamine/dichloromethane from 0:1:99 to 5:1:94 (v/v). For 1 mmol crude sample, collect 100 mL at 0% methanol, 100 mL at 1% methanol, and so on. Always add at least 0.5% of triethylamine to the chromatographic solvent mixtures used for elution and preconditioning of the column in order to avoid hydrolysis of the amidites to the H-phosphonates during purification. Step gradients indicated are only intended as guidelines and should be adjusted according to various factors including amount of crude product applied and flow rate of mobile phase.

10. Monitor fractions by TLC (APPENDIX phosphoramidite.

3D;

step 3) and pool those containing pure

11. Evaporate the phosphoramidite on a vacuum evaporator and dry under high vacuum to obtain a stable foam. If a foam does not form, try co-evaporation of the pure product with diethyl ether. Typically, a colorless or pale yellow foamy product forms easily. This can be stored in a vacuum desiccator over phosphorus pentoxide or at −20°C in a freezer for several months without decomposition. For araF-A (S.1a), the typical yield after chromatographic purification is ∼70% to 75%. TLC using 5:45:50 (v/v/v) triethylamine/hexanes/dichloromethane results in two spots corresponding to two isomers at Rf 0.22 and 0.33. 31P NMR (202.3 MHz, acetone-d6): 150.9 and 151.1, 19F NMR (470.27 MHz, acetone-d6, without external reference): yields −197.3 and –197.7 (ddd, J1′-F = 19 Hz, J′2′-F = 52 Hz, J3′-F = 19 Hz), and FAB-MS using an NBA matrix yields 846.34 [M+]. For araF-G (S.1b), the typical yield after chromatographic purification is ∼70% to 75%. TLC using 20:1 (v/v) chloroform/methanol yields a spot of Rf 0.54; 31P NMR (200.06 MHz, acetone-d6, 85% ortho-phosphoric acid as external reference): 151.8 and 151.0; 19F-NMR (282.32 MHz, acetone-d6, 99% trifluoroacetic acid as external reference): −119.7 and –119.4; and FAB-MS using an NBA matrix yields 822 [M+]. For araF-C (S.1c), the typical yield after chromatographic purification is ∼80% to 85%. TLC using 5:45:50 (v/v/v) triethylamine/hexanes/dichloromethane yields two spots, corresponding to two isomers, at Rf 0.26 and 0.34; 31P NMR (500 MHz, acetone-d6): 151.3 and 150.8; 19F NMR (470.27 MHz, acetone-d6, without external reference): −199.0 and –199.2 (ddd, J1′-F = 18 Hz, J′2′-F = 52 Hz, J3′-F = 18 Hz); and FAB-MS using an NBA matrix yields 828 [M+]. For araF-T (S.1d), typical yield after chromatographic purification is ∼90% to 95%. TLC using 10:9:1 (v/v/v) chloroform/ethyl acetate/ethanol yields two spots, corresponding to two isomers, at Rf 0.76 and 0.83; 31P NMR (500 MHz, acetone-d6): 151.3 and 150.8; 19F NMR (470.27 MHz, acetone-d6, without external reference): −199.0, –199.2 (ddd, J1′-F = 18 Hz, J′2′-F = 52 Hz, J3′-F = 18 Hz); and FAB-MS using an NBA matrix yields 822 [M+].




SOLID-PHASE ASSEMBLY OF PROTECTED araF PHOSPHORAMIDITES This protocol describes the setup and step-by-step synthesis of araF oligonucleotides. The methodology has been optimized for use with the Expedite 8909 DNA synthesizer equipped with a workstation (PerSeptive Biosystems), but can easily be adapted to other automated DNA synthesizers. All syntheses have been performed on 1-µmol scales, both for phosphodiester- and phosphorothioate-containing araF oligonucleotides. The coupling time for araF nucleosides requires a 15-min “wait” step, as compared to 1.5 min for deoxynucleosides on the Expedite instrument. Likewise, the detritylation time has been extended to 150 sec to allow for effective removal of the monomethoxytrityl (MMTr) protection as opposed to the more labile dimethoxytrityl (DMTr) groups that are commonly used with standard DNA monomers. For thio-araF-oligonucleotide synthesis, the overall success of oligomer synthesis was evaluated with different sulfur-transfer reagents, including 3H-1,2-benzodithiol-3-one-1,1-dioxide (Beaucage reagent; Iyer et al., 1990), 3-ethoxy-1,2,4-dithiazoline-5-one (EDITH; Xu et al., 1996), and 3-amino-1,2,4dithiazoline-5-thione (ADTT; Tang et al., 2000). In all cases, the concentration of sulfurization reagent was increased ∼5 fold as compared to customized S-DNA synthesis, and the time required for the oxidation was prolonged to 10 min as opposed to 1 to 2 min for S-DNA synthesis. Current procedures can easily be adapted to the synthesis of molecules with mixed nucleotide composition (e.g., DNA/araF chimeras of variable gap sequences).

BASIC PROTOCOL 2

Materials AraF phosphoramidites (see Basic Protocol 1) Argon gas Acetonitrile, anhydrous (see recipe) Liquid Reagent kit for the Expedite 8909 instrument (PerSeptive Biosystems): Acetonitrile wash and amidite diluent: anhydrous acetonitrile Activator solution: dissolve 1.8 g sublimed tetrazole (0.5 M) in 50 mL acetonitrile; store up to 2 wks at room temperature. Cap A (see recipe) Cap B (see recipe) Oxidizer solution (see recipe) Deblock solution: 15 g trichloroacetic acid in 500 mL dichloromethane; store up to 1 yr at room temperature Sulfurization reagent (see recipe) Amidite column (see recipe) Synthesizer vials with caps Vacuum desiccator containing phosphorus pentoxide Syringe Automated DNA synthesizer (e.g., Expedite 8909, Perseptive Biosystems) with trityl monitor Additional reagents and equipment for oligonucleotide synthesis (APPENDIX 3C) Synthesize phosphodiester and phosphorothioate araF oligonucleotides 1. Calculate the amount of araF phosphoramidites required for the synthesis of a desired sequence. The typical concentration of amidites used with the Expedite synthesizer corresponds to 50 mg/mL. Note that the volume of amidite solution delivered to the column during each coupling step is ∼240 ìL. However, an excess of 50 mg amidite (i.e., 1 mL solution) should be included when preparing each amidite solution in order to purge the lines of the synthesizer.



Supplement 10

Table 4.15.1 Automated 1 µmol Synthesis Cycle for the Synthesis of Phosphodiester araF Oligonucleotides on the Expedite 8909 Instrumenta

Functionb Amountc Time (sec) Description Detritylation of the support bound nucleoside: Trityl monitor on/off

—

1

Deblock

10

0

Deblock

70

150

Diverted wash A

40

0

Trityl mon. on/off Diverted wash A

— 40

1 0

Coupling of phosphoramidite: Wash 5 Activator 5

0 0

A + activator

6

0

A + activator

9

500

Wash

8

400

Wash

7

0

Wash A

20

0

Cap A and B

7

0

Cap A and B

6

15

Wash A

6

15

Wash A

14

0

Trityl analysis monitor is turned on for data collection Detritylation solution is rapidly delivered to column Slow, prolonged delivery of detritylation solution to column Lines are flushed with anhydrous acetonitrile (wash A) Trityl quantitation is stopped Lines are flushed with anhydrous acetonitrile Lines are flushed with acetonitrile Lines are flushed with tetrazole solution Rapid addition of the phosphoramidite (A) and tetrazole solutions to the column Coupling of the free activated phosphoramidite to the support-bound terminal nucleoside Slow delivery of acetonitrile to the column to purge remaining phosphoramidite solution Lines and column are washed with acetonitrile

Capping of the column: System is flushed with anhydrous acetonitrile Equal volumes of cap A and cap B solutions are rapidly delivered to column Slower delivery of cap A and cap B to column (to maximize capping efficiency) Slow delivery of acetonitrile to column to purge remaining cap A and B solutions Lines and column are flushed with anhydrous acetonitrile

Oxidation of phosphoramidite: Oxidizer

20

0

Wash A

15

0

Cap A and B

7

0

Wash A

30

0

Oxidant is delivered to the column to oxidize the newly formed phosphite triester linkages Lines and column are flushed with anhydrous acetonitrile

Capping of the column:


Equal volumes of cap A and B solutions are rapidly delivered to column Lines and column are flushed with anhydrous acetonitrile

aStandard DNA coupling cycles supplied by the manufacturer have been optimized for 2′-araF oligonucleotide assembly. bAll entries are as described in a typical cycle sequence for the Expedite instrument; see manufacturer’s directions for

further information.

cRepresents the number of pulses required for the corresponding step; each pulse has approximately 16 µL volume.



Table 4.15.2 Automated 1-µmol Synthesis Cycle for the Synthesis of Phosphorothioate araF Oligonucleotides on Expedite 8909 Instrumenta

Function

Amountb

Time (sec)

Description

Detritylation of the support-bound nucleoside: Trityl monitor on/off Deblock

1 10

1 0

Trityl analysis monitor is turned on for data collection Detritylation solution is rapidly delivered to column

Deblock

70

150

Slow, prolonged delivery of detritylation solution to column

Diverted wash A

40

0

Lines are flushed with anhydrous acetonitrile (wash A)

Trityl monitor on/off

0

1

Trityl quantitation is stopped

Diverted wash A

40

0

Lines are flushed with anhydrous acetonitrile

Coupling of phosphoramidite: Wash

5

0

Lines are flushed with acetonitrile

Activator

5

0

Lines are flushed with tetrazole solution

A + activator

6

0

Rapid addition of the phosphoramidite (A) and tetrazole solutions to the column

A + activator

9

500

Coupling of the free activated phosphoramidite to the support-bound terminal nucleoside

Wash

8

400

Slow delivery of acetonitrile to the line to purge remaining phosphoramidite solution

Wash

7

0

Lines and column are rapidly washed with acetonitrile

Wash A

20

0

System is flushed with anhydrous acetonitrile

Cap A and B

7

0

Equal volumes of cap A and cap B solutions are rapidly delivered to column

Cap A and B

6

15

Slower delivery of cap A and cap B to column (to maximize capping efficiency)

Wash A

6

15

Slow delivery of acetonitrile to column to purge remaining cap A and B solutions

Wash A

14

0

Lines and column are rapidly flushed with anhydrous acetonitrile


Sulfurization of phosphoramidite: Sox

15

0

Sulfurizing reagent is rapidly pulsed through column to sulfurize the newly formed phosphite triester linkages

Sox

25

400

Slower delivery of sulfurizing reagent through column (to maximize sulfur transfer efficiency)

Wash A

15

200

Slow delivery of acetonitrile to column to purge remaining sulfurizing solution from system

Wash A

15

0

System is rapidly flushed with acetonitrile

Cap A and B

7

0

Equal volumes of cap A and B solutions are rapidly delivered to column

Wash A

30

0

Lines and column are flushed with anhydrous acetonitrile


aStandard DNA coupling cycles supplied by the manufacturer have been optimized for 2′-araF oligonucleotide assembly. bRepresents the number of pulses required for the corresponding step; each pulse has approximately 16 µL volume.


Supplement 10

2. Weigh out the calculated amounts of phosphoramidites into the appropriate synthesizer vials and leave these to dry in a vacuum desiccator over phosphorus pentoxide overnight. 3. Flush the desiccator with argon gas before opening, and carefully remove the preweighed amidites. 4. Dissolve the phosphoramidites in anhydrous acetonitrile to a final concentration of 50 mg/mL. Use of syringe to transfer the acetonitrile through the rubber septum of each capped synthesizer vial to ensure the reagents are protected from humidity. Do not open the phosphoramidite vials until they are ready to be placed directly on the synthesizer.

5. Connect the solutions from the Liquid Reagents kit for the Expedite 8909 instrument, the phosphoramidite solutions, and the sulfurization reagent to the automated DNA synthesizer according to the manufacturer’s directions. The Expedite 8909 synthesizer is equipped with an extra position for the sulfur transfer reagent. If a synthesizer without this option is to be used, the sulfurization reagent may be placed at the position of the oxidizer, provided that the line has previously been washed with acetonitrile to avoid cross-contamination of the reagents.

6. Purge the lines of the synthesizer with all the solutions and solvents. 7. Install the appropriate amidite column. 8. Modify the synthetic cycles for araF phosphoramidites. Examples of synthesis cycles for phosphodiester and phosphorothioate araF oligonucleotides are given in Table 4.15.1 and Table 4.15.2. 9. Enter the sequence to be synthesized. 10. Carry out the assembly on the instrument according to the manufacturer’s instructions, choosing the “DMTr-off” (trityl-off) option. 11. Check the coupling efficiency periodically using the trityl monitor equipped with the instrument (also see APPENDIX 3C). 12. When the synthesis is complete, remove the column from the synthesizer and dry it under vacuum. 13. Deprotect and purify to obtain the desired oligonucleotide (see Basic Protocol 3). BASIC PROTOCOL 3


DEPROTECTION AND PURIFICATION OF araF OLIGONUCLEOTIDES This protocol describes the steps necessary for cleavage of araF sequences, assembled as described above (see Basic Protocol 2), from the solid support and removal of the protecting groups from the heterocyclic bases and phosphates. The procedure for the quantitation of isolated oligonucleotides is also described. AraF oligonucleotides can be successfully analyzed and purified by anion-exchange HPLC or denaturing gel electrophoresis (PAGE; UNIT 10.4 and APPENDIX 3B), as described below. Analytical amounts of material can easily be isolated by gel electrophoresis, while large amounts require HPLC purification. Desalting of the oligomers after gel or chromatographic isolation completes the purification procedure. There are two different methods that may be employed to accomplish this, either size-exclusion chromatography or reversed-phase SepPak C18 cartridge purification.



Materials Fully protected oligonucleotides, attached to the solid support of a synthesis column (see Basic Protocol 2) Ethanol, anhydrous 29% ammonium hydroxide Denaturing acrylamide gel stock solution (see recipe) Loading buffer (e.g., formamide/dye mix, UNIT 10.4) Sephadex G-25 column (Amersham Pharmacia Biotech) 1 M NaClO4 Anhydrous and 25% or 50% (v/v) acetonitrile 1.5 mL microcentrifuge tubes or screw-cap microcentrifuge tubes with O-ring seal Platform shaker 55°C heating block or water bath (optional) Vacuum evaporator (e.g., Savant Speedvac) with low and high vacuum sources UV spectrophotometer and cuvette 0.75- and 1.5-mm-thick gel plates TLC Merck silica plate (Kieselgel 60 F-254) Hand-held 254-nm UV lamp Camera with UV filter (optional) 0.45 µm hydrophilic fluid filter (Creative Medical) 0.22 µm membrane filter (Millipore; optional) Anion-exchange high-performance liquid chromatograph (HPLC) with: Gradient maker 0.5 mm × 7.5 cm Protein Pak DEAE 5PW column (Waters) Column heater Sep-Pak C18 cartridges (Waters Chromatography) 10 mL syringe Additional reagents and equipment for denaturing polyacrylamide gel electrophoresis (APPENDIX 3B and UNIT 10.4) Deprotect oligonucleotides 1. Remove the fully protected oligonucleotides attached to the solid support of the synthesis column, and place in a 1.5 mL microcentrifuge tube or a screw-cap microcentrifuge tube with O-ring seal (seal tightly) if deprotecting at increased speed and temperature (step 3). 2. Add 0.25 mL anhydrous ethanol and 0.75 mL of 29% ammonium hydroxide. 3. Place the tube on a platform shaker for 48 hr at room temperature. Alternatively, deprotect by incubating in a 55°C heating block or waterbath for 15 hr. 4. Once the deprotection is complete, cool the microcentrifuge tube ∼1 hr in the freezer (i.e., −20°C) before opening. 5. Microcentrifuge the samples briefly so that the CPG beads are allowed to settle, then transfer the supernatant to another microcentrifuge tube. Wash the CPG with ethanol twice and place the supernatant from each wash in additional microcentrifuge tubes. 6. Dry the samples in a vacuum evaporator, applying low vacuum first to remove all traces of ammonia. Introducing high vacuum to the ammonia/ethanol solution may cause bumping of the liquid and loss of the sample. Synthesis of Modified Oligonucleotides and Conjugates


Supplement 10

7. When the ammonia has been evaporated and sample volumes are decreased enough, combine the supernatants from each microcentrifuge tube. Evaporate the final sample to dryness under high vacuum. Quantitate oligonucleotide 8. Dissolve the oligonucleotide from the previous step in 1 mL deionized water. Remove a 10-µL aliquot and dilute this to 1 mL with deionized water in a UV cuvette. 9. Measure the absorption of the sample at 260 nm. Calculate the amount of crude oligonucleotide obtained in A260 units. Typically, a 1-ìmol synthesis may give yields of 100 to 180 A260 units of crude material, depending on the sequence length and composition.

Analyze crude oligonucleotide by denaturing PAGE 10. Prepare a 24% denaturing polyacrylamide gel on a 0.75-mm-thick analytical gel plate for gel electrophoresis (APPENDIX 3B and UNIT 10.4). 11. Prepare samples of oligonucleotides to be analyzed (e.g., 0.5 to 1 A260 units per analysis). Dissolve each of these in 10 µL loading buffer. 12. Load the samples into wells and run these alongside loading buffer alone, which is placed in the first and last lanes of the gel. Connect the gel to the power supply and start running at 200 V until the dye has fully diffused into the gel. Increase the voltage to 500 V and maintain at this setting until the faster running dye marker has traveled 3⁄ down the length of the gel. 4 13. Turn off the current and dismantle the gel apparatus. Put the gel in plastic wrap. Place a fluorescing TLC Merck silica plate under the gel and examine by illumination with a hand-held 254-nm UV lamp. If necessary, photograph the gel using a camera with a UV filter. Oligonucleotides absorb UV light and appear as dark bands in the gel against a fluorescent background. Typically, crude FANA oligonucleotides, both diester and thio, consist of one intense band. The presence of multiple bands below the desired product usually indicates poor coupling efficiency.

Purify oligonucleotides by denaturing PAGE and desalt using Sephadex G-25 14. Prepare a 24% to 16% denaturing polyacrylamide gel on a 1.5-mm-thick gel plate for preparative analysis (APPENDIX 3B and UNIT 10.4). Compared to analytical separations, a lower percentage of acrylamide may be used in preparative analyses (from 16% to 24%), depending on the quality of the synthesis. For example, a crude oligonucleotide which migrates as one main band in analytical compositions can be sufficiently purified using 16% acrylamide in the preparative run. When multiple bands corresponding to failed sequences are present in the gel, a higher percentage of acrylamide solution should be used.

15. Prepare and dry the oligonucleotide sample to be purified (∼20 to 30 A260 units). Dissolve the sample in 100 µL loading buffer. Sample loading to the gel may vary, depending on the quality of the synthesis.

16. Load the sample on the gel and repeat steps 12 and 13 above. 17. Using a razor blade, cut out the segment of gel containing the band of interest. Synthesis of 2′-Deoxy-2′-fluoroβ-D-oligoarabinonucleotides (2′F-ANA)

18. Transfer the excised band to a 15-mL conical plastic tube with screw cap and crush with a spatula. Add 10 mL water and leave the tube on a platform shaker overnight to extract the oligonucleotide from the gel.



19. Filter the supernatant through a 0.45-µm hydrophilic fluid filter and concentrate on a vacuum evaporator. 20. Desalt the sample using a Sephadex G-25 column according to the manufacturer’s directions. 21. Quantitate the amount of oligonucleotide isolated via UV spectroscopy. 22. Lyophilize the oligonucleotide. The purified, dry oligonucleotide can be stored for long periods (e.g., >1 yr) at −20°C. If the oligonucleotide is to be used in combination with microbially sensitive materials (e.g., cell culture), filtration through a 0.22-ìm filter is recommended. Collect the filtrate in previously autoclaved tubes and lyophilize.

Analyze and purify oligonucleotide by anion-exchange HPLC and desalt on SepPak C18 cartridge 23. Prepare the anion-exchange HPLC by setting the detector wavelength to 260 nm, heating the column to 50°C, and equilibrating with buffer A (deionized water) using a flow rate of 1 mL/min or as specified by the manufacturer. See UNIT 10.5 and legends to Figures 4.15.2 to 4.15.4 for more details.

A

0.35 0.30

AU

0.25 0.20 0.15 0.10 0.05 0.00 0.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00

80.00

90.00

60.00

70.00

80.00

90.00

Minutes

B 0.50 0.40

AU

0.30 0.20 0.10 0.00 0.00

10.00

20.00

30.00

40.00

50.00

Minutes

Figure 4.15.2 Ion-exchange HPLC analysis of crude S-oligonucleotide ATA TCC TTg TCg TAT CCC (cap letters represent araF nucleotides, DNA residues in lowercase). (A) Sulfurization with Beaucage reagent. The small peak at ∼58 min represents S-oligonucleotides with one P-O insertion. The main peak at ∼62 min represents full S-FANA. (B) Sulfurization with ADTT. System: Waters 600E Multisolvent Delivery System with Waters 486 Tunable Absorbance detector and oven, driven by Millennium (V 3.20) software; column: Waters 0.5-mm × 7.5-cm Protein Pak DEAE 5PW, 50°C; solvent A: H2O; solvent B: 1 M NaClO4; flow rate: 1 mL/min; detection: 260 nm; gradient conditions: 0% B for 9 min, 0% to 15% B in 3 min, 15% to 50% B in 60 min, 50% to 80% B in 2 min, hold 80% B 10 min, 80% to 0% B in 2 min, hold 0% B 10 min.



Supplement 10

24. Load ∼1 A260 unit crude oligonucleotide sample (steps 8 and 9) to the column. Wash 5 min with buffer A, then apply a linear gradient of 0% to 25% buffer B (1M NaClO4) in 60 min (i.e., 0.42%/min) to elute diester araF oligonucleotides, or of 0% to 50% buffer B in 60 min (i.e., 0.83%/min) for S-FANA oligonucleotides. Wash the column and re-equilibrate. The desired fraction should elute close to the end of the applied gradient.

25. Optimize the conditions for the next run. Examples of suitable chromatographic gradients that may be used are given in the legends to Figures 4.15.2 to 4.15.4. 26. Continue purification of the next portions of oligonucleotide using the previously optimized conditions. Increase the amount of oligonucleotide loaded to the column and set the detector at 280 to 290 nm depending on the amount of the sample to be loaded. Use higher wavelengths for larger samples to avoid saturating the UV detector. Sample loading to the column depends on the quality of the reaction mixture and type of oligonucleotide. Typical loading amounts for phosphodiester FANA oligonucleotides range from 10 to 30 A260 units of crude reaction mixture using the conditions described above. Loading for S-FANA oligonucleotides can be increased considerably as a result of their stronger adsorption on the ion-exchange column. The authors have been able to purify ∼100 A260 units of crude mixture in one loading.

27. Quantitate the total amount of oligonucleotide obtained.

A 0.060 0.050

AU

0.040 0.030 0.020 0.010 0.000 0.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00

80.00

50.00

60.00

70.00

80.00

Minutes

B 1.00

AU

0.80 0.60 0.40 0.20 0.00 0.00

10.00

20.00

30.00

40.00 Minutes


Figure 4.15.3 Ion-exchange HPLC analysis of (A) crude and (B) purified S-oligonucleotide CTC TAg cgt ctT AAA (cap letters represent araF nucleotides, lowercase letters represent DNA residues). System: Waters Binary HPLC Pump with Waters 2487 Dual λ Absorbance detector, equipped with in-line degasser and oven, driven by Breeze (V 3.20) software; column: Waters 0.5-mm × 7.5-cm Protein Pak DEAE 5PW, 50°C; solvent A: H2O; solvent B: 1 M NaClO4; flow rate: 1 mL/min; detection: 260 nm; gradient conditions: 0% B for 9 min, 0% to 15% B in 3 min, 15% to 50% B in 45 min, 50% to 80% B in 2 min, hold 80% B 10 min, 80% to 0% B in 2 min, hold 0% B 10 min. Peaks at 2′F-ANA/ANA >> bc-ANA (rigid). These trends, together with the activity enhancements observed upon introducing relatively flexible deoxy residues in the more rigid FANA strand (vide supra), prompted the authors to investigate the potential synergistic attributes of intermingling sites of flexibility within the 2′FANA polymers. Indeed, the incorporation of acyclic linkers with high degrees of flexibility in the AON strand, either a 2′,5′-linked secouridine residue S.3 or a butanediol linker S.4 (Figure 4.15.5), appear to better accommodate the stringent conformational requirements of the enzyme without disrupting other fundamental recognition elements within the enclosing AON segments (Mangos et al., unpub. observ.). Sizable increases in enzyme-mediated scission of target RNA occur upon substituting an acyclic linker in the middle of a known RNase H–competent analog (i.e., DNA or 2′F-ANA). Moreover, the susceptibility of the hybrid to RNA hydrolysis remains fully operative in other sequence contexts with up to eight-fold enhancements over strands lacking the acyclic insertion. For example, in sequences wherein the acyclic inserts are moved along the 2′F-ANA scaffold, a significant change in activity is observed that depends specifically on the insertion site. The greatest activity occurs in constructs with a centrally placed acyclic linker, although differences between 5′- versus 3′-end insertion are also observed, with a greater amount of enzyme proc-



essing occurring when the linker is placed near the 3′-end of the AON. The authors therefore speculate that RNase H induction can remarkably be improved by introducing subtle changes in local 2′F-ANA strand dynamics that enable better adherence and/or processing of the hybrid substrate by the enzyme. These and future studies with 2′F-ANA and their constructs should provide considerable insights as to the structural factors that comprise the “optimal” AON/RNA substrate. For these reasons, the protocols of this unit provide a method of quickly and conveniently preparing 2′-fluoroarabino oligonucleotides with any one or all of the structural modifications described herein. Although linker technologies are not discussed, these are quite simple to implement and exemplify the diversity with which the protocols can be applied to new AON designs that further exploit the interesting properties of 2′F-ANA.

Critical Parameters The synthesis of araF-phosphoramidites, as for other phosphoramidites, is very moisture and acid sensitive. Therefore, all reagents and apparatus should be anhydrous. If the product has to be isolated by flash column chromatography (APPENDIX 3E), it is absolutely necessary to include triethylamine in all solvent systems used to equilibrate the column and for subsequent sample elution. Heating the product above 40°C is not recommended during evaporation. If the phosphoramidite has been stored for a long period of time prior to oligomer synthesis, its purity should be verified by TLC (APPENDIX 3D) or 31P NMR. If necessary, the phosphoramidite can be repurified by flash chromatography. Phosphoramidite oligonucleotide chemistry is extremely water sensitive. All coupling reagents should be absolutely dry, and fresh reagents should be prepared as these are critical to the success of the synthesis. It is not recommended to keep the reagents on the synthesizer for longer than 1 week. Furthermore, the synthesizer should be in good working condition. All lines should be purged with each reagent and solvent prior to starting a synthesis and cleaned with acetonitrile afterward to prevent crystallization and blocking of the lines. If synthesis is performed on the Expedite instrument, a relatively low concentration of phosphoramidites can be used without loss of coupling efficiency. For example, commercial RNA protocols require 50 mg/mL phosphoramidite concentration, which is almost 3-

fold lower than for RNA synthesis on older instruments (e.g., 1.5 M amidite solutions are required for use with the ABI 381A DNA synthesizer). Phosphoramidite and activator solutions are continuously delivered to the column during the coupling step. Such permanent delivery is performed by slow pulses of reagents to the column, which permanently renew them and thereby increase the efficiency of the synthesis. If a synthesizer with stationary delivery of reagents is to be used, the amidite concentration should be increased to that used for RNA synthesis (or slightly lower). The authors were able to successfully synthesize diester araF oligonucleotides on the ABI 381A DNA instrument using 0.1 to 0.15 M concentrations of phosphoramidites. Many different sulfur-transfer reagents for S-DNA have been documented in the literature over the last several years. However, none of these reagents are able to give 100% P-S oxidation. As a consequence, a small percentage of phosphodiester bonds are inevitably present in S-oligonucleotides. The diester fragments are much less stable to hydrolysis, making them more susceptible to nucleolytic cleavage in cell culture, which limits certain biological applications of these antisense oligonucleotides. Fortunately, pure S-oligonucleotides can easily be analyzed and purified from P-O containing S-oligonucleotides by anion-exchange HPLC (Bergot and Egan, 1992). This method enables the isolation of S-oligonucleotides that contain even a single P-O linkage within the strand. The percentage of P-O bonds in the product is dependent on the source of sulfurizing reagent used, as different sources may lead to different percentages in the fully synthesized oligomers. Although undesired oxidation is not very problematic in S-DNA synthesis, 2′F-ANA oligonucleotides are intrinsically more difficult to sulfurize as a consequence of the electron-withdrawing effects of the fluorine substituent in the sugars. This limitation may be overcome by increasing the concentration of the sulfur-transfer reagent, accompanied by extended reaction times in order to obtain S-oligonucleotides of better quality. The authors of the current protocol applied several sulfurization reagents to S-araF synthesis, namely the Beaucage reagent, EDITH, and ADTT, and have found that the former normally gives a higher percentage of P-O bonds relative to the latter two reagents. EDITH and ADTT work very similarly and give oligomers of low P-O compositions. Figure 4.15.2 shows a chromatographic profile for a crude S-DNA/S-araF chimera synthesized



Supplement 10

using either the Beaucage reagent or ADTT. As is shown, the S-FANA oligonucleotide synthesized with ADTT displays more uniform P-S incorporation (i.e., P-O insertions are minimal in this oligomer as detected by chromatography). However, EDITH and ADTT reagents are very expensive and not widely available commercially, which makes the Beaucage reagent a suitable alternative. It should also be emphasized that the limitations observed with the Beaucage reagent are reserved only for 2′FANA polymers; experimentally, sulfurization with this reagent proceeds much more efficiently with other types of modified oligonucleotides (e.g., S-DNA), thereby making it the reagent of choice for those applications. The amount of contaminating P-O linkages within S-FANA oligonucleotides does not typically exceed 1% per phosphate, even when using the Beaucage reagent, but may sometimes be significantly larger. Unfortunately, PO insertions cannot be detected by PAGE. Oligonucleotides with multiple phosphodiester bonds will migrate as a single narrow band through the gel, but present multiple peaks upon anion-exchange chromatographic analysis. Consequently, chromatographic purification of S-FANA oligonucleotides is strongly recommended. In general, both HPLC and PAGE purification can be applied to araF oligonucleotides. Sequences that consist entirely of phosphodiester backbones may be purified either by HPLC or PAGE, depending upon the scale used in the synthesis. If 1 to 20 A260 units are required, PAGE purification is sufficient; for preparative amounts ranging from 100 to 200 A260 units of purified oligomer, HPLC is recommended. For S-FANA oligonucleotides, ion-exchange purification is most convenient, regardless of the amount to be purified.

Troubleshooting


Low coupling efficiency. Several conditions can cause poor monomer coupling yields. (1) The reagents and acetonitrile may contain water. This is minimized by drying the phosphoramidites and tetrazole in a vacuum desiccator over phosphorus pentoxide for one or two days prior to use. (2) Lines in the synthesizer may be partially blocked. Check delivery of each reagent to the lines by monitoring flow rates. Perform routine DNA synthesis first and monitor coupling efficiencies by trityl analysis. HPLC analysis of S-araF oligonucleotide gives multiple product peaks. This problem is usually caused by incomplete sulfurization, es-

pecially if the Beaucage reagent has been used. If so, the product may still migrate as one band under PAGE conditions. This can be circumvented by increasing the concentration of the Beaucage reagent or using another sulfurizing reagent. Beaucage reagent precipitates. The Beaucage reagent is known to be extremely sensitive to the quality of solvent and is decomposed rather easily. If this process is rapid, a switch to an acetonitrile source of better quality is recommended. Furthermore, the glass bottle for Beaucage reagent should be silanized by the investigator (APPENDIX 2A; Iyer et al., 1990) or obtained in silanized form from various commercial sources. Alternatively, acetonitrile solutions of Beaucage reagent can be placed in a 15-mL plastic conical tube, which is then deposited within an appropriate glass bottle and attached to the gene-machine to prevent direct contact of the reagent with the glass.

Anticipated Results Using araF phosphoramidites as building blocks in conjunction with the methods presented here, it should be possible to routinely obtain araF oligonucleotides with yields ranging from 120 to 180 A260 units of crude material for 1-µmol scale syntheses, and isolated yields of ∼30 to 100 A260 units. The sulfurization step with EDITH or ADTT reagents helps to minimize the extent of P-O insertions in S-oligonucleotides. These considerations, together with the optimized ion-exchange HPLC chromatographic procedures, should enable the facile isolation of araF oligonucleotides of high purity.

Time Considerations Preparation of araF phosphoramidites from the appropriately protected nucleosides usually takes ∼1 day per monomer. Depending on the type of chemistry (diester- versus thio-araF-oligonucleotides, or DNA/araF oligonucleotide chimeras), the time required for assembly of a 20-mer oligonucleotide (typical for antisense applications) on the Expedite instrument is from 6 to 12 hr. An additional 2 days are required to perform deprotection, followed by 3 to 5 days for isolation, desalting, and final analysis as described (see Basic Protocol 3), if purification is required.



Literature Cited Altmann, K.-H., Kesselring, R., Francotte, E., and Rihs, G. 1994a. 4′,6′-Methano carbocyclic thymidine: A conformationally constrained building block for oligonucleotides. Tetrahedron Lett. 35:2331-2334. Altmann, K.-H., Imwinkelried, M., Kesselring, R., and Rihs, G. 1994b. 1′,6′-Methano carbocyclic thymidine: Synthesis, X-ray crystal structure, and effect on nucleic acid duplex stability. Tetrahedron Lett. 35:7625-7628. Berger, I., Tereshko, V., Ikeda, H., Marquez, V.E., and Egli, M. 1998. Crystal structures of B-DNA with incorporated 2′-deoxy-2′-fluoro-arabinofuranosyl thymines: Implications of conformational preorganization for duplex stability. Nucl. Acids Res. 26:2473-2480. Bergot, B.J. and Egan, W. 1992. Separation of synthetic phosphorothioate oligonucleotides from their oxygenated (phosphodiester) defect species by strong-anion-exchange high-performance liquid chromatography. J. Chromatogr. 599:35-42. Christensen, N.K., Petersen, M., Nielson, P., Jacobsen, J.P., Olsen, C.E., and Wengel, J. 1998. A novel class of oligonucleotide analogues containing 2′-O,3′-C-linked [3.2.0]bicycloarabinonucleoside monomers: Synthesis, thermal affinity studies and molecular modeling. J. Am. Chem. Soc. 120:5458-5463. Cook, P.D. 1998. Second generation antisense oligonucleotides: 2′-modifications. Annu. Rep. Med. Chem. 33:313-325. Crouch, R.J. and Toulmé, J.J. (eds.) 1998. Ribonucleases H. INSERM, Paris. Damha, M.J., Meng, B., Yannopoulos, C.G., Wang, D., and Just, G. 1995. Structural basis for the RNA selectivity of oligonucleotides containing alkylsulfide internucleoside linkages and 2′-Osubstituted 3′-deoxyribose. Nucl. Acids Res. 19:3967-3973. Damha, M.J., Wilds, C.J., Noronha, A., Brukner, I., Borkow, G., Arion, D., and Parniak, M.A. 1998. Hybrids of RNA and arabinonucleic acids (ANA and 2′F-ANA) are substrates of ribonuclease H. J. Am. Chem. Soc. 120:12976-12977. Damha, M.J., Noronha, A.M., Wilds, C.J., Trempe, J.-F., Denisov, A., and Gehring, K. 2001. Properties of arabinonucleic acids (ANA & 2′FANA): Implications for the design of antisense therapeutics that invoke RNase H cleavage of RNA. Nucleosides Nucleotides 20:429-440. Giannaris, P.A. and Damha, M.J. 1994. Hybridization properties of oligoarabinonucleotides. Can. J. Chem. 72:909-918. Iyer, R.P., Phillips, L.R., Egan, W., Regan, J.B., and Beaucage, S.L. 1990. The automated synthesis of sulfur-containing oligodeoxyribonucleotides using 3H-1,2-benzodithiol-3-one 1,1-dioxide as a sulfur-transfer reagent. J. Org. Chem. 55:46934699.

Lebedeva, I. and Stein, C.A. 2001. Antisense oligonucleotides: Promise and reality. Annu. Rev. Pharmacol. Toxicol. 41:403-419. Lima, W.F. and Crooke, S.T. 1997. Binding affinity and specificity of Escherichia coli RNase H1: Impact on the kinetics of catalysis of antisense oligonucleotide-RNA hybrids. Biochemistry 36:390-398. Lok, C.-N., Viazovkina, E., Min, K-L., Nagy, E., Wilds, C.J., Damha, M.J., and Parniak, M.A. 2002. Potent gene-specific inhibitory properties of mixed-backbone antisense oligonucleotides comprised of 2′-deoxy-2′-fluoro-D-arabinose and 2′-deoxyribose nucleotides. Biochemistry 41:3457-3467. Mangos, M.M. and Damha, M.J. 2002. Flexible and frozen sugar-modified nucleic acids: modulation of biological activity through furanose ring dynamics in the antisense strands. Curr. Topics Med. Chem. 2:1145-1169. Manoharan, M. 1999. 2′-Carbohydrate modifications in antisense oligonucleotide therapy: Importance of conformation, configuration and conjugation. Biochim. Biophys. Acta 1489:117130. Minasov, G., Teplova, M., Nielsen, P., Wengel, J., and Egli, M. 2000. Structural basis of cleavage by RNase H of hybrids of arabinonucleic acids and RNA. Biochemistry 39:3525-3532. Myers, N.M. and Dean, K.J. 2000. Sensible use of antisense: How to use oligonucleotides as research tools. TIPS 21:19-23. Nakamura, H., Oda, Y., Iwai, S., Inoue, H., Ohtsuka, E., Kanaya, S., Kimura, S., Katsuda, C., Katayanagi, K., Morikawa, K., Miyashiro, H., and Ikehara, M. 1991. How does RNase H recognize a DNA:RNA hybrid? Proc. Natl. Acad. Sci. U.S.A. 88:11535-11539. Noronha, A. and Damha, M.J. 1998. Triple helices containing arabinonucleotides in the third (Hoogsteen) strand: Effects of inverted stereochemistry at the 2′-position of the sugar moiety. Nucl. Acids Res. 26:2665-2671. Noronha, A.M., Wilds, C.J., Lok, C.-N., Viazovkina, K., Arion, D., Parniak, M.A., and Damha, M.J. 2000. Synthesis and biophysical properties of arabinonucleic acids (ANA): Circular dichroic spectra, melting temperatures and ribonuclease H susceptibility of ANA:RNA hybrid duplexes. Biochemistry 39:7050-7062. Oda, Y., Iwai, S., Ohtsuka, E., Ishikawa, M., Ikehara, M., and Nakamura, H. 1993. Binding of nucleic acids to E. coli RNase HI observed by NMR and CD spectroscopy. Nucl. Acids Res. 21:46904695. Plavec, J., Thibaudeau, C., and Chattopadhyaya, J. 1994. How does the 2′-hydroxy group drive the pseudorotational equilibrium in nucleoside and nucleotide by the tuning of the 3′-gauche effect? J. Am. Chem. Soc. 116:6558-6560. Synthesis of Modified Oligonucleotides and Conjugates


Supplement 10

Sangvhi, Y.S. 1998. Synthesis of nitrogen containing linkers for antisense oligonucleotides. In Carbohydrate Mimics (Y. Chapleur, ed.) pp. 523536. Wiley-VCH, Germany. Schmit, C., Bèvierre, M-O., De Mesmaeker, A., and Altmann, K.-H. 1994. The effects of 2′- and 3′-alkyl substituents on oligonucleotide hybridization and stability. Bioorg. Med. Chem. Lett. 4:1969-1974. Shen, L.X., Kandimalla, E.R., and Agrawal, S. 1998. Impact of mixed-backbone oligonucleotides on target binding affinity and target cleaving specificity and selectivity by E. coli RNase H. Bioorg. Med. Chem. 6:1695-1705. Sørensen, M.D., Kvaernø, L., Bryld, T., Håkansson, A.E., Verbeure, B., Gaubert, G., Herdewijn, P., and Wengel, J. 2002. Alpha-L-ribo-configured locked nucleic acid (alpha-L-LNA): Synthesis and properties. J. Am. Chem. Soc. 124:21642176. Still, W.C., Kahn, M., and Mitra, A. 1978. Rapid chromatographic technique for preparative separation with moderate resolution. J. Org. Chem. 43:2923-2925. Tang, J.-Y., Han Y., Tang, J.X., and Zhang, Z. 2000. Large scale synthesis of oligonucleotide phosphorothioates using amino-1,2,4-dithiazoline-5thione as an efficient sulfur-transfer reagent. Org. Proc. Dev. 4:194-198.

Wilds, C.J. and Damha, M.J. 1999. Duplex recognition by oligonucleotides containing 2′-Deoxy2′-fluoro-D-arabinose and 2′-deoxy-2′-fluoroD-ribose. Intermolecular contacts versus sugar puckering in the stabilization of triple helical complexes. Bioconjug. Chem. 10:299-305. Wilds, C.J., and Damha M.J. 2000. 2′-deoxy-2′fluoro-β-D-arabinonucleosides and oligonucleotides (2′F-ANA): Synthesis and physicochemical studies. Nucl. Acids Res. 28:36253635. Xu, Q., Musier-Forsyth, K., Hammer, R.P., and Barany, G. 1996. Use of 1,2,4-dithiazolidine3,5-dione (DtsNH) and 3-ethoxy-1,2,4-dithiazoline-5-one (EDITH) for synthesis of phosphorothioate-containing oligodeoxyribonucleotides. Nucl. Acids Res. 24:1602-1607.

Key References Crouch and Toulmé, 1998. See above A comprehensive book on ribonucleases H, their sources, properties, biological utility, and antisense applications.

Thibaudeau, C. and Chattopadhyaya, J. 1997. The discovery of intramolecular stereoelectronic forces that drive the sugar conformation in nucleosides and nucleotides. Nucleosides Nucleotides 16:523-529.

Damha et al., 2001. See above.

Thibaudeau, C., Plavec, J., Garg, N., Papchikhin, A., and Chattopadhyaya, J. 1994. How does the electronegativity of the substituent dictate the strength of the gauche effect? J. Am. Chem. Soc. 116:4038-4043.

Freier, S.M. and Altmann, K.-H. 1997. The ups and downs of nucleic acid duplex stability: Structurestability studies on chemically-modified DNA:RNA duplexes. Nucl. Acids Res. 25:44294443.

Trempe, J.F., Wilds, C.J., Denisov, A.Y., Pon, R.T., Damha, M.J., and Gehring, K. 2001. NMR solution structure of an oligonucleotide hairpin with a 2′F-ANA/RNA stem: Implications for RNase H specificity toward DNA/RNA hybrid duplexes. J. Am. Chem. Soc. 123:4896-4903.

An extensive resource that relates duplex stabilities to nucleotide structure with 197 examples of oligonucleotide modifications.

Uhlmann, E. and Peyman, A. 1990. Antisense oligonucleotides: A new therapeutic principle. Chem. Rev. 90:543-584. Venkateswarlu, D. and Ferguson, D.M. 1999. Effects of C2′-substitution on arabinonucleic acid structure and conformation. J. Am. Chem. Soc. 121:5609-5610. Walder, R.Y. and Walder, J.A. 1988. Role of RNase H in hybrid-arrested translation by antisense oligonucleotides. Proc. Natl. Acad. Sci. U.S.A. 85:5011-5015.


Wengel, J., Koshkin, A., Singh, S.K., Nielsen, P., Meldgaard, M., Rajwanshi, V.K., Kumar, R., Skouv, J., Nielsen, C.B., Jacobsen, J.P., Jacobsen, N., and Olsen, C.E. 1999. LNA (Locked nucleic acid). Nucleosides Nucleotides 18:13651370.

Wang, J., Verbeure, B., Luyten, I., Lescrinier, E., Froeyen, M., Hendrix, C., Rosemeyer, H., Seela, F., Aerschot, A.V., and Herdewijn, P. 2000. Cyclohexene nucleic acids (CeNA): Serum stable oligonucleotides that activate RNase H and increase duplex stability with complementary RNA. J. Am. Chem. Soc. 122:8595-8602.

Highlights the role of the 2′-sugar position in ANA and 2′F-ANA conformations, and the origins of RNase H activity.

Kvaernø, L. and Wengel, J. 2001. Antisense molecules and furanose conformations—is it really that simple? Chem. Commun. 1419-1424. A concise review briefly comparing RNA binding versus cleavage with promising antisense candidates. Lebedeva and Stein, 2001. See above. Insightful discussion of antisense design that focuses on in vivo toxicity from nonspecific interactions and potential irrelevant cleavage of nontargeted RNA.

Contributed by Ekaterina Viazovkina, Maria M. Mangos, Mohamed I. Elzagheid, and Masad J. Damha McGill University Montreal, Canada



Chemistry of CpG DNA The innate immune system of vertebrates has evolved to recognize specific pathogenassociated molecular patterns (PAMPs) present in invading microorganisms through pattern recognition receptors (PRRs; Lien and Ingalls, 2002). One of several such PAMPs is the unmethylated CpG dinucleotide present in specific sequence contexts (CpG motifs) in pathogenic microorganisms (Hemmi et al., 2000). On sensing a CpG motif, PRRs trigger complex signal transduction pathways that ultimately activate a number of transcription factors, including NF-κB and AP-1. These induce specific patterns of gene expression associated with the development and maintenance of immune responses. The immune responses to bacterial and synthetic oligonucleotides containing CpG motifs include proliferation of B cells, production of cytokines IL-12, γ-IFN, IL-6, and TNF-α (Klinman et al., 1996; Zhao et al., 1997), and production of costimulatory molecules by monocytes/macrophages, B cells, dendritic cells (DCs), and natural killer (NK) cells.

HISTORY Tokunaga and coworkers were the first to report that DNA from Mycobacterium bovis induces production of interferons α, β, and γ (IFN-α, -β, and -γ), augments NK cell activity, and shows antitumor activity (Tokunaga et al., 1984). The same authors, using short synthetic DNA fragments, showed that palindromic sequences containing CpG dinucleotides efficiently induce NK cell activity and induce IFNs (Yamamoto et al., 1992a). Subsequent studies demonstrated that bacterial DNA, but not mammalian DNA, induces murine B cell proliferation and cytokine secretion (Messina et al, 1991; Yamamoto et al., 1992b). Later studies showed that unmethylated CpG dinucleotides in specific sequence contexts present in bacterial DNA, synthetic oligodeoxyribonucleotides, and DNA vaccines are responsible for the observed immune responses (Krieg et al., 1995; Sato et al., 1996). Compared to bacterial DNA, the occurrence of CpG motifs in vertebrate DNA is sparse. In addition, the vertebrate C residues usually carry a 5-methyl substituent (Bird, 1986), which enables the immune system to distinguish between its own DNA and bacterial DNA that signals an infection.

UNIT 4.16

RECEPTORS The recognition of CpG DNA has been shown to occur through Toll-like receptor 9 (TLR9), which belongs to a family of proteins called TLRs (Hemmi et al., 2000). The TLRs function as PRRs to initiate the innate immune response against invading microorganisms. However, direct evidence for binding of CpG DNA to TLR9 has not yet been documented. Although most other TLRs are membrane receptors, growing evidence suggests that TLR9 is localized in the cytoplasm (Takeshita et al., 2001; Ahmad-Nejad et al., 2002). Therefore, cellular uptake and endosomal localization seem to be prerequisites for CpG DNA recognition (Stacey et al., 2000).

Sequence and Structural Specificity of TLR9 TLR9 proteins from different vertebrates vary in the CpG DNA sequences they recognize (Bauer et al., 2001a). Therefore, a CpG motif that is active in one species may not be in another. For example, mouse TLR9 prefers an unmethylated CpG dinucleotide flanked by two purine bases on the 5′ side and two pyrimidine bases on the 3′ side, such as GACGTT (Krieg et al., 1995). CpG dinucleotides preceded by C or followed by G are generally less active in mice. Human immune cells respond optimally to GTCGTT or TTCGTT motifs (Hartmann et al., 2000). Certain other sequences, such as a palindromic AACGTT sequence, are known to induce immune responses in both mouse and human systems (Yamamoto et al., 1992a; Van Uden and Raz, 2000). Thus, TLR9 variants from different species recognize CpG dinucleotides flanked by a variety of sequences, though to different extents (Rankin et al., 2001; Zhang et al., 2001; Wernette et al., 2002).

Signaling Events Though the evidence suggests that TLR9 is the receptor for phosphorothioate CpG DNAs, the role of TLR9 in mediating the observed effects of phosphodiester CpG oligonucleotides and bacterial DNA is not yet clear. The signaling events initiated by TLR9 in response to CpG DNA include recruitment of MyD88, IRAK, and TRAF6, which leads to activation of IκB kinase, MAP kinase, the stress kinases JNK-1 and -2, and p38, which in turn causes activation of transcription factors ATF-

Contributed by Ekambar R. Kandimalla and Sudhir Agrawal Current Protocols in Nucleic Acid Chemistry (2003) 4.16.1-4.16.14 Copyright © 2003 by John Wiley & Sons, Inc.



2, AP-1, and NF-κB (Fig. 4.16.1). The activated transcription factors induce the synthesis of several regulatory cytokines and costimulatory molecules.

SIGNIFICANCE OF CpG DINUCLEOTIDES AND CHEMISTRY OF CpG DNA A CpG dinucleotide present in specific sequence contexts is essential for immunostimulatory activity, whereas the inverted dimer, GpC, is inactive. As discussed above, the flanking sequences play an important role in determining activity of CpG DNA (Yamamoto et al., 1992a; Krieg et al., 1995; Pisetsky, 1999). Recently, other structural features that influence the activity of synthetic oligonucleotides have been reported, including the nature of the internucleotide linkage, the nature and conformation of the sugar ring, modification or removal of nucleobases, accessibility of the 5′ end, and the nature and size of any 5′-terminal blocking group.

DNA Backbone Shorter CpG DNA molecules with unmodified phosphodiester backbones may elicit potent immune stimulation in vitro (Sonehara et al., 1996; Iho et al., 1999). However, phosphorothioate-modified oligonucleotides are used more commonly to prevent rapid degradation by nucleases present in cells. Although phosphodiester and phosphorothioate CpG DNAs elicit superficially similar immune re-

sponses, recent studies showed certain distinct differences in the actions of these two backbones (Ballas et al., 2001; Rothenfusser et al., 2001; Verthelyi et al., 2001, 2002; Dalpke et al., 2002; Gursel et al., 2002), which are not well understood (Fig. 4.16.2 and Table 4.16.1). Phosphodiester backbone Phosphodiester CpG DNAs containing palindromic structures and/or poly(dG) sequences effectively activate NK cells (Yamamoto et al., 1992a, 2000; Iho et al., 1999) and induce IFN-α/β production from plasmacytoid DCs (Bauer et al., 2001b; Kadowaki et al., 2001; Krug et al., 2001a,b). The effects of phosphodiester CpG DNAs are generally similar to those of bacterial DNA, and the induction of type I IFN is far greater than with phosphorothioate CpG DNAs. However, B cell activation by phosphodiester CpG DNA appears minimal compared with bacterial and phosphorothioate CpG DNAs. Phosphodiester CpG DNAs induce high levels of IFN-γ and IL-12, but produce IL-6 only minimally. Phosphorothioate substitutions on either end of the phosphodiester CpG DNA improve resistance to nucleases (Dalpke et al., 2002). Incorporation of poly(dG) sequences at the ends of the phosphodiester DNA has been reported to enhance nuclease stability and increase cellular uptake through scavenger receptors (Pearson et al., 1993; Kimura et al., 1994; Agrawal et al., 1 996). CpG DNAs containing phosphorothioate poly(dG) sequences at the 5′ and

cytosol

CpG DNA bacterial DNA (CpG) CpG oligonucleotides plasmid DNA (CpG)

TIR domain

MAPK (p38, ERK, JNK)

TLR9

AP-1

TRAF6 IRAK ds viral RNA [poly(I⋅C)]

nucleus

TIR domain TLR3 MyD88

IκB NF-κB

Chemistry of CpG DNA

Figure 4.16.1 The two members of the Toll-like receptor (TLR) family that recognize bacterial CpG DNA and double-stranded (ds) viral RNA. Key components involved in the signaling pathways are shown. The activated transcription factors ultimately upregulate the expression of a number of cytokines and costimulatory molecules. Abbreviations: AP-1, activator protein 1; ERK, extracellular-signal-regulated protein kinase; IκB, inhibitor of κB; IRAK, IL-1 receptor associated kinase; JNK, c-jun N-terminal kinase; MAPK, mitogen activated protein kinase; MyD88, myeloid differentiation factor 88; NF-κB, nuclear factor κB; TIR, Toll/IL-1 receptor; TLR, Toll-like receptor.



R = O− phosphodiester

O

B

O O

R P O O

R = S− O

B

phosphorothioate

O

R = CH3 methylphosphonate

Figure 4.16.2 Chemical structure of a typical dinucleotide showing the internucleotide phosphate linkage (boxed) and the three backbone chemistries studied in CpG DNA for immunostimulatory activity. Some of the important immunostimulatory effects observed with each backbone modification are given in Table 4.16.2.

3′ ends and a palindromic phosphodiester in the middle are potent adjuvants in vitro and in vivo (Dalpke et al., 2002; Verthelyi et al., 2002). However, it is unclear whether activity is due to the phosphodiester backbone, the poly(dG) sequences, or secondary structures formed by palindromic and poly(dG) sequences. Recently, for the first time, the authors showed the use of phosphodiester CpG DNA without requiring palindromic structures and poly(dG) sequences for immunostimulatory activity (Yu et al., 2002a). For this study, two phosphodiester CpG oligodeoxynucleotides were attached via a glyceryl linker through their 3′ ends. These 3′-3′-linked CpG DNAs are referred to as immunomers. Phosphodiester immunomers showed remarkable stability against nucleases in medium containing 10% FBS that was not heat inactivated. Surprisingly, phosphodiester immunomers induced increased IL-12 secretion and minimal amounts of IL-6 secretion in mouse spleen cell cultures (Yu et al., 2002a). These studies suggest that it would be possible to modulate cytokine secretion profiles induced by CpG DNAs by using different backbone chemistries. In J774 cell cultures they activated NF-κB and induced cytokine secretion comparable to that of an unmodified 18-mer phosphorothioate CpG DNA containing the same CpG motif. Moreover,

phosphodiester immunomers showed antitumor activity in nude mice bearing human breast (MCF-7) and prostate (DU145) cancer xenografts, suggesting that single-stranded phosphodiester CpG DNA can be a potent pharmacological agent. Phosphorothioate backbone In contrast, phosphorothioate CpG DNAs do not require palindromic sequences to induce im mu ne respo nses. I n g en er al, phosphorothioate CpG oligonucleotides strongly stimulate B cell proliferation, as well as activate monocytes/macrophages, DCs, and B cells to produce cytokines and immunoglobulins (Branda et al., 1993; Krieg et al., 1995; Zhao et al., 1997; Stacey et al., 2000). Phosphorothioate CpG DNAs are the most extensively studied to date, and several are in clinical trials. Stereoenriched phosphorothioate backbone Substitution of a nonbridging oxygen with sulfur creates a chiral phosphorus center with Rp- and Sp-diastereomers. If there are n phosphorothioate linkages in the backbone, conventional automated synthesis yields 2n stereoisomers. Studies using stereoenriched all-Rp and all-Sp phosphorothioate CpG DNAs showed that the all-Rp analog induced lower cell pro-



Supplement 12

liferation than all-Sp or racemic CpG DNA analogs (Yu et al., 2000a). However, it is not clear if this is because of their differential susceptibility to nucleases (Tang et al., 1995; Yu et al., 2000a) or because one of the diastereomers is preferentially recognized by the receptor. Role of backbone charge and nonionic methylphosphonate linkages Substitution of a nonbridging oxygen on the internucleotide phosphate with a methyl group produces a nonionic phosphorus center (Fig. 4.16.2). Recent studies show that an uncharged

Table 4.16.1

methylphosphonate internucleotide linkage between C and G of the CpG dinucleotide diminishes the immune response (Zhao et al., 1996). Further, methylphosphonate substitutions within the three internucleotide linkages to the 5′-side of the CpG dinucleotide also suppress activity (Yu et al., 2001b). It appears that a negative charge on these internucleotide linkages is important for recognition and/or interaction between the CpG DNA and its receptor. Surprisingly, substituting the fifth or sixth linkage on the 5′ side of the CpG dinucleotide significantly enhances immunostimu-

Properties of DNA Backbones with Different Linkagesa

Linkage

Property

Phosphodiester

Usually require longer length palindromic sequences to be stable against nucleases May require higher concentrations to compensate for nuclease degradation Often require phosphorothioate-end modifications Poly(dG) end modifications are extensively used Produce high levels of IL-12, IFN-γ, and IFN-α/β, and induce minimal IL-6 Potent activators of NK and dendritic cells, but weak activators of B cells Produce TH1-type immune responses. Potent adjuvants, and antitumor and antiasthmatic agents Stable against nucleases Strong activators of B cells. Also directly activate macrophages/monocytes and DCs, but do not directly activate NK and T cells. Produce a number of cytokines, including IL-12 Produce Ig Produce TH1 immune responses. Currently investigated for their potential in clinical trials as anticancer, antiallergic, and antiinfectious agents Only site-specific modifications have been studied CH3 between C and G of the CpG dinucleotide suppresses the immune response CH3 within three internucleotide linkages to 5′ of CpG suppresses the response CH3 at five or six linkages from 5′ of CpG enhances the response CH3 in 3′ flanking sequences has minimal effect May be possible to alter cytokine secretion by site-specific incorporation of one or two linkages in CpG DNA

Phosphorothioate

Methylphosphonate

aRefer to Figure 4.16.2 for graphical depiction of the linkages.




latory activity, which may reflect tighter binding to the receptor. In contrast, the presence of nonionic internucleotide linkages to the 3′ side of the CpG dinucleotide has an insignificant effect on activity.

2′-Sugar modifications Replacing either nucleoside in the CpG dinucleotide with one that is modified at the 2′-hydroxyl group (R in Fig. 4.16.3A) impedes immunostimulatory activity (Zhao et al., 1996). However, 2′-O-methyl- or 2′-O-methoxyethylribonucleosides in the sequences flanking the CpG dinucleotide have different effects depending on the position of substitution (Zhao et al., 1999; 2000). In general, 2′-O-alkylribonucleosides adjacent to the CpG dinucleotide on the 5′ side impair activity, while the same substitution on the 3′ side has minimal effect. Importantly, activity increases when the substitutions are incorporated distal to the CpG dinucleotide on either side (Zhao et al., 1999, 2000; Agrawal and Kandimalla, 2001). Hence, the effects on receptor binding due to conformational changes within ribose are strongly dependent on the position of substitution relative to the CpG dinucleotide.

Sugar Modifications While TLR9 has been shown to be involved in CpG DNA immune activation, a closely related family member, TLR3, has been reported to specifically recognize viral and synthetic double-stranded RNA and induce immune responses (Alexopoulou et al., 2001). Ribose and deoxyribose sugar moieties adopt 2′- and 3′-endo conformations (Fig. 4.16.3) that attribute distinct structural, physicochemical, and biological properties to RNA and DNA, respectively. It appears that the vertebrate innate immune system has evolved to recognize both of these nucleic acid structures from invading microorganisms, but through different receptors (Fig. 4.16.1). The authors of this unit have extensively studied the effects of sugar modifications in CpG DNA, as discussed below.

A

3′-Sugar modifications The incorporation of unnatural 3′-deoxynucleosides results in the formation of 2′-5′-inter-

B 5′

O

5′

B

O

O

O

O O

P

S−

O

O

B

O

O O

B

O

P

S−

O

R

O2′ O

O

B

B

O

R

S−

O

P

P O

O3′ O

P O

S− O

B

O3′ S−

O

P

S−

O

Figure 4.16.3 (A) Chemical structure of a DNA chain containing a single ribonucleoside (boxed). R is –OH (RNA), –OCH3 (2′-O-methyl RNA), or –OCH2CH2OCH3 (2′-O-methoxyethyl RNA). (B) Chemical structure of a DNA chain with a 3′-deoxyribonucleoside (R = –H) or 3′-O-methylribonucleoside (R = –OCH3) (boxed). Note that the incorporation of a 3′-deoxy- or 3′-O-methylribonucleoside results in a 2′-5′-linkage in an otherwise 3′-5′-linked DNA.



Supplement 12

3000

IL-12

2000

Cytokine (pg/mL)

1000

0 15000

IL-6

10000

5000 0 225

IL-10

150 75 0 2

1

3

M

5

4

6

CpG DNA

Figure 4.16.4 Effect of site-specific incorporation of a 3′-deoxyribonucleoside in different CpG DNA molecules (1 and 4) either in the 5′-flanking (2 and 5) or 3′-flanking sequence (3 and 6) on induction of cytokine (IL-12, -6, and -10) secretion in BALB/c mouse spleen cell cultures. Data for specific compounds is taken from Yu et al. (2002c). BALB/c mouse spleen cell cultures were incubated for 24 hrs with 1.0 µg/mL of CpG DNA. IL-12 secretion is not affected by either 5′ or 3′ modification. IL-6 and -10 secretion is increased when the modification is in the 5′-flanking sequence and either decreased or unaffected when the modification is in the 3′-flanking sequence. M is medium alone. Sequences of CpG DNA are: (1) 5′-d(CCTACTAGCGTTCTCATC)-3′; (2) 5′d(CCTAC*TAGCGTTCTCATC)-3′; (3) 5′-d(CCTACTAGCGTTCTC*ATC)-3′; (4) 5′-d(CTATCTGACG TTCTCTGT)-3′; (5) 5′-d(CTATC*TGACGTTCTCTGT)-3′; (6) 5′-d(CTATCTGACGTTCTC*TGT)-3′. CpG motifs are underlined. Italics indicate the 3′-deoxyribonucleosides. Asterisks indicate the position of the 2′-5′-linkages.


nucleotide linkages in an otherwise 3′-5′-linked DNA (Fig. 4.16.3B). The presence of a 3′-deoxynucleoside either within the CpG dinucleotide or adjacent to it abrogates immunostimulatory activity (Yu et al., 2002b). However, the same modification distal to the CpG dinucleotide in the 5′-flanking sequence potentiates activity. In the 3′-flanking sequence, 3′deoxynucleosides have an insignificant effect on immunostimulation. Interestingly, 3′-deoxynucleosides in either the 5′- or the 3′-flanking sequence distal to the CpG dinucleotide result in different cytokine secretion profiles compared with unmodified CpG DNA (Fig. 4.16.4). It seems that changes in recognition

and/or interaction with the receptor that are brought about by introducing chemical changes in CpG DNA are reflected in downstream events such as cytokine secretion. However, it is not clear yet if these distinct effects are a consequence of the differences in structural and/or kinetic interactions between the modified CpG DNA and its receptor. Nonetheless, by incorporating appropriate chemical modifications into CpG DNA, it may be possible to modulate cytokine secretion in a desirable fashion for specific disease indications. Similar effects have been reported with 3′-O-methylribonucleosides (Zhao et al., 2000; Agrawal and Kandimalla, 2001).



Significance of Deoxycytidine and Deoxyguanosine in CpG Dinucleotides: Synthetic Nucleosides Although flanking sequences strongly influence the interaction of CpG DNA and its receptors, the principal determinant in receptor recognition of a CpG DNA molecule (single- or double-stranded) is the unmethylated CpG dinucleotide itself. Any chemical modification introduced within the CpG dinucleotide that changes the DNA conformation—such as substitution of deoxyribose with ribose, 2′-O-substituted ribose, or 3′-deoxyribose, neutralization of the anionic phosphate charge between C and G, or deletion of the C or G nucleobase— completely abolishes recognition by the receptor and the subsequent immune responses. In addition, a methyl substitution at the 5 position of cytosine results in loss of activity (Zhao et al., 1996). Synthetic pyrimidines By replacing cytosine with synthetic pyrimidines, the authors have carried out an extensive study to delineate the importance for immunostimulation of each functional group on the pyrimidine ring of a CpG dinucleotide (Fig. 4.16.5). This study showed that deletion of any of the functional groups of cytosine (i.e., 2-keto, 3-imino, and 4-amino) resulted in loss of activity, suggesting that all three groups are

important for recognition and/or interaction with the receptor (Kandimalla et al., 2001). Unlike a hydrophobic methyl group, a hydroxyl substituent at the 5 position of cytosine in CpG does not suppress immunostimulatory activity (Kandimalla et al., 2001). In addition, while the 4-amino group of cytosine is absolutely required, an alkyl substitution on the amino group does not interfere with recognition and subsequent activity (Kandimalla et al., 2001). Synthetic purines Similarly, a number of synthetic purine analogs substituted for guanine in a CpG dinucleotide have been studied (Fig. 4.16.6). Deletion or modification of functional groups at the 1, 2, and 6 positions result in the loss of immunostimulatory activity (Kandimalla et al., 2001). The deletion of nitrogen at the 7 position, however, does not, suggesting that N7 is not required for receptor recognition. These studies have provided important clues regarding the functional groups of guanine in a CpG dinucleotide that are required for recognition and/or binding to the receptor (Kandimalla et al., 2001). Importantly, these studies identified new, trademarked, synthetic dinucleotides (i.e., YpG, CpR, and YpR, where Y and R are the synthetic pyrimidine and purine analogs in Figs. 4.16.5 and 4.16.6, respectively) that alter cytokine secretion profiles compared with the

NH2 H3 C NH2 4

O

O

N

NH2

N

O

N

N

1 2

N

O

5-methylcytosine

O

O O P

O

HO

N

3

5 6

N N

NH2

O H 3C

S−

O

5-methylisocytosine

NHCH2CH3

uracil

O

O

N

N

NH N

5-hydroxycytosine

N

NH O

4-N -ethylcytosine

N

O

P-base

Figure 4.16.5 Structure of 2′-deoxycytidine showing hydrogen bond acceptor (inward arrows) and donor (outward arrow) groups on cytosine. Chemical structures of some of the synthetic pyrimidine analogs studied are shown.



Supplement 12

O

O N

N

NH

N

N

N

hypoxanthine

NH

N N

NH2

N

7-deazaguanine

2-aminopurine

NH2

NH2 N N

O 7N 5 8

O

O

6

N 4 N 3

O

N

N

N

N

N

O

N H

N NH2

N H

1

NH purine

2

9

N

N

NH2

N

2,6-diaminopurine

isoguanine

NH2 CH3O

O

N

N N

N

NH2

N

O

N

NH

NH

O P S−

O

N H

N

O

7-deazaxanthine

K-base

H N N

N

NH2

8-bromoguanine O

O NH

O

NH

Br

N

8-oxoguanine

NH

N NH2

N

N

NH2

7-deaza-8-azaguanine

Figure 4.16.6 Structure of 2′-deoxyguanosine showing hydrogen bond acceptor (inward arrows) and donor (outward arrows) groups on guanine. Chemical structures of some of the synthetic purine analogs studied are shown.

natural CpG dinucleotide. Additionally, these studies indicate divergent synthetic nucleotide motif recognition patterns of the receptor and the possibility of modulating downstream cytokine secretion profiles using synthetic motifs placed appropriately in oligonucleotide sequences.

Role of Nucleobases


Recently, the need for each nucleobase in a CpG DNA for immune stimulation in mice by using abasic or 1′,2′-dideoxynucleosides (Fig. 4.16.7A) has been reported. The presence of a nucleobase is absolutely required in both C and G positions of the CpG dinucleotide for activity. However, deletion of a nucleobase in the 5′-flanking region at a distance of three or more nucleosides from the CpG dinucleotide increases immunostimulatory activity (Yu et al.,

2001a). A similar deletion in the 3′-flanking sequence does not significantly affect activity, suggesting that not all nucleobases are involved in recognition. Possibly, nucleobases in certain positions cause steric hindrance when binding to the receptor. Deletion of one or two of them might relieve this strain, improving recognition and/or binding to the receptor. Moreover, nucleobase deletions in the 5′-flanking sequence increased IL-6 production compared with parent CpG DNA, while those in the 3′-flanking sequence had the opposite effect (Fig. 4.16.7B).

Requirement of Nucleosides Recently it has been shown that non-nucleoside linkers (Fig. 4.16.8) could replace certain nucleosides in the 5′- and 3′-flanking sequences (Yu et al., 2002c). An alkyl linker in the flanking sequence 5′ to the CpG dinucleotide potenti-



A

B 3000

5′

O

B

O

IL-12 2250

O P

1500

S−

O

Cytokine (pg/mL)

O

O

O O

P

0 4000

IL-6

3000

S−

O

750

O

B

2000 1000

O3′ O

P

0

S−

M

7

1

8

CpG DNA

O

Figure 4.16.7 (A) Chemical structure of a DNA chain with a 1′,2′-dideoxyribonucleoside (boxed). B indicates base. (B) Effect of site-specific incorporation of a 1′,2′-dideoxyribonucleoside in a CpG DNA molecule (1) in either in the 5′-flanking (7) or 3′-flanking (8) sequence on induction of cytokine (IL-12 and IL-6) secretion in BALB/c mouse spleen cell cultures. Data for specific compounds are taken from Yu et al. (2001a). BALB/c mouse spleen cell cultures were incubated for 24 hr with 1.0 µg/mL of CpG DNA. Note that IL-12 secretion is not affected by either 5′ or 3′ modification. IL-6 secretion is increased when the modification is in the 5′-flanking sequence and decreased when the modification is in the 3′-flanking sequence. M is medium alone. Sequences of CpG DNAs are: (1) 5′-d(CCTACTAGCGTTCTCATC)-3′; (7) 5′-d(CCTXCTAGCGTTCTCATC)-3′; (8) 5′-d(CCTACTAGCGTTCXCATC)-3′. CpG motifs are underlined. X indicates the position of 1′,2′-dideoxyribonucleoside.

ated immunostimulatory activity. Interestingly, the same substitution in the 3′-flanking sequence did not affect immunostimulatory activity compared with parent CpG DNA. While a C3-linker optimally improved activity, longer ethyleneglycol and branched alkyl linkers (Fig. 4.16.8) were also beneficial. A linker in the 5′-flanking sequence increased IL-6 secretion several fold (Yu et al., 2002c). However, it is not clear whether the differences observed resulted from altered recognition/binding events with the receptor or initiation of different downstream signaling and transcriptional events compared with the parent CpG DNA.

Accessibility to the 5′ End of CpG DNA is Required for Immunostimulatory Activity The findings discussed above clearly suggest that the sequence 5′ to the CpG dinucleotide plays a major role in immune stimulation, while that on the 3′ side has an insignificant effect. The authors’ recent studies using 3′-3′- and 5′-5′-linked CpG DNAs suggest that an accessible 5′ end of CpG DNA is required for immunostimulatory activity (Yu et al., 2000b). Increased activity is observed when two CpG DNAs are tethered through their 3′ ends, while little or none is seen when they are tethered through their 5′ ends (Yu et al., 2000b). Synthesis of Modified Oligonucleotides and Conjugates


Supplement 12

O

B

O

alkyl linkers O(CH2)n O

O O

S−

P

n = 2, 3, 4, 6, or 9

O linker

ethylene glycol linkers

O O

P

S−

O

O(CH2CH2O)n CH2CH2O O

branched alkyl linkers

O O

P

n = 2 or 5

B

S−

NH2

OH

O O

O

or

O

O

Figure 4.16.8 Chemical structure of a DNA chain containing an alkyl linker. Structures of various linkers studied are shown. B indicates base.

Size of the 5′- or 3′-Attached Ligand Influences CpG DNA Activity


Subsequent studies showed that accessibility at the 5′ end depends on the size of the ligand or moiety conjugated to this end of CpG DNA (Kandimalla et al., 2002). Conjugation of a small residue, such as a phosphorothioate group, at the 5′ end has an insignificant effect on immunostimulatory activity. However, conjugation of larger groups—including fluorescein, a mononucleotide, a tetramer, or a longer oligonucleotide (5′-5′-linked)—significantly interferes with activity. Surprisingly, conjugation of an oligonucleotide or a ligand to the 3′ end of CpG DNA (3′-3′-linked) has either an insignificant effect on activity or increases activity. Studies of cellular uptake and activation of transcription factor NF-κB in J774 cells using fluorescein-conjugated CpG DNAs suggest that both 5′- and 3′-conjugates have similar cellular uptake, but only the 3′-conjugate activates NF-κB, not the 5′-conjugate. CpG DNA has been shown to efficiently mediate antigen uptake and presentation by DCs only when antigen or allergen is conjugated to the CpG DNA (Shirota et al., 2000, 2001; Tighe et al., 2000a). CpG DNA conju-

gates are currently in clinical trials for allergies (Tighe et al., 2000a,b; Horner et al., 2002). Routinely, ligands are conjugated to oligonucleotides at the 5′ end because of convenient synthetic protocols. However, the conjugation of a macromolecule or incorporation of G-rich sequences at the 5′ end could interfere with the recognition of the CpG DNA by its receptor, thereby reducing activity. The authors’ studies suggest that the conjugation of functional ligands (e.g., antigen, antibody, allergen, another CpG DNA) at the 3′ end of CpG DNA not only contributes to increased stability against nuclease digestion but also increased immunostimulatory potency of CpG DNA in vivo.

CONCLUSIONS CpG DNA is a powerful tool that can be used to modulate the immune system for treatment of a wide variety of disease indications. Several first-generation CpG DNA molecules are in clinical trials for a number of diseases either as monotherapies, in combination with antigens, vaccines, or monoclonal antibodies, or as conjugates with antigens (Gurunathan et al., 2000; Agrawal and Kandimalla, 2002; Kandimalla and Agrawal, 2002; Krieg, 2002). Extensive



safety data is available from clinical studies of antisense oligonucleotides. Antisense oligonucleotides, which are used at several-fold higher concentrations than CpG DNA, have been administered to humans without serious adverse safety concerns. To date, several hundred people have been treated for up to two years without any evidence of anti-DNA antibody formation. Many of these molecules that are in clinical trials contain CpG dinucleotides. The presence of CpG dinucleotides in antisense oligonucleotides induces immune responses (Agrawal and Kandimalla, 2000; Lewis et al., 2000; Agrawal and Kandimalla, 2001; Jahrsdorfer et al., 2002), sometimes resulting in uncontrolled cytokine secretion, causing toxicity concerns (Agrawal and Zhao, 1998a,b; Agrawal, 1999a,b). A number of second-generation chemical modifications of CpG motifs in antisense oligonucleotides have been reported to suppress immune-related side effects (Agrawal and Zhao, 1998a,b; Agrawal, 1999a,b; Agrawal and Kandimalla, 2000). The study of CpG DNA chemistry is a step closer towards understanding the biological effects and development of CpG DNA for human therapies. The chemical studies discussed in this review suggest that, in addition to the CpG dinucleotide, a number of other factors in CpG DNA molecules influence recognition by receptors, secretion of cytokines, and the resulting immunological effects. The ability to regulate cytokine induction through the use of second-generation CpG DNA modifications discussed in this review is an important factor in advancing the use of CpG DNA for specific disease indications with reduced toxicity concerns. Further understanding of the biological effects of second-generation CpG DNAs will lead to the design of more potent CpG DNA pharmacological agents.

LITERATURE CITED Agrawal, S. 1999a. Factors affecting the specificity and mechanism of action of antisense oligonucleotides. Antisense Nucleic Drug Dev. 9:371375. Agrawal, S. 1999b. Importance of nucleotide sequence and chemical modifications of antisense oligonucleotides. Biochim. Biophys. Acta 1489:53-68.

Agrawal, S. and Kandimalla, E.R. 2002. Medicinal chemistry and therapeutic potential of CpG DNA. Trends Mol. Med. 8:114-121. Agrawal, S. and Zhao, Q. 1998a. Antisense therapeutics. Curr. Opin. Chem. Biol. 2:519-528. Agrawal, S. and Zhao, Q. 1998b. Mixed backbone oligonucleotides: Improvement oligonucleotide-induced toxicity in vivo. Antisense Nucleic Acid Drug Dev. 8:135-139. Agrawal, S., Iadarola, P.L., Temsamani, J., Zhao, Q., and Shaw, D. 1996. Effect of G-rich sequences on the synthesis, purification, binding, cell uptake, and hemolytic activity of oligonucleotides. Bioorg. Med. Chem. Let. 6:2219-2224. Ahmad-Nejad, P., Hacker, H., Rutz, M., Bauer, S., Vabulas, R.M., and Wagner, H. 2002. Bacterial CpG-DNA and lipopolysaccharides activate Toll-like receptors at distinct cellular compartments. Eur. J. Immunol. 32:1958-1968. Alexopoulou, L., Holt, A.C., Medzhitov, R., and Flavell, R.A. 2001. Recognition of doublestranded RNA and activation of NF-κB by Tolllike receptor 3. Nature 413:732-738. Ballas, Z.K., Krieg, A.M., Warren, T., Rasmussen, W., Davis, H.L., Waldschmidt, M., and Weiner, G.J. 2001. Divergent therapeutic and immunologic effects of oligodeoxynucleotides with distinct CpG motifs. J. Immunol. 167:48784886. Bauer, S., Kirschning, C.J., Hacker, H., Redecke, V., Hausmann, S., Akira, S., Wagner, H., and Lipford, G.B. 2001a. Human TLR9 confers responsiveness to bacterial DNA via species-specific CpG motif recognition. Proc. Natl. Acad. Sci. U.S.A. 98:9237-9242. Bauer, M., Redecke, V., Ellwart, J.W., Scherer, B., Kremer, J.P., Wagner, H., and Lipford, G.B. 2001b. Bacterial CpG-DNA triggers activation and maturation of human CD11c-, CD123+ dendritic cells. J. Immunol. 166:5000-5007. Bird, A.P. 1986. CpG-rich islands and the function of DNA methylation. Nature 321:209-213. Branda, R.F., Moore, A.L., Mathews, L., McCormack, J.J., and Zon, G. 1993. Immune stimulation by an antisense oligomer complementary to the rev gene of HIV-1. Biochem. Pharmacol. 45:2037-2043. Dalpke, A.H., Zimmermann, S., Albrecht, I., and Heeg, K. 2002. Phosphodiester CpG oligonucleotides as adjuvants: Polyguanosine runs enhance cellular uptake and improve immunostimulative activity of phosphodiester CpG oligonucleotides in vitro and in vivo. Immunology 106:102-112.

Agrawal, S. and Kandimalla, E.R. 2000. Antisense therapeutics: Is it as simple as complementary base recognition? Mol. Med. Today 6:72-81.

Gursel, M., Verthelyi, D., Gursel, I., Ishii, K.J., and Klinman, D.M. 2002. Differential and competitive activation of human immune cells by distinct classes of CpG oligodeoxynucleotides. J. Leukoc. Biol. 71:813-820.

Agrawal, S. and Kandimalla, E.R. 2001. Antisense and/or immunostimulatory oligonucleotide therapeutics. Current Cancer Drug Targets 1:197-209.

Gurunathan, S., Klinman, D.M., and Seder, R.A. 2000. DNA vaccines: Immunology, application, and optimization. Annu. Rev. Immunol. 18:927974.



Supplement 12

Hartmann, G., Weeratna, R.D., Ballas, Z.K., Payette, P., Blackwell, S., Suparto, I., Rasmussen, W.L., Waldschmidt, M., Sajuthi, D., Purcell, R.H., Davis, H.L., and Krieg, A.M. 2000. Delineation of a CpG phosphorothioate oligodeoxynucleotide for activating primate immune responses in vitro and in vivo. J. Immunol. 164:1617-1624. Hemmi, H., Takeuchi, O., Kawai, T., Kaisho, T., Sato, S., Sanjo, H., Matsumoto, M., Hoshino, K., Wagner, H., Takeda, K., and Akira, S. 2000. A Toll-like receptor recognizes bacterial DNA. Nature 408:740-745. Horner, A.A., Takabaysahi, K., Zubeldia, J.M., and Raz, E. 2002. Immunostimulatory DNA-based therapeutics for experimental and clinical allergy. Allergy 57:24-29. Iho, S., Yamamoto, T., Takahashi, T., and Yamamoto, S. 1999. Oligodeoxynucleotides containing palindrome sequences with internal 5′-CpG-3′ act directly on human NK and activated T cells to induce IFN-γ production in vitro. J. Immunol. 163:3642-3652. Jahrsdorfer, B., Jox, R., Muhlenhoff, L., Tschoep, K., Krug, A., Rothenfusser, S., Meinhardt, G., Emmerich, B., Endres, S., and Hartmann, G. 2002. Modulation of malignant B cell activation and apoptosis by bcl-2 antisense ODN and immunostimulatory CpG ODN. J. Leukoc. Biol. 72:83-92. Kadowaki, N., Antonenko, S., and Liu, Y.J. 2001. Distinct CpG DNA and polyinosinic-polycytidylic acid double-stranded RNA, respectively, stimulate CD11c- type 2 dendritic cell precursors and CD11c+ dendritic cells to produce type I IFN. J. Immunol. 166:2291-2295. Kandimalla, E.R. and Agrawal, S. 2002. Towards optimal design of second-generation immunomodulatory oligonucleotides. Curr. Op. Mol. Ther. 4:122-129. Kandimalla, E.R., Yu, D., Zhao, Q., and Agrawal, S. 2001. Effect of chemical modifications of cytosine and guanine in a CpG-motif of oligonucleotides: Structure-immunostimulatory activity relationships. Bioorg. Med. Chem. 9:807-813. Kandimalla, E.R., Bhagat, L., Yu, D., Cong, Y., Tang, J., and Agrawal, S. 2002. Conjugation of ligands at the 5′-end of CpG DNA affects immunostimulatory activity. Bioconj. Chem. 13:966-974. Kimura, Y., Sonehara, K., Kuramoto, E., Makino, T., Yamamoto, S., Yamamoto, T., Kataoka, T., and Tokunaga, T. 1994. Binding of oligoguanylate to scavenger receptors is required for oligonucleotides to augment NK cell activity and induce IFN. J. Biochem. 116:991-994. Klinman, D.M., Yi, A.K., Beaucage, S.L., Conover, J., and Krieg, A.M. 1996. CpG motifs present in bacterial DNA rapidly induce lymphocytes to secrete interleukin 6, interleukin 12, and interferon γ. Proc. Natl. Acad. Sci. U.S.A. 93:28792883. Chemistry of CpG DNA

Krieg, A.M., Yi, A.K., Matson, S., Waldschmidt, T.J., Bishop, G.A., Teasdale, R., Koretzky, G.A., and Klinman, D.M. 1995. CpG motifs in bacterial DNA trigger direct B-cell activation. Nature 374:546-549. Krug, A., Rothenfusser, S., Hornung, V., Jahrsdorfer, B., Blackwell, S., Ballas, Z.K., Endres, S., Krieg, A.M., and Hartmann, G. 2001a. Identification of CpG oligonucleotide sequences with high induction of IFN-α/β in plasmacytoid dendritic cells. Eur. J. Immunol. 31:2154-2163. Krug, A., Towarowski, A., Britsch, S., Rothenfusser, S., Hornung, V., Bals, R., Giese, T., Engelmann, H., Endres, S., Krieg, A.M., and Hartmann, G. 2001b. Toll-like receptor expression reveals CpG DNA as a unique microbial stimulus for plasmacytoid dendritic cells which synergizes with CD40 ligand to induce high amounts of IL-12. Eur. J. Immunol. 31:3026-3037. Lewis, E.J., Agrawal, S., Bishop, J., Chadwick, J., Cristensen, N.D., Cuthill, S., Dunford, P., Field, A.K., Francis, J., Gibson, V., Greenham, A.K., Kelly, F., Kilkuskie, R., Kreider, J.W., Mills, J.S., Mulqueen, M., Roberts, N.A., Roberts, P., and Szymkowski, D.E. 2000. Non-specific antiviral activity of antisense molecules targeted to the E1 region of human papillomavirus. Antiviral Res. 48:187-196. Lien, E., and Ingalls, R.R. 2002. Toll-like receptors. Crit. Care Med. 30:S1-S11. Messina, J.P., Gilkeson, G.S., and Pisetsky, D.S. 1991. Stimulation of in vitro murine lymphocyte proliferation by bacterial DNA. J. Immunol. 147:1759-1764. Pearson, A.M., Rich, A., and Krieger, M. 1993. Polynucleotide binding to macrophage scavenger receptors depends on the formation of basequartet-stabilized four-stranded helices. J. Biol. Chem. 268:3546-3554. Pisetsky, D.S. 1999. The influence of base sequence on the immunostimulatory properties of DNA. Immunol. Res. 19:35-46. Rankin, R., Pontarollo, R., Ioannou, X., Krieg, A.M., Hecker, R., Babiuk, L.A., and van den Hurk, S.v.d.L. 2001. CpG motif identification for veterinary and laboratory species demonstrates that sequence recognition is highly conserved. Antisense Nucleic Acid Drug Dev. 11:333-340. Ro thenfusser, S., Hornung, V., Krug, A., Towarowski, A., Krieg, A.M., Endres, S., and Hartmann, G. 2001. Distinct CpG oligonucleotide sequences activate human gamma delta T cells via interferon-α/-β. Eur. J. Immunol. 31:3525-3534. Sato, Y., Roman, M., Tighe, H., Lee, D., Corr, M., Nguyen, M.D., Silverman, G.J., Lotz, M., Carson, D.A., and Raz, E. 1996. Immunostimulatory DNA sequences necessary for effective intradermal gene immunization. Science 273:352-354.

Krieg, A.M. 2002. CpG motifs in bacterial DNA and their immune effects. Annu. Rev. Immunol. 20:709-760.



Shirota, H., Sano, K., Kikuchi, T., Tamura, G., and Shirato, K. 2000. Regulation of murine airway eosinophilia and Th2 cells by antigen-conjugated CpG oligodeoxynucleotides as a novel antigen-specific immunomodulator. J. Immunol. 164:5575-5582. Shirota, H., Sano, K., Hirasawa, N., Terui, T., Ohuchi, K., Hattori, T., Shirato, K., and Tamura, G. 2001. Novel roles of CpG oligodeoxynucleotides as a leader for the sampling and presentation of CpG-tagged antigen by dendritic cells. J. Immunol. 167:66-74. Sonehara, K., Saito, H., Kuramoto, E., Yamamoto, S., Yamamoto, T., and Tokunaga, T. 1996. Hexamer palindromic oligonucleotides with 5′-CG3′ motif(s) induce production of interferon. J. Interferon Cytokine Res. 16:799-803. Stacey, K.J., Sester, D.P., Sweet, M.J., and Hume, D.A. 2000. Macrophage activation by immunostimulatory DNA. Curr. Top. Microbiol. Immunol. 247:41-58. Takeshita, F., Leifer, C.A., Gursel, I., Ishii, K.J., Takeshita, S., Gursel, M., and Klinman, D.M. 2001. Role of Toll-like receptor 9 in CpG DNAinduced activation of human cells. J. Immunol. 167:3555-3558. Tang, J.Y., Roskey, A.R., Li, Y., and Agrawal, S. 1995. Enzymatic synthesis of stereoregular (all Rp) oligonucleotide phosphorothioate and its properties. Nucleosides Nucleotides 14:985989. Tighe, H., Takabayashi, K., Schwartz, D., Van Nest, G., Tuck, S., Eiden, J.J., Kagey-Sobotka, A., Creticos, P.S., Lichtenstein, L.M., Spiegelberg, H.L., and Raz, E. 2000a. Conjugation of immunostimulatory DNA to the short ragweed allergen amb a 1 enhances its immunogenicity and reduces its allergenicity. J. Allergy Clin. Immunol. 106:124-134. Tighe, H., Takabayashi, K., Schwartz, D., Marsden, R., Beck, L., Corbeil, J., Richman, D.D., Eiden, J.J. Jr., Spiegelberg, H.L., and Raz, E. 2000b. Conjugation of protein to immunostimulatory DNA results in a rapid, long-lasting and potent induction of cell-mediated and humoral immunity. Eur. J. Immunol. 30:1939-1947. Tokunaga, T., Yamamoto, H., Shimada, S., Abe, H., Fukuda, T., Fujisawa, Y., Furutani, Y., Yano, O., Kataoka, T., Sudo, T., Makiguchi, N., and Suganuma, T. 1984. Antitumor activity of deoxyribonucleic acid fraction from Mycobacterium bovis BCG. I. Isolation, physicochemical characterization, and antitumor activity. J. Natl. Cancer Inst. 72:955-962. Van Uden, J., and Raz, E. 2000. Introduction to immunostimulatory DNA sequences. Springer Semin. Immunopathol. 22:1-9. Verthelyi, D., Ishii, K., Gursel, M., Takeshita, F., and Klinman, D. 2001. Human peripheral blood cells differentially recognize and respond to two distinct CpG motifs. J. Immunol. 166:2372-2377.

Verthelyi, D., Kenney, R.T., Seder, R.A., Gam, A.A., Friedag, B., and Klinman, D.M. 2002. CpG oligodeoxynucleotides as vaccine adjuvants in primates. J. Immunol. 168:1659-1663. Wernette, C.M., Smith, B.F., Barksdale, Z.L., Hecker, R., and Baker, H.J. 2002. CpG oligodeoxynucleotides stimulate canine and feline immune cell proliferation. Vet. Immunol. Immunopathol. 84:223-236. Yamamoto, S., Yamamoto, T., Kataoka, T., Kuramoto, E., Yano, O., and Tokunaga, T. 1992a. Unique palindromic sequences in synthetic oligonucleotides are required to induce INF and augment INF-mediated natural killer activity. J. Immunol. 148:4072-4076. Yamamoto, S., Yamamoto, T., Shimada, S., Kuramoto, E., Yano, O., Kataoka, T., and Tokunaga, T. 1992b. DNA from bacteria, but not from vertebrates, induces interferons, activates natural killer cells and inhibits tumor growth. Microbiol. Immunol. 36:983-997. Yamamoto, S., Yamamoto, T., Iho, S., and Tokunaga, T. 2000. Activation of NK cell (human and mouse) by immunostimulatory DNA sequence. Springer Semin. Immunopathol. 22:35-43. Yu, D., Kandimalla, E.R., Roskey, A., Zhao, Q., Chen, L., Chen, J., and Agrawal, S. 2000a. Stereo-enriched phosphorothioate oligodeoxynucleotides: Synthesis, biophysical and biological properties. Bioorg. Med. Chem. 8:275-284. Yu, D., Zhao, Q., Kandimalla, E.R., and Agrawal, S. 2000b. Accessible 5′-end of CpG-containing phosphorothioate oligodeoxynucleotides is essential for immunostimulatory activity. Bioorg. Med. Chem. Lett. 10:2585-2588. Yu, D., Kandimalla, E.R., Zhao, Q., Cong, Y., and Agrawal, S. 2001a. Modulation of immunostimulatory activity of CpG oligonucleotides by site-specific deletion of nucleobases. Bioorg. Med. Chem. Lett. 11:2263-2267. Yu, D., Kandimalla, E.R., Zhao, Q., Cong, Y., and Agrawal, S. 2001b. Immunostimulatory activity of CpG oligonucleotides containing non-ionic methylphosphonate linkages. Bioorg. Med. Chem. 9:2803-2808. Yu, D., Zhu, F.G., Bhagat, L., Wang, H., Kandimalla, E.R., Zhang, R., and Agrawal, S. 2002a. Potent CpG oligonucleotides containing phosphodiester linkages: In vitro and in vivo immunostimulatory properties. Biochem. Biophys. Res. Commun. 297:83-90. Yu, D., Kandimalla, E.R., Zhao, Q., Cong, Y., and Agrawal, S. 2002b. Immunostimulatory properties of phosphorothioate CpG DNA containing both 3′-5′- and 2′-5′-internucleotide linkages. Nucl. Acids Res. 30:1613-1619. Yu, D., Kandimalla, E.R., Cong, Y., Tang, J., Tang, J.Y., Zhao, Q., and Agrawal, S. 2002c. Design, synthesis, and immunostimulatory properties of CpG DNAs containing alkyl-linker substitutions: Role of nucleosides in the flanking sequences. J. Med. Chem. 45:4540-4548.



Supplement 12

Zhang, Y., Shoda, L.K., Brayton, K.A., Estes, D.M., Palmer, G.H., and Brown, W.C. 2001. Induction of interleukin-6 and interleukin-12 in bovine B lymphocytes, monocytes, and macrophages by a CpG oligodeoxynucleotide (ODN 2059) containing the GTCGTT motif. J. Interferon Cytokine Res. 21:871-881. Zhao, Q., Temsamani, J., Iadarola, P.L., Jiang, Z., and Agrawal, S. 1996. Effect of different chemically modified oligodeoxynucleotides on immune stimulation. Biochem. Pharmacol. 51:173-182. Zhao, Q., Temsamani, J., Zhou, R.Z., and Agrawal, S. 1997. Pattern and kinetics of cytokine productio n fo llowing administration of phosphorothioate oligonucleotides in mice. Antisense Nucleic Acid Drug. Dev. 7:495-502.

Zhao, Q., Yu, D., and Agrawal, S. 1999. Site of chemical modifications in CpG containing phosphorothioate oligodeoxynucleotide modulates its immunostimulatory activity. Bioorg. Med. Chem. Lett. 9:3453-3458. Zhao, Q., Yu, D., and Agrawal, S. 2000. Immunostimulatory activity of CpG containing phosphorothioate oligodeoxynucleotide is modulated by modification of a single deoxynucleoside. Bioorg. Med. Chem. Lett. 10:1051-1054.

Contributed by Ekambar R. Kandimalla and Sudhir Agrawal Hybridon, Inc. Cambridge, Massachusetts




Synthesis of Phosphorothioate Oligonucleotides with Stereodefined Phosphorothioate Linkages

UNIT 4.17

Phosphorothioate analogs of oligonucleotides (PS-oligos) constitute an important tool for studying the metabolism of nucleic acids (Eckstein, 2000, and references therein) and have been evaluated as potential therapeutics in the so-called “antisense” (Stein and Krieg, 1998) and “antigene” strategies (Thuong and Helene, 1993). In 1998, the U.S. Food and Drug Administration (FDA) approved the first PS-oligo, Fomirvirsen (trade name, Vitravene), for therapeutic application against cytomegalovirus (CMV) retinitis (Manoharan, 1999). Most of the second-generation antisense compounds that are currently undergoing clinical trials are PS-oligos (e.g., Isis Pharmaceuticals, Hybridon; Maier et al., 2000). PS-oligos are isoelectronic with natural oligonucleotides and, importantly, they are much more resistant towards intra- and extracellular nucleases (Wickstrom, 1986). These features are important with respect to their therapeutic applications. However, substitution of sulfur for one nonbridging oxygen in the internucleotide phosphate group induces asymmetry at the phosphorus atom, and standard chemical methods for the synthesis of oligo(deoxyribonucleoside phosphorothioate)s provide a mixture of 2n diastereomers, where n is the number of phosphorothioate linkages (Wilk and Stec, 1995). Therefore, even for relatively short PS-oligos (10- to 15-mers), thousands of diastereomers would be involved in interactions with other chiral biomolecules (e.g., DNA, RNA, or proteins) and, in principle, each diastereomer might interact in a slightly different manner. The enzymatic synthesis of PS-oligos allows for the preparation of PS-oligonucleotides of RP-configuration at each phosphorothioate linkage (all-RP-PS-oligos) due to the stereoselectivity of all DNA and RNA polymerases identified to date (Hacia et al., 1994; Lackey and Patel, 1997; Tang et al., 1995). The first method for stereocontrolled chemical synthesis of PS-oligos, which was elaborated in the authors’ laboratory (Stec et al., 1991), is based on a new chemistry employing P-diastereomerically pure nucleoside monomers possessing the 2-thio-1,3,2-oxathiaphospholane moiety attached to appropriately protected nucleosides at the 3′-O position (S.1; Fig. 4.17.1). Further studies resulted in the synthesis of monomers with the oxathiaphospholane ring substituted at position 4 with either two methyl groups (S.2; Stec et al., 1995) or a spiro pentamethylene ring (S.3; Stec et al., 1998). These substituents enhance a differentiation in chromatographic mobility of diastereomers, rendering their separation less laborious. The oxathiaphospholane monomers react in the presence of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) with the 5′-OH group of a nucleoside (or growing oligonucleotide attached at the 3′ end to a DBU-resistant solid support) to yield a dinucleotide (or an elongated oligomer) with an internucleotide phosphorothioate diester bond, as depicted in Figure 4.17.2. The process is fully stereospecific and occurs with retention of configuration at the phosphorus atom. The chemical yield of the condensation process is not as efficient as that of the phosphoramidite or H-phosphonate methods (UNITS 3.3 & 3.4), but repetitive yields of 92% to 94% allow syntheses of medium-sized oligomers (up to 15-mers). Longer oligonucleotides were obtained in poor yields and the syntheses were not reproducible. Oxathiaphospholanes that are 18O-labeled at the endocyclic position allowed for the synthesis of PS-oligos with internucleotide PS[18O]-phosphorothioate moieties of predetermined chirality (Guga et al., 2001). Synthesis of Modified Oligonucleotides and Conjugates Contributed by Piotr Guga and Wojciech J. Stec Current Protocols in Nucleic Acid Chemistry (2003) 4.17.1-4.17.28 Copyright © 2003 by John Wiley & Sons, Inc.


For the synthesis of stereodefined PS-oligos via the oxathiaphospholane methodology presented in this unit, pure P-diastereomers of nucleoside oxathiaphospholane monomers are required. They are not commercially available, but can be efficiently obtained by phosphitylation of widely available 5′-O-DMTr-N-protected deoxyribonucleosides with oxathiaphospholane phosphitylating reagent followed by sulfurization. The methodology of their synthesis and use in solid-phase synthesis of PS-oligos is presented in consecutive protocols. Basic Protocol 1 describes a detailed procedure for the synthesis of the phosphitylating reagent 2-chloro-spiro-4,4-pentamethylene-1,3,2-oxathiaphospholane. The procedure is general and may be applied to other analogs, depending on the aldehyde (or mercaptoalcohol) used. Alternate Protocol 1 describes a procedure for synthesis of 18O-labeled mercaptoalcohol, which is used to synthesize labeled phosphitylating reagent and, subsequently, 18O-labeled nucleoside monomers. These can be used for synthesis of stereodefined PS[18O]oligos, which are useful compounds in studying the mechanism(s) of enzymatic reactions. Support Protocol 1 describes a method for transfer of dry solvents, required for this procedure. Basic Protocol 2 outlines the synthesis of 5′-O-DMTr-N-protected-deoxyribonucleoside3′-O-(2-thio-spiro-4,4-pentamethylene-1,3,2-oxathiaphospholane)s (S.3) and their chromatographic separation into P-diastereomers. This method, although described for dA, dC, dG, and T derivatives, can be also used for derivatizing other appropriately protected nucleosides. For example, in the authors’ laboratory, N6-benzoyl-7-deaza-5′-O(4,4′-dimethoxytrityl)-3′-O-(2-thio-spiro-4,4-pentamethylene-1,3,2-oxathiaphospholane)2′-deoxyadenosine was obtained and separated into diastereomers (unpub.). Similarly, this method is suitable for phosphitylation of protected nucleosides with other oxathiaphospholane reagents containing different substituents, although the separation of diastereo-mers may be very difficult. Alternate Protocol 2 describes the conversion of 5′-O-DMTr-N-protected-deoxyribonucleoside-3′-O-(2-thio-spiro-4,4-pentamethylene-1,3,2-oxathiaphospholane)s to their 2oxo-analogs with selenium dioxide. These monomers can be used to elongate stereodefined PS-oligos and generate segments of unmodified nucleotide units possessing phosphate internucleotide linkages. This goal cannot be achieved with the phosphoramidite or H-phosphonate methods, because the phosphorothioate linkages already

DMTrO

O

B

O X P 1

O 1 2 3 4 Synthesis of Phosphorothioate Oligonucleotides with Stereodefined Linkages

X = S, X = S, X = S, X = O,

S

3

R

5

R

R=H R = CH 3 R,R = −(CH 2)5− R,R = −(CH 2)5−

Figure 4.17.1 Structural features of deoxyribonucleoside oxathiaphospholane derivatives. Abbreviations: B, thymin-1-yl or N-protected nucleobase; DMTr, 4,4′-dimethoxytrityl. Adapted from Stec et al. (1998) with permission from the American Chemical Society.



−

S O

OR′

P OR S DBU

O S

OR

P

S

ψ O

−

S

OR′

P

S

−

S

R′O RO

OR′

OR

DMTrO R=

O

P

−

S

O

B

O

−

O

S R′O RO

P

−

S

B

R′ = O Sar

Figure 4.17.2 Mechanism of base-promoted oxathiaphospholane ring-opening condensation. Abbreviations: B, thymin-1yl or N-protected nucleobase; DBU, 1,8-diazabicyclo[5.4.0]undec-7-ene; DMTr, 4,4′-dimethoxytrityl; ψ, pseudorotation; Sar, sarcosinylated or DBU-resistant solid support.

generated by the oxathiaphospholane method are diesters and would be oxidized in the I2/water/pyridine routinely used for conversion of phosphites to phosphates. Basic Protocol 3 outlines details of manual solid-phase synthesis of PS-oligos using oxathiaphospholane monomers. In principle, this synthesis can be performed on an automatic synthesizer, but the necessary modification of the manufacturer’s protocols is impossible for the majority of synthesizers. The protocol for 1-µmol-scale automated solid-phase synthesis using an ABI 391 synthesizer (Applied Biosystems) has been published (Stec et al., 1998), but in many instances the software does not allow for any changes in the protocol. Also, the instrument should be able to deliver an additional solvent (methylene chloride) to the column in order to wash delivery lines after coupling. This is necessary to avoid formation of deposits inside the tubing and valves, which may lead to major failure of the instrument and expensive replacement of the clogged valve blocks. One also has to consider that, in using an automated synthesizer, significant amounts of monomer solutions are wasted during optimization of the protocol, and due to the dead volumes of the system. Therefore, manual synthesis of a limited number of oligomers may be economically more justified. Support Protocol 2 describes preparation of solid supports for the synthesis of PS-oligos, which must be DBU-resistant because this strong base is necessary for the coupling step. This requirement is fulfilled by the use of Brown’s sarcosinyl-succinoyl linker (Brown et al., 1989).



Supplement 14

H

O

O S2Cl2

H

H

S

S

O

−HCl

HO NaBH4

S

propan-2-ol 6 (70%)

5

OH S

OH SH

7 (85%)

O

LiAlH4

PCl3

diethyl ether

pyridine

P Cl S

8 (70%)

9 (70%)

Figure 4.17.3 Synthesis of 2-chloro-spiro-4,4-pentamethylene-1,3,2-oxathiaphospholane (S.9) starting from cyclohexanecarboxaldehyde (S.5). Adapted from Stec et al. (1998) with permission from the American Chemical Society.

CAUTION: It is imperative that all reactions be run in a suitable fume hood with efficient ventilation. Many of the reactions in this unit are highly exothermic; safety glasses and reagent-impermeable protective gloves should be worn. BASIC PROTOCOL 1

SYNTHESIS OF PHOSPHITYLATING REAGENT: 2-CHLORO-spiro-4,4-PENTAMETHYLENE-1,3,2-OXATHIAPHOSPHOLANE The most simple oxathiaphosphitylating reagent, 2-chloro-1,3,2-oxathiaphospholane, can be obtained from the reaction of 2-mercaptoethanol with phosphorus trichloride in the presence of two molar equivalents of triethylamine (Martynov et al., 1969; Willson et al., 1975; Stec et al., 1991). Condensation of an appropriately protected 3′-OH-nucleoside with 2-chloro-1,3,2-oxathiaphospholane in pyridine solution, performed in the presence of dry elemental sulfur, provides nucleoside 3′-O-(2-thiono-1,3,2-oxathiaphospholane)s (S.1; Fig. 4.17.1). However, their separation as P-diastereomers is very laborious and requires several consecutive silica gel chromatographic runs of partially enriched fractions. Therefore, it is recommended to synthesize 2-chloro-spiro-4,4-pentamethylene1,3,2-oxathiaphospholane (S.9) starting from cyclohexanecarboxaldehyde (S.5), as depicted in Figure 4.17.3. Using isobutyraldehyde in the same sequence of reactions, 2-chloro-4,4-dimethyl-1,3,2-oxathiaphospholane can be obtained. However, the phosphitylating reagent S.9, when used for synthesis of nucleotide monomers, provides much more useful compounds in terms of chromatographic separability as pure P-diastereomers. It is important to note that the fast-eluting isomers of 4,4-dimethyl- and 4,4-pentamethylene-oxathiaphospholane monomers are precursors of RP internucleotidic phosphorothioate bonds. Conversely, analog internucleotide phosphorothioate bonds of RP configuration are formed from the slow-eluting isomer of S.1.

Synthesis of Phosphorothioate Oligonucleotides with Stereodefined Linkages

Using this methodology, different analogs can be synthesized. However, it is important to obtain a phosphitylating reagent without additional centers of asymmetry, as the number of diastereomers will double with each new center, rendering separation of P-diastereomers very difficult or impossible. The relationship between chromatographic mobility and absolute configuration of the monomers must be checked for each new analog.



Materials 4 to 5 M and 1.5 M sodium hydroxide (NaOH) Sulfur monochloride (S2Cl2), freshly distilled over 2 g elemental sulfur (S8; dried ≥12 hr under vacuum) per 50 mL S2Cl2 Argon (or, optionally, nitrogen), dry Cyclohexanecarboxaldehyde (Fluka) Methylene chloride Diethyl ether, anhydrous Sodium borohydride (NaBH4) Isopropyl alcohol Anti-bumping granules 20% (w/v) hydrochloric acid Chloroform Magnesium sulfate, anhydrous Hexane Lithium aluminum hydride Ethyl acetate, dry Tetrahydrofuran (THF), with traces of added moisture 10% (v/v) H2O/THF Phosphorus trichloride (PCl3) Benzene, anhydrous Pyridine Dry molecular sieves (4A, 4- to 6-µm-o.d. beads, Aldrich) 250-mL absorber with safety flask (see Fig. 4.17.4) 250-mL four-neck round-bottom flask Heated oil bath capable of magnetic stirring Thermometer (capable of reading 150°C) 100-mL dropping funnel Reflux condensers Glass gas inlet adapter (preferred) or syringe needle and rubber septum Rotary evaporator with a water aspirator and a diaphragm vacuum pump (10 to 15 mmHg; optional) 500-mL Erlenmeyer flask (29/42 joint) 1-L two-neck round-bottom flask (two 29/42 joints) Stopcock, 29/42 Flexible adapter (glass M/F joints, 29/42, on corrugated Teflon tubing; optional) 500-mL separatory funnel Filter funnel and Whatman no.1 filter paper (or equivalent) High-vacuum fractional distillation apparatus High-vacuum oil pump (0.01 mmHg) NOTE: Upon storage, cyclohexanecarboxaldehyde undergoes polymerization. Order only the amount required for use within 2 to 3 weeks. NOTE: Within this unit, evaporation of solvents is performed using a rotary evaporator connected to a water aspirator, unless otherwise specified.



Supplement 14

to the hood

4 M NaOH safety flask

absorber

thermometer NOTE: glass gas adapter not shown

aldehyde sulfur monochloride

magnetic stir bar

oil bath 567 4 8 3 2 1 119

4 567 3 8 2 9 1 1

magnetic stirrer

Figure 4.17.4 System assembly for synthesis of S.6. The glass gas adapter (or septum and needle) for delivery of dry argon should be mounted in the fourth neck of the flask (not shown).

Synthesize 2,2′-dithiobis(cyclohexanecarbaldehyde) (S.6) 1. Prepare a 250-mL absorber containing ∼150 mL of 4 to 5 M NaOH (see Fig. 4.17.4). 2. Assemble a reactor consisting of a 250-mL four-neck, round-bottom flask (to be heated in an oil bath with magnetic stirring) equipped with a thermometer, a 100-mL dropping funnel, a condenser, and a magnetic stir bar. Connect the outlet of the condenser via a safety flask to the absorber and be sure that the absorber vents into the hood. The capacity of the safety flask must be sufficient to accommodate the solution of sodium hydroxide from the absorber.

3. Add 30 g (17.8 mL, 0.22 mol) of freshly distilled sulfur monochloride to the flask. Make sure that the thermometer is in contact with the liquid. Deliver dry argon close to the bottom of the flask through the fourth joint either with a glass gas inlet adapter or with a syringe needle inserted through a rubber septum. Since gaseous hydrogen chloride is liberated from the reaction, the use of a glass inlet adapter, rather than a syringe needle, is recommended.

4. Apply heating until sulfur monochloride reaches 60°C. Stabilize the temperature. Synthesis of Phosphorothioate Oligonucleotides with Stereodefined Linkages

5. Add, dropwise through the dropping funnel, 50 g (56 mL, 0.44 mol) of cyclohexanecarboxaldehyde over a 60-min period with stirring, keeping the temperature at 60°C. Remove the oil bath and allow the mixture to continue stirring another 10 min.



6. Stop stirring and remove the stir bar. Leave the reaction mixture until it reaches room temperature and solidifies (∼1 hr). 7. Dissolve the solid residue in 150 mL of methylene chloride, then evaporate to dryness using a rotary evaporator connected to a water aspirator. CAUTION: During evaporation gaseous hydrogen chloride is liberated.

8. Add 150 mL of diethyl ether and gently reflux until the solid residue is dissolved. Transfer the solution into a 500-mL Erlenmeyer flask and close the stopcock. Cool down the mixture (containing the synthesized S.6) and keep overnight in a refrigerator (4°C) to allow crystallization. Collect crystalline S.6 by filtration. Approximately 35 g of 2,2′-dithiobis(cyclohexanecarbaldehyde) (S.6) should be collected as a white solid (∼70% yield, m.p. 88° to 89°C). 1H NMR (CDCl3, δ): 8.98 ppm (s, 1H, CHO), 1.2-2.1 ppm (m, 10H). FAB MS (positive mode, Cs+, 13 keV, matrix NBA) m/z 286, [M]+, 25%; m/z 111, [C6H10CHO]+, 100%.

Synthesize 2,2′-dithiobis(cyclohexanemethanol) (S.7) 9. In a 1-L two-neck flask (two 29/42 joints), equipped with a reflux condenser and a stopcock, suspend 5.67 g (0.15 mol) of NaBH4 in 500 mL isopropyl alcohol. Add anti-bumping granules and heat to boiling. 10. While gently refluxing, remove the stopcock momentarily and add, with a chemical spoon, ∼2 g of S.6 every 3 to 5 min in 10 to 12 portions for a total of 21.5 g (0.075 mol). CAUTION: The addition of each portion of 2,2′-dithiobis(cyclohexanecarboxaldehyde) results in enhanced boiling and emission of vapors of isopropyl alcohol through the open neck. The stopcock should thus be closed as soon and possible. Alternatively, one can use a flexible adapter (glass M/F joints on corrugated Teflon tubing) for stepwise delivery of 2,2′-dithiobis(cyclohexanecarboxaldehyde) without opening the reactor.

11. Reflux the mixture for 1 hr, then evaporate to dryness and add 200 mL of 1.5 M sodium hydroxide. 12. Cautiously neutralize the mixture with 20% hydrochloric acid, checking pH with indicator strips. 13. Transfer the mixture to a 500-mL separatory funnel and extract the solution twice, each time with 150 mL chloroform. Dry the organic layer with anhydrous magnesium sulfate and evaporate the solvent. 14. Dissolve the product in 200 mL of diethyl ether and add, with stirring, a few 3- to 5-mL aliquots of hexane until the mixture becomes translucent. Leave in a refrigerator (4°C) overnight for crystallization. Collect crystalline S.7 by filtration. Approximately 19 g of 2,2′-dithiobis(cyclohexanemethanol) (S.7) should be collected as a colorless crystalline material (85% to 88% yield, m.p. 49° to 50°C). 1H NMR (CDCl3, δ): 3.54 ppm (s, 2H, CH2OH), 2.23 ppm (s, 1H, CH2OH), 1.2-1.8 ppm (m, 10H). 13C NMR (CDCl3, δ): 21.9, 25.73, 32.45, 56.18, 67.99 ppm. FAB MS (positive mode, Cs+, 13 keV, matrix NBA) m/z 290, [M]+, 45%; m/z 273, [M-OH]+, 25%; m/z 113, [C6H10CH2OH]+, 100%.

Synthesize 2-mercaptocyclohexanemethanol (S.8) 15. In a 1-L two-neck flask equipped with a reflux condenser and a dropping funnel (atmosphere of dry argon or nitrogen) suspend 5.9 g (0.16 mol) lithium aluminum hydride in 500 mL of dry diethyl ether. CAUTION: The suspension of lithium aluminum hydride in diethyl ether is highly flammable. Advise coworkers of the hazard and keep an appropriate fire extinguisher at hand.



Supplement 14

16. Add dropwise through the dropping funnel a solution of 4.6 g (0.16 mol) S.7 in 150 mL of diethyl ether over 60 min with magnetic stirring. Continue stirring for an additional 60 min. The reaction is exothermic and mild reflux occurs.

17. Cautiously decompose excess reducing agent by adding dropwise through the funnel 3 mL of dry ethyl acetate, followed by 10 mL THF containing traces of moisture, and then ∼10 mL of 10% water/THF, until the solid suspension becomes gray and finally white. 18. Filter off inorganic salts using a filter funnel and Whatman no. 1 filter paper, and dry the filtrate over anhydrous magnesium sulfate. Filter off the drying agent and evaporate the filtrate to dryness. 19. Distill the residue in a high-vacuum fractional distillation apparatus under reduced pressure (0.05 mmHg, provided by high-vacuum oil pump). Collect the fraction boiling between 74° and 76°C, which contains S.8. CAUTION: Avoid overheating the vessel. Keep pressure below 0.1 mmHg. Approximately 3.2 g of 2-mercaptocyclohexanemethanol (S.8) should be collected as a colorless oil (70% yield, nD20 = 1.5188). 1H NMR (CDCl3, δ): 3.49 ppm (s, 2H, CH2OH), 2.15 ppm (bs, 1H, CH2OH), 1.31 ppm (s, 1H, CHSH), 1.15-1.85 ppm (m, 10H). 13C NMR (CDCl3, δ): 22.0, 26.07, 36.06, 52.34, 73.12 ppm. FAB MS (negative mode, Cs+, 13 keV, matrix GLY) m/z 145, [M]–, 100. FAB MS (positive mode, Cs+, 13 keV, matrix GLY) m/z 113, [C6H10CH2OH]+, 45%; m/z 129, [C6H10SHCH2]+, 20%.

Synthesize 2-chloro-spiro-4,4-pentamethylene-1,3,2-oxathiaphospholane (S.9) 20. In a 1-L two-neck flask equipped with a thermometer and a dropping funnel, add 28.2 g (0.21 mol) PCl3 to 500 mL of dry benzene under an argon (optionally nitrogen) atmosphere. Cool the flask to 5°C with an ice bath. Add, dropwise through the funnel, a solution of 20 g (0.14 mol) S.8 and 22 mL (0.27 mol) pyridine in 35 mL dry benzene over a 15-min period with magnetic stirring. Keep the temperature of the reaction mixture below 10°C. 21. Continue stirring at room temperature for 30 min and filter off pyridine hydrochloride with exclusion of moisture. Load the reaction mixture in a filter funnel inside a bag filled with dry argon (or nitrogen) and gently apply suction to keep the bag slightly inflated with continuous delivery of dry gas. 22. Evaporate the solvent under reduced pressure with exclusion of moisture (preferably in a rotary evaporator equipped with a diaphragm vacuum). If a water aspirator must be used, insert a drying tube filled with blue indicator silica gel between the rotary evaporator and the aspirator to reduce the risk of hydrolysis of the product. Apply vacuum gently.

23. Distill the product in a high-vacuum fractional distillation apparatus under reduced pressure (0.01 mmHg, provided by a high-vacuum oil pump). Collect the fraction boiling between 82° and 84°C, which contains S.9. CAUTION: Avoid overheating the vessel. Keep pressure at 0.01 mmHg. When overheated, spontaneous decomposition of the crude product may occur, leading to destruction of the apparatus.

Synthesis of Phosphorothioate Oligonucleotides with Stereodefined Linkages

Approximately 20 g of 2-chloro-spiro-4,4-pentamethylene-1,3,2-oxathiaphospholane (S.9) should be collected as a colorless liquid (70% to 75% yield). 31P NMR (C6D6; δ) 217.7 ppm; EI (electron impact) MS: (70 eV) m/z 210, [M]+, 12%; m/z 175, [M-Cl]+, 5.8%; m/z 90, 100%.

24. Store S.9 in a tightly closed vessel inside another tightly closed container filled with several grams of dry 4A molecular sieves at –20°C (stable for at least 1 year).



SYNTHESIS OF 2,2′-DITHIOBIS([18O]CYCLOHEXANECARBOXALDEHYDE) The 18O-labeled 2,2′-dithiobis(cyclohexanecarboxaldehyde), which can be further transformed into the corresponding phosphitylating reagent as described in Basic Protocol 1, is obtained by hydrolysis of the N-phenylimine derivative of 2,2′-dithiobis(cyclohexanecarboxaldehyde) with H2[18O], catalyzed with gaseous hydrogen chloride (Fig. 4.17.5). The N-phenylimine derivative is obtained from 2,2′-dithiobis(cyclohexanecarboxaldehyde) (S.6; see Basic Protocol 1, step 8) upon treatment with aniline.


Additional Materials (also see Basic Protocol 1) 2,2′-Dithiobis(cyclohexanecarboxaldehyde) (S.6; see Basic Protocol 1, step 8) Aniline, freshly distilled in inert atmosphere 95:5 (v/v) chloroform/hexane H2[18O] (95 atom%) Hydrogen chloride, anhydrous Tetrahydrofuran (THF), dried over sodium hydride 250-mL two-neck round-bottom flasks Azeotropic trap (e.g., mini Dean-Stark trap, Aldrich) 8 × 40–cm chromatography column packed with silica gel 60, 230 to 400 mesh (Merck) Rubber septum TLC silica gel plates with UV indicator (Merck; also see APPENDIX 3D) High vacuum valve (e.g., Rotaflo, Quickfit) Drying tube (8 × 5⁄8 in. with connectors, Aldrich) 2- to 5-mL gas-tight syringe Rotary evaporator with water aspirator or membrane pump Buchner funnel with glass frit Additional reagents and equipment for column chromatography (APPENDIX 3E), thin-layer chromatography (TLC; APPENDIX 3D), and high-vacuum transfer of solvent (see Support Protocol 1)

O

H S

H S

O

PhN

NPh S

PhNH2

S

−H2O 6

10 18O

18

[ O ]-H2O

H S

H S

HCl

18O

18O

1. NaBH 4

P Cl S

2. LiAlH 4 3. PCl 3 11

12

Figure 4.17.5 Synthesis of 2-chloro-spiro-4,4-pentamethylene-1,3,2-[18O]oxathiaphospholane (S.12) starting from 2,2′-dithiobis(cyclohexanecarboxaldehyde) (S.6).



Supplement 15

Synthesize N-phenylimine derivative of 2,2′-dithiobis(cyclohexanecarboxaldehyde) (S.10) 1. In a 250-mL two-neck flask equipped with an azeotropic trap, reflux condenser, and dropping funnel, dissolve 14.3 g (0.05 mol) of 2,2′-dithiobis(cyclohexanecarboxaldehyde) (S.6) in 150 mL benzene. Add anti-bumping granules and heat to boiling. 2. Add, dropwise through the dropping funnel, a solution of 10.0 mL (10.2 g, 0.11 mol) aniline in 25 mL benzene over a 30-min period. Continue the reaction for 30 min, keeping gently boiling with azeotropic removal of liberated water. The end of the reaction is confirmed by disappearance of a resonance line of the aldehyde proton in the 1H NMR spectrum.

3. Cool the reaction mixture to room temperature and evaporate the solvent under reduced pressure using a rotary evaporator with a water aspirator. 4. Dissolve the residue in 15 to 20 mL of benzene and apply to an 8 × 40–cm chromatography column packed with ∼200 g of 230 to 400 mesh silica gel. 5. Elute the column with chloroform and collect the eluate in 12- to 15-mL fractions. 6. Analyze fractions by TLC on silica gel plates (APPENDIX 3D). Develop TLC plates with 95:5 (v/v) chloroform/hexane. 7. Combine all fractions that contain the desired product (S.10; Rf = 0.55). Evaporate the solvent under reduced pressure. Typically 18 g (80% yield) of the N-phenylimine derivative (S.10; see Fig. 4.17.5) should be obtained. 1H NMR (CDCl3, δ): 6.67-6.80 ppm (m, 1H), 7.07-7.55 ppm (m, 4H), 3.6 ppm (very broad singlet, 1H, CH=NPh), 1.29-2.18 ppm (m, 10H). 13C NMR (CDCl3, δ): 22.8, 24.88, 25.22, 30.05, 30.15, 33.23, 56.13, 56.793, 60.13, 60.62, 76.38, 77.01, 77.65, 114.84, 118.14, 120.63, 120.69, 125.42, 125.69, 128.11, 128.78, 129.00, 146.28, 150.99, 151.33, 164.78, 165.62, 194.09, 194.77. This number of resonances in the 13C NMR spectrum reflects the presence of cis- and trans-isomers of the N-phenylimine derivative.

Synthesize labeled 2,2′-dithiobis([18O]cyclohexanecarboxaldehyde) (S.11) 8. Place 16 g (0.037 mol) S.10 in a 250-mL two-neck round-bottom flask with a magnetic stir bar inside, with a high-vacuum valve in one joint and a rubber septum in the other, and dry overnight at high vacuum (0.01 mmHg; provided by high-vacuum oil pump). At the end of drying, deliver dry argon gas to the flask through the septum. 9. Prepare an absorber containing ∼150 mL of 4 to 5 M NaOH (see Basic Protocol 1, step 1). 10. Using the vacuum line technique (see Support Protocol 1), transfer ∼100 mL of dry THF to the flask containing S.10. 11. Connect the vacuum valve with the absorber through an 8 × 5⁄8–in. drying tube (filled with anhydrous magnesium sulfate) and a safety flask (Fig. 4.17.4). 12. Add 1.8 mL (0.9 mol) H2[18O] (20% excess) with a 2- to 5-mL gas-tight syringe. 13. Flush the apparatus continously with dry argon and cool the mixture to 70% in both cases.

Time Considerations The preparation of the phosphoramidites S.3 or S.7 requires ∼13 days including drying glassware and reagents, setting up and working up reactions, and purifying and drying products. DNA synthesis, 5′-biotinylation, and postsynthetic cleavage/deprotection can be achieved in 1 or 2 days depending on the length of the sequence. Affinity purification can be carried out within 6 hr, but drying the product may require ∼12 hr.

Literature Cited Agrawal, S., Christodoulou, C., and Gait, M.J. 1986. Efficient methods for attaching non-radioactive labels to the 5′ ends of synthetic oligodeoxyribonucleotides. Nucl. Acids Res. 14:6227-6245. Cadet, J. and Vigny, P. 1990. The photochemistry of nucleic acids. In Bioorganic Photochemistry (H. Morrison, ed.) Vol. 1, pp. 170-184. John Wiley & Sons, New York.

Reversible Biotinylation of the 5′-Terminus of Oligodeoxyribonucleotides

Dawson, B.A., Herman, T., Haas, A.L., and Lough, J. 1991. Affinity isolation of active murine erythroleukemia cell chromatin: Uniform distribution of ubiquitinated histone H2A between active and inactive fractions. J. Cell. Biochem. 46:166-173.

De Vos, M.J., Van Elsen, A., and Bollen, A. 1994. New non-nucleosidic phosphoramidites for the solid phase multi-labeling of oligonucleotides: Comb- and multifork-like structure. Nucleosides Nucleotides 13:2245-2265. Fang, S. and Bergstrom, D.E. 2003a. Reversible biotinylation phosphoramidite for 5′-end-labeling, phosphorylation and affinity purification of synthetic oligonucleotides. Bioconjugate Chem. 14:80-85. Fang, S. and Bergstrom, D.E. 2003b. Fluoridecleavable biotinylation phosphoramidite for 5′end-labeling, affinity purification of synthetic oligonucleotides. Nucl. Acids Res. 31:708-715. Gildea, B.D., Coull, J.M., and Koster, H. 1990. A versatile acid-labile linker for modification of synthetic biomolecules. Tetrahedron Lett. 31:7095-7098. Greenberg, M.M. 1995. Photochemical release of protected oligodeoxyribonucleotides containing 3′-glycolate termini. Tetrahedron 51:29-38. Greenberg, M.M. and Gilmore, J.L. 1994. Cleavage of oligonucleotides from solid-phase supports using o-nitrobenzyl photochemistry. J. Org. Chem. 59:746-753. Guzaev, A., Salo, H., Azhayev, A., and Lönnberg, H. 1995. A new approach for chemical phosphorylation of oligonucleotides at the 5′-terminus. Tetrahedron 51:9375-9384. Langer, P.R., Waldrop, A.A., and Ward, D.C. 1981. Enzymatic synthesis of biotin-labeled polynucleotides: Novel nucleic acid affinity probes. Proc. Natl. Acad. Sci. U.S.A. 78:6633-6637. McInnes, J.L. and Symons, R.H. 1989. Preparation and detection of nonradioactive nucleic acid and oligonucleotide probes. In Nucleic Acid Probes (R.H. Symons, ed.) pp. 33-80. CRC Press, Boca Raton, Fla. Neuner, P. 1996. New non nucleosidic phosphoramidite reagent for solid phase synthesis of biotinylated oligonucleotides. Bioorg. Med. Chem. Lett. 6:147-152. Olejnik, J., Krzymanska-Olejnik, E., and Rothschild, K.J. 1996. Photocleavable biotin phosphoramidite for 5′-end-labelling, affinity purification and phosphorylation of synthetic oligonucleotides. Nucl. Acids Res. 24:361-366. Pon, R.T. 1991. A long chain biotin phosphoramidite reagent for the automated synthesis of 5′-biotinylated oligonucleotides. Tetrahedron Lett. 32:1715-1718. Reddy, M.P., Hanna, N.B., and Farooqui, F. 1994. Fast cleavage and deprotection of oligonucleotides. Tetrahedron Lett. 35:4311-4314. Shimkus, M., Levy, J., and Herman, T. 1985. A chemically cleavable biotinylated nucleotide: Usefulness in the recovery of protein-DNA complexes from avidin affinity columns. Proc. Natl. Acad. Sci. U.S.A. 82:2593-2597.



Urdea, M.S., Warner, B.D., Running, J.A., Stempien, M., Clyne, J., and Horn, T. 1988. A comparison of non-radioisotopic hybridization assay methods using fluorescent, chemiluminescent and enzyme labeled synthetic oligodeoxyribonucleotide probes. Nucl. Acids Res. 16:49374956. Wincott, F., DiRenzo, A., Shaffer, C., Grimm, S., Tracz, D., Workman, C., Sweedler, D., Gonzalez, C., Scaringe, S., and Usman, N. 1995. Synthesis, deprotection, analysis and purification of RNA and ribozymes. Nucl. Acids Res. 23:2677-2684.

Contributed by Shiyue Fang and Donald E. Bergstrom Purdue University West Lafayette, Indiana Walther Cancer Institute Indianapolis, Indiana



Supplement 14

Uridine 2′-Carbamates: Facile Tools for Oligonucleotide 2′-Functionalization

UNIT 4.21

This unit contains procedures for synthesis of uridine 2′-carbamate phosphoramidites and oligonucleotides thereof. 3′,5′-Silyl-diprotected uridine can be converted into the corresponding 2′-carbamate by reaction with 1,1′-carbonyldiimidazole followed by treatment with an aliphatic amine. The 2′-carbamate can then be converted in several steps into a 3′-phosphoramidite suitable for machine-assisted oligonucleotide synthesis. The preparation of eleven different uridine 2′-carbamates and their phosphoramidites from several primary and secondary amines is described in the first two methods (see Basic Protocol 1 and Alternate Protocol). Their use in oligonucleotide synthesis is then described (see Basic Protocol 2). 2′-Carbamate modification is stable to conditions of standard phosphoramidite oligonucleotide synthesis. Although 2′-carbamate modification is somewhat destabilizing for DNA-DNA and DNA-RNA duplexes, it is suitable for the direction of ligands into the minor groove or into non-base-paired sites (e.g., loops, bulges) of oligo- and polynucleotides. Pyrene-modified oligonucleotide 2′-carbamates show a considerable increase in fluorescence intensity upon hybridization to a complementary RNA, and have interesting binding properties when hybridized to a mismatched DNA. CAUTION: Carry out all operations involving organic solvents and reagents in a wellventilated fume cupboard, and wear gloves and protective glasses. PREPARATION OF URIDINE 2′-CARBAMATE PHOSPHORAMIDITES FROM PRIMARY AND SECONDARY AMINES Nucleoside 2′-carbamates have been used previously in the syntheses of various modified nucleosides (McGee et al., 1996; Zhang et al., 2003) and oligonucleotides (Freier and Altmann, 1997; Seio et al., 1998; Dubey et al., 2000; Prhavc et al., 2001). The present protocol is based primarily on Korshun et al. (2002) and is illustrated in Figure 4.21.1. St able 2′-O-(imidazol-1-ylcarbonyl)-3′,5′-O-(tetraisopropyldisiloxan-1,3-diyl)uridine (S.3) is first prepared from uridine (S.1). It is then treated with primary or secondary aliphatic amines to give uridine 2′-carbamates (S.4) in high yield. Preparation of 2′-carbamates bearing a variety of N-substituents is described (S.4a-f; see Fig. 4.21.2). After 3′,5′-O-deprotection with triethylamine trihydrofluoride, 5′-O-dimethoxytritylation, and 3′-O-phosphinylation with bis(N,N-diisopropylamino)-2-cyanoethoxyphosphine, the corresponding phosphoramidites (S.7a-f) are obtained. They are used in machine-assisted synthesis of modified oligodeoxynucleotides containing uridine-2′-carbamate residues bearing these N-substituents. Strategic planning Formation of 2′-carbamates. In many cases, the imidazolide S.3 need not be isolated, but instead can be reacted in situ with an excess of the appropriate amine in dry dichloromethane (CH2Cl2). In the case of an amine hydrochloride (e.g., for S.4d), 1.1 eq triethylamine (TEA) or N,N-diisopropylethylamine (DIPEA) is added to the reaction mixture to liberate the free base. Reaction times differ vastly from one amine to another. Whilst propargylamine reacts rapidly in CH2Cl2 (99.9%, Romil) 2,6-Dihydroxyacetophenone (DHAP, ≥99.0%, Fluka) Methanol Diammonium hydrogen citrate (≥99.0%, Fluka)



Millipore water or double-distilled deionized water 2,4,6-Trihydroxyacetophenone (THAP, ≥99%, Fluka) 1.5-mL solvent-resistant microcentrifuge tubes MALDI-TOF mass spectrometer Prepare matrix For peptides, PNA oligomers, and peptide-PNA conjugates 1a. Dissolve 10 mg CHCA in 0.5 mL acetonitrile in a 1.5-mL solvent-resistant microcentrifuge tube. 2a. Add 0.5 mL of 3% aq. TFA solution. 3a. Mix thoroughly. Continue with step 4.

For oligonucleotides and peptide-oligonucleotide conjugates (low salt content in sample) 1b. Dissolve 20 mg DHAP in 0.5 mL methanol in a 1.5-mL solvent-resistant mirocentrifuge tube. 2b. Dissolve 40 mg diammonium hydrogen citrate in 0.5 mL water in a separate tube. 3b. Mix the two solutions together thoroughly. Continue with step 4.

For oligonucleotides and peptide-oligonucleotide conjugates (high salt content in sample; also used to check for contamination by peptides) 1c. Add 0.5 mL acetonitrile to 60 mg THAP in a 1.5-mL solvent-resistant microcentrifuge tube and vortex for 1 min. 2c. Dissolve 5.7 mg diammonium hydrogen citrate in 0.5 mL water and add to THAP/acetonitrile mixture. 3c. Vortex for 30 sec and centrifuge 3 min at 16,060 × g (13,000 rpm), room temperature. Use the supernatant and continue with step 4. Excess THAP will be in pellet form at the bottom of the tube.

Prepare sample 4. Spot 1 µL of the matrix onto the plate. 5. Immediately (within 5 sec) add 0.1 µL of the sample and mix by pipetting the solution up and down five times with a pipet. The concentration of the sample is critical for good mass acquisition (see UNIT 10.1). Sample concentrations exceeding 1 mM can give inferior spectra. For THAP matrix (which is saturated), immediate crystallization can occur upon dispensing the sample. In this case, do not mix. In general, cease mixing the sample/matrix mixture if crystallization commences.

6. Allow to crystallize and dry completely. 7. Acquire spectra in positive ion mode (see UNIT 10.1).



Supplement 24

SUPPORT PROTOCOL 2

DETERMINATION OF THIOL CONTENT BY THE ELLMAN’S TEST The Ellman’s test is a widely used and established technique for the quantitation of thiol content in a given solution.

Materials Ellman’s reagent: 2 mM dithio-bis-2-nitrobenzoic acid (DTNB) in 50 mM sodium acetate (NaOAc) 2.0 M Tris·Cl, pH 8.0 UV spectrometer with 1-mL, 1-cm path length quartz glass cuvette (Suprasil, Hellma) 1. Add 50 µL Ellman’s reagent, 100 µL of 2.0 M Tris·Cl, and water (850 µL minus the sample volume) to a quartz cuvette. 2. Mix thoroughly by carefully pipetting 200 µL of solution up and down ten times with a pipet. 3. Zero the UV spectrometer at 412 nm. 4. Repeat step 2 and check that the UV spectrometer still reads zero. If not, repeat steps 2 and 3 until it does. 5. Add the sample to the cuvette, mix carefully ten times, and take the reading. 6. Calculate absorbance as follows:

A(sample) = [total volume (µL)/sample volume (µL)] × A412 7. Repeat steps 1 through 6 two times, varying the amount of sample used (e.g., 10, 15, 20 µL), and take the average. 8. Calculate the thiol content as follows:

thiol (M) = average A(sample)/13,600 where 13,600 M−1 cm−1 is the extinction coefficient of the reagent. Due to the sensitivity of the test, when fluorophores are present (e.g., fluorescein, 492 nm maximum) ensure that a background check at pH 8.0 (i.e., no Ellman’s reagent present) at 412 nm is carried out so that the absorbance value reading can be adjusted if necessary.


Disulfide Conjugation of Peptides to Oligonucleotides and Their Analogs

Antisense oligonucleotides and siRNA The antisense oligonucleotide field emerged some 25 years ago with the pioneering work of Zamecnik (Zamecnik and Stephenson, 1978). However, a severe limitation to biological activity of oligonucleotides and their analogs has been the poor cellular uptake and delivery into the right cell compartment. Most oligonucleotides and analogs are not taken up into cells in culture without use of an additional carrier, of which cationic lipids (such as Lipofectamine 2000) are the most popular (Bennett et al., 1992). The same problems of poor cell uptake extend

to siRNA and PNA. For therapeutic use, cationic lipids have disadvantages because of the need for careful formulation and difficulties in maintaining stability. Major efforts have been made to find alternative methods of oligonucleotide and siRNA drug delivery that do not require formulation, but merely involve chemical modification of the oligonucleotide itself. Peptide conjugates of oligonucleotides to enhance cell delivery One idea that has been pursued for several years is the covalent conjugation of cellpenetrating peptides (CPP), also known as protein transduction domains (PTD) (Lindgren



et al., 2000; Lindsay, 2002; Wadia and Dowdy, 2002; Lochmann et al., 2004; Zorko and Langel, 2005). Such peptides have been shown to translocate into mammalian cells in culture, but it has been controversial as to what extent such peptides are taken up by endocytotic pathways or by non-energy-dependent processes. It appears that some cationic peptides, such as Tat, are predominantly taken up by endocytosis (Richard et al., 2003), but even now the uptake mechanism is disputed (Potocky et al., 2003; Kaplan et al., 2005; Ziegler et al., 2005). It is also less clear for more hydrophobic peptides such as Transportan (Zorko and Langel, 2005). Methods of synthesis of peptide conjugates of oligonucleotides and analogs have been reviewed (Zubin et al., 2002; Zatsepin et al., 2005) and their applications for cell delivery covered extensively (Stetsenko et al., 2000; Gait, 2003; Thierry et al., 2003; Shi and Hoekstra, 2004; Juliano, 2005). Very recently there have been a couple of papers suggesting that peptide conjugates of siRNA are also useful for carrier-free delivery (Chiu et al., 2004; Muratovska and Eccles, 2004). Such peptide carrier–oligonucleotide or siRNA cargo conjugates can be synthesized either with a stable linkage (such as amide, thioether, oxime, or hydrazine) or with an unstable linkage (such as disulfide). Disulfide-linked peptide-oligonucleotide and peptide-PNA conjugates Very little information had been available as to whether unstable disulfide linkages may have any advantage over more stably linked conjugates. Recent evidence from the PNA-peptide field suggests, at least for this class of conjugates, that disulfide-linked conjugates may have higher activity (Koppelhus et al., 2002; Turner et al., 2005b). However, this has not been proven in the case of peptide conjugates of negatively charged oligonucleotides, where sufficient biological activity has been hard to achieve (Antopolsky et al., 1999; Astriab-Fisher et al., 2000; Turner et al., 2005a). Potentially, the disufide bond may be cleaved by reduction within a mammalian cell to allow cargo release linkage, but the rates of such intracellular cleavage are unknown and may vary depending on the cargo as well as the peptide type. Disulfide conjugates have become popular with laboratories studying the uptake and activity of peptideantisense cargo conjugates because the chemistry is simple and reasonably specific, and the thiol functional groups needed on each component are readily introduced. However,

with regards to the synthesis of highly cationic peptide-oligonucleotide conjugates, reliable protocols have been missing, and there have even been reports suggesting that such conjugations are impossible to achieve (Prater and Miller, 2004). The protocols provided here for synthesis, purification, and analysis of peptide-oligonucleotide (Turner et al., 2005a) and peptide-PNA (Turner et al., 2005b) conjugates should fill this gap and encourage others to explore a wider variety of peptides and oligonucleotides for conjugation, toward the goal of improved cell delivery. The authors’ results in this area have shown that conjugation by several types of CPP does indeed enhance cellular uptake into HeLa cells and fibroblasts in cell culture, but that biological activity levels in a model system involving steric block of HIV-1 Tat-dependent transactivation in the nucleus are limited in most cases by slow release from endosomal compartments (Turner et al., 2005a,b). However, recent exciting results suggest that disulfidelinked PNA-peptides targeting the HIV-1 TAR region have great potential as antiviral and virucidal agents (Tripathi et al., 2005) and also in the targeting of neuropeptide receptor mRNAs (Kilk et al., 2004). Disulfidelinked peptide-oligonucleotide conjugates are also useful for delivery into living cells for the detection of mRNA by a molecular beacons approach (Nitin et al., 2004).

Critical Parameters and Troubleshooting For the synthesis of peptides using the APEX or Pioneer machines, problematic sequences (often indicated by repeating motifs or large sections of hydrophobicity) should be identified and the use of double couplings or double-double couplings employed. Although the APEX machine can in theory synthesize 96 peptides in parallel, it is best to operate with smaller numbers due to the variation in exposure time to reagents in the different steps (e.g., piperidine). The authors have found that groups of four to six peptides at a time are manageable. Although the machine is by definition automated, it should be monitored as much as possible to catch any errors that may occur. Manual intervention at this stage can rescue the synthesis and prevent loss of reagents and time. The programming of the machine should also be double-checked before commencing the synthesis. The Pioneer peptide synthesizer, which is a flow-through instrument where one peptide is synthesized at a time, is less



Supplement 24

prone to error, but is no longer commercially available. Reagent solutions should not be used if older than 3 days. This is especially true for PyBop. Any solutions exhibiting discoloration should be discarded. Masses of +170 and +50 are occasionally observed as a result of degraded resin. Replacing the resin solves this. For PNA synthesis, much of what has been described for the synthesis of peptides on the APEX machine also holds true (i.e., programming and monitoring the machine during synthesis). Several additional factors are also critical to the synthesis of the PNA oligomers. (1) A short piperidine treatment should be used for cleavage of the Fmoc protecting group. The Bhoc nucleobase protecting group is susceptible to cleavage when longer (standard peptide) reaction times are used. (2) The solubility of the PNA monomers and, in general, the growing chain can be maintained by the addition of N-methylpyrrolidone (NMP) to the reaction mixture, if necessary. (3) The quality of the coupling reagent is of greater importance in PNA synthesis than in peptide synthesis. Although an amide bond is formed in both cases, peptide synthesis and solutions are more tolerant in general to the presence of small amounts of water than PNA synthesis. Poor yields can in some cases be rectified by adding molecular sieves to the solutions for PNA synthe-


sis. Stock solutions of thiol-containing peptides that are not highly cationic and of PNA oligomers should be stored in 0.1% aqueous TFA to inhibit dimerization. For both, acidic media may be required for solvation. Oligonucleotide synthesis is generally straightforward. With LNA, longer coupling times and a stronger activator (e.g., 5-ethylthiotetrazole) are used. For conjugations, it is important that the quality of both starting materials be high. Pure oligonucleotide or PNA and peptide suitable for conjugation will simplify the purification of the conjugate. It has been observed occasionally that the desalting procedure after the liberation of the oligonucleotide thiol with DTT can be incomplete. Traces of DTT in the final solution can diminish the coupling efficiency with activated peptides. Desalting again will resolve this problem. Alternatively, begin with a bigger Sephadex NAP column. It is not advisable to dialyze the thiol oligonucleotide, as dimer formation will occur. Maintaining solubility is vital for the smooth formation and purification of conjugates. The addition of a highly cationic peptide to a solution of oligonucleotide without formamide present will result in immediate aggregation. This is best visualized with fluorescently labeled materials. A high percentage of the solution should be formamide to ensure solubility. This percentage can rise to >90%

Figure 4.28.2 HPLC chromatogram of conjugate formation from peptide RQIKIWFQNRRMKWKKGGC with (pys)S-(CH2 )6 -5 -2 -O-Me/LNA[CUC CCA GGC UCA]-3 -fluorescein; peaks (i) salts and formamide, (ii) excess peptide, (iii) conjugate product, (iv) unconjugated oligo(pys). The solid trace is at 280 nm and the dashed trace is at 480 nm, which identifies the fluorescein label on the oligonucleotide.



Figure 4.28.3 MALDI-TOF mass spectra of GRKKKRRQRRRPC(S-)-S-(CH2 )6 -5 -2 -O-Me/LNA[CUC CCA GGC UCA]-3 -fluorescein conjugate: (A) purified and (B) pure conjugate with 0.5 eq. peptide added.

for reactions involving highly cationic peptides. It is best to keep highly concentrated stocks of peptides and oligonucleotides, so that the amount of formamide needed does not prevent single injection purification by HPLC. A high concentration is advisable to ensure quick reaction times. When purifying the conjugate, care must be taken when formamide and 2 M TEAA are used to resolubilize aggregated material. It has been observed that HPLC runthrough can occur when large amounts have been used. This material can be collected and re-injected in the following run.

Care must be taken when loading Slidea-Lyzer cassettes so as not to puncture the dialysis membrane. In addition, for highly basic peptide-oligonucleotide conjugates, there is a risk of self-aggregation onto the membrane during dialysis, resulting in a reduction in yield. The Slide-a Lyzer cassette can be washed with appropriate buffer (0.1 to 2 M TEAA) to solubilize the conjugate. This can then be lyophilized/concentrated or transferred into the appropriate buffer with a NAP column. Alternatively, dialysis can be carried out using an appropriate buffer instead



Supplement 24

Figure 4.28.4 RP-HPLC of (A) the synthesis of RRRRRRRQIKIWFQNRRMKWKKGGCCK[CTCCCAGGCTCAGATC]PNA KKK and (B) analysis of the purified product.


of water. Lyophilized conjugates can require a modest amount of salt to aid their solution. Often, adding a small amount of 2 M TEAA is sufficient—typically, enough to make a 10 mM solution, but in some cases up to 1 M TEAA is needed. For these latter, difficult cases, it is necessary to add concentrated

salt solution first and then dilute the solution with water. MALDI-TOF mass spectral analysis of conjugates dissolved in high salt may be problematic due to poor crystallization. Use THAP matrix (see Support Protocol 1), diluting a small amount of the sample, if necessary.



Figure 4.28.5

MALDI-TOF mass spectrum of the conjugate described in Figure 4.28.4.

The conjugation reaction and purification of PNA-peptide conjugates is less problematic, since aggregation with cationic peptides does not occur. However, the overall solubility of conjugates should be considered (e.g., the conjugation of hydrophobic peptides to PNA). Self-aggregation of PNA oligomers can occur at neutral pH. This is circumvented by the addition of formamide. A heated HPLC column is advisable to obtain good resolution with all conjugates synthesized. Provided that the oligonucleotide (or analog) and peptide remain soluble at the concentrations described, disulfide formation should be complete within 30 min. Reaction times significantly greater than this can cause scrambling of the disulfide bond and hence a reduction in yield. Due to the short reaction times, it is not necessary to remove oxygen from the aqueous solution.

purity ≥90%. Oligonucleotide-peptide conjugations proceed >90% according to HPLC (see Fig. 4.28.2 for an example). Due to the self-aggregation properties of some conjugates (discussed above), yields are often lower. Yields may vary between 12% for selfaggregating conjugates and 75% for minimally self-aggregating conjugates. The presence of excess peptide can be verified by MALDI-TOF mass spectral analysis using the THAP matrix (see Fig. 4.28.3). Note that mass M/2 is often observed with THAP matrix. Conversion of a PNA oligomer into a conjugate is typically 80% to 90% (see Fig. 4.28.4). Yields obtained are typically 45% to 70%. Purity of all conjugates is >90%. MALDITOF mass spectra of PNA conjugates typically show molecular mass of the conjugate, molecular mass of conjugate/2, PNA thiol, and peptide thiol (see Fig. 4.28.5).

Time Considerations Anticipated Results Good yields and good purity can be expected for both peptides and oligonucleotides. Activation of peptide or oligonucleotide thiol groups proceeds near quantitatively (as judged by HPLC analysis). However, after isolation of the desired activated material, a typical yield is ≥85%. Liberation of the thiol from oligonucleotides typically proceeds >90%, sometimes quantitatively. For PNA synthesis, yields of ∼20% are expected after purification, depending on the sequence and length, with

The assembly of peptides and PNA can be completed within hours depending on the length of the oligomer. Allow 1 to 3 hr to program and set up the machine as well as prepare reagents, and then ∼1 hr per cycle. Oligonucleotide synthesis is more rapid, with cycles of 15 to 30 min. Conjugations are typically finished within 30 min. HPLC purification for oligo-peptide conjugates is 30 min per run. Dialysis is usually achieved with one change of water allowing at least 6 hr between (resulting in



Supplement 24

≥1 × 106 dilution of salt content). Purification of PNA-peptide conjuates takes 45 min. Starting with synthesis of the peptide and oligonucleotide (or analog), allow 2 to 3 weeks for the synthesis of conjugates altogether. Allow longer if experience is short in these areas and/or more than one peptide or oligonucleotide is being synthesized. If the oligonucleotide and peptide are obtained commercially, allow at least 1 week for the synthesis and analysis of conjugates.

Literature Cited Antopolsky, M., Azhayeva, E., Tengvall, U., Auriola, S., Jäa¨ skeläinen, I., Rönkkö, S., Honkakoski, P., Urtti, A., Lönnberg, H., and Azhayev, A. 1999. Peptide-oligonucleotide phosphorothioate conjugates with membrane translocation and nuclear localization properties. Bioconjug. Chem. 10:598-606. Astriab-Fisher, A., Sergueev, D.S., Fisher, M., Shaw, B.R., and Juliano, R.L. 2000. Antisense inhibition of P-glycoprotein expression using peptide-oligonucleotide conjugates. Biochem. Pharmacol. 60:83-90. Bennett, C.F., Chiang, M.-Y., Chan, H., Shoemaker, J.E.E., and Mirabelli, C.K. 1992. Cationic lipids enhance cellular uptake and activity of phosphorothioate antisense oligonucleotides. Mol. Pharmacol. 41:1023-1033. Bongartz, J.P., Aubertin, A.M., Milhaud, P.G., and Lebleu, B. 1994. Improved biological activity of antisense oligonucleotides conjugated to a fusogenic peptide. Nucl. Acids Res. 22:4681-4688. Braun, K., Peschke, P., Pipkorn, R., Lampel, S., Wachsmuth, M., Waldeck, W., Friedrich, E., and Debus, J. 2002. A biological transporter for the delivery of peptide nucleic acids (PNAs) to the nuclear compartment of living cells. J. Mol. Biol. 318:237-243. Chiu, Y.-L., Ali, A., Chu, C., Cao, H., and Rana, T.M. 2004. Visualizing a correlation between siRNA, localization, cellular uptake and RNAi in living cells. Chem. Biol. 11:1165-1175. Corey, C.R. 1995. 48000-fold acceleration of hybridisation by chemically modified oligonucleotides. J. Amer. Chem. Soc. 117:93739374. Eritja, R., Pons, A., Escarceller, M., Giralt, E., and Albericio, F. 1991. Synthesis of defined peptide-oligonucleotide hybrids containing a nuclear transport signal sequence. Tetrahedron 47:4113-4120.


Gait, M.J. 2003. Peptide-mediated cellular delivery of antisense oligonucleotides and their analogues. Cell. Mol. Life Sci. 60:1-10.

enters cells by macropinocytosis. J. Control Release 102:247-253. Kaushik, N., Basu, A., Palumbo, P., Nyers, R.L., and Pandey, V.N. 2002. Anti-TAR polyamide nucleotide analog conjugated with a membranepermeating peptide inhibits Human Immunodeficiency Virus Type I production. J. Virol. 76:3881-3891. Kilk, K., Elmquist, A., Saar, K., Pooga, M., Land, T., Bartfai, T., Soomets, U., and Langel, U. 2004. Targeting of antisense PNA oligomers to human galanin receptor type 1 mRNA. Neuropeptides 38:316-324. Koppelhus, U., Awasthi, S.K., Zachar, V., Holst, H.U., Ebbeson, P., and Nielsen, P.E. 2002. Celldependent differential cellular uptake of PNA, peptides and PNA-peptide conjugates. Antisense & Nucleic Acid Drug Dev. 12:51-63. Lindgren, M., Hällbrink, M., Prochiantz, A., and Langel, U. 2000. Cell-penetrating peptides. Trends Pharmacol. Sci. 21:99-103. Lindsay, M.A. 2002. Peptide-mediated cell delivery: Application in protein target validation. Curr Opin Pharmacol 2:587-594. Lochmann, D., Jauk, E., and Zimmer, A. 2004. Drug delivery of oligonucleotides by peptides. Eur. J Pharm. Biopharm. 58:237-251. Muratovska, A. and Eccles, M.R. 2004. Conjugate for efficient delivery of short interfering RNA (siRNA) into mamalian cells. FEBS Lett. 558:63-68. Nitin, N., Santangelo, P.J., Kim, G., Nie, S., and Bao, G. 2004. Peptide-linked molecular beacons for efficient delivery and rapid mRNA detection in living cells. Nucl. Acids Res. 32:e58. Pooga, M., Soomets, U., Hällbrink, M., Valkna, A., Saar, K., Rezaei, K., Kahl, U., Hao, J.-X., Xu, X.-J., Wiesenfeld-Hallin, Z., Hökfelt, T., Bart¨ 1998. Cell penetrating fai, T., and Langel, U. PNA constructs regulate galanin receptor levels and modify pain transmission in vivo. Nat. Biotechnol. 16:857-861. Potocky, T.B., Menon, A.K., and Gellman, S.H. 2003. Cytoplasmic and nuclear delivery of a TAT-derived peptide and a β-peptide after endocytic uptake into HeLa cells. J. Biol. Chem. 278:50188-50194. Prater, C.E. and Miller, P. 2004. 3 Methylphosphonate-modified oligo-2 -Omethytlribonucleotides and their Tat peptide conjugates: Uptake and stability in mouse fibroblasts in culture. Bioconjug. Chem. 15:498-507. Richard, J.-P., Melikov, K., Vivès, E., Ramos, C., Verbeure, B., Gait, M.J., Chernomordik, L.V., and Lebleu, B. 2003. Cell-penetrating peptides. A re-evaluation of the mechanism of cellular uptake. J. Biol. Chem. 278:585-590.

Juliano, R.L. 2005. Peptide-oligonucleotide conjugates for the delivery of antisense and siRNA. Curr. Opin. Mol. Ther. 7:132-138.

Shi, F. and Hoekstra, D. 2004. Effective intracellular delivery of oligonucleotides in order to make sense of antisense. J. Control Release 97:189-209.

Kaplan, I.M., Wadia, J.S., and Dowdy, S.F. 2005. Cationic TAT peptide transduction domain

Stetsenko, D.A., Arzumanov, A.A., Korshun, V.A., and Gait, M.J. 2000. Peptide conjugates of



oligonucleotides as enhanced antisense agents: A review. Mol. Biol. (Russ.) 34:852-859. Thierry, A.R., Vivès, E., Richard, J.-P., Prevot, P., Martinand-Mari, C., Robbins, I., and Lebleu, B. 2003. Cellular uptake and intracellular fate of antisense oligonucleotides. Curr. Opin. Mol. Ther. 5:133-138. Tripathi, S., Chaubey, B., Ganguly, S., Harris, D., Casale, R.A., and Pandey, P.K. 2005. Anti-HIV1 activity of anti-TAR polyamide nucleic acid conjugated with various membrane transducing peptides. Nucl. Acids Res. 33:4345-4356. Turner, J.J., Arzumanov, A.A., and Gait, M.J. 2005a. Synthesis, cellular uptake and HIV1 Tat-dependent trans-activation inhibition activity of oligonucleotide analogues disulphide-conjugated to cell-penetrating peptides. Nucl. Acids Res. 33:27-42. Turner, J.J., Ivanova, G.D., Verbeure, B., Williams, D., Arzumanov, A., Abes, S., Lebleu, B., and Gait, M.J. 2005b. Cell-penetrating peptide conjugates of peptide nucleic acids (PNA) as inhibitors of HIV-1 Tat-dependent trans-activation in cells. Nucl. Acids Res. 33:6837-6849. Vivès, E. and Lebleu, B. 1997. Selective coupling of a highly basic peptide to an oligonucleotide. Tetrahedron Lett. 38:1183-1186. Wadia, J.S. and Dowdy, S.F. 2002. Protein transduction technology. Curr. Opin. Biotechnol. 13:52-56.

Zamecnik, P.C. and Stephenson, M.L. 1978. Inhibition of Rous sarcoma virus replication and cell transformation by a specific oligodeoxynucleotide. Proc. Natl. Acad. Sci. U.S.A. 75:280-284. Zatsepin, T.S., Turner, J.J., Oretskaya, T.S., and Gait, M.J. 2005. Conjugates of oligonucleotides and analogues with cell penetrating peptides as gene silencing agents. Curr. Pharm. Des. 11:3639-3654. Ziegler, A., Nervi, P., Dürrenberger, M., and Seelig, J. 2005. The cationic cell-penetrating peptide CPPTat derived from the HIV-1 protein TAT is rapidly transported into living fibroblasts: Optical, biphysical, and metabolic evidence. Biochemistry 44:138-148. Zorko, M. and Langel, U. 2005. Cell-penetrating peptides: Mechanism and kinetics of cargo delivery. Adv. Drug Deliv. Rev. 57:529-545. Zubin, E.M., Romanova, E.A., and Oretskaya, T.S. 2002. Modern methods for the synthesis of peptide-oligonucleotide conjugates. Russ. Chem. Rev. 71:239-264.

Contributed by John J. Turner, Donna Williams, David Owen, and Michael J. Gait Medical Research Council, Laboratory of Molecular Biology Cambridge, United Kingdom



Supplement 24

Methoxyoxalamido Chemistry in the Synthesis of Tethered Phosphoramidites and Functionalized Oligonucleotides

UNIT 4.29

This unit elaborates on the synthesis of specialty phosphoramidites tethered with single or multiple linkers through the use of methoxyoxalamido (MOX) precursors. The strategy is based on the reaction of dimethyl oxalate with a proper aliphatic amine (phosphoramidite base) to form a MOX precursor that, in turn, is reacted with a selected primary aliphatic amine (linker) to form a simple tethered synthon for phosphoramidite preparation (Fig. 4.29.1). The strategy is quite general since a vast number of commercially available primary aliphatic amines can be employed. Basic Protocol 1 describes syntheses of exemplary phosphoramidites tethered with single linkers. Additional flexibility of the MOX strategy comes from the possibility of reiterating the dimethyl oxalate/primary diamine treatments leading to synthons with multiple linkers (Fig. 4.29.2). Synthesis of selected phosphoramidites with tethers is outlined in the Alternate Protocol. Finally, Basic Protocol 2 describes the use of tethered phosphoramidites for 5 -derivatization of oligonucleotides.

Figure 4.29.1 Synthetic pathway for phosphoramidites tethered with a single linker. MOX, methoxyoxalamido; DMTr, dimethoxytrityl; MMTr, monomethoxytrityl.

Synthesis of Modified Oligonucleotides and Conjugates

Contributed by Alan M. Morocho and Nikolai N. Polushin

4.29.1

Current Protocols in Nucleic Acid Chemistry (2006) 4.29.1-4.29.19 C 2006 by John Wiley & Sons, Inc. Copyright

Supplement 25

Figure 4.29.2 Synthetic pathway to phosphoramidites tethered with multiple linkers. MOX, methoxyoxalamido; DMTr, dimethoxytrityl; MMTr, monomethoxytrityl; n = number of cycles.

BASIC PROTOCOL 1

PREPARATION OF PHOSPHORAMIDITES TETHERED WITH SINGLE LINKERS Synthetic procedures for preparation of specialty phosphoramidites using MOX chemistry are all straightforward and robust. All necessary chemicals are commercially available and generally inexpensive. One of the starting reagents, 5 -amino-5 deoxythymidine, is somewhat expensive but can be prepared from inexpensive thymidine as described by Bannwarth (1988). There are four major steps in the synthesis of a phosphoramidite tethered with a single linker. First, the MOX derivative is prepared from a proper amino alcohol (phosphoramidite base). Second, the MOX precursor is treated with a primary aliphatic diamine or a primary amino alcohol to form an aminated or hydroxylated tether synthon. Next, the primary amino or hydroxyl group is protected with a monomethoxytrityl (MMTr) or dimethoxytrityl (DMTr) group, respectively. Lastly, the secondary hydroxyl of the MMTr/DMTr-protected intermediate is phosphinylated to give the final phosphoramidite.

Methoxyoxalamido Chemistry for Tethered Phosphoramidites



Figure 4.29.3 depicts structures of the initial MOX precursors S.1 and S.2. Figure 4.29.4 shows phosphoramidites tethered with single aminated linkers (S.3d-S.6d) and their intermediates (S.3a,c-S.6a,c). Figure 4.29.5 shows phosphoramidites tethered with single hydroxylated linkers (S.10d and S.11d) and their precursors (S.10a,c and S.11a,c). Note that the b intermediate for S.5, S.6, and S.8 is used only for the formation of the tethered compounds described in the Alternate Protocol. The following considerations are important in strategic planning for the syntheses described in this unit: Choice of starting material (phosphoramidite base). The choice of starting materials, trans-4-aminocyclohexanol hydrochloride and 5 -amino-5 -deoxythymidine, is based on their availability and the presence of secondary hydroxyl groups, which imparts stability to the corresponding phosphoramidites. Formation of MOX precursors. At least a 2-fold excess of dimethyl oxalate over the amino alcohol should be used to suppress formation of a dimer byproduct. Addition of triethylamine, although not necessary in the case of free amines, does accelerate the reaction. In the case of trans-4-aminocyclohexanol hydrochloride, at least one equivalent of triethylamine must be added to free the protonated amino group. Reaction of MOX precursor with an aliphatic primary diamine. To suppress dimer formation, at least a 4-fold molar excess of diamine over the MOX precursor (S.1 or S.2) should be used. Typically, simple precipitation of the product (S.3a-S.6a) in ether gives purity acceptable to perform the next step. If the diamine used is of a substantial value, it can be recovered from the supernatant by distillation or crystallization. Reaction of MOX precursor with an aliphatic primary amino alcohol. Since only one amino group is available to react with a MOX precursor, only one equivalent of an amino alcohol is needed. Again, the product (S.10a, S.11a) is purified simply by precipitation in ether. Protection of primary amino/hydroxyl group. Due to poor solubility of amino alcohols S.3a-S.6a in pyridine, reaction with MMTr-Cl is carried out in 1:1 (v/v) dimethylformamide (DMF)/pyridine. Even in this solvent composition, the starting materials are not fully soluble, and the reaction mixture is stirred for 2 to 3 days to ensure complete transformation. Diols S.10a and S.11a are soluble in 1:1 DMF/pyridine, and the reaction with DMTr-Cl is generally complete within 12 hr. After extraction and concentration, the product is precipitated into 3:2 (v/v) ether/pentane and is normally used in the next step without further purification. However, if the purity of the MMTr/DMTr-protected intermediate is 50 nt, increase the incubation time to 30 to 35 min. 8. Add 20 µl of 10% (w/v) LiClO4 in EtOH and vortex. Add 700 µL of EtOH, vortex, and leave at room temperature for at least 15 to 20 min to ensure full precipitation of the modified oligonucleotide. 9. Microcentrifuge 5 min at 13,000 rpm. Discard the supernatant. Wash the residue with 700 µL EtOH, microcentrifuge 5 min at 13,000 rpm, and discard the wash. 10. Add 200 µL deionized water, vortex, and microcentrifuge briefly at 13,000 rpm. Transfer the solution to a new 1.7-mL tube. Wash the remaining solid support with 200 µL deionized water, microcentrifuge, and transfer the wash to the main solution. 11. Concentrate the solution to dryness in vacuo on a Speedvac evaporator and dissolve the residue in 7 M urea. 12. Purify by denaturing polyacrylamide gel electrophoresis (PAGE; UNIT 10.4 & APPENDIX 3B) using 15% polyacrylamide and 7 M urea in 0.5× TBE electrophoresis buffer. 13. Cut out the band(s) and elute the product with 0.25 M TEAB. Desalt using a Sephadex G-25 NAP-10 column and lyophilize. 14. Dissolve the oligonucleotide in deionized water and quantitate by measuring UV absorbance at 260 nm. Store the solution at –20◦ C. 15. Analyze the products by MALDI-TOF-MS (UNIT mass.

10.1)

to determine molecular



Use of specialty phosphoramidites, such as amino modifiers and spacers, is widespread in modern oligonucleotide synthesis. A number of phosphoramidites with aminated and hydroxylated linkers are commercially available (Glen Research), but they generally possess at least three limitations. First, these phosphoramidites are mainly derived from aliphatic primary alcohols and are thus not

very stable even at low temperatures due to susceptibility to Arbuzov rearrangement (Polushin, 2000). This instability leads to preparation and storage complications and also necessitates fairly rapid phosphoramidite consumption on automated DNA synthesizers, thus making the commercially available amino modifiers and spacers less than ideal for reliable large-scale oligonucleotide production. Second, the available amino modifier



and spacer phosphoramidites are quite limited in variety. They are primarily flexible in nature, since they contain methylene or oxyethylene units within the tether chain, and are rather short, not exceeding 19 atoms in length. Third, the production chemistry for these phosphoramidites is not amenable to constructing a vast number of distinctly different monomers, each possessing varying degrees of flexibility, hydrophobicity, and length. This unit elaborates on a general synthetic route toward phosphoramidite linkers. The method is based on methoxyoxalamido (MOX) chemistry, which is very robust in the synthesis of modified oligonucleotides (Polushin, 2000). MOX chemistry is easy to implement and offers great control over tether composition, rigidity, and length. All βcyanoethyl phosphoramidites in this unit are derived from secondary alcohols and can be used on the synthesizer for at least two weeks without any loss in coupling efficiency. Phosphoramidite synthesis The general synthetic routes towards novel phosphoramidites tethered with single or multiple linkers are outlined in Figures 4.29.1 and 4.29.2, respectively. Phosphoramidite preparations proceed by a single linker addition (for a single tether) or by serial linker additions (for a compound tether). In either case, the reactions proceed in a straightforward manner and the synthesis proves to be robust. The stepwise yields are high, and crystallization or precipitation effectively cleans up the majority of the reactions leading up to the phosphoramidite. Column chromatography is used only at the final phosphinylation step and to purify the MOX intermediates (S.5b, S.6b, and S.8b). The choice of starting compound (phosphoramidite base) for the synthesis of tethered phosphoramidites is based on two major considerations. First, it must have a primary amino group as a starting point for MOX chemistry. Second, the hydroxyl group for attachment of the phosphoramidite moiety should be secondary to impart greater stability to the phosphoramidite (Polushin, 2000). It is also essential that the starting material be commercially available and inexpensive, or can be easily prepared from an inexpensive reagent. The trans-4-aminocyclohexanol hydrochloride and 5 -amino-5 -deoxythymidine used as starting compounds in this unit both satisfy all the above demands.

Structural versatility The practical advantage of the described synthetic approach lies in its amenability to structural versatility by utilizing an arsenal of distinctly different, commercially available diamines and amino alcohols. It is possible to tailor the synthesis to suit practically any requirement in chemical or physical properties of the tethers. The selection of tethers for this unit encompasses only a short range of hydrophobicity, rigidity, and length, but the synthetic strategy can easily produce a much wider variety of chain features. Further flexibility comes from the possibility of assembling compound tethers through the use of oxalamido bonding. This modular assembly allows different combinations of available diamines to be used, and thus allows selective modification of the physical nature of the tether at each step. For example, if more rigidity is needed at the end of the tether, a diamine with a constrained configuration can be used at the last step of tether construction. By exploiting the serial addition strategy and a fairly long diamine such as the 15atom TTDD, one can easily prepare specialty phosphoramidites with extraordinarily long tethers. Through three iterations of MOX chemistry, phosphoramidites with a 56- or a 48-atom tether were synthesized (S.9d or S.14d, respectively). Theoretically, there is no limit to the length that can be achieved using this strategy. Recently biotin amino phosphoramidites with tethers up to 106 atoms in length have been synthesized (Polushin, unpub. observ.). It is essential to stress that all tethered compounds prepared through the MOX strategy are monodispersed, unlike previously described polydispersed tethers prepared from low-molecular-weight polyethylene glycol (PEG) polymer (Jaschke et al., 1994). Synthesis of modified oligonucleotides: Coupling and deprotection Most of the described phosphoramidites are solids and thus easier to handle. All are freely soluble in acetonitrile. The average coupling efficiency of the phosphoramidites at 0.1 M, when using ETT as a catalyst over a 3 to 5 min coupling time, was greater than 95%. The tethered phosphoramidites presented here have been used to synthesize a number of 5 -modified oligonucleotides. The modified oligonucleotides were observed as distinct bands on polyacrylamide gels, with mobilities



Supplement 25

Figure 4.29.6 PAGE analysis of crude 5 -modified T10 oligonucleotides. Lane 1 of each gel corresponds to the unmodified T10 oligonucleotide. The other lanes correspond to oligonucleotides modified at the 5 terminus with the following phosphoramidites: S.3d (A2), S.4d (A3), S.5d (A4), S.6d (A5), S.8d (A6), S.7d (A7), S.9d (A8), S.10d (B2), S.11d (B3), S.13d (B4), S.12d (B5), S.14d (B6).

Table 4.29.1 MALDI-TOF-MS Data for Modified T10 Oligonucleotides

Starting oligonucleotide

Modifier

Lanea

MW (calculated)

MW (experimental)

(Tp)9 T

none

A1, B1

2978

2980

(Tp)9 T

3d

A2

3271

3274

(Tp)9 T

4d

A3

3485

3488

(Tp)9 T

5d

A4

3431

3433

(Tp)9 T

6d

A5

3557

3560

(Tp)9 T

8d

A6

3706

3709

(Tp)9 T

7d

A7

3832

3834

(Tp)9 T

9d

A8

3980

3983

(Tp)9 T

10d

B2

3316

3318

(Tp)9 T

11d

B3

3328

3330

(Tp)9 T

13d

B4

3590

3592

(Tp)9 T

12d

B5

3717

3719

(Tp)9 T

14d

B6

3865

3867

a Lane numbers correspond to Figure 4.29.6.




less than those of unmodified counterparts. As an example, Figure 4.29.6 shows a series of crude T10 oligonucleotides modified by the addition of tethered phosphoramidites. MS data for these oligomers are presented in Table 4.29.1. To determine the effect of oligonucleotide deprotection conditions on tethered compounds, several MMTr-protected intermediate amino alcohols and DMTr-protected intermediate diols were subjected to neat ethanolamine, ammonia/methylamine (30% aq. ammonium hydroxide/2 M methanolic methylamine, 1:1, v/v), and aqueous NaOH (0.2 M NaOH in water/methanol, 1:1, v/v). Exposure to NaOH for even 1 hr at ambient temperature was too harsh, and some degradation of all tethers was detected under these conditions. However, the tethers did hold up to ammonia, methylamine, and ethanolamine, and they were stable at the temperature and duration required for complete deprotection of the modified oligonucleotides.

Critical Parameters and Troubleshooting The synthesis of tethered phosphoramidites is straightforward and efficient. However, familiarity with general chemical laboratory techniques such as filtration, extraction, concentration, TLC, and column chromatography is a must. General knowledge of standard analytical techniques (1 H NMR, 31 P NMR, UV, and mass spectroscopies) is necessary for characterization of products. The efficiency of specialty phosphoramidite incorporation may be influenced by the particular model of the DNA/RNA synthesizer; thus, some optimization of critical parameters (e.g., coupling time, delivery time) might be needed to maximize the coupling yields. Practical knowledge of gel electrophoresis is required for oligonucleotide purification. Laboratory safety must always be a primary concern. The amidites are quite stable in the presence of triethylamine and no special precautions are used prior to removal of triethylamine. Although triethylamine is known to reduce phosphoramidite coupling efficiency, the final toluene co-evaporations are sufficient to remove triethylamine completely.

prepared in 40% to 50% yields starting from the MOX precursors. Coupling efficiency of all phosphoramidites should be greater than 95%.

Time Considerations Each of the initial MOX precursors S.1 and S.2 can be prepared in 2 days. Since these precursors are the starting materials for all other compounds, at least 10-g preparations are recommended. The synthesis of a phosphoramidite with a single hydroxylated linker (S.10d-S.11d) generally takes 3 to 5 days, while synthesis of a phosphoramidite with a single aminated linker (S.3d-S.6d) takes 2 days longer. One round of tether elongation (diamine treatment followed by dimethyl oxalate treatment) can be accomplished in 2 to 3 days. Typically, only one or a few incorporations of specialty phosphoramidites are required in the synthesis of a modified oligonucleotide. Thus, the length of the oligonucleotide and the synthesizer throughput dictate the time needed for synthesis. On the ASM-800 DNA/RNA synthesizer, eight 5 -modified 20-mers can be synthesized in ∼2 hr. EA deprotection and desalting (by precipitation) takes less than an hour. PAGE purification and desalting requires 1 to 2 days.

Literature Cited Bannwarth, W. 1988. Solid phase synthesis of oligonucleotides containing phosphoramidate internucleotide linkages and their specific chemical cleavage. Helv. Chim. Acta 71:1517-1527. Jaschke, A., Furste, J.P., Nordhoff, E., Hillenkamp, F., Cech, D., and Erdmann, V.A. 1994. Synthesis and properties of oligodeoxyribonucleotidepolyethylene glycol conjugates. Nucl. Acids Res. 22:4810-4817. Polushin, N.N. 2000. The precursor strategy: Terminus methoxyoxalamido modifiers for single and multiple functionalization of oligodeoxyribonucleotides. Nucl. Acids Res. 28:31253133. Polushin, N.N., Morocho, A.M., Chen, B.C., and Cohen, J.S. 1994. On the rapid deprotection of synthetic oligonucleotides and analogs. Nucl. Acids Res. 22:639-645. Reddy, M.P., Hanna, N.B., and Farooqui, F. 1994. Fast cleavage and deprotection of oligonucleotides. Tetrahedron Lett. 35:4311-4314.

Anticipated Results Good to moderate yields of the final specialty phosphoramidites from the corresponding MOX precursors are expected. Phosphoramidites tethered with single linkers (S.3d-S.6d, S.10d, and S.11d) have been

Contributed by Alan M. Morocho and Nikolai N. Polushin Fidelity Systems, Inc. Gaithersburg, Maryland



Supplement 25

Using Morpholinos to Control Gene Expression

UNIT 4.30

Morpholino oligos (Morpholinos) are synthetic uncharged P-chiral analogs of nucleic acids. They are typically constructed by linking together 25 subunits, each bearing one of the four nucleic acid bases. Figure 4.30.1 illustrates the structure of three Morpholino subunits joined by inter-subunit linkages. The morpholino phosphorodiamidate backbone of Morpholinos consists of morpholine rings that bear methylene groups and are bound through modified phosphates in which the anionic oxygen is replaced by a nonionic dimethylamino group. The substituted phosphate is bound through an oxygen atom to the morpholine’s exocyclic methylene group, and through a phosphorous-nitrogen bond to the nitrogen atom of another morpholine ring. One standard DNA nucleobase (adenine, guanine, cytosine, or thymine) is bound to each morpholine ring. The ends of Morpholinos are conventionally named 3 and 5 by analogy with the nomenclature for nucleic acids (though if one were to number the atoms of a morpholino oligonucleotide backbone by IUPAC rules, the numbers assigned to the ends would be different). The secondary amine of the morpholine ring at the end of an unmodified morpholino oligonucleotide is called the 3 end of the oligo, whereas the 5 end terminates with a chiral carboxamidated phosphorodiamidate group (Fig. 4.30.1). Antisense Morpholinos block the interactions of macromolecules with mRNA by base pairing with the targeted mRNA in a complementary fashion, thus preventing initiation

Figure 4.30.1

Structure of a Morpholino 3-mer.

Contributed by Jon D. Moulton Current Protocols in Nucleic Acid Chemistry (2006) 4.30.1-4.30.24 C 2006 by John Wiley & Sons, Inc. Copyright



complex read-through or modifying splicing in cells ranging from bacterial (Geller et al., 2005) to human (Suwanmanee et al., 2002). In particular, antisense Morpholinos have become a standard tool for developmental biologists to manipulate gene expression in embryos such as zebrafish and Xenopus sp. (Ekker and Larson, 2001). These modified oligonucleotides combine efficacy, specificity, stability, lack of non-antisense effects, and good water-solubility properties. This unit presents three protocols for the design of a knockdown experiment using Morpholinos (Basic Protocol 1), preparation of Morpholino solutions (Basic Protocol 2), and introduction of Morpholinos into cells by endocytosis in the presence of an amphiphilic peptide (Basic Protocol 3). The Commentary provides a thorough discussion of conditions and considerations for the application of Morpholinos. BASIC PROTOCOL 1

DESIGN OF A MORPHOLINO KNOCKDOWN EXPERIMENT This protocol outlines the choices commonly encountered while designing a Morpholino knockdown experiment. Considerations for the steps are addressed in the Commentary. 1. Choose the target gene. 2. Choose the cells or organism into which the oligo will be delivered. 3. Choose between splice blocking or translation blocking, which determines the molecular assays available for measuring antisense activity. 4. Obtain the sequence of the target RNA. Use the mRNA 5 -UTR and the first 25 coding bases for translation blockers, or pre-mRNA with introns and exons defined for splice blockers. 5. Choose a delivery method. 6. Select control oligos. 7. Decide whether end-modification of any oligos is necessary. 8. For blocking splicing, select which pre-mRNA splice boundary (intron-exon) to block. 9. Select the oligo target (following the targeting rules described in the Commentary) and determine the Morpholino sequence (the inverse complement of the target). 10. Use a transcript database and a homology search tool such as BLAST to test the selected target for homologies with other RNAs. If the selected target is too homologous with a region of an off-target mRNA, a partially complementary Morpholino might affect the expression of that mRNA. Another target on the desired mRNA must be selected to prevent off-site Morpholino interaction.

11. Order the synthesis of the selected Morpholinos. BASIC PROTOCOL 2

PREPARATION AND VERIFICATION OF MORPHOLINO STOCK SOLUTIONS This protocol describes the preparation of stock aqueous solutions of Morpholinos at concentrations of 1 mM or 500 µM, if necessary.

Materials Using Morpholinos to Control Gene Expression

Lyophilized Morpholino oligonucleotide (Gene Tools) Distilled autoclaved water without DEPC, sterile 0.1 M HCl



Glass or polypropylene/polyethylene tubes with labels 65◦ C water bath Quartz spectrophotometer cell (1 cm path length) Parafilm Lint-free lab tissues UV spectrophotometer (or UV colorimeter) capable of measurements at 265 nm Morpholino product information sheet Prepare Morpholino solution 1. Read the amount of Morpholino given on the vial label and, using sterile technique, add the appropriate volume of distilled sterile water to make a 1 mM stock solution (i.e., 0.1 mL water for a vial containing 100 nmol Morpholino). The aqueous solubility of Morpholinos is sequence-dependent, but most Morpholino sequences with G content below 36% will dissolve in water at the recommended stock concentration of 1 mM. Do not keep Morpholino solutions of 2, the sample should be diluted and remeasured. BASIC PROTOCOL 3

DELIVERY OF MORPHOLINOS INTO CELLS USING ENDO-PORTER Endo-Porter is an amphiphilic peptide. After co-endocytosis with Morpholinos, EndoPorter permeabilizes endosomal membranes, releasing the Morpholino from the endosomes to the cytosol (Summerton, 2005). Endo-Porter was optimized using a HeLa cell line. Because tolerance of other cell types toward Endo-Porter often varies, a range of Endo-Porter concentrations should be tested before beginning knockdown experiments.

Materials 1 mM Endo-Porter solution (aqueous or DMSO formulation; Gene Tools) Cell cultures in plates or flasks at 80% to 100% confluence 1 mM Morpholino stock solution (Gene Tools) 1 mM fluoresceinated dextran, 10 kDa Cell culture medium with 10% serum Fluorescence microscope Select amount of Endo-Porter for cell type 1. Prepare concentrations of 2, 4, 6, and 8 µM Endo-Porter by pipetting 2, 4, 6, and 8 µL of a 1 mM Endo-Porter solution into 1-mL aliquots of cell culture.


2. Add 10 µM fluorescently labeled Morpholino (10 µL of 1 mM stock per 990 µL cell culture) or 10 kDa fluoresceinated dextran (10 µL of 1 mM stock per 990 µL cell culture). 3. Allow endocytotic uptake to proceed over a period of 24 hr.



4. Observe intracellular fluorescence using a fluorescence microscope. See discussion on assessing delivery in the Commentary section.

5. Observe cells 72 hr after delivery to determine any cellular toxicity. For subsequent Morpholino delivery to the selected cell type, use the concentration of Endo-Porter that gave the best delivery without toxicity.

Deliver Morpholinos to cells 6. Using a cell culture not previously exposed to Endo-Porter, replace spent culture medium with fresh medium (with up to 10% serum). 7. Add the Morpholino stock solution to produce the desired concentration and swirl well to mix. For functional experiments (e.g., gene knockdown, splice blocking), Morpholinos are typically effective at concentrations as low as 1 µM. However, it is recommended that a range of concentrations be tested (such as 1, 4, and 10 µM Morpholino) to define optimal conditions.

8. Add Endo-Porter to produce the optimized concentration for the cell type and immediately swirl to mix. 9. Place the plates or flasks in the incubator. Wait at least 16 hr before assessing uptake by fluorescence, and at least 24 hr before measuring knockdowns by molecular assays. The delay needed to assay a knockdown depends on the stability of any preexisting protein encoded by the targeted mRNA; a protein with a long half-life will take longer to disappear from the cells.

COMMENTARY Background Information The morpholino phosphorodiamidate backbone of a Morpholino oligonucleotide has no significant ionic charge at neutral pH, in contrast with the polyanionic phosphodiester backbone of a natural nucleic acid. This favors the interaction of Morpholinos with nucleic acids, since there is no repulsion between anionic backbones as there is in duplexes of natural nucleic acids. Dissolved in pure water, nucleic acids lose their ability to form stable Watson-Crick bonds due to anionic repulsion between strands, whereas Morpholinos will still bind to complementary sequences (Summerton, 2004). Because Morpholinos are uncharged, they have no strong electrostatic interactions with proteins. Unmodified Morpholinos have little or no affinity for bovine or human serum albumin (H.M. Moulton, unpub. observ.). In contrast, interactions of anionic phosphorothioate oligos with proteins cause multiple physiological, non-antisense effects (Lebedeva and Stein, 2001). Proteins that bind nucleic acids generally interact electrostatically with the anionic phosphates of nucleic acids, stabilizing binding. Morpholinos appear to have little or no interaction with nucleic acid–binding proteins (Hudziak et al., 1996).

Morpholinos are very stable to nucleolytic enzymes. There are no known enzymes that degrade Morpholinos. Specifically, Morpholinos have been exposed to a range of nucleases (e.g., DNase I, DNase II, Benzonase, S1 nuclease, mung bean nuclease, Bal 31 nuclease, RNase A, RNase T1, phosphodiesterase I, and phosphodiesterase II) and proteases (e.g., pronase E, proteinase K, and pig liver esterase) under conditions where lytic enzymes would degrade their substrates. In no case was degradation of the Morpholinos detected (Hudziak et al., 1996). Morpholinos were incubated in serum and in liver homogenate without degradation (Summerton and Weller, 1997). When peptide-Morpholino conjugates were extracted from cells and analyzed by MALDITOF mass spectrometry, the Morpholino oligo entity was not degraded in the cells (Nelson et al., 2005). No crystal structure or high-resolution NMR structural analysis of phosphorodiamidate Morpholinos has been published. However, the study of a morpholino phosphorodiamidate ApA dimer using circular dichroic spectroscopy showed stacking of bases in aqueous phosphate buffer (Kang et al., 1992). On the basis of molecular modeling, the bases



Supplement 27

Table 4.30.1 RNA Binding Affinity of Various Oligo Types Ranked by Dissociation Temperature in Physiological Isotonic Buffers

Figure 4.30.2


Affinity

Type of oligo

Strongest

RNA:RNA, PNA:RNA, 2 -O-methyl-RNA:RNA (all very similar)

Strong

Morpholino:RNA

Medium

DNA:RNA

Weakest

Phosphorothioate:RNA

Comparison of RNase H–dependant, RISC-dependant, and steric blocking oligos.

of Morpholinos should stack in a fashion analogous to those of natural nucleic acids, allowing strong interactions with complementary nucleic acid sequences by Watson-Crick base pairing. A 400 MHz 1 H NMR analysis of a carbamate-linked Morpholino found the morpholine ring in the chair conformation (Stirchak et al., 1989). Molecular modeling of a Morpholino with the morpholine rings in the chair conformation suggests that a Morpholino and an RNA form an A-form heteroduplex with a helical pitch similar to that of an A-form RNA-RNA duplex (J.E. Summerton, unpub. observ.). Various types of antisense oligos are ranked by their affinity for binding to single strands of sense RNA based on their dissociation temperatures in physiological salt buffers (Table 4.30.1; Stein et al., 1997). The affinity of RNA for RNA is greater than the affinity of Morpholinos for RNA. However, single strands of mRNA folded into secondary structures contain single-stranded regions, such as the loops of stem-loops, with which Morpholinos can readily hybridize. Given that double-stranded regions of most RNA secondary structures are shorter than 25 base pairs, the overall binding affinity of Morpholinos for RNA is suffi-

cient to invade and displace those short doublestranded regions (Summerton, 1999). Antisense oligos such as DNA, RNA, and phosphorothioate (S-DNA) oligos recruit RNase H to degrade their mRNA targets (Summerton, 1999). RNAi and siRNA also employ an antisense mechanism to recognize a sense mRNA through interaction with a RISC complex, which leads to enzymatic degradation of complementary mRNA and translational blocking of partially complementary mRNA (Scacheri et al., 2004). In contrast, instead of degrading mRNA, antisense Morpholinos were designed to block the translation of mRNA into protein (Summerton and Weller, 1997). Figure 4.30.2 compares steric blocking, RNase H–dependant, and RISCdependant oligos. When comparing an RNase H– dependant oligo (a methylphosphonate diester/phosphodiester chimera) with a Morpholino, a CpGNNN motif was shown to induce apoptosis and cell cycle arrest when present in the RNase H–dependent oligo but not when present in the Morpholino (Tidd et al., 2001). There have been no reports of Morpholinos inducing either interferon production or induction of NF-κB



mediated inflammation, and Morpholinos containing CpG motifs do not stimulate immune responses (J.E. Summerton and A. Krieg, unpub. observ.), suggesting that Morpholino-RNA heteroduplexes do not stimulate Toll-like receptors. Morpholinos complementary to sequences in the 5 -UTR and the first 25 coding bases of an mRNA can halt the progression of the initiation complex toward the start codon, preventing assembly of the entire ribosome. This inhibits the translation of the mRNA sequence into a polypeptide. Morpholinos targeted downstream of the start codon are usually ineffective for blocking translation (Summerton, 1999). In addition to their application to knock down gene expression, because stericblocking oligos do not trigger degradation of RNA, Morpholinos are also widely used to block splicing of pre-mRNA. Splicing in eukaryotes is directed by snRNPs that bind to introns and mark the intron-exon boundaries. Morpholinos targeted to these snRNP-binding sites can modify splicing (Sazani et al., 2001), either preventing splicing and causing an intron inclusion (Giles et al., 1999) or redirecting splicing and causing an exon excision (Draper et al., 2001). Blocking a splice site can cause activation of a cryptic splice site, complicating interpretation of the splice modification by producing partial deletions of exons (Draper et al., 2001) or partial inclusions of introns. Morpholinos stimulate site-specific ribosome frameshifting when bound just downstream of a shift site on an mRNA, and they do so with far higher efficiency than RNA, phosphorothioate oligos, or 2 -O-methyl RNA oligos (Howard et al., 2004). Although Morpholinos are most often used to block the translation initiation complex or the snRNPs that direct splicing, there are other mRNA sequences that are attractive targets for steric blocking. Specifically, Morpholinos can block miRNA activity by binding to the miRNA and preventing it from binding its mRNA target (Kloosterman et al., 2004), or by binding to the site on the mRNA where the miRNA would otherwise bind. Along similar lines, Morpholinos targeted across the cleavage site of a hammerhead ribozyme inhibited auto-cleavage, leading to over two orders of magnitude increase in the expression of a downstream reporter gene (Yen et al., 2004). While Morpholinos have also been shown to block intronic splice silencers (Bruno et al., 2004) and exonic splice enhancers (McClorey

et al., 2006), no publications have yet explored other potential regulatory targets such as zipcode binding sites, riboswitches, or binding sites for elements of the nonsense-mediated mRNA decay pathway. Morpholinos are commonly microinjected into embryos at the single-cell or few-cell stages to block genes involved in development (Heasman et al., 2000; Nasevicius and Ekker, 2000; Nutt et al., 2001). Morpholinos are also commonly used in cell cultures (Tyson-Capper and Europe-Finner, 2006). Applications in intact adult organisms have until recently been limited by poor in vivo delivery into the cytosol of cells (Summerton, 1999; Sazani et al., 2002). However, very recent advances in conjugating Morpholinos to cell-penetrating peptides (Nelson et al., 2005) now allow effective systemic delivery into adult organisms (Alonso et al., 2005; Kinney et al., 2005; Neuman et al., 2005; Enterlein et al., 2006). Combinations of several oligonucleotide sequences can bind to several different RNA targets simultaneously if introduced together into embryos (Ekker, 2000) or cell cultures (Summerton, 2005), allowing multiple knockdowns or synergistic targeting of a single messenger. Targeting of viral RNA with Morpholinos has been reported for hepatitis C (Jubin et al., 2000; McCaffrey et al., 2003), dengue virus (Kinney et al., 2005), ebola virus (Enterlein et al., 2006; Warfield et al., 2006), SARS virus (Neuman et al., 2005), West Nile virus (Deas et al., 2005), equine arterivirus (van den Born et al., 2005), mouse hepatitis virus (Neuman et al., 2004), novirhabdovirus (Alonso et al., 2005), and vesivirus (Stein et al., 2001). In addition to translation start sites, successful targets for inhibition of viral replication include cyclization sequences (Deas et al., 2005), terminal stem loops (Deas et al., 2005), and internal ribosomal entry sites (IRES; Jubin et al., 2000). Radioisotope delivery into organisms can be pretargeted using Morpholinos (Mang’era et al., 2001). Practitioners of nuclear medicine strive to minimize radiation exposure of a patient while delivering radionuclides to target tissues for imaging or for therapeutic applications. By attaching radioisotopes to antibodies that are specific for target tissues, the antibodies can anchor isotopes on these tissues. Because the large antibody molecules diffuse slowly, the isotopes must be maintained in the plasma at high concentrations or for long durations to achieve good delivery of radioisotope-linked antibodies to their targets.



Supplement 27

Pretargeting with Morpholinos involves introducing an antibody-Morpholino conjugate into the bloodstream; this can be done using high concentrations or re-dosing to saturate the target without exposing the patient to radiation during this pretargeting stage. Next, a conjugate of a radioisotope (possibly chelated) with a complementary Morpholino is added to the blood. Because the Morpholino has a much smaller molecular mass than an antibody, the radionuclide-Morpholino conjugate diffuses relatively quickly and is captured at the target tissue more rapidly through MorpholinoMorpholino pairing. Unbound radionuclideMorpholino conjugate is rapidly eliminated through the kidneys. This technique allows delivery of radioisotopes to the targeted tissue while exposing the organism to lower doses of radiation away from the targeted region. In the process of developing these techniques, pharmacokinetics of Morpholino-radionuclide conjugates have been studied in vivo (Liu et al., 2002a,b; He et al., 2003). In a recent modification, signals are amplified by binding a polymer bearing many complementary Morpholinos to each Morpholino-conjugated antibody fragment, followed by delivering radioisotopelabeled Morpholino complementary to the polymer-linked Morpholinos (He et al., 2003, 2004).

Critical Parameters Choosing Morpholino sequences The parameters considered when selecting oligonucleotide target sequences include CG%, G%, self-complementarity, tetra-G

moieties, length of the oligo, and the intended temperature at which the oligo will be used. The targeting recommendations are summarized below and in Table 4.30.2. CG range. A range of 40% to 60% CG is considered ideal for 25-base Morpholinos in 37◦ C systems. Oligos with 60% CG are more likely to interact with off-target messengers through high-affinity subsequences. G content. G content affects aqueous solubility of an oligo, with higher G contents being less soluble, particularly when the oligo is dissolved in isotonic salt solutions. Oligos with G contents up to 36% should be soluble in the millimolar range in pure water or aqueous buffer. However, freeze-thaw cycles are likely to cause high-G oligos to precipitate and the oligos must be heated to redissolve (see Basic Protocol 2). Self-complementarity. Self-complementary sequences can cause either intramolecular interactions, forming stem-loops, or intermolecular interactions, forming dimeric Morpholinos. When a short sequence of one part of an oligo is complementary to another short sequence separated by an intervening sequence, stem-loops can form. If small selfcomplementary sequences are separated by zero to a few bases, formation of a stable stemloop is unlikely because a hairpin with a small loop is not energetically favored. To prevent loss of oligo activity through competition between self-pairing and target binding, it is prudent to limit self-complementary sequences in oligo designs to 16 contiguous hydrogen bonds

Table 4.30.2 Summary of Targeting Recommendations for 37◦ C Systems


Parameter

Recommendation

Comments

CG range

40%-60%

At lower GC, affinity may be too low to block processes; higher GC favors nonspecific binding of subsequences.

G content

Up to 36% G

Higher G causes loss of water solubility; avoid upper end of acceptable range, if possible.

Selfcomplementarity

16 contiguous H-bonds maximum

For intermolecular (complementary palindrome) and intramolecular (stem loop) binding. Example: AGCGCT has 16 H-bonds (2+3+3+3+3+2 = 16). Check for non-Watson-Crick G-T pairing, which can participate in self-complementarities.

Consecutive G

3 consecutive Gs maximum

Runs of ≥4 G can associate through Hoogsteen bonding to form oligo tetramers.

Oligo length

25 bases or shorter by only a few bases

Using shorter oligos can decrease the chance of off-target interaction for high CG oligos.



or less, where CG pairs contribute 3 hydrogen bonds and AT pairs contribute 2 hydrogen bonds. For instance, the short sequences ATGGC and GCCAT can form 13 contiguous hydrogen bonds (2+2+3+3+3 = 13). When analyzing sequences for self-complementarity, check for both Watson-Crick base-pairing and for GT base-pairing. Like an AT pair, a GT pair also forms two hydrogen bonds. However, because the overall stability of the GT pair is far lower than an AT pair, a GT pair can be scored as a single hydrogen bond when calculating its contribution to the stability of a self-complementary moiety (Aboul-ela et al., 1985). An oligo containing a self-complimentary sequence can form dimers. To prevent loss of oligo activity through competition between dimer formation and target binding, it is prudent to limit complimentary palindromes to 16 contiguous hydrogen bonds or less. For instance, if two oligos bearing the self-complimentary sequence ATGCATGCGT encounter each other, they can form 22 contiguous hydrogen bonds (2+1+3+3+2+2+3+3+1+2 = 22, taking into account the GT pairs) and would likely have poor antisense activity. G tetrads. Nucleic acids containing GGGG moieties can interact through Hoogsteen bonding to form oligo tetramers (Cheong and Moore, 1992). Morpholinos containing G tetrads have reduced activity, likely through the same mechanism. Because of this, contiguous stretches of four or more G bases should be avoided when designing Morpholinos. MIL and oligo length. The minimum inhibitory length (MIL) of an antisense oligo is the length needed to achieve 50% reduction in translation of a targeted gene at a concentration typically achieved in cells. The MIL of Morpholinos varies somewhat between targets, but averages about 14 bases for 37◦ C cell cultures (Summerton, 1999). To ensure good affinity between Morpholinos and their RNA targets, the oligos are usually synthesized as 25-mers. CG content can influence the MIL of an oligo, with a higher CG oligo having a shorter MIL. Oligos with high CG content might interact with off-target RNA; these oligos can be shortened by a few bases to lessen the likelihood of off-target interactions. The marginal loss of affinity resulting from shortening a high-CG oligo will not ruin activity but will slightly improve specificity. A more effective way to improve specificity is to choose a target with a lower CG content.

Temperature and oligo selection The targeting guidelines were developed for oligos to be used at 37◦ C. Many embryos are grown at lower temperatures. When temperatures are decreased appreciably, stability of base-pairing increases. The ideal CG content for oligos designed for use at lower temperatures is lower than the 40% to 60% CG recommended for 37◦ C systems. The ideal CG content for colder systems (e.g., fish and frogs) must be determined experimentally. Similarly, the allowable number of base pairs in self-complementary sequences should be reduced for colder systems. Solubility is also decreased at lower temperatures, so it is prudent to select oligos with lower G contents for use in colder systems. Targetable region for translation blockers To block translation, a 25-mer Morpholino can target anywhere between the 5 cap to 25 nucleotides into the coding sequence. The target can extend downstream into the coding sequence as long as the translational start codon is covered. In the first steps of translation, the initiation complex forms at the 5 cap and then scans through the UTR to the start codon (Fig. 4.30.3A). At the start codon, the large ribosomal subunit binds, the initiation factors dissociate, and translation proceeds through the coding region. If a Morpholino gets in the way of the initiation complex before the initiation complex reaches the start codon, it prevents assembly of the ribosome and translation of the mRNA. Nonetheless, it is preferable to target the start codon instead of upstream for two reasons. First, the quality of sequence deposited in public databases is often poor in the UTR, especially for older sequence records. Second, though rare in vertebrate genomes, internal ribosome entry sites (IRES) do exist and can allow a ribosome to enter and assemble downstream of a Morpholino bound in the 5 -UTR. Targetable region for splice blockers To block splicing, Morpholinos are targeted to pre-mRNA across or near the boundaries between exons and introns. A pre-mRNA that undergoes splicing has two flanking exons (the first and last exon) and an arbitrary number of internal exons. The first exon has a single splice site, a splice donor, where it contacts intron 1. The internal exons have two junctions each, a splice acceptor at the upstream end and a splice donor at the downstream end. The last exon has only a splice acceptor at its upstream end. Targeting the splice sites of the internal exons usually causes exon excision, resulting in an mRNA missing the exon with



Supplement 27

Figure 4.30.3


Targetable regions for translation blocking (A) and splice blocking (B).

the blocked splice site (Fig. 4.30.3B). Targeting splice sites of the flanking (first or last) exons usually causes intron inclusion, resulting in an mRNA containing the first or last intron. Sometimes blocking a splice site activates a cryptic splice site, resulting in an mRNA with an unexpected mass. The snRNPs that direct splicing bind at the intronic sides of the splice junctions, so Morpholinos are chosen that are complementary to more intronic sequence than exonic sequence. Morpholinos can have good activity if targeted entirely to intronic sequence near the splice junction, but activity decreases as the target is moved farther into the intron (P.A. Morcos, unpub. observ.). Splicing can also be modified by preventing excision of an arbitrary intron by blocking the nucleophilic adenosine that closes the splicing lariat (P.A. Morcos, unpub. observ.) or by targeting splice-regulatory sequences. It is often the goal of a splice-blocking experiment to eliminate activity of a protein. If the active site of the protein is known, a straightforward strategy is to target a Morpholino to the exon encoding the active site, causing the loss of that exon and of the active site. When the active site is not known, other useful strategies are available. One is to eliminate an upstream exon that has a number of nucleotides not evenly divisible by three, causing downstream translation to be frameshifted. Another is to trigger inclusion of the first intron, especially useful if it contains an in-frame stop codon or if its number of nucleotides is not evenly divisible by three. Sometimes causing a random exon exclusion or intron inclusion is

sufficient to eliminate activity of a protein, perhaps due to a resulting change in the protein’s tertiary structure. Quality of sequence Since a few mismatches can seriously decrease the activity of a Morpholino, the quality of the target sequence is an important consideration when designing Morpholinos. There are sometimes errors in sequence database files. Variations in sequence between strains of an organism can also present a problem. The most definitive way to ensure the correct target sequence is to sequence the targeted gene in the strain that will be used in the experiments. Mismatched unintentional targets and Morpholinos When a 25-base Morpholino is used near its lowest effective concentration, its effects are very specific. Under such conditions the oligo can also interact with sequences containing one or two mismatches when compared to the oligo’s perfectly complementary target, though even a single mismatch can decrease activity (Khokha et al., 2002). However, few to no such sequences are expected to occur randomly in a base pool the size of the Morpholino-targetable sites in the human transcriptome (Summerton, 1999). Effect of concentration on specificity When the concentration of any antisense oligo is increased well above its minimum effective concentration, it can interact with targets containing more mismatches; at some concentration a Morpholino will begin knocking down expression of off-target mRNAs.



Therefore it is important that the oligo concentration be kept as low as practicable while still eliciting the desired targeted knockdown. The concentration at which off-target effects occur, the concentration at which targeted knockdown occurs, and the ratio of these concentrations are all sequence-specific and therefore unknown for each new oligo sequence. In most cases, an effective and specific concentration window exists such that, for complementary mRNA and off-target mispaired mRNA at similar concentrations, the onset of the targeted knockdown will occur at a lower concentration than the onset of the offtarget knockdown. However, knocking down high-copy-number mRNAs requires higher oligo concentrations, increasing the probability of knocking down low-copy-number offtarget mRNAs; such a situation can narrow or even close the effective and specific concentration window. Acceptable off-target homology A single mismatch in a Morpholino 25-mer may cause a significant decrease in antisense activity (Khokha et al., 2002), though many single-mismatched oligos have retained good activity. When used near the concentration at which a perfectly complementary oligo elicits a knockdown, five mismatches distributed throughout a 25-mer usually decreases activity of the mismatched oligo to near undetectable levels (S.T. Knuth, unpub. observ.). It is prudent to check the target sequence of a proposed oligo against a nucleotide sequence database in order to identify regions where the Morpholino might bind to off-target mRNA. When searching for homologous targets, keep in mind that 25-base Morpholinos will only block translation when targeted to the 5 -UTR and first 25 bases of coding sequence. Morpholinos can modify splicing if targeted in introns near intron-exon boundaries. If the Morpholino has homology to an off-target mRNA outside of these limited regions, binding of the oligo to the mRNA is not likely to affect expression of the off-target mRNA (though blocking miRNA targets or regulatory sequences such as exonic splice enhancers may affect expression). When comparing a 25-base Morpholino against an off-target sequence in a region where a Morpholino might have a biological effect, the fraction of homologous bases should always be below 80%. However, that percentage ignores important considerations about the distribution of the mismatches throughout the oligo. About 14 contiguous

bases of homology is the minimum inactivating length for a Morpholino (Summerton, 1999). However, if 10 bases of perfect homology are flanked with a mismatch at either side and some runs of homologous bases are just beyond the flanking single mismatches, the oligo may still bind sufficiently to block translation or splicing. High CG content can make shorter homologous sequences active, since CG pairs are more stable than AT pairs. Distributing five mismatches throughout a 25mer almost always results in loss of knockdown at low concentrations, so 5-mispair oligos are commonly used as specificity controls. If all five mismatches are at one end of the oligo, there are still 20 contiguous complementary bases in a 25-mer, and those 20 bases would retain considerable antisense activity. When checking a Morpholino target against a sequence database and finding a partially homologous region, following a rule of thumb like “30% G). Primary amine. Morpholinos may be modified with a primary amine to provide a reactive site for attachment of other moieties to the oligo. An unmodified Morpholino has a secondary amine on the 3 end, the pKa of which is 6.5. The primary amine, with its pKa of 10.2, provides a more reactive site. When a primary amine is attached to the 3 end of the oligo, this converts the 3 -secondary amine of the morpholine ring to a tertiary amine as a consequence of the attachment of a short spacer tethering the new primary amine. When a primary amine is attached to the 5 end of the oligo, the 3 -secondary amine is acetylated so that a reagent added to react with the primary amine will not react with the 3 end of the oligo. When reacting a primary amine with a derivatizing reagent, it is prudent to include an additional short spacer to prevent steric hindrance between the moiety being added and the Morpholino. Morpholino stock solutions and reconcentrating Morpholinos Morpholino stock solutions in distilled water should be kept sterile and can be autoclaved. Do not use water containing diethylpyrocarbonate. Morpholino stock solutions can be dissolved in buffers such as Ringer’s solution or Danieau buffer, but this can cause problems later if the stock solution must be reconcentrated, since lyophilization can be more difficult from a buffer. Also, Morpholinos are substantially more soluble in distilled water than in isotonic salt solutions. A solution of Morpholino in water can be concentrated by using a Speedvac or by lyophilization (freeze-drying). Lyophilizing Morpholinos from water produces a fluffy solid that dissolves fairly readily if the sequence has good solubility properties. However, dissolution of Morpholinos concentrated with a Speedvac may be more difficult and will likely require patience and heating to 65◦ C. Temperature during handling Morpholinos are not degraded by nucleolytic enzymes. Solutions of DNA and RNA are normally kept on ice during experiments to prevent nucleolytic degradation, but this is not

a concern with Morpholinos. However, some Morpholino solutions have low enough solubility that icing a solution may cause a loss of activity, due to the oligo coming out of solution. Therefore, icing Morpholino solutions is not only unnecessary but can cause problems; Morpholinos should be kept at room temperature during experiments. Material affinity Morpholinos have some affinity for plastics, so passaging very dilute (submicromolar) solutions through plastic containers may cause appreciable decreases in activity. Similarly, filter sterilization may cause Morpholino solutions to lose some activity as some oligo binds to the filter. When put through the same procedures with the same exposure to plastic surfaces, high-concentration Morpholino solutions have a smaller fractional decrease in concentration than low-concentration Morpholino solutions. Therefore, if exposure to plastic surfaces is required, it is best to do the procedures with Morpholinos in a relatively concentrated state (>1 µM). Similarly, if Morpholinos are to be stored in solution for more than a few days, it is best to store them at high concentration. Since solutions of Morpholinos at very low concentrations (1500 publications have reported experiments with Morpholino oligos in a broad range of systems. Citations and many abstracts are searchable here.




Solid-Phase Oligonucleotide Labeling with DOTA

UNIT 4.31

This unit contains protocols for the synthesis of a nucleosidic phosphoramidite tethered to a protected DOTA ligand and a description of its use in the preparation of oligonucleotide conjugates. Commercial 5 -O-(4,4 -dimethoxytrityl)-2 -deoxyuridine is covalently bound to N-protected 1,4,7,10-tetraazacyclododecane by an N3 -aminoalkyl linker and then converted to its phosphoramidite derivative, allowing normal oligonucleotide chain elongation. Upon completion of the oligonucleotide syntheses, the conjugates are deprotected and converted to the corresponding gadolinium(III) chelates by treatment with gadolinium(III) citrate, and purified by polyacrylamide gel electrophoresis (PAGE). The Gd-DOTA-labeled oligonucleotides can be used in applications based on magnetic resonance imaging. Synthesis of the nucleosidic phosphoramidite building block is described in Basic Protocol 1, and its use in the preparation of oligonucleotide conjugates is detailed in Basic Protocol 2. CAUTION: Perform all operations involving organic solvents and reagents in a wellventilated chemical fume hood, and wear gloves and protective glasses.

PREPARATION OF THE NUCLEOSIDIC PHOSPHORAMIDITE TETHERED TO DOTA

BASIC PROTOCOL 1

The synthesis of the DOTA-labeled phosphoramidite building block S.8 is based on the procedure described in Jaakkola et al. (2006) and is outlined in Figure 4.31.1. First, S.1 (prepared as described in Heppeler et al., 1999) is converted to the tetraester S.2 by reaction with bromomethyl acetate. Subsequent hydrogenolysis gives S.3. Synthesis of the nucleoside S.6 starts from commercially available 5 -O-(4,4 -dimethoxytrityl)2 -deoxyuridine (S.4), which is initially converted to S.5 by Mitsunobu reaction with N-trifluoroacetyl-6-aminohexan-1-ol (Hovinen and Hakala, 2001). Ammonolysis of the trifluoroacetate S.5 gives rise to S.6. Condensation of S.3 and S.6 in the presence of O(7-azabenzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate (HATU) and N,N-diisopropylethylamine gives the nucleoside derivative S.7. Finally, conventional phosphitylation yields the phosphoramidite S.8.

Materials 1,4,7,10-Tetraazacyclododecane-1-carboxymethylbenzyl ester (S.1; Heppeler et al., 1999) Dry acetonitrile, ≥99.5% pure, ≤0.005% H2 O (Merck) Anhydrous potassium carbonate, p.a. (Merck) Methyl bromoacetate, 99% pure (Acros Organics) Silica gel 60, 0.063 to 0.200 nm (Merck) Methanol (MeOH), ≥99.9% pure (Merck) Dichloromethane (CH2 Cl2 ), 99% pure (Lab-Scan) 10% palladium on activated carbon (Pd/C; Aldrich) Hydrogen gas Celite 521 (Aldrich) 5 -O-(4,4 -Dimethoxytrityl)-2 -deoxyuridine (S.4; Sigma) N-Trifluoroacetyl-6-aminohexan-1-ol (Sinha and Striepeke, 1991) Triphenylphosphine, 99% pure (Aldrich) Dry tetrahydrofuran (THF), ≥99.5% pure, ≤0.0075% H2 O (Merck)

Contributed by Lassi Jaakkola, Alice Ylikoski, and Jari Hovinen Current Protocols in Nucleic Acid Chemistry (2007) 4.31.1-4.31.11 C 2007 by John Wiley & Sons, Inc. Copyright



Figure 4.31.1 Synthesis of the DOTA-labeled nucleosidic phosphoramidite (see Basic Protocol 1). Bn, benzyl; DIAD, diisopropyl azodicarboxylate; DIPEA, N,N-diisopropylethylamine; DMF, N,N-dimethylformamide; DMTr, 4,4 -dimethoxytrityl; HATU, O-(7-azabenzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate. Modified from Jaakkola et al. (2006) with permission from the American Chemical Society.

Diisopropyl azodicarboxylate (DIAD), 95% pure (Aldrich) Diethyl ether, ≥99.7% pure (Merck) Aqueous ammonia, p.a. 28% to 30% (Merck) Sodium sulfate (Na2 SO4 ), anhydrous, ≥99.0% pure (Merck) N,N-Diisopropylethylamine (DIPEA), ≥99.5% pure (Aldrich) N,N-Dimethylformamide (DMF), dry, 99.8% pure, ≤0.01% H2 O (Lab-Scan) O-(7-Azabenzotriazol-1-yl)-N,N,N ,N -tetramethyluronium hexafluorophosphate (HATU; Applied Biosystems) Sodium hydrogen carbonate (NaHCO3 ) solutions, saturated and 5% (w/v) 2-Cyanoethyl-N,N,N ,N -tetraisopropylphosphordiamidite, 97% pure (Aldrich) 0.45 M 1H-tetrazole in acetonitrile (Applied Biosystems) Triethylamine (TEA), ≥99% pure (Merck) Glass filters (3-µm pore size) Rotary evaporator equipped with an oil pump Chromatography columns: 18 × 4 cm, 5 × 30 cm, 4 × 15 cm, and 4 × 2.5 cm TLC plate: silica-coated glass plate with fluorescent indicator (Merck silica gel 60 F254 ) Hydrogenation apparatus (Parr Instruments Company) Ultrasonic bath 254-nm UV lamp Reflux condenser Solid-Phase Oligonucleotide Labeling with DOTA

Additional reagents and equipment for TLC (APPENDIX 3D) and column chromatography (APPENDIX 3E)



Alkylate S.1 1. Dissolve 0.45 g (1.4 mmol) S.1 in 8 mL dry acetonitrile with stirring. 2. Add 0.79 g (5.7 mmol) anhydrous potassium carbonate. 3. Prepare a solution of 0.54 mL (5.7 mmol) methyl bromoacetate in 2 mL dry acetonitrile. Add this to the stirring S.1 solution over a 30-min period. 4. Stir for 2 hr at room temperature (∼25◦ C). 5. Filter off all solid material using a glass filter with a 3-µm pore size. 6. Concentrate the filtrate in vacuo on a rotary evaporator equipped with an oil pump. 7. Apply the residue onto a silica gel column (18 × 4–cm column) and elute with 1:9 (v/v) methanol/CH2 Cl2 . The typical volume for elution is 1 L and the size of the fractions is ∼20 mL.

8. Monitor fractions by TLC with 1:9 (v/v) MeOH/CH2 Cl2 . Visualize the desired product by staining the plates in an iodine chamber. The product S.2 can be seen as a dark spot with Rf = 0.41.

9. Combine the fractions containing the product and evaporate to dryness using a rotary evaporator connected to an oil pump. 10. Characterize by 1 H NMR spectroscopy and mass spectrometry. 1,4,7,10-Tetraazacyclododecane-4,7,10-tricarboxymethylmethyl ester 1-carboxymethylbenzyl ester (S.2): Yield of pale yellow oil: 0.45 g (60%). 1 H NMR (CDCl3 ): 7.35 (5H, m); 5.20 (2H, s); 3.76 (6H, s); 3.74 (3H, s); 3.49-2.35 (24 H). +ESI-TOF-MS: calcd. for C26 H40 N4 NaO8 + (M+Na)+ , 559.27; found, 559.27.

Reduce S.2 11. Dissolve 0.35g (0.64 mmol) S.2 in 50 mL methanol. 12. Add 55 mg of 10% Pd/C. 13. Hydrogenate in a hydrogenation apparatus overnight at atmospheric pressure. 14. Filter through Celite 521 and concentrate the filtrate in vacuo on a rotary evaporator. The product can be used for the next step without further purification.

15. Monitor the product by TLC in 2:8 (v/v) MeOH/CH2 Cl2 . Visualize by staining the plates in an iodine chamber. The product S.3 can be seen as a dark spot with Rf = 0.52.

16. Characterize the product by mass spectrometry. 1,4,7,10-Tetraazacyclododecane-1,4,7,10-tetraacetic acid trismethyl ester (S.3): +ESITOF-MS: calcd. for C19 H34 N4 NaO8 + (M+Na)+ , 469.23; found, 469.22.

Synthesize S.5 17. Dissolve 5.0 g (9.51 mmol) S.4, 2.61 g (12.2 mmol) N-trifluoroacetyl-6-aminohexan1-ol, and 2.98 g (11.46 mmol) triphenylphosphine in 50 mL dry THF. 18. Add 2.22 mL (11.2 mmol) DIAD to the stirring solution over a 15-min period. Continue stirring for 2.5 hr at room temperature. 19. Concentrate in vacuo using a rotary evaporator to obtain an oil. 20. Add 50 mL diethyl ether and keep the reaction mixture in an ultrasonic bath until all oil has been dissolved.



Supplement 29

21. Allow to stand 1 hr at room temperature. 22. Remove the triphenylphosphine oxide that forms by filtration using a glass filter with a 3-µm pore size. 23. Concentrate the filtrate to half volume using a rotary evaporator connected to an oil pump. 24. Remove any remaining triphenylphosphine oxide by repeating the filtration. 25. Concentrate the filtrate to an oil in vacuo on a rotary evaporator. 26. Purify the oily crude residue on a silica gel column (5 × 30–cm column), eluting with 1:9 (v/v) MeOH/CH2 Cl2 . The typical volume for elution is 1 to 2 L, and the size of the fractions is ∼25 mL.

27. Monitor fractions by TLC in 1:9 (v/v) MeOH/CH2 Cl2 . The product bearing the DMTr group can be visualized, in addition to normal UV detection, by heating up the fluorescent indicator plate on a hot plate. The product can be seen as an orange spot on a white background (Rf = 0.49).

28. Combine fractions containing the product and evaporate in vacuo using a rotary evaporator to afford S.4 as a white amorphous solid. 29. Characterize the compound by 1 H and 13 C NMR spectroscopy and mass spectrometry. N3 -(6-Trifluoroacetamidohex-1-yl)-5 -O-(4,4 -dimethoxytrityl)-2 -deoxyuridine (S.5): Yield: 6.2 g (91%). 1 H NMR (DMSO-d6 ): 9.36 (1H, br, NH); 7.68 (1H, d, J = 8.1 Hz, H6); 7.35 (2H, DMTr); 7.25 (7H, DMTr); 6.87 (4H, d, DMTr); 6.17 (1H, t, J = 6.2 Hz, H1 ), 5.47 (1H, d, J = 8.1 Hz, H5), 5.38 (1H, d, J = 4.7 Hz, 3 -OH), 4.31 (1H, m, H3 ), 3.90 (1H, m, H4 ), 3.5 (1H, dd, H5 ), 3.80 (2H, t, NCH2 ), 3.71 (6H, s, 2 × OMe), 3.25 (1H, m, H5 ), 3.20 (1H, dd, H5 ); 3.15 (2H, q, CH2 NH), 2.23 (2H, m, H2 , H2 ), 1.47 (4H, m); 1.25 (4H, m). 13 C NMR (DMSO-d6 ): 161.7 (C4), 158.0 (C=O), 156.5 (q, CF3 ); 150.3 (C2); 144.8 (DMTr); 138.8 (C6); 129.7, 127.8, 127.7, 126.7, 113.1 (DMTr); 100.7 (C5), 85.7 (DMTr); 85.5 (C4 ); 85.2 (C1 ); 69.8 (C3 ); 63.3 (C5 ); 55.5 (2 × OMe); 40.1 (NCH2 ); 39.7 (C2 ); 39.0 (CH2 NHCO); 28.0, 25.9, 25.8 (CH2 ). +ESI-TOF-MS: calcd. for C38 H42 F3 N3 NaO8 + (M+Na)+ , 748.28; found, 748.27.

Ammonolyze S.5 30. Suspend 1.41 g (1.95 mmol) S.5 in a mixture of 50 mL methanol and 50 mL aqueous ammonia. Heat overnight at reflux at 80◦ C with stirring. 31. Allow to cool to room temperature and then remove all volatiles in vacuo using a rotary evaporator connected to an oil pump. 32. Add 100 mL water and 100 mL CH2 Cl2 . Shake vigorously and isolate the organic phase. 33. Dry over Na2 SO4 , filter off the drying agent, and concentrate the filtrate in vacuo using a rotary evaporator connected to an oil pump. The product can be used for the next step without further purification.

34. Characterize the product by 1 H NMR spectroscopy and mass spectrometry.

Solid-Phase Oligonucleotide Labeling with DOTA

N3 -(6-Aminohex-1-yl)-5 -O-(4,4 -dimethoxytrityl)-2 -deoxyuridine (S.6): 1 H NMR (CDCl3 ): 7.75 (1H, d, J = 8.3 Hz, H6); 7.40-7.23 (9H, DMTr); 6.84 (4H, d, J = 8.9 Hz, DMTr); 6.31 (1H, t, J = 6.2 Hz, H1 ); 5.45 (1H, d, J = 8.3 Hz, H5); 4.53 (1H, m, H3 ); 4.00 (1H, m, H4 ); 3.89 (2H, m); 3.78 (6H, s, OMe); 3.49 (1H, dd, J = 3.0, 10.6 Hz, H5 ); 3.41 (1H, dd, J = 3.3, 10.6 Hz, H5 ); 2.64 (2H, t, J = 6.5 Hz); 2.42 (1H, m, H2 ); 2.24 (1H, m, H2 ); 2.19 (3H, br); 1.62 (2H, qv, J = 6.4 Hz); 1.40 (2H, qv, J = 6.7 Hz); 1.34 (4H, m). +ESI-TOF-MS: calcd. for C36 H44 N3 O7 + (M+H)+ , 630.31; found, 630.34.



Introduce DOTA to S.6 35. Dissolve 0.26 g (0.58 mmol) S.3 and 100 µL DIPEA in 9 mL dry DMF. 36. Add 220 mg (0.58 mmol) HATU and stir the mixture for 30 min at room temperature. 37. Add 0.37 g (0.58 mmol) S.6 and stir the mixture for 4 hr at room temperature. 38. Concentrate the filtrate in vacuo using a rotary evaporator connected to an oil pump. 39. Dissolve the residue in 50 mL CH2 Cl2 and wash two times with 25 mL sat. NaHCO3 . 40. Dry over Na2 SO4 , filter off the drying agent, and concentrate the filtrate in vacuo. 41. Purify the oily crude residue on a silica gel column (4 × 15–cm column), eluting with 1:9 (v/v) MeOH/CH2 Cl2 . The typical volume for elution is 300 to 400 mL, and the size of the fractions is ∼15 mL.

42. Monitor fractions by TLC in 1:9 (v/v) MeOH/CH2 Cl2 . The product bearing the DMTr group can be visualized, in addition to normal UV detection, by heating up the fluorescent indicator plate on a hot plate. The product can be seen as an orange spot on a white background (Rf = 0.51).

43. Combine fractions containing the product and evaporate in vacuo using a rotary evaporator to afford S.6 as a white amorphous solid. 44. Characterize the compound by 1 H NMR spectroscopy and mass spectrometry. DOTA-labeled nucleoside (S.7): Yield 0.43 g (82%). 1 H NMR (CDCl3 ): 7.75 (1H, d, J = 8.3 Hz, H6); 7.40-7.22 (9H, DMTr); 6.84 (4H, d, J = 8.8 Hz); 6.47 (1H, br t, J = 4.7 Hz, NH); 6.32 (1H, t, J = 6.3 Hz, H1 ); 5.43 (1H, d, J = 8.0 Hz, H5); 4.59 (1H, m, H3 ); 4.05 (1H, m, H4 ); 3.89 (2H, m); 3.79 (6H, s, OMe); 3.74 (6H, s); 3.73 (3H, s); 3.42 (2H, d, J = 2.9 Hz, H5 and H5 ); 3.20 (8H); 2.60 (16H); 2.46 (1H, m, H2 ); 2.27 (1H, m, H2 ); 1.70 (1H, br); 1.62 (2H, m); 1.50 (2H, m); 1.35 (4H, m). +ESI-TOF-MS: calcd. for C55 H76 N7 O14 + (M+H)+ , 1058.54; found, 1058.54.

Phosphitylate S.7 45. Dry 300 mg (0.28 mmol) S.7 by coevaporating two times with 5 mL dry acetonitrile, and dissolve in 10 mL dry acetonitrile. 46. Add 127 mg (0.42 mmol) 2-cyanoethyl-N,N,N ,N -tetraisopropylphosphorodiamidite and 0.63 mL of 0.45 M 1H-tetrazole in dry acetonitrile. Stir the mixture for 30 min at room temperature. 47. Pour the reaction mixture into 25 mL of 5% NaHCO3 with vigorous stirring. 48. Extract twice with 25 mL CH2 Cl2 . 49. Dry the combined organic layers over Na2 SO4 , filter off the drying agent, and concentrate the filtrate in vacuo. 50. Purify the residue on a silica gel column (4 × 2.5–cm column), eluting with 5:95 (v/v) MeOH/CH2 Cl2 containing 1% (v/v) TEA as the eluent. Typical volume for elution is 300 to 400 mL and the size of the fractions is ∼10 mL.

51. Monitor fractions by TLC in 5:95 (v/v) MeOH/CH2 Cl2 containing 1% (v/v) TEA. The product bearing the DMTr group can be visualized, in addition to normal UV detection, by heating up the fluorescent indicator plate on a hot plate. The product can be seen as an orange spot on a white background (Rf = 0.32).

52. Combine fractions containing the phosphoramidite, evaporate and dry in vacuo using a rotary evaporator to afford S.8 as a white solid foam.



Supplement 29

53. Characterize the compound by 31 P NMR spectroscopy and mass spectrometry. Phosphoramidite S.8 is stable for prolonged storage (at least 1 year) at −20◦ C. DOTA-labeled phosphoramidite (S.8): Yield of solid white foam: 230 mg (85%). 31 P NMR (CDCl3 ): 149.60; 149.20. +ESI-TOF-MS: calcd. for C64 H93 N9 O15 P [M + H]+ , 1258.65; found 1258.66. BASIC PROTOCOL 2

SYNTHESIS OF OLIGONUCLEOTIDE-DOTA CONJUGATES The standard phosphoramidite method can be applied to incorporate S.8 into oligonucleotides, except that a prolonged coupling time (10 min) should be used to achieve an acceptable 98% coupling efficiency. The oligodeoxyribonucleotides can be assembled on any automated DNA/RNA synthesizer. The overall strategy is outlined in Figure 4.31.2, and the synthesis of 5 -d(XTAATGTAGCCCCTGAA)-3 , where X stands for S.7, is presented as an example. This sequence was assembled on a 0.2-µmol scale using recommended protocols and DMTr-off synthesis. After chain assembly, the oligonucleotide conjugate is treated as described previously (Hovinen and Hakala, 2001). The support-bound oligonucleotide S.9 is treated with 0.1 M sodium hydroxide for 4 hr at room temperature to ensure complete hydrolysis of the ester protecting groups and to cleave the oligonucleotide conjugate from the support. After evaporation in the presence of ammonium chloride and treatment with aqueous ammonia for the final global deprotection, the oligonucleotide conjugate is treated with gadolinium(III) citrate. The chelate is then desalted, concentrated in vacuo, and purified by urea-PAGE. The purified product S.11 is concentrated using butanol, desalted, and characterized by UV spectroscopy and ESI-TOF-MS. Familiarity with automated DNA synthesis (APPENDIX 3C) and purification of oligonucleotides by urea-PAGE (APPENDIX 3B) is required to carry out this protocol. Characterization of the oligonucleotide conjugate requires knowledge of ESI-TOF-MS techniques (UNIT 10.2).

Figure 4.31.2 Synthesis of the oligonucleotide conjugate tethered to DOTA (see Basic Protocol 2). CPG, controlled-pore glass; DMTr, 4,4 -dimethoxytrityl. Modified from Jaakkola et al. (2006) with permission from the American Chemical Society. Solid-Phase Oligonucleotide Labeling with DOTA



Materials DOTA-labeled phosphoramidite (S.8; see Basic Protocol 1) Dry acetonitrile, ≥99.5%, ≤0.005% H2 O (Merck) Deoxyribonucleoside phosphoramidites (e.g., Proligo) 0.1 M NaOH 1 M NH4 Cl Aqueous ammonia, p.a. 28% to 30% (Merck) Gadolinium(III) citrate: 0.1 M Gd(III)Cl3 /0.2 M citric acid 20% polyacrylamide gel containing 7 M urea (APPENDIX 3B) 1 M Na2 CO3 , pH 9.8 1-Butanol, ≥99.5% pure (Merck) Sterile, nuclease-free, deionized water Automated DNA/RNA synthesizer (e.g., Applied Biosystems) ◦ 0.2-µmol DNA synthesis column (1000 A CPG; Applied Biosystems) 1-mL syringes 2-mL vials Speedvac evaporator (e.g., SPD121P, Savant) NAP5 gel filtration columns (GE Healthcare) Centrifuge (e.g., BR4i, JOUAN) Additional reagents and equipment for automated DNA synthesis (APPENDIX 3C) and urea-PAGE (APPENDIX 3B) Incorporate S.7 into oligodeoxyribonucleotide chain 1. Prepare a 0.2 M solution of S.8 in anhydrous acetonitrile. Dissolve the standard deoxyribonucleoside phosphoramidites in anhydrous acetonitrile according to the manufacturer’s instructions. 2. Start the automated solid-phase oligonucleotide synthesis according to the manufacturer’s instructions. Typically, use a 0.2-µmol synthesis column packed with a 1000 ◦ A CPG support, the recommended synthesis protocol for a 0.2-µmol scale, and the DMTr-off mode. Adjust the coupling time to 600 sec for S.8. 3. Upon completion of chain assembly (S.10), remove the synthesis column.

Cleave oligonucleotide conjugate from support and hydrolyze methyl ester protecting groups 4. Attach a 1 mL-syringe to one end of the synthesis column. Take up 1 mL of 0.1 M NaOH using another syringe and attach it to the other end of the synthesis column. Transfer the 0.1 M NaOH solution back and forth through the column every 30 min for 4 hr at ambient temperature. 5. Collect the NaOH solution containing the oligonucleotide conjugate into a clean 2-mL vial, add 100 µL of 1 M NH4 Cl, and evaporate to dryness using a Speedvac evaporator.

Ammonolyse and introduce gadolinium(III) 6. Dissolve the oligonucleotide conjugate in 1 mL of concentrated aqueous ammonia and incubate 16 hr at 55◦ C. Ammonium hydroxide is used for final global deprotection.

7. Add 15 mol equiv per ligand of gadolinium(III) citrate and keep at least overnight at ambient temperature.



Supplement 29

Purify oligonucleotide conjugate S.11 8. Desalt the oligonucleotide conjugate containing gadolinium(III) using a NAP5 gel filtration column and sterile deionized water according to manufacturer’s instructions. 9. Concentrate the oligonucleotide conjugate in vacuo using a Speedvac evaporator. 10. Purify the oligonucleotide conjugate using a 20% polyacrylamide gel containing 7 M urea. 11. Inspect the gel under UV illumination and collect the area of the gel that contains the gadolinium(III) oligonucleotide conjugate. The main product with highest molecular weight is collected.

12. Elute the oligonucleotide conjugate passively from the gel pieces by soaking overnight in 10 mM aqueous Na2 CO3 , pH 9.8, at ambient temperature. 13. Collect the aqueous solution containing the oligonucleotide conjugate and concentrate the solution by adding 2 vol of 1-butanol per aqueous volume. Mix well and centrifuge for 2 min at 90 × g (700 rpm), room temperature. Discard the butanol layer and repeat the butanol concentration until the volume of the aqueous phase is 98%). Labels attached at the N3 position of uracil residues naturally weaken hydrogen bonds in a duplex. Thus, these labels should only be used upstream or downstream of the coding sequence. Knowledge of automated DNA synthesis (APPENDIX 3C) and isolation of the products by gel electrophoresis (UNIT 10.4, APPENDIX 3B) is needed. Characterization of the compounds demands knowledge of 1 H, 13 C, and 31 P NMR spectroscopies and electrospray ionization mass spectrometry (ESI-MS; UNIT 10.2). Solid-Phase Oligonucleotide Labeling with DOTA

Metal ion chelation It is essential to treat the protected oligonucleotide with aqueous sodium hydroxide for at least 4 hr at room temperature (∼25◦ C) prior to

conventional ammonolysis to avoid formation of carboxamides that will result from reaction of ammonia with the DOTA ester groups. Although DOTA amides do chelate various metal ions, the amide chelates are known to be less stable than the corresponding carboxylic acid– derived chelates (Paul-Roth and Raymond, 1995). Furthermore, if saponification is incomplete, ammonolysis will result in the formation of a mixture containing mono-, di-, and triamides, which are complicated to separate by gel electrophoresis. The metal must be introduced as a citrate. Citric acid forms rather stable and soluble complexes with several metal ions, which do not dissociate even under the basic conditions needed for oligonucleotide deprotection. This is extremely important with lanthanide ions, since trivalent lanthanide(III) ions are powerful catalysts in the hydrolysis of phosphate esters (Butcher and Westheimer, 1955). Furthermore, a prolonged reaction time (overnight at room temperature) and an excess of gadolinium(III) citrate are required to ensure complete chelate formation. According to gel electrophoresis, the crude oligomer bearing the tethered DOTA is completely stable to the reaction conditions required for deprotection and chelate formation. The desired oligonucleotide conjugate (>95% of the crude reaction mixture) is easily isolated from the gel and characterized by mass spectrometry. The observed molecular weight is in accordance with the proposed structure. As shown here, the phosphoramidite S.8 is suitable for site-specific incorporation of MRI contrast agents into oligonucleotides. Although only the preparation of Gd3+ chelates is presented here, the method could also be used in applications based on PET and SPECT (single photon emission tomography) using, for example, 67/68 Ga3+ and 111 In3+ as central ions, respectively. Naturally, for PET and SPECT applications, the metal should be introduced just prior to analysis. The slow kinetics of DOTA chelate formation should be taken into account when short-lived radioisotopes must be used. However, as proposed by Velikyan et al. (2004), the complexation can be accelerated using microwave radiation.

Anticipated Results Good to moderate yields of the individual steps of the total synthesis of the DOTAphosphoramidite are expected. The building block allows convenient synthesis of oligonucleotide conjugates on a solid support. The



target compound is expected to be clearly the main product, and its isolation by gel electrophoresis is therefore fairly straightforward.

Time Considerations The synthesis of phosphoramidite S.8 starting from the cyclen S.3 and 5 -O-(4,4 dimethoxytriyl)-2 -deoxyuridine (S.4) can be accomplished in ∼10 days (including two overnight reactions). The time needed for automated oligonucleotide chain assembly differs from the standard protocol only by the extended coupling time of 600 sec per DOTA phosphoramidite unit. The deprotection and subsequent conversion of the oligonucleotide to the corresponding metal chelate increases the total time of the conjugate synthesis by 24 hr. The times needed for purification and characterization are the same as with standard methods.

Literature Cited Aime, S., Botta, M., Fasano, M., and Terreno, E. 1999. Lanthanide(III) chelates for NMR biomedical applications. Chem. Soc. Rev. 27:1929. Anderson, C.J. and Welch, M.J. 1999. Radiometallabeled agents (non-technetium) for diagnostic imaging. Chem. Rev. 99:2219-2234. Butcher, W.W. and Westheimer, F.H. 1955. The lanthanum hydroxide gel promoted hydrolysis of phosphate esters. J. Am. Chem. Soc. 77:24202424. Caravan, P., Ellison, J.J., McMurry, T.J., and Lauffer, R.B. 1999. Gadolinium(III) chelates as MRI contrast agents: Structure, dynamics, and applications. Chem. Rev. 99:2293-2352. Heppeler, A., Froilevaux, S., Mäcke, H.R., Jermann, H.E., Béhé, M., Powell, P., and Hennig, M. 1999. Radiometal-labeled macrocyclic chelator-derivatised somatostatin analogue with superb tumor-targeting properties and potential for receptor-mediated tumor therapy. Chem. Eur. J. 5:1974-1981.

Hovinen, J. and Takalo, H. 2005. Oligonucleotide labeling reactants and their use. US Pat. 6,949,639. Jaakkola, L., Ylikoski, A., and Hovinen, J. 2006. Simple synthesis of a building block for solid phase labeling of oligonucleotides with 1,4,7,10-tetraazacyclododecane-1,4,7,10tetraacetic acid (DOTA). Bioconjug. Chem. 17:1105-1107. Mitsunobu, O. 1981. The use of diethyl azodicarboxylate and triphenylphosphine in synthesis and transformation of natural products. Synthesis 1-28. Paul-Roth, C. and Raymond, K.N. 1995. Amide functional group contribution to the stability of gadolinium(III) complexes: DTPA derivatives. Inorg. Chem. 34:1408-1412. Runge, V.M. 2000. Safety of approved MR contrast media for intravenous injection. J. Magn. Reson. Imaging 12:205-213. Sinha, N.D. and Striepeke, S. 1991. Oligonucleotides with reporter groups attached to the 5 terminus. In Oligonucleotides and Analogues. A Practical Approach (F. Eckstein, ed.) p. 189. Oxford University Press, Oxford. Velikyan, I., Lendvai, G., Välilä, M., Roivainen, A., Yngve, U., Bergström, M., and L˚angström, B. 2004. Microwave-accelerated 68 Ga-labelling of oligonucleotides. J. Labeled Comp. Radiopharm. 47:79-89. Volkert, W.A. and Hoffman, T.J. 1999. Therapeutic radiopharmaceuticals. Chem. Rev. 99:22692292. Woods, M., Kovacs, Z., and Sherry, A.D. 2002. Targeted complexes of lanthanide(III) ions as therapeutic and diagnostic pharmaceuticals. J. Supramol. Chem. 2:1-15.

Contributed by Lassi Jaakkola, Alice Ylikoski, and Jari Hovinen PerkinElmer Life and Analytical Sciences Turku, Finland

Hovinen, J. and Hakala, H. 2001. Versatile strategy for oligonucleotide derivatization. Introduction of lanthanide(III) chelates to oligonucleotides. Org. Lett. 3:2474-2476.



Supplement 29

CHAPTER 5 Methods for Cross-Linking Nucleic Acids INTRODUCTION

C

haracterization of synthetic oligonucleotides shapes much of our understanding of native, higher-molecular-weight DNA and RNA molecules. Although of immense utility, oligonucleotides usually possess lower structural and thermal stability and have greater end effects than the larger nucleic acid constructs they are intended to model. Hence, the physiochemical and biological properties of oligonucleotides may not always compare favorably to those of larger nucleic acids. One of the most successful strategies to stabilize oligonucleotides is to connect the strands that comprise elements of helical structure with an interstrand cross-link. Cross-linked oligonucleotides are often very resistant to denaturation (induced thermally or by changes in pH, salt, or oligonucleotide concentration) relative to their unmodified constructs. A wide variety of novel chemistries exist to introduce cross-links into nucleic acids. In nearly all of these methods, solid-phase synthesis of oligomers site-specifically labeled with modified nucleosides bearing the appropriate reactive functional groups is used to generate cross-links in high yield. In addition to providing increased structural stability, cross-links have been used to probe the geometry and conformational dynamics of both medium- and large-sized nucleic acids and have been used to explore the topology of protein-ligand complexes. In the units presented in this chapter, along with those in forthcoming supplements, the reader will be provided with state-of-the art protocols to form cross-links within nucleic acids and nucleic acid–ligand complexes. To provide the reader with a comprehensive array of techniques, the chapter will present the chemistry to synthesize both interstrand and intrastrand cross-links, cyclic nucleic acids, and nucleic acid–ligand complexes.

UNIT 5.1 presents protocols to postsynthetically modify 2-amino-containing oligoribonucleotides with either an alkyl-phenyl disulfide or an alkyl thiol group. These groups react under mild conditions to form disulfide cross-links by thiol-disulfide interchange. When incorporated on opposite faces of a short, continuous RNA helix, these reactants, as expected, do not form a disulfide bond. In contrast, when these reactive groups are placed in proximity, disulfide cross-links form rapidly. In addition, by incorporating these groups at various positions of large RNAs through semisynthesis, the dynamics of thermal motions can be detected. Such motions are believed to be linked to biological function, and the protocols presented are among the few simple ways to assess such dynamics.

Methods to synthesize small circular oligonucleotides for use in diagnostic, therapeutic, and laboratory operations are presented in UNIT 5.2. These systems have gained considerable attention in recent years because they form unusually strong and specific complexes with RNA and DNA strands. In addition to their properties as molecular recognition agents, synthetic circular DNAs 20 to 200 nucleotides in size can also serve as catalysts for the amplified synthesis of DNA and RNA, a process termed “rolling circle synthesis.” One of the most convenient methods for generating oligonucleotides possessing either intrastrand or interstrand cross-links is through incorporation of oligo(ethylene glycol) bridges by solid-phase synthesis as described in UNIT 5.3. Many of the reagents needed are either commercially available or can be prepared in a few easy synthetic steps. Unlike many other DNA and RNA cross-links, aspects of the structural and thermodynamic impact of modifying nucleic acids with oligo(ethylene glycol) has been studied. Contributed by Gary D. Glick Current Protocols in Nucleic Acid Chemistry (2003) 5.0.1-5.0.2 Copyright © 2003 by John Wiley & Sons, Inc.

Methods for Cross-Linking Nucleic Acids

5.0.1 Supplement 13

In UNIT 5.4, a second group of methods for incorporating disulfide cross-links within RNA structure is presented. These protocols describe methods for the synthesis of alkylthiolmodified ribonucleosides, their incorporation into synthetic RNA, and the formation of intramolecular disulfide bonds in RNA by air oxidation. The disulfide bonds can be formed in quantitative yields between thiols positioned in close proximity in either RNA secondary or tertiary structure. Disulfide cross-links are useful tools to probe solution structures of RNA, to monitor dynamic motions, to stabilize folded RNAs, and to study the process of tertiary structure folding. In UNIT 5.5, site-specific cross-links are introduced into oligodeoxyribonucleotides by electrophilic substitution. A nucleophilic base (deoxythiouridine) is incorporated into an oligodeoxyribonucleotide, and an electrophilic tether is used to cross-link that base to a complementary DNA strand. A variety of different electrophilic DNA strands can be generated from the same deoxythiouridine-containing oligodeoxyribonucleotide by changing the nature of the electrophilic tether. In UNIT 5.6, the preparation of short endcapped DNA duplexes is presented. Using this approach, the 5′ terminus of one strand of a duplex is cross-linked to the complementary 3′ strand with a hydrophobic or hydrophilic linker. Several different linkers are presented along with methods for their incorporation into DNA during solid-phase synthesis. Finally, in UNIT 5.7, terminal disulfide cross-links are generated by substituting the terminal bases of oligodeoxyribonucleotides with a modified thymidine residue. The usefulness of this approach is discussed, with a demonstration that the cross-links do not modify the structure of the oligodeoxyribonucleotides, and examples of applications. Gary D. Glick

Introduction

5.0.2 Supplement 13


Engineering Disulfide Cross-Links in RNA Using Thiol-Disulfide Interchange Chemistry

UNIT 5.1

This unit presents methods for incorporating disulfide cross-links within RNA structures using thiol-disulfide interchange chemistry (Fig. 5.1.1). The alkyl phenyl disulfide S.1 and the alkyl thiol S.2 are incorporated at specific ribose 2′ positions within RNA, and then react through thiol-disulfide interchange to form the more stable dialkyl disulfide (see Basic Protocol 4). Such disulfide cross-linking can be used to prepare a simple covalent conjugate of two RNA molecules or, in more complex systems, to exert a specific conformational constraint to a dynamic RNA molecule. Compared to the approach previously used to obtain disulfide cross-linking of RNA—oxidation of two thiols (Goodwin et al., 1996; Sigurdsson et al., 1995)—thiol-disulfide interchange has the advantages that it proceeds under mild conditions without an oxidative catalyst and can be kinetically characterized. Preparation of disulfide cross-linking precursors S.1 and S.2 begins with a site-specific incorporation of a 2′-amino-2′-deoxy residue within each of the two RNA species that are to be cross-linked, achieved through standard solid-phase RNA synthesis with a protected 2′-amino-2′-deoxy nucleotide phosphoramidite (see Basic Protocol 1). After deprotection of the synthetic RNA, the 2′ amine is modified with N-succinimidyl-3-(2pyridyldithio) propionate (S.3), affording the alkyl pyridyl disulfide S.4 (Fig. 5.1.2). Thiol-disulfide interchange of S.4 with thiophenol affords the alkyl phenyl disulfide S.1 (see Basic Protocol 2); reduction of S.4 with dithiothreitol yields the alkyl thiol S.2 (Fig. 5.1.2; see Basic Protocol 3). The RNA system to be cross-linked by thiol-disulfide interchange is limited to a two-piece system. Because both cross-linking precursors arise from the same intermediate (S.4, Fig. 5.1.2), S.1 and S.2 cannot be differentiated synthetically on the same RNA molecule.

O

O

O

O

B HS

O N H O P O O O

H

N

O P O O

+

S S

O

B

O

2

1

O

O

B O

O

O N H O P O O O

S S

N

B

H

O

O P O O

O

HS

Figure 5.1.1 Disulfide cross-linking of RNA through thiol-disulfide interchange. B, nitrogenous base.

Methods For Cross-Linking Nucleic Acids

Contributed by Scott B. Cohen and Thomas R. Cech

5.1.1


O

O

B

O +

NH2 O O P O O

O

O

N O

S S

O

O

N

3

B

N O H O P O O O

S S

N

4 DTT HS

O

O

B

O

N O H O P O O O

S S

B

O

N O H O P O O O

SH

2

1 S

N H

Figure 5.1.2 Preparation of the alkyl phenyl disulfide S.1 and the alkyl thiol S.2 as precursors for thiol-disulfide interchange. B, nitrogenous base.

Therefore, the cross-linking system must be designed such that two components of the RNA molecule (two complementary strands, a ribozyme-substrate complex, etc.), containing either S.1 or S.2 separately, can be associated through base pairing or other noncovalent interaction. After the full RNA molecule is associated, initiation of crosslinking and the rate at which it proceeds are controlled by manipulating the pH of the reaction mixture. The RNA component containing the alkyl thiol S.2 is present in saturating excess over the RNA component containing the alkyl phenyl disulfide S.1 to discourage spurious cross-linking resulting from the presence of unassociated S.1. NOTE: Experiments involving RNA require careful precautions to prevent contamination and RNA degradation; see APPENDIX 2A (do not use DEPC; this should be unnecessary, and is inadvisable with 2′ amine chemistry). BASIC PROTOCOL 1

PREPARATION OF RNA OLIGONUCLEOTIDES CONTAINING A SITE-SPECIFIC 2′ AMINE GROUP

RNA Cross-Linking Using Thiol-Disulfide Interchange Chemistry

Solid-phase synthesis of the two RNA oligonucleotides containing a unique 2′ amine employs the same procedures used for standard RNA phosphoramidites with the inclusion of a 2′-amino-2′-deoxy nucleotide phosphoramidite. The 2′-amino-2′-deoxy C-, U-, and G-phosphoramidites are synthesized according to the procedures of Verheyden et al. (1971), Imazawa and Eckstein (1979), and Benseler et al. (1992) and are produced commercially by Nexstar Pharmaceuticals. The following protocol describes deprotection and purification procedures for the solid support–bound product. Treatment with ammonium hydroxide to remove the exocyclic and 2′ amine protecting groups remaining from the synthesis procedure is followed by removal of the 2′-O-silyl protecting groups with fluoride to liberate the 2′ hydroxyl. One RNA will then be radiolabeled and modified with


an alkyl phenyl disulfide group (S.1) as described in Basic Protocol 2. The other RNA will be modified to contain an alkyl thiol group (S.2; Basic Protocol 3). For a general overview of oligonucleotide synthesis, see APPENDIX 3C. Materials Solid support–bound product of automated RNA synthesis (1-µmol synthesis scale) 3:1 (v/v) concentrated ammonium hydroxide (NH4OH)/absolute ethanol 24:46:30 (v/v/v) triethylamine/1-methyl-2-pyrrolidinone/triethylamine trihydrofluoride (see recipe) TE buffer, pH 7.5 (APPENDIX 2A) NAP-25 Sephadex column (Amersham Pharmacia Biotech) 1 M sodium chloride Absolute ethanol TBE buffer (APPENDIX 2A) 80% formamide/TBE solution (see recipe) Denaturing polyacrylamide gel: 20% polyacrylamide/8 M urea in TBE buffer, dimensions 20 cm long × 26 cm wide × 0.3 cm thick (see APPENDIX 3B, CPMB UNIT 7.6 or Sambrook et al., 1989) TEN buffer (see recipe) 4-mL screw-cap vial Teflon tape Water baths, 55° and 65°C Rotary drying system for microcentrifuge tubes (Savant) 40-mL Oak Ridge centrifuge tube Preparative centrifuge (Sorvall or Beckman) UV lamp (hand held) 50-mL polypropylene centrifuge tube 0.45-µm cellulose acetate filter Additional reagents and equipment for denaturing polyacrylamide gel electrophoresis and UV shadowing (APPENDIX 3B, CPMB UNIT 7.6 or Sambrook et al., 1989) Remove the RNA protecting groups 1. Transfer the solid support–bound RNA product into a 4-mL screw-cap vial. 2. Add 1.5 mL of 3:1 (v/v) NH4OH/ethanol. Seal the vial with Teflon tape and incubate 8 hr at 55°C. 3. Cool the vial to –20°C and decant the supernatant into a 1.5-mL microcentrifuge tube. Concentrate the supernatant under vacuum with a rotary drying system. 4. Dissolve the dry residue in 400 µL of 24:46:30 (v/v/v) triethylamine/1-methyl-2pyrrolidinone/triethylamine trihydrofluoride. Incubate 1.5 hr at 65°C. Heating at 65°C (as suggested by Wincott et al., 1995) will help dissolve the dry residue. For RNAs shorter than 10 nt, reduce the amount of fluoride solution to 100 mL and skip step 5.

Purify the crude RNA 5. Transfer the solution to 5 mL TE buffer, pH 7.5, and mix thoroughly. 6. Load 2.5 mL of the resulting solution onto each of two NAP-25 Sephadex columns. Elute the RNA from each column with 3.5 mL water, collecting the first 3.5 mL of eluate from each (7 mL total product solution).



7. Add 1 mL of 1 M sodium chloride to the product solution and transfer to a 40-mL Oak Ridge centrifuge tube. 8. Precipitate the RNA by adding 24 mL absolute ethanol and then centrifuging 30 min at 10,000 × g, 2°C. Decant the supernatant and allow the RNA pellet to air dry. 9. Dissolve the RNA pellet in 150 µL TE buffer and 150 µL of 80% formamide/TBE (300 µL total). 10. Purify the RNA to single-nucleotide resolution by denaturing polyacrylamide gel electrophoresis at 25 W until the full-length product has migrated about two-thirds of the way down the gel, as indicated by the dye markers (see APPENDIX 3B, CPMB UNIT 7.6 or Sambrook et al., 1989, for standard procedures). A gel 20 cm long × 26 cm wide × 0.3 cm thick is sufficient for three separate RNA oligonucleotide preparations.

Isolate the pure RNA 11. Identify the full-length RNA band by UV shadowing and excise the gel slice. 12. Place the gel slice in a 50-mL polypropylene centrifuge tube, crush thoroughly, and add 15 mL TEN buffer. Place the suspension on a shaker for 24 hr at 2°C. 13. Pellet the polyacrylamide by centrifugation for 10 min at 2000 × g, 2°C. 14. Filter the supernatant through a 0.45-µm cellulose acetate filter. 15. Precipitate the RNA by adding 3 vol absolute ethanol and then centrifuging 30 min at 10,000 × g, 2°C. 16. Dissolve the RNA in 100 µL TE buffer, pH 7.5. Determine RNA concentration spectrophotometrically by measuring A260, and adjust to 100 µM. Store up to 6 months at −20°C or indefinitely at –80°C. BASIC PROTOCOL 2

PREPARATION OF 32P-LABELED RNA CONTAINING AN ALKYL PHENYL DISULFIDE GROUP The RNA component of the cross-linking system containing the alkyl phenyl disulfide S.1 is prepared as a 32P-labeled reagent to allow monitoring of the cross-linking reaction, indicated by a shift to a product that migrates more slowly under denaturing electrophoresis conditions. Labeling of the RNA at the 5′ end with 32P is followed by chemical modification of the 2′ amine with N-succinimidyl-3-(2-pyridyldithio) propionate (S.3), yielding the corresponding amide S.4. Thiol-disulfide interchange of the pyridyl disulfide moiety of S.4 with thiophenol yields the alkyl phenyl disulfide S.1 (Fig. 5.1.2).


Materials RNA oligonucleotide with 2′ amine group (100 µM in TE buffer; see Basic Protocol 1) ≥0.1 Ci/µL [γ-32P]ATP (6000 Ci/mmol) 10 U/µL T4 polynucleotide kinase and 10× buffer (New England Biolabs) 1 M sodium chloride Absolute ethanol 1 M sodium borate buffer, pH 8 500 mM N-succinimidyl-3-(2-pyridyldithio) propionate (S.3; Pierce Chemicals) in N,N-dimethylformamide (prepare solution just before use) 70 mM thiophenol in absolute ethanol TE buffer, pH 7.5 (APPENDIX 2A)


1× TBE buffer (APPENDIX 2A) 80% formamide/TBE solution (see recipe) Denaturing polyacrylamide gel: 20% polyacrylamide/8 M urea in TBE buffer, dimensions 20 cm long × 10 cm wide × 0.05 cm thick (see APPENDIX 3B, CPMB UNIT 7.6 or Sambrook et al., 1989) TEN buffer (see recipe) 10 mM sodium acetate buffer, pH 4.5 (APPENDIX 2A) Preparative centrifuge (Sorvall or Beckman) X-ray film for autoradiography Water bath, 37°C Additional reagents and equipment for denaturing polyacrylamide gel electrophoresis (APPENDIX 3B, CPMB UNIT 7.6 or Sambrook et al., 1989) CAUTION: Step 8 should be performed in a fume hood because of the stench of thiophenol. Any labware that comes in contact with thiophenol should be soaked in bleach solution. Label the RNA 1. Mix the following in a 1.5-mL microcentrifuge tube (9 µL total volume): 1 µL 100 µM RNA oligonucleotide with 2′ amine 5 µL ≥0.1 Ci/µL [γ-32P]ATP (6000 Ci/mmol; ≥0.5 mCi total) 1 µL 10× T4 polynucleotide kinase buffer 2 µL H2O. Warm the solution to 37°C. 2. Add 1 µL of 10 U/µL T4 polynucleotide kinase and incubate 30 min at 37°C. 3. Add 200 µL water and 50 µL of 1 M sodium chloride; mix well. 4. Precipitate the labeled RNA by adding 900 µL absolute ethanol and centrifuging 20 min at 16,000 × g, 2°C. 5. Decant the supernatant and allow the RNA pellet to air dry. Modify the 2′ amine 6. Dissolve the labeled RNA in: 140 µL water 20 µL 1 M sodium borate buffer, pH 8 20 µL 1 M sodium chloride. Warm the solution to 37°C. 7. Add 20 µL freshly dissolved 500 mM N-succinimidyl-3-(2-pyridyldithio) propionate (S.3) in N,N-dimethylformamide and mix well. Incubate 20 min at 37°C. Upon addition of S.3, the solution will become cloudy because of the compound’s limited water solubility but will clear during the course of the reaction as a result of hydrolysis.

8. In a fume hood, add 200 µL of 70 mM thiophenol in ethanol. Incubate 2 min at 23°C. The reaction should display a light yellow color from formation of pyridine-2-thione.

9. Precipitate the RNA by adding 700 µL absolute ethanol and centrifuging 20 min at 16,000 × g, 2°C. 10. Decant the supernatant and allow the RNA pellet to air dry.



Purify the modified RNA 11. Dissolve the crude product in 10 µL TE buffer and 10 µL of 80% formamide/TBE solution (20 µL total). 12. Purify the RNA by denaturing polyacrylamide gel electrophoresis at 15 W (see APPENDIX 3B, CPMB UNIT 7.6 or Sambrook et al., 1989, for standard procedures). A gel 20 cm long is sufficient to provide clean separation from the faster-migrating RNA that remains unmodified after treatment with S.3 (∼25%).

13. Identify the band by autoradiography and excise the gel slice. 14. Crush the gel slice thoroughly in a 1.5-mL microcentrifuge tube and add 500 µL TEN buffer. Incubate the suspension 10 min on ice, vortexing occasionally. 15. Pellet the polyacrylamide by centrifugation for 2 min at 16,000 × g, 2°C, and carefully withdraw the supernatant into a fresh 1.5-mL microcentrifuge tube. 16. Precipitate the RNA by adding 4 vol absolute ethanol and centrifuging 30 min at 16,000 × g, 2°C. 17. Dissolve the modified RNA in 100 µL 10 mM sodium acetate buffer, pH 4.5. Determine radioactivity level to confirm that it is ≥105 cpm/µL. Store in 20-µL aliquots at −80°C. BASIC PROTOCOL 3

PREPARATION OF RNA CONTAINING ALKYL THIOL GROUP The RNA component of the cross-linking system containing the alkyl thiol S.2 is prepared using similar chemistry as for S.1. The 2′ amine is modified with N-succinimidyl-3-(2pyridyldithio) propionate (S.3), yielding the corresponding amide S.4. Reduction of the pyridyl disulfide moiety of S.4 with dithiothreitol yields the alkyl thiol S.2 (Fig. 5.1.2). Materials 20 nmol RNA oligonucleotide with 2′ amine group (see Basic Protocol 1) 1 M sodium borate buffer, pH 8 1 M sodium chloride 500 mM N-succinimidyl-3-(2-pyridyldithio) propionate (S.3; Pierce Chemicals) in N,N-dimethylformamide (prepare just before use) Absolute ethanol 1 M dithiothreitol (DTT) in water (APPENDIX 2A) TEN buffer (see recipe) TE buffer, pH 7.5 (APPENDIX 2A) Water bath, 37°C Modify the 2′ amine 1. In a 1.5-mL microcentrifuge tube, precipitate 20 nmol of the RNA to be modified by mixing 200 µL of the 100 µM 2′-amine-modified RNA with 50 µL of 1 M sodium chloride followed by 750 µL absolute ethanol, then centrifuging 20 min at 16,000 × g, 2°C. 2. Dissolve the RNA pellet in:


120 µL H2O 40 µL 1 M sodium borate buffer, pH 8 20 µL 1 M sodium chloride. Warm the solution to 37°C.


3. Add 20 µL freshly dissolved 500 mM S.3 in N,N-dimethylformamide and mix well. Incubate 20 min at 37°C. This reaction may be scaled up as needed, maintaining the following final concentrations: 200 mM NaB(OH)3, 100 mM NaCl, 100 mM RNA, and 50 mM S.3.

4. Precipitate the RNA by adding 600 µL absolute ethanol and centrifuging 20 min at 16,000 × g, 2°C. 5. Repeat steps 2 to 4. Double treatment with S.3 should modify ≥95% of the 2′ amines.

Liberate the thiol 6. Dissolve the RNA pellet in 140 µL water and 40 µL of 1 M sodium borate buffer, pH 8. Liberate the thiol by adding 20 µL of 1 M DTT. Incubate 30 min at 37°C. 7. Precipitate the RNA by adding 50 µL of 1 M sodium chloride and 750 µL absolute ethanol, and centrifuging 20 min at 16,000 × g, 2°C. 8. Dissolve the modified RNA in 300 µL TEN buffer. Precipitate the RNA by adding 900 µL absolute ethanol and then centrifuging again as in step 7. 9. Dissolve the modified RNA in 100 µL TE buffer. Determine RNA concentration spectrophotometrically by measuring A260, and adjust to 100 µM. Store up to 6 months at −20°C or indefinitely at –80°C. CROSS-LINKING OF RNA THROUGH THIOL-DISULFIDE INTERCHANGE The two RNA components containing S.1 and S.2 are associated at pH 4.5, where the nucleophilicity of the alkyl thiol S.2 is attenuated such that cross-linking before association is discouraged. Cross-linking is then initiated by increasing the pH, usually to the range of pH 7 to 8. The cross-linking reaction is conducted with the thiol component S.2 in saturating excess (up to 10 times the equilibrium dissociation constant) over the phenyl disulfide component S.1 to discourage spurious cross-linking resulting from the presence of unassociated S.1.

BASIC PROTOCOL 4

The following protocol is from original ribozyme cross-linking experiments by Cohen and Cech (1997). The concentrations of RNA components, cross-linking pH, and other experimental conditions can be adjusted to accommodate other experimental systems.

Materials 32 P-labeled RNA modified with phenyl disulfide S.1 (≥105 cpm/µL in 10 mM sodium acetate buffer, pH 4.5; see Basic Protocol 2) RNA modified with alkyl thiol S.2 (1 µM in TE buffer; see Basic Protocol 3) 100 mM sodium acetate buffer, pH 4.5 (APPENDIX 2A) 100 mM magnesium chloride 5 M sodium chloride Formamide quenching solution (see recipe) 1 M sodium HEPES buffer, pH 7.5 (APPENDIX 2A) 0.65-mL microcentrifuge tube Water bath, 30°C Additional reagents and equipment for denaturing polyacrylamide gel electrophoresis (APPENDIX 3B, CPMB UNIT 7.6 or Sambrook et al., 1989)



1. In a 0.65-mL microcentrifuge tube, mix (90 µL total volume): 10 µL 1 µM S.2-modified RNA 20 µL 100 mM sodium acetate buffer, pH 4.5 10 µL 100 mM magnesium chloride 20 µL 5 M sodium chloride 30 µL H2O. Warm the solution to 30°C. 2. Add 10 µL of ≥105 cpm/µL 32P-labeled S.1-modified RNA. Incubate 30 min at 30°C. 3. Remove 5 µL of the reaction and transfer to a fresh microcentrifuge tube containing 25 µL formamide quenching solution. Store frozen on dry ice. This aliquot serves to measure the small amount of cross-linking (if any) that occurs during the association incubation at pH 4.5.

4. Initiate cross-linking by adding 10 µL of 1 M sodium HEPES buffer, pH 7.5, to attain a final reaction pH of ∼7.2. The absolute rates of cross-linking can be controlled by manipulating the pH of the cross-linking reaction (∆log k/∆pH ∼1 in the pH range of 4.5 to 8.0).

5. At the desired time intervals, transfer 5 µL of the reaction to separate tubes containing 25 µL formamide quenching solution. Store the time aliquots frozen on dry ice until analysis by denaturing polyacrylamide gel electrophoresis (see APPENDIX 3B, CPMB UNIT 7.6 or Sambrook et al., 1989). For example, in the original ribozyme cross-linking experiments, 3- to 5-min intervals were used.


Formamide, 80%, in TBE To 40 mL formamide add 25 mg bromphenol blue and 25 mg xylene cyanol; mix well. Add 10 mL of 5× TBE buffer (APPENDIX 2A). Store up to 6 months at 23°C. The resulting 50 mL contains 80% formamide, 0.05% bromphenol blue, 0.05% xylene cyanol, 90 mM Tris-borate, and 1 mM EDTA.

Formamide quenching solution 800 µL formamide 0.5 mg bromphenol blue 0.5 mg xylene cyanol 100 µL 1 M sodium acetate buffer, pH 4.5 100 µL 0.5 M disodium EDTA Prepare fresh This must be prepared fresh on the day of use because EDTA will precipitate during extended storage at pH 4.5. The resulting 1.0 mL contains 80% formamide, 0.05% bromphenol blue, 0.05% xylene cyanol, 100 mM sodium acetate buffer (pH 4.5), and 50 mM EDTA. RNA Cross-Linking Using Thiol-Disulfide Interchange Chemistry

5.1.8

TEN buffer 10 mM Tris⋅Cl, pH 7.5 (APPENDIX 2A) 1 mM disodium EDTA (APPENDIX 2A) 250 mM sodium chloride Store indefinitely at room temperature (e.g., 23°C). Current Protocols in Nucleic Acid Chemistry

Triethylamine/1-methyl-2-pyrrolidinone/triethylamine trihydrofluoride, 24/46/30 (v/v/v) To a stirred solution of 12 mL triethylamine and 23 mL 1-methyl-2-pyrrolidinone (Aldrich), add dropwise 15 mL triethylamine trihydrofluoride (Aldrich). Continue stirring until the solution is homogeneous (1.4 M fluoride ion). Store in 1-mL aliquots up to 6 months at −20°C. CAUTION: Wear gloves when working with fluoride solutions. This solution is derived from Wincott et al. (1995).

COMMENTARY Background Information The authors’ goal was to use site-specific disulfide cross-linking to measure long-range conformational dynamics within a large (>300nt) catalytic RNA molecule (ribozyme). Existing methods of forming disulfide bonds had relied on chemical oxidation of two thiols, a reaction that is prohibitively slow in the absence of a catalyst. Redox-active metal complexes, such as copper(II) phenanthroline, are effective oxidants (Patai, 1974) but are incompatible with RNA because they promote oxidative cleavage of the ribose backbone (Chen and Sigman, 1988). In addition, large structured RNAs often contain binding sites for multivalent metal ions, further complicating analysis. Sulfoxide catalysis has been used to form disulfides in RNA (Sigurdsson et al., 1995), but the experimental conditions required (≥50% dimethyl sulfoxide) are not compatible with a folded and active ribozyme structure. The new disulfide cross-linking procedure (Cohen and Cech, 1997) that is presented in this unit can be applied to large RNAs and allows kinetic characterization of the cross-linking reaction. Incorporation of cross-linking precursors S.1 and S.2 within RNA relies on the presence of a 2′ amine, incorporated by use of the corresponding phosphoramidite during solid-phase RNA synthesis (Verheyden et al., 1971; Imazawa and Eckstein, 1979; Benseler et al., 1992). Modification of the 2′ amine is then performed after the RNA is deprotected, labeled with 32P, or otherwise manipulated. For example, a 2′ amine is compatible with T4 DNA ligase; ligation of a 41-nt synthetic RNA containing a 2′ amine to a 269-nt transcribed RNA as described by Moore and Sharp (1992) provided a route for preparing a 310-nt ribozyme containing the alkyl thiol S.2 (Cohen and Cech, 1997). Thiol-disulfide interchange chemistry (Fig. 5.1.1) proceeds under mild conditions (aqueous

solution, pH ∼7) without an oxidative catalyst and has been used to measure conformational dynamics between helical domains within a 310-nt ribozyme (Cohen and Cech, 1997). Association of the substrate domain containing S.1 with a set of ribozymes, each containing S.2 at a different position, afforded substrateribozyme cross-linking representing interhelical displacements of at least 50 Å. The kinetic profile of the cross-linking revealed the distribution of motions between the two domains. Cross-linking was achieved under a variety of experimental conditions (temperature 30° to 50°C; pH 6 to 8; NaCl concentration 0 to 1.0 M; MgCl2 concentration 0 to 100 mM) and allowed preparation of a series of conformationally constrained substrate-ribozyme complexes.

Critical Parameters Optimization of the modification reaction of the 2′ amine with S.3 revealed a strong temperature dependence: the efficiency of the reaction was significantly lower at 23°C than at 37°C. Allow ample time for the solution to warm to 37°C before addition of S.3. The absolute rates of cross-linking can be controlled by manipulating the pH of the crosslinking reaction (∆log k/∆pH ~1). The pH may require optimization. For example, if the crosslinking precursors S.1 and S.2 are in close proximity within the associated RNA molecule, then a reaction pH of 8 may result in cross-linking that is too fast to measure (150 nmol involve rather large volumes for lyophilization and can become quite costly. NOTE: If the cyclized oligonucleotides are to be transcribed, all solutions and equipment coming into contact with DNA must be autoclaved to be free of RNase contaminants. All water used should be ultrapure (at least distilled and deionized). BASIC PROTOCOL

DOUBLE-HELICAL SPLINT COMPLEX–ASSISTED ENZYMATIC CYCLIZATION OF OLIGONUCLEOTIDES USING T4 DNA LIGASE The basic protocol describes synthesis and purification of ssDNA circles using a complementary short ssDNA template (a splint) to juxtapose the reactive 3′-hydroxyl and 5′-phosphate groups of precursor segments for ligation by T4 DNA ligase. The protocol is applicable to circles in the size range of 28 to at least 188 nucleotides. This “one-pot” synthesis consists of a two-step ligation of two approximately equal-length precursor segments, which are first enzymatically ligated into a full-length precircle, without isolation, and subsequently enzymatically cyclized into a ssDNA circle (Fig 5.2.3). Circles 140 nt by ligation of more than two precursor segments; see Strategic Planning for more information. The products are separated by preparative denaturing polyacrylamide gel electrophoresis (PAGE), and circularity is confirmed by endonuclease cleavage. CAUTION: Acrylamide and N,N′-methylenebisacrylamide are neurotoxins. Preparation of solutions with these compounds should be performed in a well-ventilated fume hood, and extreme precautions taken to minimize contact with solids or solutions. NOTE: To synthesize larger quantities, the first ligation reaction can be scaled up. The second ligation reaction can also be scaled up, but multiple reaction tubes should be used rather than a larger-scale one-pot synthesis. NOTE: For cyclization using a single precursor segment, skip steps 4 to 7 (i.e., omit the first ligation and proceed directly from DNA quantitation to the second ligation described).

Chemical and Enzymatic Methods

Materials 20 nt splint ssDNA(s) Precursor segments containing 5′-phosphate and 3′-hydroxyl groups Ultrapure (e.g., distilled, deionized) water 2× ligation buffer: 20 mM MgCl2/100 mM Tris⋅Cl (pH 7.5) 1 M dithiothreitol (DTT; APPENDIX 2A) 25 mM and 100 mM adenosine triphosphate (ATP) 400 U/µL T4 DNA ligase (New England Biolabs) 1× TBE (APPENDIX 2A) 10% denaturing polyacrylamide gel mix (see recipe or purchase) 10% (w/v) ammonium persulfate (APS) TEMED or TMEDA (Life Technologies) Formamide loading buffer (see recipe)


0.2 N NaCl 5× S1 nuclease buffer: 50 mM NaCl/50 mM NaOAc/5 mM ZnCl2, pH 4.6 332 U/µL S1 nuclease (from Aspergillus oryzae; Amersham Pharmacia Biotech) Stop solution (see recipe) Stains-all dye solution (see recipe) UV spectrophotometer 1.5-mL microcentrifuge tubes and cap locks, autoclaved 15-mL screw-top centrifuge tubes, autoclaved 90°C heat block or thermal cycler Glass wool Dialysis tubing, MWCO 1000 (e.g., SpectraPor) Gel electrophoresis equipment: Vertical (sequencing) gel stand 2000-V power supply Glass plates: 13 × 15.5 cm, 13 × 16.5 cm, 6.5 × 15 cm, and 6.5 × 16 cm Gel combs: 1.5-mm-thick with 1.5-cm-wide wells and 0.4 mm thick with 5-mm-wide wells Spacers: 1.5 mm thick and 0.4 mm thick UV shadow box or light source Lyophilizer or SpeedVac Saran Wrap (or other UV-transparent plastic wrap) Razor blades, sterile Glass stir rod 50-mL filter tubes with 0.45-µm cellulose acetate filter (e.g., Spin-X II, Corning Costar) Obtain splint DNAs 1. Construct or purchase ssDNAs of the desired splint and precursor-segment sequences. Splint sequences should bridge, by 10 nt on each side, the gap between precursor segments for ligation (Fig. 5.2.3A). For ssDNAs purchased commercially, splints should be fully deprotected and precursor segments should be fully deprotected and 5′-phosphorylated. Alternatively, ssDNAs may be constructed on a DNA synthesizer using the standard DNA cycle and a 0.2-µmol scale; see Support Protocol 1 for synthesis of modified precursor segments, see Support Protocol 4 for deprotection, and see Support Protocol 5 for (optional) purification.

Quantitate DNA 2. Redissolve each DNA in 1.00 mL water. Prepare a 100× dilution sample by combining 10.00 µL of this stock solution with 990 µL water. Synthesized DNA after deprotection and lyophilization should be powdery and white and should readily redissolve in water. If the solution does not appear homogeneous, mild heating (∼60°C) and/or brief sonication (1 min) can be used to help dissolve the DNA. Insoluble material may cause some turbidity but is not cause for concern.

3. Use a properly calibrated UV spectrophotometer to obtain a measurement of the absorbance at 260 nm (A260) of the 100× dilution sample prepared in step 2. Calculate the concentration (c) of DNA for the stock sample prepared in step 2 in moles per milliliter (recalling the dilution factor from step 2 of 102) using Beer’s Law: A260 = ε × b × c b is generally 1, so c (in mol/mL) = (A260 × 102)/ε

Methods for Cross-Linking Nucleic Acids


Extinction coefficients (ε) for each sequence can be calculated using the nearest-neighbor method (Borer, 1985). See also UNIT 7.3. IMPORTANT NOTE: If a single precursor segment is to be used (i.e., synthesis of a ssDNA circle 140 nt), use all precursor segments and all but one splint in step 4. It may not matter which splint is chosen for the first ligation here; however, more is required for the second ligation, so if one is limiting, it should be used here. If the combined volume of DNA (two precursor segments and splint) exceeds 119 mL, one or more DNAs must be lyophilized to reduce the volume.

5. Calculate the volume of water to add according to the following equation: volume water = 250 µL − (125.0 µL + volume DNA [step 4] + 5.40 µL) Add the calculated amount of water to the microcentrifuge tube, vortex briefly, and secure closed with a cap lock. The 250 mL represents the total reaction volume, 125.0 mL represents the 2× ligation buffer, volume DNA represents the combined volumes of precursor segments and splint, and 5.40 mL represents volumes of reagents to be added in step 7.

6. Incubate the capped tube 10 min in a 90°C heat block. After 10 min, turn the heat block off and allow it to slowly cool, with its contents, to room temperature (∼25°C). Step 6 takes ∼2 to 2.5 hr to complete. Alternatively, a PCR thermal cycler can be used. Set the cycler to heat for 10 min at 90°C, then slowly cool (0.5°C/min) to room temperature. The slow cooling allows the splints to bind their complementary targets efficiently, aligning reactive ends for ligation by the enzyme.

7. When the block and its contents have reached room temperature (∼25°C), add 2.50 µL of 1 M DTT, 1.00 µL of 25 mM ATP, and 1.88 µL of 400 U/µL T4 DNA ligase to the tube. Gently invert the tube and let stand 4 to 6 hr at room temperature. Final concentrations: 10 mM DTT, 100 mM ATP, 3 U/mL T4 DNA ligase. Perform second ligation

8. In autoclaved 15-mL centrifuge tube, combine 5.0 mL of 2× ligation buffer with either 167.0 µL of reaction from step 7 (2⁄3 vol) or 10 nmol full-length synthesized precursor segment (if cyclization is from a single precursor segment), plus 30 nmol of the remaining splint (splint 2). Final concentrations: 10 mM MgCl2, 50 mM Tris⋅Cl, 1 mM precircle, 3 mM splint 2, 1.2 mM splint 1. IMPORTANT NOTE: For circles >140 nt, the final splint is added in this step.

Chemical and Enzymatic Methods

The 167.0 mL of reaction from step 7 represents, in theory, a maximum yield of 10 nmol precircle from ligated precursor segments. Alternatively, if the circle size is 70 nt) should be run almost off the gel.

19. Disassemble the gel apparatus. Sandwich the prep gel between two pieces of Saran Wrap. View the gel under ultraviolet light with fluorescent white background (glass silica plates work well) to visualize the bands. The reaction lanes should show several bands. Figure 5.2.4 is a representation of a typical gel, with lane 6 containing second ligation reaction mixture. Depending on how far the gel is run, the splints (lane 1) will probably be run off, but the precursor segments may be visible and their identity can be confirmed by their co-migration with the precursor-segment markers (step 2 solutions) run alongside (lanes 2 and 3). The next band up (migrating more slowly) from the precursor segments should be the precircle band; this is confirmed by its co-migration with the largest band from the first ligation reaction (step 7) run alongside (lane 4). Although the secondary structure present in many DNA circles may cause them to migrate more slowly than their linear counterparts (Serwer and Allen, 1984), identification of the circle (lane 8) can be difficult, as there may be one or more products in between the linear precircle and monomer circle. These are likely to be ligation of an odd number of precursor segments (lane 7), and results are especially complex if precursor segments are of different sizes. Migration of circles is not always consistent and can be temperature dependent. Therefore, most bands migrating more slowly than confirmed precircle should be cut out for analysis.

Isolate purified DNAs after gel electrophoresis 20. Cut out the desired band with a clean, sterile razor blade and place the pieces in a 15-mL centrifuge tube. Using a glass stir rod, crush the gel pieces thoroughly. Combine crushed pieces with ∼5 to 10 mL of 0.2 N NaCl to make an easily shaken slurry. Repeat, in separate tubes, for each band of interest isolated. Shake each slurry for ∼12 hr. Precircle and higher bands should be isolated for reasons stated in step 19. Be sure all crushed gel pieces are transferred to the slurry. Conversion of precircle to circle may appear high by UV shadowing (step 19), but isolated yields after step 22 are often 3000 Ci/mmol) T4 polynucleotide kinase 1:1 phenol/chloroform (APPENDIX 2A) Chloroform 4.0 M ammonium acetate Taq DNA polymerase TE buffer, pH 8.0 (APPENDIX 2A) PCR amplification buffer (APPENDIX 2A) 2× formamide loading buffer (APPENDIX 2A) 15 × 17–cm denaturing polyacrylamide gel (APPENDIX 3B) Thermal cycler Phosphor imager plate and phosphor imager Additional reagents and equipment for quantitation of DNA (e.g., CPMB APPENDIX 3D), end-labeling of DNA (e.g., CPMB UNIT 3.10), phenol/chloroform and chloroform extraction of DNA ( APPENDIX 2A), PCR amplification (e.g., CPMB Chapter 15), and denaturing polyacrylamide gel electrophoresis (APPENDIX 3B) 1. Quantitate DNA by UV absorption assuming that A260 of 1.0 indicates ∼37 µg/ml of single stranded DNA. Also see, e.g., CPMB APPENDIX 3D.

2. Label the 5′ end of the 3′ PCR primer with [γ-32P]ATP by preparing the following reaction mixture. For 30-ml reaction (volume of reaction and concentration of DNA and [g32 P]ATP will vary depending on application): 50 mM Tris⋅Cl, pH 7.5 10 mM MgCl2 5 mM DTT 1 to 50 pmol dephosphorylated DNA, 5′ ends 50 pmol (150 µCi) [γ-32P]ATP 50 µg/ml BSA 20 U T4 polynucleotide kinase Incubate 60 min at 37°C, then stop reaction by adding 1 µl of 0.5 M EDTA. Phenol/chloroform and chloroform extract the labeled oligonucleotide (see recipe for phenol/chloroform/isoamyl alcohol in APPENDIX 2A), and precipitate by adding an equal volume of 4.0 M ammonium acetate and 2 vol ethanol. Microcentrifuge to collect the pellet, remove the supernatant, and redissolve the labeled DNA pellet in 10 µL of TE buffer, pH 8.0.

Combinatorial Methods in Nucleic Acid Chemistry


This procedure ensures that most of the unincorporated label remains in the supernatant.

3. Incubate ∼50 pmol of labeled primer with a 2- to 5-fold molar excess of pool in a 50-µL extension reaction, under the same conditions that will be used in the final amplification, in a thermal cycler as follows (see, e.g., CPMB UNIT 15.1 for PCR). a. Denature and anneal the primer and template DNA in PCR amplification buffer (usually 94°C for the denaturation step and ∼50°C for the annealing step). b. Add Taq or other DNA polymerase (scaled to the anticipated enzyme concentration to be used in the large-scale amplification), then ramp the temperature to 72°C for 20 min. It may be useful to take time points to determine whether the reaction has gone to completion.

c. Finally, terminate the reaction by the addition of 2× formamide loading buffer. 4. Heat the extension reaction to 90°C for 3 min and load the reaction on a 15 × 17–cm denaturing polyacrylamide gel with appropriate radiolabeled size markers. Electrophorese until the dye is at or near the bottom of the gel, but do not let the radiolabeled primers run off. It is also useful to load a separate well with an aliquot of the primer alone. Choose an acrylamide percentage that allows efficient separation of small primers from larger extended products.

5. Dry and expose the gel to a phosphor imager plate. Using a phosphor imager, quantify the control primer band and the extended product band. There may be a smear leading up to the extended band. One should use one’s best judgement in determining how much near-full-length material will be included in the quantitation. Calculate the percent extension by dividing counts of labeled, extended product by counts of labeled primer. Percent extension for a gel-purified ssDNA pool can range from 10% to 30%. The complexity of the pool is then the yield (determined in step 1) multiplied by the extension efficiency (percent extension determined above). If the complexity of the pool is insufficient for planned experiments, then the pool must be resynthesized. SUPPORT PROTOCOL 2

DETERMINING THE POOL BIAS

SUPPORT PROTOCOL 3

SMALL-SCALE PCR OPTIMIZATION OF POOL AMPLIFICATION

Following extension, the reaction should be repeated using a cold primer and the nonradioactive double-stranded DNA pool should be amplified in a PCR reaction, cloned (e.g., using a TA cloning kit from Invitrogen), and individual members sequenced to determine the degree of partial or completely randomness. The cloning step could also be carried out following PCR optimization (see Support Protocol 3). From 20 to 30 clones should be sequenced to determine the base composition of the starting pool. The random region should be composed of roughly 25% of each base. A pool with the random region skewed toward one or more bases (>30%) should be resynthesized.

To enhance yield and further avoid bias, the amplification conditions for a pool should be optimized prior to carrying out a large-scale amplification. Moreover, since amplifying a pool is costly in terms of both time and money, any optimization of the PCR should first take place on a small scale. The more involved large-scale amplification can then be carried out with confidence.

DNA Pools for In Vitro Selection


Materials dNTPs (APPENDIX 2A) Taq DNA polymerase (e.g., Boehringer Mannheim) PCR amplification buffer containing 1.5 mM Mg2+ (APPENDIX 2A) dsDNA mass markers (e.g., Life Technologies) 4% Nu Sieve agarose gell (FMC Bioproducts) Thermal cycler Densitometer Additional reagents and equipment for agarose gel electrophoresis (e.g., CPMB UNIT 2.5) 1. Carry out a 0.1 mL PCR reaction using 2 nM of synthetic pool oligonucleotide as template, 2 µM primers, and PCR buffer with 1.5 mM magnesium. Use the manufacturer’s suggested quantity of Taq (e.g., 2.5 U of Boehringer Mannheim Taq) in a reaction containing 200 µM dNTPs. A suggested temperature regime is: 10 to 15 cycles: 2 min 1 min 3 min

95°C (denaturation) 55°C (annealing) 72°C (extension).

After 10 to 15 cycles of amplification, check the length and purity of the amplified DNA on a 4% Nu Sieve agarose gel in 1× TBE buffer (e.g., CPMB UNIT 2.5). Annealing temperature may need to be adjusted to as low as 45°C depending on primer composition (e.g., for a small or AU-rich primer). A 0.1 mL reaction typically yields ∼1 ìg, but the amount can vary from 0.1 to 10 ìg. A fuzzy band may indicate that too many cycles of PCR have been carried out. In this case, set up the reaction again and perform fewer cycles.

2. Dilute the double-stranded PCR DNA product 1:128, and repeat the PCR reaction, removing a 5- to 10-µL aliquot during the last 10 sec of the cycle-7 extension step. Serially dilute the amplified product 1:2, 1:4, ... 1:128. Electrophorese all of the samples on a large agarose gel. Note that it is quite difficult to accurately pipet solutions at 72°C. It may therefore be desirable to pipet an amount slightly larger than that intended for use in the serial dilution.

3. Calculate the average PCR efficiency by identifying to what extent the cycle-7 PCR reaction is the result of progressive doublings of the original synthetic DNA. Determine which dilution lanes lack detectable DNA. The largest dilution that lacks detectable DNA is also the dilution that is a minimum estimate of the number of doublings. For example, if the 1/64 dilution is the largest dilution without detectable DNA, this implies that 6 “doublings” of the synthetic DNA yielded at least 64-fold more DNA. This is expressed as follows:

(average efficiency)no. of theoretical doublings (i.e., PCR cycles) = fold increase in DNA Thus, if 7 cycles of PCR were performed, then the average number of doublings per cycle is ∼1.81 [from (∼1.81)7 = 64].

4. Modulate PCR conditions to enhance PCR efficiency. If the pool’s average number of doublings per cycle is 5 ìM are generally not helpful). It may be useful to scan both above and below 2.5 ìM in 0.5-ìM increments. Magnesium concentration affects both primer annealing and the fidelity of Taq (which decreases with increasing magnesium concentration). Starting at the magnesium supplied in the PCR buffer (usually 1.5 mM), scan in 1-mM increments toward 5 mM as a maximal concentration. DNA denaturation at temperatures above 95°C is usually impractical since this greatly reduces Taq’s half-life. While other thermostable polymerases can be more resistant to higher temperatures, they usually have a lower extension efficiency and are more expensive than Taq. Annealing temperatures are dependent upon both primer sequence and length. The primer annealing temperatures should already be known from the primer design p rocess, o r may be calculated via an algorithm that can be found at http://paris.chem.yale.edu/extinct.html. This algorithm takes into account nucleotide composition, stacking energies (according to Turner’s rules), and empirical data. An annealing temperature ∼5°C less than the calculated annealing temperature is a good place to begin optimization. The amplification is more efficient at a lower annealing temperature, but mispriming and secondary structural problems are more pronounced. Higher temperatures improve the specificity, but decrease the overall yield of the reaction. To determine the optimum annealing temperature for a given primer and magnesium concentration, one should scan in both directions around the annealing temperature in 5°C increments. Finally, extension temperatures are modulated by the properties of Taq, which will extend (although inefficiently) at temperatures as low as 65°C. When extending at temperatures above Taq’s optimum temperature (70° to 75°C) somewhat more polymerase may be required; scanning of the enzyme quantity should be done in 2.5-U increments. However, too much Taq may be harmful to structured single-stranded nucleic acids (Lyamichev et al., 1993).

5. Confirm the results of the extension reaction described in Support Protocol 1 by the optimization method as follows. After optimizing pool PCR conditions for >1.8 average number of doublings per cycle, determine the pool complexity by performing another 0.1-ml PCR reaction with 2 nM of the original, synthetic pool oligonucleotide under the now optimized reaction conditions. After 7 or more cycles of PCR, perform agarose gel electrophoresis on serial dilutions of the PCR reaction adjacent to serial dilutions of dsDNA mass markers. Calculate the amount of amplified DNA using either a densitometer or by estimating which dilutions are most similar. Calculate the approximate pool complexity as follows: g of PCR DNA afterN cycles of PCR = g of starting extendable ssDNA g avg no. of doublings per cycle (see step 4) g of starting extendable ssDNA 330 g / mole × (no. of bases in fullM length product) = mol starting extendable ssDNA mol starting extendable ssDNA × (6.02 × 1023) = molecules of star ting extendablessDNA molecules of starting extendable ssDNA = fraction of extendable ssDNA starting molecules fraction of extendable ssDNA × no. of synthetic pool molecules = pool complexity DNA Pools for In Vitro Selection

9.2.16 Supplement 8


PCR efficiency should be optimized to balance the average number of doublings per cycle against the total reaction volume. A pool of 1 × 1015 molecules (∼1.7 × 109 mol) at a starting template concentration of 2 nM will require 0.85 L for amplification. Therefore, it is greatly desirable to amplify the pool at the highest template concentration that still gives a reasonable number of doublings per cycle. The amplification should generate at least 8 copies of pool DNA if the pool complexity is to be archived and preserved (see Basic Protocol 2).

LARGE-SCALE PCR AMPLIFICATION OF POOL DNA Very long and complex pools often require PCR amplification on a multiple-milliliter scale. Large-scale PCR differs from conventional PCR in that it is typically conducted in water baths using 15 mL, 17 × 120–mm, screw-capped (Sarstedt) thermostable tubes to accommodate the larger volumes. Amplification reactions of up to 2.5 L have been carried out in this way. Medium-scale amplifications can sometimes be carried out in thermal cyclers that can accommodate multiple samples (e.g., 96-well PCR plates).

BASIC PROTOCOL 2

Materials Purified ssDNA pool and primers EDTA 1:1 phenol/chloroform (APPENDIX 2A) Chloroform 4 M ammonium acetate Ethanol TE buffer, pH 8.0, containing 50 mM of a salt such as potassium chloride Thermal cycler or three water baths (one must be a circulating water bath) 96-well PCR plate or 13-mL thermostable tubes (Sarstedt) Thermometer Styrofoam racks Spectrophotometer or fluorometer Additional reagents and equipment for PCR amplification ( CPMB UNIT 15.1; see Support Protocol 3 for determination of conditions on a small scale) and phenol/chloroform and chloroform extraction of DNA (APPENDIX 2A) Plan the reaction Since large-scale reactions are quite expensive in terms of nucleotides and enzyme, preparedness and planning for the large-scale amplification cannot be overemphasized. Primers 2′-NH2 (Brieba

and Sousa, 2000). Several 2′-modified ribonucleoside triphosphates are now commercially available (Jena BioScience, Trilink); Y639F T7 polymerase is available from Epicentre Technologies. The replacement of the ribose 2′-OH group with other chemical moieties interferes with the primary mechanism for nuclease cleavage of RNA, attack of the 2′-hydroxyl on the bridging phosphate. However, substitutions on the backbone, such as replacing the phosphate with a phosphorothioate, have also been shown to increase oligonucleotide stability in the presence of nucleases (Zon and Geiser, 1991). An additional benefit is that phosphorothioate nucleotides have been shown to be incorporated into an elongating transcript by T7 polymerase with little or no increase in KM (Griffiths et al., 1987). While DNA is not as vulnerable to hydrolysis as RNA, it is nonetheless susceptible to cleavage by a variety of deoxyribonucleases and phosphodiesterases. The stability of DNA can also be increased by the incorporation of phosphorothioate nucleotides, and these can be readily incorporated by Taq DNA polymerase (Nakamaye et al., 1988). These modified ribonucleoside and deoxyribonucleoside triphosphates are also commercially available (NEN Life Sciences, Sigma, Life Technologies). Nucleic acid selection experiments have generated a wide variety of binding species (aptamers) and catalysts (ribozymes and deoxyribozymes). However, nucleic acids are by and large not as functional as proteins: aptamers bind their ligands less well than antibodies, for the most part, while ribozymes are slower than

9.6.16 Supplement 7


protein enzymes, for the most part. While it is possible that these limitations are a function of the relative newness of aptamer and catalyst selections relative to a well-established field like protein engineering, it is also possible that functional nucleic acids are inherently limited by the diversity of the genetic alphabet. The binding interactions and chemical reactions performed by a nucleic acid biopolymer may be constrained by the functional groups they contain. The ability to expand the functional groups available to a DNA or RNA polymer through the incorporation of modified nucleotides could potentially open up realms of chemistry and binding interactions that were previously inaccessible. Researchers are currently trying to assess this hypothesis and demonstrate the utility of modified bases in functional nucleic acids. For example, the first RNA capable of catalyzing the formation of carbon-carbon bonds utilized 5-pyridylmethylcarboxamid-UTP in place of UTP (Tarasow et al., 1997). When the most active isolate from this selection was transcribed with unmodified UTP, it was inactive. Similar results were obtained for a ribozyme that catalyzed amide bond formation (Wiegand et al., 1997) whose activity was dependent on the incorporation of 5-imidizole-UTP. Additionally, two sequence-specific RNase deoxyribozymes have been selected that were dependent on the incorporation of a 5-imidizole-TTP (Santoro et al., 2000), and both 8-2-(4-imidazolyl)ethylamino-2′-dATP and 5-(3-aminoallyl)-2′-dUTP (Perrin et al., 2001). However, none of the sequences, motifs, or activities found in these selection experiments was directly compared with ribozymes that contained canonical nucleotides and that were sieved from the same pool using the same selection conditions. Therefore, at present time, it is unclear whether modified nucleotides truly improve RNA catalysis. Surprisingly, though, there are at least a few counterexamples that would suggest that modified nucleotides do not greatly contribute to binding or catalysis relative to natural nucleotides. Santoro and colleagues (Santoro and Joyce, 1997; Santoro et al., 2000) selected deoxyribozymes with RNA hydrolysis activity from different aliquots of the same, unamplified random sequence pool. The selection performed with natural nucleotides produced a much faster catalyst. Ultimately, it is unknown whether this indicates the superiority of natural nucleotides for this pool and this function, or whether the fraction of the original pool used

for the selection of the natural catalyst just contained a “jackpot” sequence. An additional example of a catalytic activity selected for with both modified and unmodified pools is the Diels-Alder catalyzing ribozymes (Tarasow et al., 1997; Seelig and Jaschke, 1999, respectively). The modified selection yielded a catalyst with a kcat/KM of ∼4 M−1sec−1, while the unmodified selection yielded a catalyst with a kcat/KM of 167 M−1sec−1. However, these selections were performed by different research groups with different pools, and thus are not directly comparable. For a final example, see Critical Parameters, discussion of multiple approaches.

Critical Parameters Optimizing preparative reactions A good portion of this protocol has been devoted to testing reactions and optimizing conditions. Although this process will likely take several days, it is time wisely spent. The first round of selection is by far the most important, as this is the round when selection conditions will query the greatest number of possible answers. Concomitantly, creating the largest possible number of unique sequences (1013 to 1015) for the first round of selection is a critical task. However, because creating a large library requires a large amount of effort and large volumes of relatively expensive reagents, initial optimization and practice should always be performed on a small scale. Similarly, it is expected that only a small number of the input molecules will survive any given round; therefore, it is essential that these successful sequences be efficiently carried to the next round. Inefficient reverse transcription or amplification reactions may lead to a selection for molecules that can be efficiently replicated (amplification artifacts) rather than to molecules that are highly functional. Archiving reactions In vitro selection experiments are particularly grueling because the ultimate outcome, the successful isolation of binding species or catalysts, may not be known for days, weeks, or even months. This is especially true when working with modified nucleotides, because of the strong possibility that one or more amplification reactions or selection steps will not work at some point during the course of the selection. Therefore, it is desirable to keep an archived copy of each round. After the initial round, there should be multiple copies of each winning

Combinatorial Methods in Nucleic Acid Chemistry


Supplement 7

sequence present in the selected population. Thus, it is not necessary to use all of the RNA, cDNA, or double-stranded DNA template in a given selection or amplification step. Rather, some portion of the pool should be saved at any convenient point (e.g., as a double-stranded PCR product). In the event an unsuccessful round of selection should occur, it will not be necessary to repeat the entire experiment. Rather, selection can continue from the last successful round, with a minimal loss of time and effort.

In Vitro Selection Using Modified or Unnatural Nucleotides

Understanding modified nucleotide chemistry The incorporation of additional functional groups into the context of an RNA backbone is expected to increase the diversity of its available interactions, both with itself and with its desired substrates. Thiol groups, for example, can participate directly in catalysis as nucleophiles. Additionally, disulfide bonds could be formed intramolecularly between thiols. This may add to the structural diversity of nucleic acids, normally limited to hydrogen bonding, salt bridges with metals, and stacking interactions. Charged groups, such as a lysine-like side chain, could potentially add to the structural repertoire of nucleic acids by allowing the formation of electrostatic interactions and salt bridges. The inclusion of chemical moieties with a pKa closer to neutrality, such as an imidizole group, is also expected to benefit nucleic acid chemistry. Natural nucleotide bases contain no functional groups with unperturbed pKa values between 3.5 and 9.2, inherently limiting the proton “push-pull” chemistry found in so many protein enzymes. The introduction of these and other functional groups can also potentially increase the abilities of nucleic acids to bind metals. While attempts to associate nucleotide functionality with binding or catalytic properties are at best still guesswork, it is nonetheless true that failing to appreciate the chemical properties of modified nucleotides can potentially adversely affect how a selection experiment proceeds. For example, the inclusion of thiolated nucleotides can potentially lead to the unwanted formation of disulfide-bonded products if care is not taken to include a reducing agent in enzyme and/or selection reactions. For example, deoxyribozyme ligases that form an unnatural internucleotide linkage between a 5′-iodinated pool and an oligonucleotide substrate with a 3′ phosphorothioate have been selected from random sequence pools (Levy and Ellington, 2001a,b).

Unless reducing agents are kept in the selection reaction, the oligonucleotide substrates will dimerize, reducing their effective concentration. Similarly, the inclusion of modified nucleotides that have altered pKas could very easily change the pH of a concentrated stock solution, and care should be taken to make sure that all such stocks are at the desired pH and appropriately buffered. The accidental alteration of pH in enzyme or selection reactions can of course lead to decreases in product yield. Finally, the inclusion of modified nucleotides that have new metal binding or chelating properties may alter the available metal concentrations in an enzyme or selection reaction. In particular, the chelation of magnesium can lead to large changes in the efficiency of product formation by many different polymerases. Multiple approaches As has been apparent throughout, the in vitro selection of functional nucleic acids that contain modified nucleotides is still in its infancy, and thus there are few hard and fast rules regarding what will and will not work. Because of this, researchers need to be somewhat versatile in their approach to selection experiments involving modified nucleotides. If a given approach does not work, this does not mean that the selection experiment inherently has no chance of working, but instead indicates that it may be necessary to alter one or more parameters. In particular, to improve the incorporation of modified nucleotides into a pool there are three variables that should be adjusted as needed. First, the buffer conditions for incorporation can be adjusted. Padilla and Sousa (1999) have systematically investigated several buffer conditions that aid the incorporation of nucleotides modified specifically at the 2′ position. These authors find better incorporation upon supplementing a transcription reaction with 0.5 mM MnCl2, 1 U/µL pyrophosphatase, and either 8 mM spermidine for plasmid templates or 8 mM spermine for short DNA templates. Second, different polymerases clearly have different potentialities for the incorporation of modified nucleotides. For example, the Benner group has tested several thermostable DNA polymerases for their ability to incorporate a variety of modified nucleotides (Lutz et al., 1998, 1999). Finally, the sequence of the pool itself can have a surprisingly large effect on selection experiments. The authors originally selected a relatively fast ribozyme ligase from an RNA pool that contained 90 random

9.6.18 Supplement 7


sequence positions (Robertson and Ellington, 1999; Robertson et al., 2001). A variant of this pool was generated in which only four positions in the constant region were altered, substituting a GAAA tetraloop in place of a UUCG tetraloop. Selection experiments for ribozyme ligases were initiated from amplified aliquots of this pool; one aliquot was used for the selection of ribozymes containing canonical nucleotides, other aliquots of identical complexity were used for the selection of ribozymes containing one of three different modified nucleotides (Robertson, 2001). After six rounds of selection and amplification, all of the pools had collapsed to contain a relatively few winning sequences. However, the rates of all of these winning sequences were remarkably slow (500-fold slower) compared to the ribozyme ligase originally selected from the slightly different pool. There was no discernable difference between the ribozymes that contained canonical or modified nucleotides; they were all very slow. The authors hypothesized that the pools had somehow become artificially narrowed during the course of the selection, and therefore repeated the selection experiments with new aliquots of the original, amplified pool. Again, only very slow ribozymes were obtained.

be incorporated between rounds in a selection to more efficiently explore sequence space. However, it should be noted that in the latter case, one does not want the mutation rate to be so high that a few of each winning sequence from one round do not survive unchanged into the next round. One benefit of using modified nucleotides rather than standard mutagenic PCR is that it is easy to control the amount of mutation introduced into the sample simply by adjusting the relative rate of modified to unmodified nucleotide in the amplification process. As such, a higher rate of mutation can be achieved than with mutagenic PCR, which would be particularly useful for diversifying an initial pool. Zaccolo et al. (1996) describe such a system using both a purine and pyrimidine analog in a PCR reaction, and have quantitated the frequencies of each base transition. It should be mentioned that when using mutagenic modified nucleotides, one would not want them to be included in the active pool, as their decreased fidelity would make it less likely that they would be in the same position during the next round. For example, if a DNA selection were being performed, a second amplification of the pool would be required using only natural nucleotides.

Modified nucleotides and mutations Although a selection starts with a large number of sequences, this number is usually a small fraction of the total number of sequences possible for the length of the random region. Additionally, with each round, the number of sequences within the pool is diminished. As such, it may be useful to explore the sequence space around the selected winners in order to discover functional variants. To some extent, this occurs in any selection due to the inherent mutations that arise in the amplification process; however, this background mutation rate is small, and one may wish to increase the frequency of mutations. For these purposes, the mutagenic potential of a nucleotide analog that serves as a nonspecific template can be used to increase the diversity of a pool. Similar to mutagenic PCR (UNIT 9.4), these techniques can be used at the outset of a selection to mutate an existing ribozyme or aptamer for either optimization or reselection for altered specificity. For example, Kore et al. (2000) have used modified nucleotides to create a degenerate pool based on the hammerhead ribozyme, from which they selected a variant that cleaves at an alternate sequence. Additionally, a mutagenic step can

Anticipated Results As with any selection experiments (UNITS 9.3 is virtually impossible to anticipate the outcome of any given experiment. This is especially true when considering the incorporation of modified nucleotides, since relatively few selection experiments have so far been carried out with modified residues. However, if the protocols for the production of RNA and DNA pools that have been outlined are followed, it can be anticipated that it should be possible to generate and purify upwards of at least 1014 different nucleic acid sequences (∼10 µg) that contain a particular modified nucleotide. For successful selection experiments, the population should be sieved by a factor of 100 to 1000 each round. That is,

Current Protocols in Nucleic Acid Chemistry

Nucleic Acid Transfection (Topics in Current Chemistry)

The Nucleic Acid Protocols Handbook

Nucleic Acid Transfection (Topics in Current Chemistry 296)

New Nucleic Acid Techniques

Giáo trình Nucleic Acid

New Nucleic Acid Techniques

Drug-Nucleic Acid Interactions

Current Protocols in Food Analytical Chemistry

Nucleic Acid Targeted Drug Design

Principles of Nucleic Acid Structure

Protocols for Nucleic Acid Analysis by Nonradioactive Probes

Principles of Nucleic Acid Structure

Protocols for Nucleic Acid Analysis by Nonradioactive Probes

Fluorescent Energy Transfer Nucleic Acid Probes: Designs And Protocols

Nucleic Acid and Peptide Aptamers Methods and Protocols

Species Diagnostics Protocols: Pcr and Other Nucleic Acid Methods

Current Protocols in Cytometry

Current Protocols in Toxicology

Nucleic Acid Targeted Drug Design

Principles of Nucleic Acid Structure

Principles of Nucleic Acid Structure

Molecular Beacons: Signalling Nucleic Acid Probes, Methods, and Protocols

Molecular Beacons. Signalling Nucleic Acid Probes, Methods and Protocols

Fluorescent Energy Transfer Nucleic Acid Probes. Designs and Protocols

Species Diagnostics Protocols: PCR and Other Nucleic Acid Methods

Protocols for Nucleic Acid Analysis by Nonradioactive Probes

Amino Acid Analysis Protocols

Amino Acid Analysis Protocols

Current Protocols in Molecular Biology

Current Protocols in Nucleic Acid Chemistry

Nucleic Acid Transfection (Topics in Current Chemistry)

The Nucleic Acid Protocols Handbook

Nucleic Acid Transfection (Topics in Current Chemistry 296)

New Nucleic Acid Techniques

Giáo trình Nucleic Acid

New Nucleic Acid Techniques

Drug-Nucleic Acid Interactions

Current Protocols in Food Analytical Chemistry

Nucleic Acid Targeted Drug Design

Principles of Nucleic Acid Structure

Protocols for Nucleic Acid Analysis by Nonradioactive Probes

Principles of Nucleic Acid Structure

Protocols for Nucleic Acid Analysis by Nonradioactive Probes

Fluorescent Energy Transfer Nucleic Acid Probes: Designs And Protocols

Nucleic Acid and Peptide Aptamers Methods and Protocols

Species Diagnostics Protocols: Pcr and Other Nucleic Acid Methods

Current Protocols in Cytometry

Current Protocols in Toxicology

Nucleic Acid Targeted Drug Design

Principles of Nucleic Acid Structure

Principles of Nucleic Acid Structure

Molecular Beacons: Signalling Nucleic Acid Probes, Methods, and Protocols

Molecular Beacons. Signalling Nucleic Acid Probes, Methods and Protocols

Fluorescent Energy Transfer Nucleic Acid Probes. Designs and Protocols

Species Diagnostics Protocols: PCR and Other Nucleic Acid Methods

Protocols for Nucleic Acid Analysis by Nonradioactive Probes

Amino Acid Analysis Protocols

Amino Acid Analysis Protocols

Current Protocols in Molecular Biology

Recommend Documents