Plasmids for Therapy and Vaccination
Edited by M. Schleef
Plasmids for Therapy and Vaccination Edited by M.Schleef
@WILEY-VCH Weinheim - New York - Chichester - Brisbane - Singapore - Toronto
Editor. Or. Martin Schleef Plasmid Factory GmbH & Co. KG Meisenstrasse 96 D-33607 Bielefeld Germany
1
This book was careful produced. Nevertheless, authors, editor, and publisher do not warrant the information contained therein to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items inadvertently be inaccurate. Libraly of Congress Card No: applied for British Library Cataloguing-in-Publication Data: A catalogue record for this book is available from the British Library Die Deutsche Bibliothek -
CIP-Cataloguingin-PublicationData A catalogue record for this publication is available from Die Deutsche Bibliothek
0 WILEY-VCH Verlag GmbH, D-694G9 Weinheim (Federal Republic of Germany), 2001 All rights reserved (including those of translation into other languages). No part of this book may be reproducted in any form nor transmitted or translated into machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not specifically marked as such, are not to be considered unprotected by law. Printed in the Federal Republic of Germany Printed on acid-free paper. Composition Hagedorn Kommunikation, Viernheim Printing betz-dmck, Darmstadt Bookbinding Wilh. Osswald + Co., Neustadt ISBN
3-527-30269-7
I"
Preface Gene therapy and vaccination with nucleic acids is one of the most striking innovations in medical and veterinary sciences. The treatment of a disease on the level of genes rather than on the phenotype level is a promising option to obtain enhanced therapeutics. The pharmaceutical application of genetic material - modified so as to become the so-called active pharmaceutical ingredient (API) - requires developments in biotechnology and pharmacy to obtain systems for the transfer and expression of the API at the right place and time, resulting in the appropriate effects on the organism or cell. The vector for such gene transfer may be as simple as a short piece of DNA. For years now geneticists have been working with plasmids, and the recent developments in vector design and manufacturing point to potent therapeutics and vaccines. In chapters 1 and 2 the background of plasmids is summarized and their structures are presented, since these well-known molecules still have had an unknown potency decades after their discovery. Their different size and structure turned out to be of importance for their function, and the characterization of an API made from DNA required a complete set of quality assurance in manufacturing and quality control. These aspects are explained in chapters 11 and 12. Detailed examples of clinical applications are presented in chapters 4-6, providing an overview of the wide range of preventive and medical applications using plasmid DNA. In chapters 3 and 5, recent overviews on DNA vaccination are presented which should help to oversee the rapid development and published literature in the application of nucleic acid vaccines. Regulatory and quality assurance aspects of such new drugs are considered in chapter 13. Chapters 7-9 describe modified vector systems based on plasmids, as well as the potency of genomic research and vector design by informatics. The link between genomics and the function of genomic information necessarily requires nucleic acids. One example of veterinary health care is presented in chapter 10. The development of veterinary vaccines still requires some effort and the recent worldwide discussion on BSE in (at least) cattle makes obvious that health care in animals is also health care for humans. However, the treatment of animals grown for food production raise further questions on the aspect of genetically modified food.
VI
I
Preface
Plasmid production on an industrial scale is necessarily linked to the question of the expected market size. Chapter 14 guides through the history of gene drug development and the expectations in todays’ and future pharmaceutical markets. The development and preclinical testing of therapeutic or preventive plasmid pharmaceuticals is right at its beginning. The vision of individualized medical treatment might become reality with this type of drug system. The option to have major improvement in comparison to conventional drugs attracts research, industry and politics and is a challenge for all disciplines involved - from genomics to clinical application. Finally, I wish to thank Karin Dembowsky from Wiley-VCH for her continuous collaboration with this project and all authors who contributed to this book, to make it what I hope it will be: A recent overview on the field and a guide through an area of useful innovation. Bielefeld, January 2001
Martin Schleef
Contents Preface
V
List of Contributors XIV
1
The Biology of Plasmids
1 2 2.1 2.2 2.3 2.4 2.4.1 2.4.2 3 3.1 3.2 3.3 3.4 4 4.1
Introduction: What are plasmids? 1 General properties of plasmids 2 Plasmid replication and its control 3 The molecular basis of incompatibility 6 Plasmid inheritance 7 Mechanisms of plasmid spread 8 Conjugation in gram-negative bacteria 9 Conjugation in gram-positive bacteria 11 Plasmid-encoded phenotypes 11 Bacteriocin production and resistance 12 The Ti plasmids 12 Heavy metal resistance 14 Other phenotypical traits 16 The clinical importance of plasmids 18 The spread of antibiotic resistance and the evolution of multiple antibiotic resistance 18 Transfer of antibiotic resistance genes 19 Mechanisms of antibiotic resistance 21 Bacterial virulence genes 23 Plasmid cloning vectors 24 Perspectives 27 References 28
4.2 4.3 4.4 5 6
2
1 2 3 4
1
DNA 29 Introduction 29 Topological structures of plasmids 30 Supercoiling of DNA 32 DNA intercalating dyes 32 Structures o f Plasmid
Vlll
I
Contents
5 5.1 5.2 5.3 5.4 6
Analysis of plasmid structures 33 Electron microscopy (EM) 33 Agarose gel electrophoresis (AGE) 35 Capillary gel electrophoresis (CGE) 38 Analytical chromatography 42 Conclusion 41 References 42
3
Genetic Vaccination with Plasmid Vectors 45
1 2 2.1 2.2 2.3 3 4
Introduction 45 Vector design 45 Plasmid DNA 45 Construction of simple transcription units 47 Construction of complex transcription units 47 Strategies for DNA delivery 49 Priming humoral and cellular immune responses by DNA vaccines 50 Experimental strategies facilitated by DNA vaccination 53 Unique advantages of DNA vaccination 54 DNA vaccines in preclinical animal models 56 DNA vaccines to control infectious diseases 56 Therapeutic tumor vaccines 57 Autoimmune disease 57 Treatment of allergy by therapeutic DNA vaccination 57 Proposed clinical applications of DNA vaccines 58 Risks of nucleic acid vaccination 59 Future perspectives 59 References 61
5 G 7 7.1 7.2
7.3 7.4 8 9 10
4
A Liposomal iNOS-Gene Therapy Approach to Prevent Neointimal Lesion Formation in Porcine Femoral Arteries 75
1 2 2.1 2.2 2.3
Introduction 75 Results and discussion 77 Therapeutic plasmid 77 The gene therapy product has a clinically acceptable format 78 Efficient gene transfer was established in a minipig femoral artery injury model 79 Transfection efficiency is dose dependent 81 Non-viral iNOS gene transfer efficiently inhibits neointimal lesion formation 82 Summary and perspectives 84 References 85
2.4 2.5
3
Contents
5 1 1.1 1.2 1.3 1.4 1.5 2 2.1 2.1.1 2.1.2 2.1.3 2.1.4 2.1.5 2.1.6 3 3.1 4
6
1 2 3 3.1 3.2 3.3 3.4 3.4.1 3.4.2 3.4.3 3.4.4 4 5
7
1 2 2.1 2.2 2.3 2.4 2.5
87
lmmunotherapy o f Chronic Hepatitis B by pCMV-SZ.S DNA Vaccine Introduction 87 Hepatitis B: the disease 87 Hepatitis B: treatments 87 Hepatitis B: immune response to infection 88 What are DNA vaccines? 88 Which DNA vaccines for hepatitis B? 89 DNA vaccines for the prevention of hepatitis B 90
The mouse model 90 Humoral response 90 Cell-mediated response 92 Mechanisms of DNA-induced immune response to HBsAg 93 The primate model 94 DNA-based vaccination of chimpanzees against HBV 95 Neonatal immunization 96 DNA-based vaccination for chronic HBV infections 96 HBsAg transgenic mice as a model for HBV chronic carriers 96 Clinical trials of DNA vaccines 98 References 99 pSG.MEMRAP - A First Generation Malaria DNA Vaccine Vector Parasite life cycle and impact of malaria 103 Concept of vaccination against malaria 104
First-generation plasmid: pSG.MEPfTFL4P 106 Vector backbone 106 Insert 206 Production and formulation 111 Preclinical testing of pSG.MEPfTRAP 112 Toxicity studies 113 Biodistribution 113 Stability testing 113 Potency testing 114 Regulatory aspects 114 Future perspectives 124 References 1 1 5 Polyvalent Vectors for Coexpression of Multiple Genes 129 Introduction 119 Polycistronic expression vectors 121 Mechanisms of translation initiation 121 Characteristics of IRES elements 124 Application of IRES elements in cells and animals 126
Polycistronic vector systems 128 Expression properties of IRES vectors 130
103
I
IX
X
I
Contents
3 3.1 3.2 3.3 4
Bidirectional promoters 130 Natural bidirectional promoters 130 Artificial bidirectional promoters 131 Combining polycistronic and bidirectional expression 132 Perspectives 133 References 133
8
Form Follows Function: The Design o f Minimalistic lmmunogenically Defined Gene Expression (MIDGE') Constructs 139 The problem 139 The solution 141 MIDGE -the concept 142
1 2 2.1 2.2 2.3 2.4 2.5
9
1 2 3
3.1 3.2 3.3 3.4 3.5 4 4.1 4.2 4.3 4.4
Simple MIDGE 143 Smart MIDGE 143 Applications 145 Practical aspects of vector sequence design 145 References 146 Synthetic Genes for Prevention and Therapy: Implications on Safety and Efficacy o f DNA Vaccines and Lentiviral Vectors Introduction 147
147
Paradoxon: HIV-derived vaccines and gene delivery systems 151 Synthetic genes: Novel tools contributing to the understanding of HIV replication 152 Construction of a synthetic, HIV-1 derived gag gene 152 Codon usage modification in the gag gene abolishes Rev dependency and increases expression yields 153 Codon usage modification in the gag gene increases nuclear RNA stability and promotes constitutive nuclear translocation 154 Codon usage modification in the gag gene alters the nuclear export pathway of otherwise CRMl dependent RNAs 1% Codon usage modification increases RNA stability, modulates nuclear RNA export and increases translational efficiency 157 Synthetic genes: Implications on the development of safe and effective DNA vaccines 158 Safety issues to be considered for DNA vaccine development 158 Codon optimization of a gag-specific candidate vaccines results in increased antibody responses 159 Enhanced in vitro cytokine release of splenocytes from mice immunized with synthetic gag plasmid DNA 160 Induction of CTL responses in mice immunized with the modified Gag expression plasmids 161
Contents
5 5.1 5.2 5.3 5.4 5.5 G
Synthetic genes: Implications on the development of safe lentiviral vectors for gene delivery into quiescent cells 162 Safety issues to be considered for lentiviral vector development 163 Construction and characterization of synthetic gugpol expression plasmids 163 Production of lentiviral vectors using synthetic gugpol genes 164 Transduction o f non-dividing cells 164 Absence of replication-competent recombinants (RCRs) 165 Future perspectives 165 References 166
10
Plasmids in Fish Vaccination
1 2 3 3.1 3.2
Introduction 169 Fish 270 Fish immunology 171 Innate defence mechanisms 171 Adaptive defence mechanisms 171 Vaccination o f fish 173 Nucleic acid vaccination of fish 174 Plasmid constructs used in fish studies 175 Routes of plasmid administration 176 Intramuscular injection of plasmid DNA 2 76 Other routes of plasmid administration 178 Fate of injected plasmid DNA 178 Magnitude, distribution and longevity of expressed antigen Responses offish to injection with plasmid DNA 181 Inflammatory responses 181 Avirulent antigens 181 Virulent antigens 182 Regulatory issues and future directions 187 References 188
4 5 G 7 7.1 7.2 8 9 10 10.1 10.2 10.3 11
169
11
Plasmid Manufacturing - An Overview
1 2 2.1 2.2 3 3.1 3.1.1 3.1.2 3.1.3
Introduction 1% Structure of nucleic acids 194 Brief structural description of DNA and RNA structures 294 DNA supercoiling 196 Plasmid DNA Manufacturing 199 Major impurities and main product specifications 201 Host nucleic acids 202 Proteins 203 Endotoxins 203
193
2 79
I
XI1
I
Contents
3.2 3.2.1 3.2.2 3.2.3 3.3 3.3.1 3.3.2 3.3.3 3.4 4
12
1 2 3 4 5 5.1 5.2 5.2.1 5.2.2 5.2.3 5.2.4 5.2.5 5.3 5.3.1 5.3.2
G
13
1 2 3 4 5
G 7
Factors influencing the production of plasmid DNA: Some considerations on the upstream processing and fermentation stages 204 The plasmid vector 204 The bacterial host strain 205 Plasmid fermentation 20G Downstream processing of plasmid DNA 208 Cell lysis 210 Pre-chromatography processing: clarification and concentration 213 Chromatographic processing: purification of supercoiled plasmid DNA 214 Purification Strategies 228 Concluding remarks 229 References 23 1 Quality Control of pDNA Introduction 237
237
Characterization and quality control of pDNA 237 Validation of test procedures 239 GLP 240 Detailed description of the characterization of pDNA (final product) 242 Sterility 241 Purity 243 Content 243 Homogeneity 244 Host DNA 246 Host cell protein impurities 248 Endotoxins 248 Identity 249 Restriction analysis 249 Determination of the DNA sequence 251 Conclusion 251 References 252 From Research Data to Clinical Trials Introduction 255 Approaching regulators 256
255
Vaccine manufacture 256 Predinical safety testing 257 Clinical trials 258 Approval for clinical trials 259 Clinical trial applications in Germany 259 References 260
Contents
262
14
Market Potential for DNA Therapeutics
1 2 3
Definition of biotechnology 261 History 262 Process of pharmaceutical development 263 Human society and technical revolution 264 From sequence to product: Applications of biotechnology 265 Milestones in biotechnology: The Human Genome Project 267 The future is now: Examples for existing therapeutic approaches using gene products 268 Gene therapy in cardiovascular diseases 268 Legal aspects of gene technology and pDNA derived products 270 The extension of patent law to living creatures and their components 270 Impacts of biomedical patents 272 Resisting corporate ownership of life forms 272 Health care in the light of biotech is different in Europe and US 274 Biotech in the US from an economical point of view 274 Emergence of new companies 274 Biotech in Germany 275 Who is the health care industry and their clients, a paradigm? 276 Health care industry share prices 277 HMO enrolment rose 278 Limitations to the access of health care services 278 US uninsured population rose 279 Economical evaluation of the biotech marked in the future 279 Conclusion 280
4 5 5.1 5.2
5.2.1 6 6.1 6.2 6.3 7 7.1 7.2 7.3 8 8.1 8.2 8.3 8.4 9 10
Index
281
I
XIII
I
List of Contributors Jens Alfken MOLOGEN Berlin Fabeclcstrasse 30 D-14195 Berlin Germany Alfl<
[email protected] Ruth Baier QIAGEN GmbH Max-Volmer-Strasse 4 D-40724 Hilden Germany
[email protected] Kurt Bieler Institut fur Medizinische Mikrobiologie und Hygiene Universitat Regensburg Franz-josef Straufi Allee 11 D-93053 Regensburg Germany
[email protected] Joaquim M. S. Cabral Centro de Engenharia Bioldgica e Qumica Instituto Superior Tcnico Av Rovisco Pais 1049-001 Lisboa Portugal gnmferreira @ist.utl.pt
Klaus Cichutek Abteilung fur Medizinische Biotechnologie Paul-Ehrlich-Institut Paul-Ehrlich-Strasse51-59 D-63225 Langen Germany
[email protected] Juergen Dobmeyer IMS HEALTH GmbH & Co. OHG Hahnstrasse 30-32 D-60528 Frankfurt/Main Germany dobmeyer@ grnx.net Rita Dobmeyer Johann-IClotz-Strasse3 D-GO5 28 Franltfurt/Main Germany dobmeyeret-online.de Julia Dorge Cardion AG Max-Planck-Strasse 15a D-40699 Erlcrath Germany
doergee cardion-ag.de
xv
XVI
I
List of Contributors
Andreas Fels NewLab Bioquality AG Max-Planck-Strasse 15A D-40699 Erlcrath Germany fels @newlab.de Guilherme N. M. Ferreira Centro de Engenharia Biologica e Qumica Instituto Superior Tcnico Av Rovisco Pais 1049-001 Lisboa Portugal
[email protected] Eiwin Flaschel Universitat Bielefeld Technische Fakultat AG Fermentationstechnik D-33501 Bielefeld Germany
[email protected] Karl Friehs Universitat Bielefeld Technische Fakultat AG Fermentationstechnik D-33501 Bielefeld Germany kfr @fermtech.techfak.uni-bielefeld.de Sarah C. Gilbert Wellcome Trust Centre for Human Genetics Roosevelt Drive Headington Oxford OX3 7BN UK
[email protected] Hansjorg Hauser Abteilung fur Genregulation and Differenzierung GBF - Gesellschaft fur Biotechnologische Forschung Mascheroder Weg 1 D-38124 Braunschweig Germany hha @ gbf.de Angela Heischmann Cardion AG Max-Planck-Strasse 15a D-40699 Erkrath Germany
[email protected] Andreas Herrmann Cardion AG Max-Planck-Strasse 15a D-40699 Erkrath Germany
[email protected] Adrian V. S. Hill Level 7 NDM John Radcliffe Hospital Headington Oxford OX3 9DU
uIZ
adrian.hil1@ Molecular-Medicine. oxford.ac.uk Simon R.M. Jones Department of Fisheries and Oceans Pacific Biological Station 3190 Hammond Bay Road Nanaimo, British Columbia V9R 5K6 Canada
[email protected] List of Contributors
Claas Junghans MOLOGEN Berlin Fabeckstrasse 30 D-14195 Berlin Germany
[email protected] Andreas Muhs Cardioii AG Max-Planck-Strasse 15a D-40699 Erltrath Germany muhs @cardion-ag.de
Sven A. IZoenig-Merediz MOLOGEN Molecular Medicines S. L. CNBF Instituto de la Salud Carlos I11 28220 Majadahonda Madrid Spain I 100 mIU per mL) were induced in a female chimpanzee with 2 mg DNA. However, at least two boost injections of DNA were required to prevent antibody levels from diminishing over time. Nevertheless, extremely high antibody titers were ultimately attained (> 14,000 mUI per mL), and high levels were sustained for at least one year. In a second chimpanzee (male) injected with a lower dose of DNA (400 ,ug) anti-HBs were only detected after the second injection of DNA. Although antibody titers of 60 mUI per mL were attained, these were transient even after three DNA boosts. Despite loss of detectable anti-HBs, this chimpanzee showed a strong anamnestic response to injection of recombinant HBsAg given one year after the initial DNA injection. As observed in mice, DNA-based vaccination of chimpanzees induces HBsAg-specific antibodies. Initially of IgM isotype, these subsequently shift to IgG antibodies, which are predominantly IgG1. The antibodies induced are predominantly against the preS2 domain, but are also specific to the group and subtype determinants of the S protein (Davis et al., 199Gb). Levels of antiHBs and protection afforded correlate well for human and chimpanzee. Based on titers higher than 10 mUI per mL observed in one chimpanzee, this animal would have been protected against HBV infection. Regarding the second animal, the strong and rapid anamnestic response to recombinant HBsAg suggests that this animal would probably resist hepatitis. Regarding the safety issues, no abnormalities were found in the weekly serum assays (routine hematology and liver enzymes), nor were any other untoward reactions noted.
I
95
96
I
M a w L o u i s e Mrchel
2.1.6
Neonatal immunization
Two chimpanzees born to HBsAg-naive mothers were immunized on the day of birth and at 6 and 24 weeks with 500 pg pCMV-Sl.SZ.S. Although only transient and low levels of anti-HBs were detected both animals were protected from subsequent live viral challenge (100 CID). The immunized animals developed anamnestic anti-HBs antibody responses; but neither developed detectable HBsAg or antibody to the core protein. In both animals small amounts of HBV DNA were produced concurrent with the anamnestic response indicating that sterilizing immunity was not achieved (Prince et al., 1997).
3
DNA-based vaccination for chronic HBV infections 3.1
HBsAg transgenic mice as a model for HBV chronic carriers
Owing to the endogenous synthesis of antigen, the processing into relevant epitopes, and the induction of CD8' CTLs, DNA vaccines may be useful for the treatment of individuals chronically infected with HBV As a model to study the possibility of inducing an immune response in HBV chronic carriers that could control the infection, we used transgenic mice that constitutively express the HBsAg in their liver only. The transgene which consists of one copy of the HBV genome, but with the core gene deleted, is expressed before birth under the control of an endogenous HBV promoter (Babinet et al., 1985; Farza et al., 1987). This results in the secretion of large amounts of HBsAg into the serum without antibody production and without apparent liver pathology. These mice may represent a model for HBV chronic carrier infected at birth without HBV replication, generally defined as asymptomatic carriers. Using the pCMV62.S plasmid encoding the small and the middle HBV envelope protein, rapid production of anti-HBs antibodies was induced in these transgenic mice which in turn induce the clearance of circulating HBsAg. Anti-preS2 and anti-HBs antibodies induced after a single immunization in HBsAg-Tg mice reach levels as high as those induced in normal C57BL/G mice after DNA-based immunization. The isotypes of the induced antibodies is comparable in Tg and in non-Tg mice and include IgM and then IgG; predominantly IgG2b and IgGl with some IgG3. Elimination of HBsAg concomitantly with anti-HBs induction occurred in most of the mice in 4-8 weeks and persisted for at least 20 weeks without any further injection of DNA. This clearance is not only due to the neutralization of antigen by antibodies, but is also correlated with a decrease or a disappearance of the HBV messenger RNA from the liver (Mancini et al., 1996b).Down-regulation of transgene expression is likely mediated by HBsAg-specificT cells as shown by adoptive transfer experiments. These experiments, involving fractionated HBsAg primed spleen cells obtained from DNA-immunized mice into Tg mice, showed that both CD4' and CD8' T cells were able to control transgene expression even in the absence of antibody production. Interest-
5 lrnrnumtherapy of Chronic Hepatitis B by pCMV-S2.S D N A Vaccine
ingly, this regulation of transgene expression occurred without increase of liver enzymes and without histological evidence of liver necrosis (Mancini et al., 1998). This suggests that the T cell effect is probably mediated through a non cythopathic mechanism. Results from HBsAg-Tg mice that are knock-out for the y-interferon receptor gene indicate that the regulation of the HBV envelope messenger RNA was mediated by type 1 cytoltines produced by the activated T lymphocytes. Indeed, in vitro activation of spleen cells from DNA-immunized mice with HBV antigens resulted in predominant production of IFN-y, TNF-a and low levels of IL-2, but not IL-4 or IL-10. It was shown in similar Tg mouse models that HBV-specific CTL can abolish viral replication or transgene expression in the hepatocytes by noncytopathic mechanisms that are mediated by IFN-y and TNF-a (Guidotti et al., 1996). The antiviral activity of the cytokines operates probably through the degradation of HBV mRNA in the liver following the induction of endoribonuclease activity cleaving the viral RNA immediately adjacent to a stem loop structure (Heise et al., 1999a, b). The mechanisms by which DNA-based immunization can break B and T cell tolerance, or anergy, in transgenic mice is still unresolved. Systemic injection of recombinant HBsAg alone is not sufficient to induce B or T cell responses able to control transgene expression, unless it contains heterologous B and T cell epitopes or is derived from an heterologous subtype of HBV (Mancini et al., 1993). In contrast, intramuscular injection of a plasmid encoding HBsAg and secreting a very limited amount of antigen in viuo is sufficient to induce a response in these Tg mice. This could be related to the presentation of different antigenic peptides to the immune system following either DNA-based or systemic immunization, or to the direct activation of antigen presenting cells via the CpG motifs presents in the plasmid backbone (Klinman et al., 1997). The non-cytolytic antiviral process described in the Tg mice, if confirmed in humans, could be of importance for the control of the viral infection without liver disease, but could also result in an incomplete elimination of the virus from the liver, thereby promoting viral persistence. These results have been further confirmed in duck, a model for DHBV infection. Three day old infected ducklings develop chronic infection with aspects mimicking what occurs in the HBV chronic carrier state. In these experiments chronically infected ducks were immunized four times with DNA plasmids encoding the large envelope protein. A significant decrease or even elimination of viral replication was observed in immunized ducks. This effect was persistent since no rebound in viral replication was observed after cessation of the treatment. More importantly, analysis of intrahepatic viral DNA showed the clearance of the cccDNA pool in two animals which is tightly associated with rebounds of viral replication after anti-viral treatments (Rollier et al., 1999). These results point to the possibility of designing more effective ways for the prevention and treatment of HBV infections. However, many basic questions of this vaccine approach still need to be answered, namely safety issues associated with the injection of DNA vaccines and potential adverse effects of a too strong cytotoxic response against the infected liver cells.
I
97
98
I
M a r k L o u i s e Miche/
4
Clinical trials o f DNA vaccines
Research efforts to develop DNA vaccine products have progressed to the clinical evaluation of several vaccine candidates. These clinical trials assess safety and immunogenicity and whether different routes and schedules of administration can improve immune responses. Vaccine candidates currently under study include those for hepatitis B, herpes simplex-2, HIV-1, influenza and malaria. Unique safety concerns include the potential for genomic integration, biodistribution, tolerance and autoimmunity. One major concern was that after injection the DNA would integrate into the recipient host’s chromosome leading to mutagenesis and potentially insertional carcinogenesis. Animal studies involving plasmid DNA injections have shown that any mutations from a potential integration event would be infrequent and result in a calculated mutation rate much lower than the spontaneous mutation rate for the mammalian genome (Martin et al., 1999). A second concern was that the immunization with plasmid DNA would induce anti-DNA antibodies and this would accelerate the development of autoimmune diseases. Animal studies have shown non-significant increase in anti-DNA antibodies and no evidence of autoimmune disease induction or acceleration after administration of plasmid DNA injections (Mor et al., 1997). In a Phase I safety and tolerability clinical trial of a malaria DNA vaccine, 28 healthy, malaria-nalveadult volunteers were immunized with plasmid DNA encoding a malaria antigen. It was found that intramuscular administration of three doses of up to 2,500 pg of plasmid DNA was safe, well tolerated and not inducing any obvious hematological or biochemical abnormalities or anti-DNA antibodies. This was the first demonstration of the induction of CD8’ CTLs by DNA vaccines in healthy nalve humans (Wang et al., 1998). However, DNA vaccination via intramuscular injection failed to induce detectable malaria-specificantibodies in any of the volunteers (Le et al., 2000). Two clinical trials of DNA vaccines for the treatment of human immunodeficiency virus-1 (HIV-1)infections were also reported. A DNA-based vaccine containing HIV-1 envelope and rev genes was tested for safety and host immune response in 15 asymptomatic HIV-infectedpatients. Vaccine administration induced no local or systemic reactions and no laboratory abnormalities were detected (no anti-DNA antibodies or muscle enzyme elevations). Injection of plasmid DNA resulted in antibody against the HIV envelope and some increases were observed in envCTL and lymphocytes proliferative activity (MacGregor et al., 1998; Ugen et a]., 1998). Another trial was performed in 9 asymptomatic HIV-infected patients using DNA encoding regulatory genes of HIV-1 (neJ;tat, rev). DNA immunization induced antigen-specific T cell proliferation which persisted up to nine months after the last injection and CD8-mediated cytolytic activities, but did not by itself reduce viral load (Calarota et al., 1998; Calarota et al., 1999). A phase I clinical trial of DNA vaccine for hepatitis B was conducted in seven adult healthy volunteers using a plasmid encoding HBsAg and the “gene gun” as delivery system. Extremely low doses of DNA (0,25 pg) coated onto gold
5 lmmunotherapy of Chronic Hepatitis B by pCMV-S2.S D N A Vaccine
beads were administered two times into the sltin. Results were very disappointing, since the only one of the six volunteers who developed high titers of HBs antibodies may have had previous exposure to hepatitis B (Tadtet et al., 1999). Nevertheless, CTL induction specific for HBV envelope has been reported for some HLA-A2 patients receiving HBsAg-encoding DNA via gene gun immunization (Swain et al., Development and clinical progress of DNA vaccines, Langen, Germany, 1999, personal communication). Altogether, the results from the first clinical trials show that DNA vaccines appear to be well tolerated. The aim of the studies described above is to develop a therapeutic vaccine against persisting HBV infection. Patients suffering from chronic-associated liver disease with the risk of developing hepatocellular carcinoma will greatly benefit from the availability of a therapeutic vaccine against HBV. Vaccination may be the therapeutic procedure with the lowest cost and the greatest potential benefit. As chronic carriers are the major source of spreading the disease, control of infection in this population will be a major advance in controlling overall HBV infection. Acknowledgements
I thank all of my colleagues who have collaborated with me on the studies reported here: H. L. Davis, D. Loirat, S. Le Borgne, M. Mancini, E. Malanchi.re, M. Schleef, R. G. Whalen. I am particularly grateful to P. Tiollais for his constant support. I greatly appreciate the help of Dr. H. Farza and Dr. P. Townsend for the critical
reading of this manuscript.
References DNA vaccination in HIV-1-infected patients, ALBERTI,A,, CAVALLETIO, D., PONTISSO,P., Lancet 351 (9112), 1320-1325. CHEMELLO, L., TAGARIELLO, G., BELUSSI,F. CALAROTA, S. A., LEANDERSSON, A. C., BRAIT, (19881, Antibody response to pre-S2 and G., HINIWLA,J., KLINMAN, D.M. et al. hepatitis B virus induced liver damage, (1999), Immune responses in asymptomatic Lancet 331, 1421-1424. HIV-1-infected patients after HIV-DNA BABINET,C., FARZA,H., MORELLO, D., HADimmunization followed by highly active CHOUEL, M., POURCEL, C. (1985), Specific antiretroviral treatment, I. bnrnunol. 163 (4), expression of hepatitis B surface antigen 2330-2338. (HBsAg) in transgenic mice, Science 230, CHISARI,F. V. (1995), Hepatitis B vims 1160-1163. immunopathogenesis, Annu. Rev. Irnrnunol. BERTOLETII, A., D’ELIOS,M. M., BONI, C., 13, 29-60. DE CARLI,M., ZIGNEGO,A. L. et al. (19971. Different cytolune profiles of intraphepatic T COUILLIN,I., POL, S., MANCINI,M., DRISS,F., BRECHOT, C. et al. (1999), Specific vaccine cells in chronic hepatitis B and hepatitis C therapy in chronic hepatitis E: induction of vims infections, Gastroenterology 112 (l), T cell proliferative responses specific for 193-199. envelope antigens [published erratum CALAROTA, S., BRAT^, G., NORDLUND, S., appears in J Infect Dis 1999 Nov;180(5):1756], HINKULA, J., LEANDERSSON,A. C. el al. (1998),Cellular cytotoxic response inducedby /. Infect. Dis. 180 (l),15-26. I
,
I
99
100
I
MarkLouise Michel
DAVIS,H. L., MANCINI,M., MICHEL, M.-L., R. G. (1996a),DNA-mediated WHALEN, immunization to hepatitis B surface antigen: Longevity of primary response and effect of boost, Vaccine 14 (9), 910-915. DAVIS,H. L., MCCLUSICIE, M. J., GERIN,J. L., PURCELL, R. H. (1996b), DNA vaccine for hepatitis B: evidence for immunogenicity in chimpanzees and comparison with other vaccines, Proc. Natl. Acad. Sci. U S A 93, 72137218. DAVIS,H. L., MICHEL, M.-L., WHALEN, R.G. (1993), DNA based immunization for hepatitis B induces continuous secretion of antigen and high levels of circulating antibody, Hum. Mol. Genet 2, 1847-1851. DAVIS,H.L., MICHEL, M.L., MANCINI, M., SCHLEEF,M., WHALEN, R. G. (1995), DNAbased immunization overcomes H-2 haplotype-restricted non-responsiveness to hepatitis B surface antigen in mice, in: Vaccines 95, Molecular Approches to the Control of Infectious Diseases (Ginsberg, H. S., Brown, F., Chanocck, R. M., Eds.), pp. 111-116. Cold Spring Harbor Laboratory Press, New York. S. C. DAVIS,H. L., MILLAN, C. L., WATKINS, (1997), Immune-mediated destruction of transfected muscle fibers after direct gene transfer with antigen-expressingplasmid DNA, Gene Ther. 4 (3), 181-188. DONNELLY, J. J., ULMER, J. B., SHIVER, J. W, LIU, M.A. (1997), DNA vaccines, Annu. Rev. Immunol. 15, 617-648. M., FARZA, H., SALMON,A,-M., HADCHOUEL, MOREAU, J.-L., BABINET,C. et al. (1987), Hepatitis B surface antigen gene expression is regulated by sex-steroids and glucocorticoids in transgenic mice, Proc. Natl. Acad. Sci. U S A 84, 1187-1191. D. (1991), Assembly of hepadnaviral GANEM, virions and subviral particles, in: Current Topics in Microbiology and Immunology: Hepadnaviruses, Molecular Biology and Pathogenesis Vol. 168 (Mason, W. S., Seeger, C., Eds.), pp. 61-83. Springer-Verlag, Berlin, Heidelberg. D. (1996), Hepadnaviridae: The Viruses GANEM, and their Replication, Third Edition ed. Virology 2. 2 Vols. (Fields, B. N., Ed.). Lippincott-Raven, Philadelphia. GUIDOTTI,L. G., CHISARI,F.V (1999), Cytoltine-induced viral purging - role in viral pathogenesis, Curr. Opin. Microbiol. 2 (4), 388-391.
GUIDOTTI, L.G., ISHIIWWA,T., HOBBS, M.V., B., SCHREIBER,R., CHISARI, F.V. MATZICE, (1996), Intracellular inactivation of the hepatitis B virus By cytotoxic T lymphocytes, Immunity 4 (l),25-36. HEISE,T., G U I D O ~L.G., I , CAVANAUGH, V J., F.V. (1999a), Hepatitis B virus CHISARI, RNA-binding proteins associated with cytoltine-induced clearance of viral RNA from the liver of transgenic mice, /. Virol. 73 (l), 474-481. HEISE,T., GUIDOITI,L. G., CHISARI, F. V. (1999b),La autoantigen specifically recognizes a predicted stem-loop in hepatitis B virus RNA, J . Virol. 73 (7), 5767-5776. J. H., DI BISCEGLIE,A.M. (1997), HOOFNAGLE, The treatment of chronic viral hepatitis, N . Engl. /. Med. 336 (5), 347-356. J. H., LAU, D. (1997),New theraHOOFNAGLE, pies for chronic hepatitis B, J. Viral Hepat. 4 (Suppl. l),41-50. Hut, J., MANCINI, M., LI, G., WANG,Y., P., MICHEL, M. L. (1999), ImmuniTIOLLAIS, zation with a plasmid encoding a modified hepatitis B surface antigen carrying the receptor binding site for hepatocytes, Vaccine 17 (13-14), 1711-1718. ITOH,Y., TAICAI, E., OHNUMA, H., KITAJIMA, K., TSUDA, F. et al. (1986), A synthetic peptide vaccine involving the product of the pre-S(2) region of the hepatitis B virus DNA: protective efficacy in chimpanzees, Proc. Natl. Acad. Sci. U S A 83, 9174-9178. JUNG, M. C., HARTMANN, B., GERLACH, J.T., DIEPOLDER, H., GRUBER, R. et al. (1999), Vims-specific lympholcine production differs quantitatively but not qualitatively in acute and chronic hepatitis B infection, Virology 261 (2), 165-172. KLINMAN,D. M., YAMSHCHIKOV,G., ISHIGATSUBO, Y. (1997), Contribution of CpG motifs to the immunogenicity of DNA vaccines, /. Immunol. 158 (8), 3635-3639. KRUSKALL,M. S., ALPER,C.A., AWDEH, Z., YUNIS, E. J., MARCUS-BAGLEY, D. (1992),The immune response to hepatitis B vaccine in humans: Inheritance patterns in families, /. Exp. Med. 175, 495-502. LAI, C. L., CHIEN,R. N., LEUNG,N. W., CHANG, T. T., GUAN,R. et al. (1998).A one-year trial of lamivudine for chronic hepatitis B. Asia Hepatitis Lamivudine Study Group, N. Engl. /. Med. 339 (2), 61-68.
5 lmmunotherapy of Chronic Hepatitis B by pCMV-52.5 D N A Vaccine
1101
LE BORGNE,S., MANCINI, M., LE GRAND, R., “nonresponse” to HBsAg, J. Med. Virol. 39, SCHLEEF, M., DORMONT,D. et al. (1998). In 67-74. vivo induction of specific cytotoxic T MARTIN,T., PARKER, S. E., HEDSTROM, R., lymphocytes in mice and rhesus macaques LE, T., HOFFMAN, S. L. et al. (1999), Plasmid immunized with DNA vector encoding an DNA malaria vaccine: the potential for HIV epitope fused with hepatitis B surface genomic integration after intramuscular antigen, Virology 240, 304-315. injection, Hum. Gene %her. 10 (S), 759-768. LE, T. P., COONAN, I 92 %) and the amount of RNA and other low molecular weight nucleic acids. Up to 75 % (w/w) plasmid purity degrees, corresponding to 100-fold purification factors, were achieved after clarifying the plasmid extracts with the chaotropes (Ferreira et al., 1999).However, the use of such substances may require
214
1 additional investment in applicable process elements for their removal from the Guiiherffle N. M . Ferreira et a/.
process stream as well as for quality assurance and may cause safety concerns. The concentration of plasmid DNA by polyethylene glycol (PEG) precipitation is commonly used after the clarification step to further remove small nucleic acids and to reduce the volume of process streams prior to chromatographic purification. While several polymer salt systems were tested at a small scale (Cole, 1991; Lis and Schleif, 1975; Nicoletti and Codorelli, 1993; Ohlsson et al., 1978), low step yields were reported when processing large volumes (Ferreira et al., 1999; Horn et al., 1995). Although plasmid precipitation with PEG is highly empirical (Prazeres et al., 1999), scaling-up of the PEG8000-NaC1 system is the common method for the concentration of plasmid streams prior to the final purification steps (Horn et al., 1995). PEG precipitation also enables a buffer exchange, preparing the plasmid extracts for the next purification step.
3.3.3
Chromatographic processing: purification o f supercoiled plasmid DNA
Purification of supercoiled plasmid DNA has been traditionally a molecular biology method, with laboratory research as the sole intended application for the final product. Saccharose or caesium chloride/ethidium bromide density gradient ultracentrifugation are standard molecular biology methods for plasmid purification (Sambrook et al., 1989). These methods are time consuming, difficult to scale up, they use toxic and mutagenic reagents and are thus unsuitable for the purification of plasmids for clinical purposes (Schleef, 1999).Table 5 compares standard laboratory and clinical manufacturing processes for producing supercoiled plasmid DNA. Chromatography is the method of choice for the large-scale purification of supercoiled plasmid DNA. The size of the target nucleic acid molecules, and by consequence the chemical properties related to the molecule size, such as charge and hydrophobicity as well as the exposition or the accessibility of the nucleotide bases within single- or double-stranded nucleic acid molecules, respectively, and the topological constraints due to supercoiling, are explored in the interaction of nucleic acids with solid supports. In any case, the objective is to selectively isolate and purify plasmid DNA from impurities, such as gDNA, endotoxins, RNA and proteins as well as removing traces of contaminants, such as isopropanol and PEG 8000, introduced in the previous downstream processing steps. Reverse-phase, hydrophobic interaction and ion-pair chromatography
Reverse-phase chromatography explores the hydrophobicity of the nucleotide bases. It has been used in molecular biology to detect partial denaturation on DNA fragments (Huber and Berti, 1996),to purify oligonucleotides and DNA fragments (Floyd et al., 1986; Huber et al., 1995; Patient et al., 1979),to purify plasmid DNA (Best et al., 1981; Colote et al., 1986) and to separate plasmid topoisomers (Kapp and Langowski, 1992). Reverse-phase chromatography is based on the interaction of hydrophobic, nonpolar molecules or regions of the molecules, with non-polar immobilized ligands. Bound molecules are eluted with decreasing polarity gradients by adding organic
7 7 Plasmid Manufacturing
- An
Overview
Comparison of laboratory methods and large-scale pharmaceutical processes for purifying supercoiled plasmid D N A
Table 5.
Process Stea
Laboratorv Method
Large-Scale Process
Cell lysis
RNase, lysozyme
no enzymes only GRAS" reagents
Removal of cell debris
centrifugation
filtration, centrifugation or expanded bed chromatography
Removal of host contaminants: (e.g., RNA, gDNA, proteins, endotoxins)
RNase, proteinase K, organic solvents (phenol, chloroform)
salting out, PEG precipitation
Plasmid enrichment
alcohol precipitation
alcohol precipitation but PEG precipitation is preferred
Plasmid purification
ultracentrifugation (mutagenic reagents) (ethidium bromide) IEC" (gravity flow columns provided in commercial kits) RPC" (organic, toxic solvents)
IECa and or SEC" (use only GRAS reagents)
a
I
I
Abbreviations: G U S , generally regarded as safe; RPC, reverse phase chromatography;IEC, ion-exchange chromatography: SEC, size-exclusionchromatography
modifiers to the column eluent (Brown and ICrstulovic, 1979; Sofer and Hagel, 1997). Therefore, the solutes are eluted in the order of decreasing polarity or increasing hydrophobicity (Brown and ICrstulovic, 1979; Sofer and Hagel, 1997). When the target molecules are polar, which is the case for nucleic acids, however, reverse-phase principles are achieved in a technique termed ion-pair reverse-phase chromatography or simply ion-pair chromatography (Huber, 1998). In this technique, the solute polarity is reduced by adding amphiphilic organic ions to the buffers. These ions establish ionic interactions with the target molecules, forming hydrophobic non-polar ion pairs that bind to reverse-phase resins (Huber, 1998; Meyer, 1994). The major factors affecting the retention time of nucleic acids in reverse-phase columns are size, base composition and secondary structure (Colote et al., 1986). Nucleic acids of increasing size are expected to elute with increasing retention times in reverse-phase chromatography (Patient et al., 1979). However, the effect of the nucleic acid size is not independent on the base composition (Colote et al., 1986; Huber and Berti, 1996; Kapp and Langowski, 1992; Patient et al.,
I
215
216
I
Cuifherme N. M. Ferrefra et a/.
1979). Experimental data on reverse-phase chromatography of nucleic acids indicate that AT-rich DNA binds more tightly to reverse-phase matrices than DNA of equivalent size which is not AT-rich (Patient et al., 1979). The explanation for this result relies in the thermolability of the AT-rich regions (Patient et al., 1979). Partial denaturation of the double helix at the AT-rich locations forming single stranded regions within the molecule leads to the exposition of the bases to the ligands, and thus to an increase of the hydrophobic interaction strength. This hypothesis is consistent with the observation that single-stranded oligonucleotides bind more tightly to reverse-phase matrices than duplex DNA fragments of the same length (Colote et al., 1986; ICapp and Langowsli, 1992; Patient et al., 1979). In ion-pair reverse-phase chromatography an opposite trend has been reported: Partial denaturation of DNA fragments at AT-rich regions decreases the retention factor when compared with native duplex molecules of similar size (Huber, 1998; Huber and Berti, 1996). This apparent contradiction is explained by the mode of molecule-ligand interaction. In ion-pair chromatography nucleic acid molecules interact with the immobilized ligands through amphiphilic ions. Formation of single-stranded regions at the AT-rich locations implies a decrease of the charge density from two to one charges per helix repeat. Thus, less amphiphilic ions bond to these regions, and, consequentially, lower hydrophobicities of the ionic pair are achieved (Huber and Berti, 1996). In either situation, ion-pair (Huber, 1998; Huber and Berti, 1996) or reversephase chromatography (Colote et al., 1986; Patient et d., 1979), the formation of single-stranded regions within the molecule depends on the denaturation conditions, such as, e. g., temperature or formamide concentration. Thus, by selecting and optimizing such conditions two different separation dimensions can be explored (Huber, 1998): one dimension is the size of the molecules and the other is the base sequence. The secondary structure of nucleic acids also influences the retention time in reverse-phase chromatography (Colote et al., 1986; Huber, 1998; ICapp and Langowski, 1992). As mentioned above, single-stranded nucleic acids are more retained in reverse-phase chromatography than equivalent size duplex molecules. On the other hand, the torsion constraints due to supercoiling render the bases more accessible to interact with the stationary phase. Therefore, the retention times are expected to be higher with increasing supercoiling (Colote et al., 1986). This hypothesis was confirmed by the separation of plasmid topoisomers by reverse-phase chromatography,were the elution profiles of a 2.7 kb plasmid followed the order of increasing superhelix density (Kapp and Langowsli, 1992). Reverse-phase and ion-pair chromatography have been used to purify supercoiled plasmid DNA from crude cell lysates (Colote et al., 1986; Green et al., 1997).In either case, low molecular weight RNA, gDNA fragments and linear plasmid DNA were completely separated from the supercoiled plasmid DNA, with the last being more strongly retained (Colote et al., 1986; Green et al., 1997). However, a second pass in the column, i. e. recycling, is suggested due to the failure in completely removing high molecular weight RNA from the plasmid peak (Colote et al.,
1 7 Plasmid Manufacturing
- An
Overview
1986). Genomic DNA and endotoxins remain bound to the column and are removed after sanitizing the column with 1 M NaOH (Green et al., 1997). The need to use organic solvents to elute the plasmid from reverse-phase columns constitutes a disadvantage of these purification techniques (reverse-phase and ion-pair chromatography). Most of the organic solvents are toxic or mutagenic, volatile, and even explosive. Thus, special safety concerns regarding the manufacturing personnel and facility may be required, such as the design of explosion proof facilities and use of appropriate protection masks (Marquet et al., 1995). Although large-scale plasmid purification schemes based on reverse-phase and ionpair chromatography are scarce, these techniques have been widely used for molecular biology applications (Huber, 1998),and they are suggested techniques for the characterization of final plasmid DNA preparations (Marquet et al., 1997a, b). Ion-exchange chromatography
The polyanionic structure of nucleic acids can be explored in ion-exchangechromatography. The strategy relies upon binding the nucleic acids to positively charged matrices, followed by selective elution with increasing ionic strength. It has been used in molecular biology (Chandra et al., 1992; Hines et al., 1992; Merion and Warren, 1989) to separate DNA restriction fragments (Drager and Regnier, 1985; Muller, 1986; Newton et al., 1983; Yamakawa et al., 1996) and plasmid topoisomers (Onishi et al., 1993),as well as in the purification of plasmid DNA for gene therapy and DNA vaccination applications (Ferreira et al., 1998; Ferreira et al., 1999; Ferreira et al., 2000a; Lahijani et al., 1996; Prazeres et al., 1998; Varley et al., 1998; Schleef, 1999). As mentioned above, nucleic acid molecules have one negative charge per base, i.e. the overall net charge equals the number of bases in the molecule. The expected elution profde thus follows the order of increasing molecule size, with the larger nucleic acid molecules being eluted at higher ionic strengths (Colpan and Riesner, 1984; Yamakawa et al., 1996).The relative retention time between single- (ss) and double-stranded (ds) nucleic acids can also be predicted by this rule depending on the overall molecule charge, therefore, on the number of bases that constitute the molecule (Colpan and Riesner, 1984). Inversions in the predicted retention factors were observed in the anion-exchange purification of dsDNA fragments on a variety of stationary phases (Huber, 1998; Muller, 1986; Yamaltawa et al., 1996). It was suggested that the retardation of some dsDNA fragments i s due to their high AT content (Muller, 1986; Yamakawa et al., 1996), therefore indicating that the ion-exchange elution profile of nucleic acids i s also sequence-dependent. A more detailed analysis revealed that peak retardation i s not always proportional to the high AT content of the dsDNA sequences (Huber, 1998; Yamakawa et al., 1996). It was found that binding of nucleic acid molecules onto anion exchangers i s favored by optimal interaction with the curvature of the particle pores (Colpan and Riesner, 1984) (Figure 9). Therefore, a conformation-dependent separation was proposed (Huber, 1998; Yamalawa et al., 1996). Nudeic acid bending promotes better fits within pore curvatures, enabling more charges to interact with the solid phase (Figure 9A, B), thus
I
217
218
I
Cuilherme
N. M.
Ferreira et a/.
A
B
C
D
Fig. 9. Representation o f possible interactions with a particle pore of linear (A) and bent (B)dsDNA fragments, supercoiled (C)and relaxed (D) plasmid DNA.
leading to higher retention factors. Another physical explanation for the stronger binding of bent dsDNA fragments relies upon the appearance of a dipole character due to the local compression of charges at bent regions (Huber, 1998). Bending increases the local charge density and, by consequence, stronger electrostatic interactions take place at these locations. Nucleic acid bending is mainly the result of poly A runs (Sinden, 1994). Therefore bent dsDNA fragments exhibit an inherently high AT content. Consequently, the observed sequence-dependent elution profile of dsDNA fragments (Miiller, 1986; Yamakawa et al., 1996) is an artifact that hides the phenomena responsible for peak retardation, which is sequence bending. The separation of plasmid topoisomers (Onishi et al., 1993), mainly the relaxed and supercoiled plasmid isoforms (Lahijani et al., 1996; Marquet et al., 1995; Prazeres et al., 1998) with anion-exchangers, confirms the conformation-dependent retention hypothesis outlined above. Supercoiled DNA is more stretched, compacted and bent than its relaxed isoform, therefore presenting higher charge densities and better fits with the particle pores’ curvature (Figure 9C, D). Selectivity and peak resolution are other factors affected by the interaction of nucleic acids with the solid phase (Colpan and Riesner, 1984; Huber, 1998; Thompson, 1986). At least 50 nm pores were needed to completely separate a mixture of RNAs with molecular weights in the range of 25,000 Da to 1,000,000 Da, in which case the elution profile followed the order of increasing size (Colpan and Riesner, 1984). Circular RNA molecules added to the RNA mixture were lastly eluted, at the highest ionic strength (Colpan and Riesner, 1984). Furthermore, a 4.2 kb plasmid was not separated from an E. coli extract with particles of less
1 1 Plasrnid Manufacturing
- An
Overview
than 400 nm pores (Colpan and Riesner, 1984).These experimental results suggest that nucleic acids must fit well into the particle pores, within their maximum extension, while matching the pore inner curvature as well in order to achieve optimal resolution and selectivity (Colpan and Riesner, 1984; Huber, 1998; Thompson, 1986). Elution buffer, gradient former, and flow rate also influence the selectivity and resolution in ion-exchange chromatography (Sofer and Hagel, 1997). While the buffer pH has no apparent effect on the separation of nucleic acids with strong anion exchangers, a decrease in pH usually requires higher ionic strengths to elute the bound nucleic acids in week anion exchange matrices (Colpan and Riesner, 1984; Huber, 1998).This result is attributed to the increase of the charge density of week anion exchangers at lower pH values (Colpan and Riesner, 1984; Sofer and Hagel, 1997), being therefore an obvious corollary when using such supports (Sofer and Hagel, 1997). Variation of the cation (lithium, sodium, potassium and caesium) while keeping chloride as the anion showed that smaller cations eluted the nucleic acids at lower ionic strengths (Huber, 1998). This observation, however, simply reflects the decrease of the activity coefficients with increasing cation radius (Atlcins, 1990; Muller, 1986),with particular activity values and therefore ionic strengths achieved at higher concentrations for larger cations (Muller, 1986). Thus, the variation observed on the retention factors by changing the gradient cation is due to gradient slope alterations which are unrelated to the system selectivity and therefore should not be confused with. It is well known in ion-exchange chromatography that longer gradients usually improve the resolution of two components, which is at maximum with gradients within the range of 5-20 column volumes (Sofer and Hagel, 1997). While the gradient slope has no significant effect in the separation of low molecular weight dsDNA fragments it is very important to maintain shallow gradients, if the separation of high molecular weight nucleic acids is desired (Huber, 1998; Yamalcawa et al., 1996). Regarding the purification of plasmid DNA, removal of impurities from the plasmid preparations is also improved with shallow gradients (Colpan and Riesner, 1984; Huber, 1998). Furthermore, the optimization and selection of shallow NaCl gradients leads to the partial separation of relaxed, supercoiled and denatured plasmid DNA (Prazeres et al., 1998) which is not accomplished with sharp NaCl steps (Ferreira et al., 1998; Ferreira et al., 1999). The optimal gradient, however, should combine acceptable resolutions with relatively fast purification times in order to enable high throughputs and maximize the purification productivity. The flow rate has no significant effect on the resolution of nucleic acids (Colpan and Riesner, 1984),with the sole disadvantage of high flow rates being the dilution of the eluted nucleic acids, However, in process chromatography the flow rate controls the residence time which is a function of target molecule diffusivity, sample viscosity, and particle size (Sofer and Hagel, 1997).The loss of plasmid DNA in the flow through the column (Prazeres et al., 1998) can thus be attributed to high flow rates and, therefore, shorter residence times than required for efficient plasmid binding.
I
219
220
I
Cuilherme N M. Ferreira et al.
A carry-over phenomenon, which i s the contamination of a nucleic acid peak by impurities that normally elute earlier, has been described (Colpan and Riesner, 1984; Huber, 1998).This phenomenon was attributed to the interaction of the nucleic acid with the impurity, either directly or promoted by di- or multivalent metal ions such as Ca2+ or Mg2+ (Colpan and Riesner, 1984; Huber, 1998). Addition of chelating agents such as EDTA to the elution buffers to complex the metal ions, and/or denaturating agents such as urea, formamide or isopropanol, to suppress intermolecular hydrophobic interactions is thus suggested (Huber, 1998). Furthermore, highly charged cationic molecules such as urea, spermidine and spermine are known to selectively compact supercoiled plasmid DNA (Murphy et al., 1999), thus improving its separation from relaxed plasmid DNA and other impurities in anion-exchangechromatography (Colpan and Riesner, 1984; Murphy et al., 1999). Although high protein and nucleic acid clearances can be achieved in the anion exchange purification of supercoiled plasmid DNA, this operation is hampered by the inevitable contamination of plasmid DNA with high molecular weight RNA, gDNA and endotoxins (Ferreira et al., 1999; Prazeres et al., 1999). While carryover phenomena contribute to this fact to some extent, unselectivity of anion exchangers (Huber, 1998; Muller, 1986) and diffusion constraints of the macromolecules in the particle pores (Huber, 1998; Lyddiatt and O'Sullivan, 1998) are better explanations for this. In anion exchange chromatography molecules with higher charge densities are expected to elute at higher ionic strengths. As previously mentioned, the elution profile of linear, unbent nucleic acids i s expected to be a function of the molecule size. If a mass action law (Bellot and Condoret, 1993; Brooks and Cramer, 1992) is considered for the adsorption of nucleic acids onto anion-exchangers, a linear relationship between the elution ionic strength and the nucleic acid size is expected. However, experimental elution profiles of linear dsDNA fragments showed a drastic compression of the elution position for fragment sizes above 180 bp (Huber, 1998; Muller, 1986), suggesting unselectivity of anion exchangers after this limit (Figure 10). Competitive binding between plasmid DNA and impurities present in the feedstocks, such as high molecular weight RNA (Figure ll), as demonstrated by batch adsorption experiments (Ferreira, 2000) provides further evidence for the unselectivity of anion exchangers. Therefore, the selective elution of molecules with high charge density such as plasmid DNA, gDNA, high molecular weight RNA and endotoxins, can thus be difficult to accomplish in a single anion-exchange step. The diffusional constraints of macromolecules in the particle pores also contribute to the recovery of contaminated plasmid DNA. Large plasmid DNA molecules are partially excluded from particle pores smaller than 400 nm (Colpan and Riesner, 1984),being adsorbed at the particle surface (Ferreira et al., 2000b; Ljunglof et al., 1999). On the other hand, RNA i s small enough to be accommodated in the particle pores. Therefore, size exclusion mechanisms can override the ionexchange principles (Thompson, 1986), with the eluted RNA diffusing slowly through the pores of the particles and through the column. It is important to
0,50
Fig. 10. Plot o f the eluting NaCl activity versus dsDNA fragment size n (bp) (Muller, 1986).
- +
0,45 h
5 m
Y
0,40
1 I
0,35
Fig. 11. Adsorption isotherm for the binding o f high molecular weight E. coli W1