Biosensors and Biodetection
Series Editor John M. Walker School of Life Sciences University of Hertfordshire Hatfield, Hertfordshire, AL10 9AB, UK
For other titles published in this series, go to www.springer.com/series/7651
METHODS
IN
MOLECULAR BIOLOGY™
Biosensors and Biodetection Methods and Protocols Volume 503: Optical-Based Detectors
Edited by
Avraham Rasooly* and Keith E. Herold† *FDA Center for Devices and Radiological Health, Silver Spring, MD, USA and National Cancer Institute, Bethesda, MD, USA † Fischell Department of Bioengineering, University of Maryland, College Park, MD, USA
Editors Avraham Rasooly FDA Center for Devices and Radiological Health Silver Spring, MD USA and National Cancer Institute Bethesda, MD USA
[email protected] Keith E. Herold Fischell Department of Bioengineering University of Maryland College Park, MD USA
[email protected] ISBN: 978-1-60327-566-8 e-ISBN: 978-1-60327-567-5 ISSN: 1064-3745 e-ISSN: 1940-6029 DOI: 10.1007/978-1-60327-567-5 Library of Congress Control Number: 2008941063 © Humana Press, a part of Springer Science+Business Media, LLC 2009 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Humana Press, c/o Springer Science + Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. Printed on acid-free paper springer.com
Preface 1. Biosensor Technologies In recent years, many types of biosensors have been developed and used in a wide variety of analytical settings, including biomedical, environmental, research, and others. A biosensor is defined by the International Union of Pure and Applied Chemistry (IUPAC) as a “device that uses specific biochemical reactions mediated by isolated enzymes, immunosystems, tissues, organelles, or whole cells to detect chemical compounds usually by electrical, thermal, or optical signals” (1). Thus, almost all biosensors are based on a two-component system: a biological recognition element (ligand) that facilitates specific binding to or biochemical reaction with a target, and a signal conversion unit (transducer). Although it is impossible to fully cover the fast-moving field of biosensing in one publication, this publication presents some of the many types of biosensors to give the reader a sense of the enormous potential for these devices. An early reference to the concept of a biosensor is from Dr. Leland C. Clark, who worked on biosensors in the early 1960s (2) developing an “enzyme electrode” for glucose concentration measurement with the enzyme glucose oxidase, a measurement that is important in the diagnosis and treatment of disorders of carbohydrate metabolism in diabetes patients. Still today, the most common biosensors used are for glucose analysis. A large number of basic biosensors, all combining a biological recognition element and a transducer, were subsequently developed. Currently, the trend is toward more complex integrated multianalyte sensors capable of more comprehensive analyses. Advances in electronics and microelectrical and mechanical systems (MEMS) have enabled the miniaturization of many biosensors and the newest generation biosensors include miniaturized multianalyte devices with high-throughput capabilities and more than 1,000 individually addressable sensor spots per square centimeter. A useful categorization of biosensors is to divide them into two groups: direct recognition sensors, in which the biological interaction is directly measured, and indirect detection sensors, which rely on secondary elements for detection. Figure 1 shows a schematic of the two groups of biosensors. In each group, there are several types of transducers including optical, electrochemical, and mechanical. For all of these technologies, the recognition ligand plays a major role. Although the most commonly used ligands are antibodies, other ligands are being developed including aptamers (protein-binding nucleic acids) and peptides. In the literature and in practice, there are numerous types of biosensors, and the choice of a suitable system for a particular application is complex, based on many factors such as the nature of the application, the label molecule (if used), the sensitivity required, the number of channels (or area), cost, technical expertise, and the speed of detection needed. A primary purpose of this book is to provide more access to the
v
vi
Preface
Fig. 1. General schematic of biosensors: (a) direct detection biosensors where the recognition element is label free; (b) indirect detection biosensors using a “sandwich” assay where the analyte is detected by labeled molecule. Direct detection biosensors are simpler and faster but typically yield a higher limit of detection compared with indirect detection systems
technical methods involved in using a variety of biosensors to facilitate such decision making. Direct detection biosensors utilize direct measurement of the biological interaction. Such detectors typically measure physical changes (e.g., changes in optical, mechanical, or electrical properties) induced by the biological interaction, and they do not require labeling (i.e., label free) for detection. Direct biosensors can also be used in an indirect mode, typically to increase their sensitivity. Direct detection systems include optical-based systems (most common being surface plasmon resonance) and mechanical systems such as quartz crystal resonators. Indirect detection sensors rely on secondary elements (labels) for detection. Examples of such secondary elements are enzymes (e.g., alkaline phosphatase or glucose oxidase) and fluorescently tagged antibodies that enhance detection of a sandwich complex. Unlike direct detectors, which directly measure changes induced by biological interactions and are “label free,” indirect detectors require a labeled molecule to bind to the target. Most indirect sensors based on optical detection are designed to measure fluorescence. The detection system can be based on a charge coupled device (CCD), photomultiplier tube (PMT), photodiode, or spectrometer. Electrochemical transducers, which measure the oxidation or reduction of an electroactive compound on the secondary ligand, are another common type of indirect detection sensor. Several types of electrochemical biosensors are in use including amperometric devices, which measure the electric current as a function of time while the electrode potential is held constant. Ligands are recognition molecules that bind specifically with the target molecule to be detected. The most important characteristics for ligands are affinity and specificity. Various types of ligands are used in biosensors. Biosensors that use antibodies as recognition elements (immunosensors) are common because antibodies are highly specific, versatile, and bind strongly and stably to the antigen. Several limitations of antibodies are long-term stability, and manufacturing costs, especially for multitarget biosensor applications where many ligands are needed.
Preface
vii
Other types of ligands that show promise for high-throughput screening and chemical synthesis are aptamers and peptides. Aptamers are protein-binding nucleic acids (DNA or RNA molecules) selected from random pools on the basis of their ability to bind other molecules with high affinity. Peptides can be selected for affinity to a target molecule by display methods (phage display and yeast display). However, in general, the binding affinity of peptides is lower than the affinity of antibodies or aptamers.
2. Biosensor Applications Biosensors have several potential advantages over other methods of biodetection, especially increased assay speed and flexibility. Rapid, essentially real-time analysis can provide immediate interactive information to users. This speed of detection is an advantage in essentially all applications. Applications of biosensors include medical, environmental, public security, and food safety areas. Medical applications include clinical, pharmaceutical and device manufacturing, and research. Biosensor-based diagnostics might facilitate disease screening and improve the rates of earlier detection and attendant improved prognosis. Such technology may be extremely useful for enhancing health care delivery in the community setting and to underserved populations. Environmental applications include spill clean-up, monitoring, and regulatory instances. Public safety applications include civil and military first responders as well as unattended monitoring. Food safety applications include monitoring of food production, regulatory monitoring, and diagnosis of food poisoning. Biosensors allow multitarget analyses, automation, and reduced costs of testing. The key strengths of biosensors are the following: • Fast or real-time analysis: Fast or real-time detection provides almost immediate interactive information about the sample tested, enabling users to take corrective measures before infection or contamination can spread. • Point-of-care detection: Biosensors can be used for point-of-care or on-site testing where state-of-the-art molecular analysis is carried out without requiring a stateof-the-art laboratory. • Continuous flow analysis: Many biosensor technologies can be configured to allow continuous flow analysis. This is beneficial in food production, air quality, and water supply monitoring. • Miniaturization: Biosensors can be miniaturized so that they can be integrated into powerful lab-on-a-chip tools that are very capable while minimizing cost of use. • Control and automation: Biosensors can be integrated with on-line process monitoring schemes to provide real-time information about multiple parameters at each production step or at multiple time points during a process, enabling better control and automation of many industrial and critical monitoring facilities.
viii
Preface
3. Aims and Approach The primary aim of this book is to describe the basic types and the basic elements of biosensors from methods point of view. We tried to include manuscripts that represent the major technologies in the field and to include enough technical detail so that the informed reader can both understand the technology and also be able to build similar devices. The target audience for this book includes engineering, chemical, and physical science researchers, who are developing biosensing technologies. Other target groups are biologists and clinicians, who are the users and developers of applications for the technologies. In addition to supporting the research community, the book may also be useful as a teaching tool for bioengineering, biomedical engineering, and biology faculty and students. To better represent the field, most topics are covered by more than one chapter. The purpose of this “redundancy” is to try to include several alternative approaches for the topics, so as to help the reader choose an appropriate design.
4. Chapter Organization This publication is divided into two volumes: Vol. 503 is focused on Optical-Based Detectors and Vol. 504 is focused on Electrochemical and Mechanical Detectors, Lateral Flow, and Ligands for Biosensors. 4.1. Volume 503: Optical-Based Detectors
Optical detection is used in a broad array of biosensor technologies, including both direct and indirect style sensors. Volume 503 is organized in two parts. Part I focuses on direct optical detectors, while Part II concentrates on indirect optical detection. Probably, the most common approach for direct optical detection is based on evanescent wave physics, where the interaction between the evanescent wave and the bound target generates a detection signal. The most common technology in this group is surface plasmon resonance (SPR) and several chapters (see Chaps. 1–5) describe biosensors based on SPR. Other important optical direct detection methods including resonant mirror (see Chap. 6), optical ring resonator (see Chap. 7), interferometric sensors (see Chaps. 8 and 9) and grating coupler (see Chap. 10) are all included in Part I. The second part of Vol. 503 describes various indirect optical detectors. As discussed earlier, indirect detectors require a labeled molecule to bind to the target generating a signal. For optical sensors, the label molecule emits or modifies light. Most indirect optical detectors are designed to measure fluorescence. However, optical detectors can also measure optical density (densitometry), changes in color (colorimetric), and chemoluminesence, depending on the type of label used. Optical signals can be measured in various ways (described in Part II) including various CCD-based detectors, which are very versatile, inexpensive, and relatively simple to construct and use (see Chaps. 11–16 and 25). Other optical detectors discussed in Part II are photodiodes (see Chaps. 17–20), photomultipliers (see Chaps. 21–23),
Preface
ix
and spectrometers (see Chaps. 24 and 25). Photomultipliers may offer higher sensitivity, smaller footprint (the size of photodiode can be few millimeters). Spectrometers offer better interrogation of changes in light wavelengths. 4.2. Volume 504: Electrochemical and Mechanical Detectors, Lateral Flow, and Ligands
Volume 504 describes various electrochemical and mechanical detectors, lateral flow devices, and ligands for biosensors. As in Vol. 503, we describe several direct measurement sensors (in Part I), indirect methods (Parts II–III). Ligands are described in Part IV and two related technologies are described in Part V. In Part I, we describe several mechanical detectors that modify their mechanical properties as a result of biological interactions. Such mechanical direct biosensors typically sense resonance of the mechanical element, which changes when the target molecule binds to the surface. Piezoelectric biosensors (see Chaps. 1–3) employ a technology that is widely used in a variety of applications (e.g., vapor deposition of metals) and is thus readily available and relatively inexpensive. Cantilever-based systems (see Chaps. 4 and 5) can be miniaturized to micrometer dimensions with attendant benefits for system and sample size. In Part II, we describe several electrochemical detectors (see Chaps. 6–11). Electrochemical biosensors were the first biosensors developed and are the most commonly used biosensors today (e.g., glucose monitoring). Part III covers lateral flow technologies (see Chaps. 12–15). Although lateral flow devices are not “classical” biosensors, with ligands and transducers, they are included in this book because of their importance for biosensing. Lateral flow assays are simple immunodetection (or DNA hybridization) devices, which utilize competitive or sandwich assays. They are used mainly for medical diagnostics, including laboratory, home and point-of-care detection. A common format is a “dipstick” in which the test sample diffuses through a porous matrix via capillary action followed by detection by a colorimetric reagent bound to a secondary antibody. The primary antibody is bound to the matrix in a line, and the assay result is a color change at a particular location on the matrix. Lateral flow assays can be dependable and inexpensive. Part IV focuses on recognition ligands, which are key elements in any biosensor (see Chaps. 16–22). The recognition ligands bind specifically with the target molecule to be detected. Various ligands described in Part IV include antibodies, aptamers, and peptides. Antibodies are the most commonly used ligands but advances in selection methods for aptamers (SELEX) and peptides (phage and yeast display) are currently providing alternatives. Part V includes two papers on protein (see Chap. 23) and DNA preparation (see Chap. 24). These papers are relevant to the subject of biosensor technologies but did not fit elsewhere into the book organization outline. References 1. IUPAC Compendium of Chemical Terminology 2nd Edition (1997). (1992), International Union of Pure and Applied Chemistry: Research Triangle Park, NC. 2. Clark LC Jr., Lyons C (1962) Electrode systems for continuous monitoring in cardiovascular surgery. Ann N Y Acad Sci 102:29–45.
Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contents of Volume 504. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
v xiii xvii
PART I: OPTICAL-BASED DETECTORS 1.
Surface Plasmon Resonance and Surface Plasmon Field-Enhanced Fluorescence Spectroscopy for Sensitive Detection of Tumor Markers . . . . . . . . Yusuke Arima, Yuji Teramura, Hiromi Takiguchi, Keiko Kawano, Hidetoshi Kotera, and Hiroo Iwata 2. Surface Plasmon Resonance Biosensor for Biomolecular Interaction Analysis Based on Spatial Modulation Phase Detection . . . . . . . . . . . . . . . . . . . Xiang Ding, Fangfang Liu, and Xinglong Yu 3. Array-Based Spectral SPR Biosensor: Analysis of Mumps Virus Infection . . . . . . Jong Seol Yuk and Kwon-Soo Ha 4. Optical Biosensors Based on Photonic Crystal Surface Waves . . . . . . . . . . . . . . . Valery N. Konopsky and Elena V. Alieva 5. Surface Plasmon Resonance Biosensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Marek Piliarik, Hana Vaisocherová, and Ji í Homola 6. Label-Free Detection with the Resonant Mirror Biosensor. . . . . . . . . . . . . . . . . Mohammed Zourob, Souna Elwary, Xudong Fan, Stephan Mohr, and Nicholas J. Goddard 7. Label-Free Detection with the Liquid Core Optical Ring Resonator Sensing Platform. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ian M. White, Hongying Zhu, Jonathan D. Suter, Xudong Fan, and Mohammed Zourob 8. Reflectometric Interference Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Guenther Proll, Goran Markovic, Lutz Steinle, and Guenter Gauglitz 9. Phase Sensitive Interferometry for Biosensing Applications . . . . . . . . . . . . . . . . Digant P. Davé 10. Label-Free Serodiagnosis on a Grating Coupler . . . . . . . . . . . . . . . . . . . . . . . . . Thomas Nagel, Eva Ehrentreich-Förster, and Frank F. Bier
3
21 37 49 65 89
139
167 179 189
PART II: INDIRECT DETECTORS 11. CCD Camera Detection of HIV Infection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . John R. Day 12. Simple Luminescence Detector for Capillary Electrophoresis . . . . . . . . . . . . . . . Antonio Segura-Carretero, Jorge F. Fernández-Sánchez, and Alberto Fernández-Gutiérrez 13. Optical System Design for Biosensors Based on CCD Detection . . . . . . . . . . . . Douglas A. Christensen and James N. Herron
xi
203 221
239
xii
Contents
14. A Simple Portable Electroluminescence Illumination-Based CCD Detector . . . . Yordan Kostov, Nikolay Sergeev, Sean Wilson, Keith E. Herold, and Avraham Rasooly 15. Fluoroimmunoassays Using the NRL Array Biosensor . . . . . . . . . . . . . . . . . . . . Joel P. Golden and Kim E. Sapsford 16. Biosensors Technologies: Acousto-Optic Tunable Filter-Based Hyperspectral and Polarization Imagers for Fluorescence and Spectroscopic Imaging . . . . . . . . Neelam Gupta 17. Photodiode-Based Detection System for Biosensors. . . . . . . . . . . . . . . . . . . . . . Yordan Kostov 18. Photodiode Array On-chip Biosensor for the Detection of E. coli O157:H7 Pathogenic Bacteria. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Joon Myong Song and Ho Taik Kwon 19. DNA Analysis with a Photo-Diode Array Sensor . . . . . . . . . . . . . . . . . . . . . . . . Hideki Kambara and Guohua Zhou 20. Miniaturized and Integrated Fluorescence Detectors for Microfluidic Capillary Electrophoresis Devices . . . . . . . . . . . . . . . . . . . . . . . Toshihiro Kamei 21. Photomultiplier Tubes in Biosensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Yafeng Guan 22. Integrating Waveguide Biosensor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Shuhong Li, Platte Amstutz III, Cha-Mei Tang, Jun Hang, Peixuan Zhu, Yunqi Zhang, Daniel R. Shelton, and Jeffrey S. Karns 23. Detection of Fluorescence Generated in Microfluidic Channel Using In-Fiber Grooves and In-Fiber Microchannel Sensors . . . . . . . . Rudi Irawan and Swee Chuan Tjin 24. Multiplex Integrating Waveguide Sensor: Signalyte™-II. . . . . . . . . . . . . . . . . . . Shuhong Li, Yunqi Zhang, Platte Amstutz III, and Cha-Mei Tang 25. CCD Based Fiber-Optic Spectrometer Detection. . . . . . . . . . . . . . . . . . . . . . . . Rakesh Kapoor Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
259
273
293 307
325 337
361 375 389
403 423 435 447
Contributors ELENA V. ALIEVA • Institute of Spectroscopy, Russian Academy of Sciences, Troitsk, Moscow Region, Russia PLATTE AMSTUTZ • Creatv MicroTech, Inc., Potomac, MD, USA YUSUKE ARIMA • Institute for Frontier Medical Sciences, Kyoto University, Kyoto, Japan FRANK F. BIER • Department of Molecular Bioanalytics & Bioelectronics, Fraunhofer Institute for Biomedical Engineering, Branch Potsdam-Golm, Potsdam, Germany Institute of Biochemistry and Biology, University of Potsdam, Potsdam, Germany DOUGLAS A. CHRISTENSEN • Department of Bioengineering and Department of Electrical & Computer Engineering, University of Utah, Salt Lake City, UT, USA DIGANT P. DAVÉ • University of Texas at Arlington, Arlington, TX, USA JOHN R. DAY • Gen-Probe Incorporated, San Diego, CA, USA XIANG DING • Department of Precision Instruments and Mechanics, Tsinghua University, Beijing, China EVA EHRENTREICH-FÖRSTER • Department of Molecular Bioanalytics & Bioelectronics, Fraunhofer Institute for Biomedical Engineering, Branch Potsdam-Golm, Potsdam, Germany SOUNA ELWARY • Biosensors Division, Biophage Pharma, Montreal, QC, Canada XUDONG FAN • Department of Biological Engineering, University of Missouri-Columbia, Columbia, MO, USA ALBERTO FERNÁNDEZ-GUTIÉRREZ • Department of Analytical Chemistry, Faculty of Sciences, University of Granada, Granada, Spain JORGE F. FERNÁNDEZ-SÁNCHEZ • Department of Analytical Chemistry, Faculty of Sciences, University of Granada, Granada, Spain GUENTER GAUGLITZ • Institute of Physical and Theoretical Chemistry, University of Tuebingen, Tuebingen, Germany NICHOLAS J. GODDARD • School of Chemical Engineering and Analytical Science (CEAS), The University of Manchester, Manchester, UK JOEL P. GOLDEN • Center for Bio/Molecular Science & Engineering, US Naval Research Laboratory, Washington, DC, USA YAFENG GUAN • Department of Instrumentation & Analytical Chemistry, Dalian Institute of Chemical Physics, Dalian, China NEELAM GUPTA • Army Research Laboratory, Adelphi, MD, USA KWON-SOO HA • Department of Molecular and Cellular Biochemistry and Nanobio Sensor Research Center, Kangwon National University College of Medicine, Chuncheon, Kangwon-do, Korea JUN HANG • Creatv MicroTech, Inc., Potomac, MD, USA
xiii
xiv
Contributors
KEITH E. HEROLD • Fischell Department of Bioengineering, University of Maryland, College Park, MD, USA JAMES N. HERRON • Fischell Department of Bioengineering and Department of Electrical & Computer Engineering, University of Utah, Salt Lake City, UT, USA JIrˇÍ HOMOLA • Institute of Photonics and Electronics, Academy of Sciences of the Czech Republic, Prague, Czech Republic RUDI IRAWAN • BioMedical Engineering Research Centre, Singapore-University of Washington Alliance, Nanyang Technological University, Singapore Department of Physics, University of Lampung, Bandar Lampung, Indonesia HIROO IWATA • Institute for Frontier Medical Sciences, Kyoto University, Kyoto, Japan HIDEKI KAMBARA • Central Research Laboratory, Hitachi Ltd., Tokyo, Japan TOSHIHIRO KAMEI • National Institute of Advanced Industrial Science and Technology (AIST), Ibaraki, Japan RAKESH KAPOOR • Department of Physics, University of Alabama at Birmingham, Birmingham, AL, USA JEFFREY S. KARNS • Environmental Microbial Safety Laboratory, U.S. Department of Agriculture-Agricultural Research Service, Beltsville, MD, USA KEIKO KAWANO • Advanced Software Technology and Mechatronics Research Institute of Kyoto, Kyoto, Japan VALERY N. KONOPSKY • Institute of Spectroscopy, Russian Academy of Sciences, Troitsk, Moscow Region, Russia YORDAN KOSTOV • Center for Advanced Sensor Technology, University of Maryland Baltimore County (UMBC), Baltimore, MD, USA HIDETOSHI KOTERA • Department of Microengineering, Graduate School of Engineering, Kyoto University, Kyoto, Japan HO TAIK KWON • Celltek Co., Ltd., Ansan-si, South Korea SHUHONG LI • Creatv MicroTech, Inc., Potomac, MD, USA FANGFANG LIU • Department of Precision Instruments and Mechanics, Tsinghua University, Beijing, China GORAN MARKOVIc • Institute of Physical and Theoretical Chemistry, University of Tuebingen, Tuebingen, Germany STEPHAN MOHR • School of Chemical Engineering and Analytical Science (CEAS), The University of Manchester, Manchester, UK THOMAS NAGEL • Department of Molecular Bioanalytics & Bioelectronics, Fraunhofer Institute for Biomedical Engineering, Branch Potsdam-Golm, Potsdam, Germany Institute of Biochemistry and Biology, University of Potsdam, Potsdam, Germany MAREK PILIARIK • Institute of Photonics and Electronics, Academy of Sciences of the Czech Republic, Prague, Czech Republic GUENTHER PROLL • Institute of Physical and Theoretical Chemistry, University of Tuebingen, Tuebingen, Germany AVRAHAM RASOOLY • FDA Center for Devices and Radiological Health, Silver Spring, MD, USA, National Cancer Institute, Bethesda, MD, USA
Contributors
xv
KIM E. SAPSFORD • Center for Bio/Molecular Science & Engineering, US Naval Research Laboratory, Washington, DC, USA ANTONIO SEGURA-CARRETERO • Department of Analytical Chemistry, Faculty of Sciences, University of Granada, Granada, Spain NIKOLAY SERGEEV • FDA Center for Devices and Radiological Health, Silver Spring, MD, USA DANIEL R. SHELTON • Environmental Microbial Safety Laboratory, U.S. Department of Agriculture-Agricultural Research Service, Beltsville, MD, USA JOON MYONG SONG • Research Institute of Pharmaceutical Sciences and College of Pharmacy, Seoul National University, Seoul, South Korea LUTZ STEINLE • Institute of Physical and Theoretical Chemistry, University of Tuebingen, Tuebingen, Germany JONATHAN D. SUTER • Biological Engineering Department, University of Missouri-Columbia, Columbia, MO, USA HIROMI TAKIGUCHI • Advanced Software Technology and Mechatronics Research Institute of Kyoto, Kyoto, Japan CHA-MEI TANG • Creatv MicroTech, Inc., Potomac, MD, USA YUJI TERAMURA • Department of Polymer Chemistry, Graduate School of Engineering, Kyoto University, Kyoto, Japan SWEE CHUAN TJIN • Photonics Research Centre, School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore HANA VAISOCHEROVÁ • Institute of Photonics and Electronics, Academy of Sciences of the Czech Republic, Prague, Czech Republic IAN M. WHITE • Biological Engineering Department, University of Missouri-Columbia, Columbia, MO, USA SEAN WILSON • University of Maryland Baltimore County (UMBC), Baltimore, MD, USA Xinglong Yu • Department of Precision Instruments and Mechanics, Tsinghua University, Beijing, China JONG SEOL YUK • Department of Molecular and Cellular Biochemistry and Nanobio Sensor Research Center, Kangwon National University College of Medicine, Chuncheon, Kangwon-do, Korea YUNQI ZHANG • Creatv MicroTech, Inc., Potomac, MD, USA GUOHUA ZHOU • Central Research Laboratory, Hitachi Ltd., Tokyo, Japan HONGYING ZHU • Biological Engineering Department, University of Missouri-Columbia, Columbia, MO, USA PEIXUAN ZHU • Creatv MicroTech, Inc., Potomac, MD, USA MOHAMMED ZOUROB • Biosensors Division, Biophage Pharma, Montreal, QC, Canada
Contents of Volume 504 Preface Contributors Contents of Volume 503
PART I: MECHANICAL DETECTORS 1. 2. 3.
4.
5.
A Set of Piezoelectric Biosensors Using Cholinesterases Carsten Teller, Jan Halámek, Alexander Makower, and Frieder W. Scheller Piezoelectric Biosensors for Aptamer–Protein Interaction Sara Tombelli, Alessandra Bini, Maria Minunni, and Marco Mascini Piezoelectric Quartz Crystal Resonators Applied for Immunosensing and Affinity Interaction Studies Petr Skládal Biosensors Based on Cantilevers Mar Álvarez, Laura G. Carrascosa, Kiril Zinoviev, Jose A. Plaza, and Laura M. Lechuga Piezoelectric-Excited Millimeter-Sized Cantilever Biosensors Raj Mutharasan
PART II: ELECTROCHEMICAL DETECTORS 6.
Preparation of Screen-Printed Electrochemical Immunosensors for Estradiol, and Their Application in Biological Fluids Roy M. Pemberton and John P. Hart 7. Electrochemical DNA Biosensors: Protocols for Intercalator-Based Detection of Hybridization in Solution and at the Surface Kagan Kerman, Mun’delanji Vestergaard, and Eiichi Tamiya 8. Electrochemical Biosensor Technology: Application to Pesticide Detection Ilaria Palchetti, Serena Laschi, and Marco Mascini 9. Electrochemical Detection of DNA Hybridization Using Micro and Nanoparticles María Teresa Castañeda, Salvador Alegret, and Arben Merkoçi 10. Electrochemical Immunosensing Using Micro and Nanoparticles Alfredo de la Escosura-Muñiz, Adriano Ambrosi, Salvador Alegret, and Arben Merkoçi 11. Methods for the Preparation of Electrochemical Composite Biosensors Based on Gold Nanoparticles A. González-Cortés, P. Yáñez-Sedeño, and J.M. Pingarrón
PART III: LATERAL FLOW 12. Immunochromatographic Lateral Flow Strip Tests Gaiping Zhang, Junqing Guo, and Xuannian Wang
xvii
xviii
Contents of Volume 504
13. Liposome-Enhanced Lateral-Flow Assays for the Sandwich-Hybridization Detection of RNA Katie A. Edwards and Antje J. Baeumner 14. Rapid Prototyping of Lateral Flow Assays Alexander Volkov, Michael Mauk, Paul Corstjens, and R. Sam Niedbala 15. Lateral Flow Colloidal Gold-Based Immunoassay for Pesticide Shuo Wang, Can Zhang, and Yan Zhang
PART IV: LIGANDS 16. Synthesis of a Virus Electrode for Measurement of Prostate Specific Membrane Antigen Juan E. Diaz, Li-Mei C. Yang, Jorge A. Lamboy, Reginald M. Penner, and Gregory A. Weiss 17. In Vivo Bacteriophage Display for the Discovery of Novel Peptide-Based Tumor-Targeting Agents Jessica R. Newton and Susan L. Deutscher 18. Biopanning of Phage Displayed Peptide Libraries for the Isolation of Cell-Specific Ligands Michael J. McGuire, Shunzi Li, and Kathlynn C. Brown 19. Biosensor Detection Systems: Engineering Stable, High-Affinity Bioreceptors by Yeast Surface Display Sarah A. Richman, David M. Kranz, and Jennifer D. Stone 20. Antibody Affinity Optimization Using Yeast Cell Surface Display Robert W. Siegel 21. Using RNA Aptamers and the Proximity Ligation Assay for the Detection of Cell Surface Antigens Supriya S. Pai and Andrew D. Ellington 22. In Vitro Selection of Protein-Binding DNA Aptamers as Ligands for Biosensing Applications Naveen K. Navani, Wing Ki Mok, and Yingfu Li
PART V: PROTEIN AND DNA PREPARATION 23. Immobilization of Biomolecules onto Silica and Silica-Based Surfaces for Use in Planar Array Biosensors Lisa C. Shriver-Lake, Paul T. Charles, and Chris R. Taitt 24. Rapid DNA Amplification Using a Battery-Powered Thin-Film Resistive Thermocycler Keith E. Herold, Nikolay Sergeev, Andriy Matviyenko, and Avraham Rasooly Index
Chapter 1 Surface Plasmon Resonance and Surface Plasmon Field-Enhanced Fluorescence Spectroscopy for Sensitive Detection of Tumor Markers Yusuke Arima, Yuji Teramura, Hiromi Takiguchi, Keiko Kawano, Hidetoshi Kotera, and Hiroo Iwata Summary Surface plasmon resonance (SPR), which provides real-time, in situ analysis of dynamic surface events, is a valuable tool for studying interactions between biomolecules. In the clinical diagnosis of tumor markers in human blood, SPR is applied to detect the formation of a sandwich-type immune complex composed of a primary antibody immobilized on a sensor surface, the tumor marker, and a secondary antibody. However, the SPR signal is quite low due to the minute amounts (ng–pg/mL) of most tumor markers in blood. We have shown that the SPR signal can be amplified by applying an antibody against the secondary antibody or streptavidin-conjugated nanobeads that specifically accumulate on the secondary antibody. Another method employed for highly sensitive detection is the surface plasmon field-enhanced fluorescence spectroscopy-based immunoassay, which utilizes the enhanced electric field intensity at a metal/water interface to excite a fluorophore. Fluorescence intensity attributed to binding of a fluorophore-labeled secondary antibody is increased due to the enhanced field in the SPR condition and can be monitored in real time. Key words: Surface plasmon resonance, Immunosensing, Tumor marker, Signal amplification, Polyclonal antibody, Surface plasmon field-enhanced fluorescence spectroscopy.
1. Introduction Surface plasmon resonance (SPR)-based sensing has been used to study interactions between biomolecules (1). The SPR method is a sensitive technique to detect changes in the local refractive index near the surface of a thin metal (typically gold) film (2).
Avraham Rasooly and Keith E. Herold (eds.), Methods in Molecular Biology: Biosensors and Biodetection, Vol. 503 © Humana Press, a part of Springer Science + Business Media, LLC 2009 DOI: 10.1007/978-1-60327-567-5_1
3
4
Arima et al.
Figure 1 shows an SPR apparatus of Kretchmann configuration (a) and its schematic representation (b). A beam of p-polarized light is used to illuminate the back side of a gold thin film on glass, and the front side of the film faces a solution of interest. When the incident angle exceeds the critical angle, total internal reflection occurs. An evanescent wave is generated on the surface facing the solution, which has a lower refractive index than glass. At a specific incident angle, the evanescent wave of the incoming light is able to couple with the free oscillating electrons (plasmons) in the metal film, and the surface plasmon is resonantly excited. This excitation causes energy from the incident light to be lost to the metal film, resulting in a reduction in the intensity of reflected light (Fig. 2a). The resonance angle is a function of the refractive index at the interface of the metal film and solution. Thus, a shift in resonance angle reflects events at the interface, such as protein adsorption on the surface and antigen–antibody interactions. SPR offers rapid, label-free, and real-time monitoring of binding events between biomolecules.
(a)
10 cm
(b) Photodiode
Lens Photodiode
Biaxial rotation stage Prism Flow cell Protein solution
Lens Iris Beam splitter Glass plate
Iris He-Ne laser (632.8 nm)
Gran-Thomson prism
Fig. 1. a A surface plasmon resonance (SPR) apparatus and b its schematic representation.
Surface Plasmon Resonance and Surface Plasmon Field-Enhanced
5
Recently, SPR-based immunoassays for the clinical detection of biomarkers in human blood have been investigated (3, 4). The SPR-based immunoassay detects the specific interaction of a biomarker with an antibody immobilized on the SPR sensor (Fig. 2b). Primary antibodies against a particular biomarker are immobilized onto a gold surface modified with a self-assembled monolayer (SAM) of alkanethiols. A blood sample is brought into contact with the sensor, and the biomarker in blood specifically binds to the antibodies immobilized on the sensor surface. If the concentration of the biomarker in blood is high and its molecular weight is large, the SPR resonance angle shift can be detected easily without further modification. However, the concentrations of most tumor markers are in the range ng/mL–pg/mL (Table 1). Therefore, amplification of SPR signal intensity is needed to detect most tumor markers in clinical samples. Several methods for amplifying SPR signal intensity are shown schematically in Fig. 3. In the method shown in Fig. 3a, the sensor surface is incubated with a solution containing
(a)
(b) Secondary antibody
Reflectance
Primary antibody Tumor marker Self-assembled monolayer
Gold thin film q
Incident light
Reflected light
Incident angle q
Fig. 2. a Relationship between incident angle q and intensity of reflected light before (solid line) and after (dashed line) protein adsorption. For real-time monitoring, the intensity of reflected light is monitored at a fixed angle throughout the measurement (arrow). b Schematic representation of SPR-based sandwich-type immunoassay.
Table 1 Concentrations of some tumor markers Molecular weight
Cut-off level (ng/mL)
Cut-off level (pmol/L)
Tumor marker
Name
PSA
Prostate-specific antigen
34,000
4
118
AFP
α-Fetoprotein
70,590
10
142
CEA
Carcinoembryonic antigen
180,000
5
28
6
Arima et al.
(a)
(b) Antibody against secondary antibody Secondary antibody Secondary antibody Tumor marker
Tumor marker
Primary antibody
(c)
Primary antibody
(d) Biotin-labeled anti-streptavidin antibody Streptavidin-conjugated magnetic beads (50 nm)
Detector for fluorescence Fluorophore-conjugated secondary antibody
Biotin-labeled secondary antibody
Tumor marker Primary antibody
Fig. 3. Signal amplification methods for detection of a minute amount of tumor marker. a Binding of secondary antibodies to tumor marker captured by immobilized primary antibodies. b Binding of polyclonal antibodies to secondary antibody. c Accumulation of streptavidin-conjugated nanobeads and biotin-labeled antistreptavidin antibodies. d Detection of fluorophore-conjugated secondary antibodies by surface-plasmon field-enhanced fluorescence spectroscopy.
secondary antibodies, which bind to the tumor marker previously captured by the primary antibody on the SPR sensor surface. As shown in Fig. 3b, further amplification can be achieved by applying polyclonal antibodies against the secondary antibody (5). Because the immobilization of nanobeads causes a large change in the refractive index at the metal/solution interface, nanobeads are expected to be useful for inducing a substantial shift in the SPR resonance angle (Fig. 3c). In this method, biotin-labeled secondary antibodies are bound to tumor markers, which are trapped by primary antibodies immobilized on a sensor surface. Streptavidin-conjugated nanobeads (50 nm in diameter) and biotin-labeled anti-streptavidin antibodies are alternately layered on the surface via the specific biotin–streptavidin interaction (6). All three of the previously described amplification methods utilize an SPR angle shift caused by changes in the local refractive index near the sensor surface. A fourth interesting method to amplify signal intensity utilizes the strong increase in surface light intensity at the metal/water interface around the SPR resonance angle, i.e., surface plasmon field-enhanced fluorescence spectroscopy (SPFS) (7). In this method, SPFS is employed to detect the
Surface Plasmon Resonance and Surface Plasmon Field-Enhanced
7
fluorescence due to binding of a fluorophore-labeled secondary antibody (Fig. 3d). Here, we introduce the methods of SPR-based and SPFSbased immunoassays for the detection of tumor markers.
2. Materials 2.1. Optical System for Surface Plasmon Resonance
1. Hemicylindrical prism (custom-made, diameter: 25 mm, width: 10 mm; Sigma Koki Co., Ltd, Tokyo, Japan). 2. Index-matching fluid (n = 1.515; Cargille Laboratories, Ceder Grove, NJ). 3. Silicone rubber tubes (inner diameter: 0.5 mm, outer diameter: 1 mm) (Fig. 4). 4. Silicone rubber heater regulated by a digital controller (E5EK; Omron Corp., Kyoto, Japan). 5. Peristaltic pump (MP-3 N; Tokyo Rikakikai Co., Ltd, Tokyo, Japan). 6. He–Ne laser beam (l = 633 nm, 05-LHP-151; Melles Griot, Carlsbad, CA). 7. Beam splitter (NPCH-10-6328; Sigma Koki). 8. Gran-Thomson prism (GTPC-08-20AN; Sigma Koki). 9. Photodiode detector (S3590-01; Hamamatsu Photonics K.K., Shizuoka, Japan). 10. Lens (SLB-30-300PM, f = 300 mm; Sigma Koki). 11. Iris (IH-30; Sigma Koki). 12. Optical rail (OBA-500SH; Sigma Koki). 13. Optical table (MB-PH; Sigma Koki).
PVC plate
Silicone rubber Glass plate
12 mm
0.5 mm Hemicylindrical prism
Fig. 4. Assembly of a flow cell.
2 mm
8
Arima et al.
14. Lens (SLB-30-60PM, f = 60 mm; Sigma Koki). 15. Biaxial rotation stage (SGSP-120YAW-W; Sigma Koki), which is operated through an intelligent driver (CSG-522R; Sigma Koki) by homemade software. 2.2. Optical System for Surface Plasmon Field-Enhanced Fluorescence Spectroscopy
The basal components of an SPFS apparatus are the same as for an SPR apparatus (Fig. 1b), except for the addition of a CCD camera on the sensor surface: 1. Laser diode (LD, VHK laser diode module l = 635 nm, 0.95 mW; Coherent, Santa Clara, CA). 2. Lens (SLB-20-25P, f = 25 mm; Sigma Koki, Tokyo, Japan) for collimation of the laser light. 3. Polarizing filter (TS0851-G; Sugitoh, Tokyo, Japan). 4. Neutral density filter (AND-20C-10, T = 10%; Sigma Koki). 5. Guide rail (OBS-200G; Sigma Koki) with appropriate holders. 6. Rotational stage (MM-40θ; Chuo Seiki, Tokyo, Japan). 7. Lens (SLB-20-25P; Sigma Koki, Japan). 8. Photodiode detector (S2281-01; Hamamatsu Photonics, Hamamatsu, Japan). 9. Two-axis stage controller (QT-CM2; Chuo Seiki). 10. Objective lens (SLWD Plan20×; Nikon, Tokyo, Japan). 11. Interference filter (l. = 670 nm, transmittance 75%, full-width half length max (FWHM) = 7 nm; Optical Coatings Japan, Tokyo, Japan). 12. Extension barrel (TS0155-H; Sugitoh, Tokyo, Japan). 13. CCD camera with a charge multiplier (MC681-SPD; Texas Instruments, Dallas, TX) (see Note 1). 14. Image capture board (MT-PCI2; Micro-Technica, Tokyo, Japan). 15. Homemade intensity scanning software (see Note 2). 16. A glass plate (S-LAL10, refractive index: 1.720; Sigma Koki) with a thin gold film (49 nm in thickness). 17. Triangular prism (25 × 25 × 25 mm, S-LAL10; Sigma Koki). 18. Index-matching fluid (n = 1.72, Cargille Laboratories).
2.3. Working Solutions
1. Ethanol is deoxygenated with bubbling nitrogen gas for 20–30 min before use. 2. (1-Mercaptoundecanoic-11-yl)tri(ethylene glycol) (TEG: HS–(CH2)11–(OCH2CH2)3–OH) and (1-mercaptoundecanoic11-yl)hexa(ethylene glycol)carboxylic acid (HEG:
Surface Plasmon Resonance and Surface Plasmon Field-Enhanced
9
HS–(CH 2 ) 11 –(OCH 2 CH 2 ) 5 –OCH 2 CH 2 OCH 2 COOH) (SensoPath Technologies, Inc., Bozeman, MT) are dissolved in deoxygenated ethanol solution at 0.9 and 0.1 mM, respectively (see Note 3). 3. α-Fetoprotein (AFP; Morinaga Institute of Biological Science, Inc., Yokohama, Japan) is dissolved in phosphate buffer at 10 μg/mL and stored in single-use aliquots at −80°C. 4. Whole blood is collected by drawing venous blood from healthy donors into Venoject® II blood collection tubes containing EDTA-2 Na (TERUMO Co., Tokyo, Japan). To separate plasma, the tubes are centrifuged at 3,000 × g at 4°C for 30 min. After centrifugation, supernatant is collected and stored at −80°C until use. 5. Working solutions of AFP at prescribed concentrations are prepared by dilution in human plasma (see Note 4). 6. A tablet of PBS-Tween® (10 mM phosphate buffer, 140 mM NaCl, 3 mM KCl, 0.05% Tween® 20; Calbiochem, Inc., Darmstadt, Germany) is dissolved in 1 L pure water (one tablet per 1 L). 7. Phosphate buffer (pH 6.6) is prepared by dissolving 33 mM Na2HPO4 and 33 mM KH2PO4 in pure water at a volume ratio of 2:1. This solution is degassed by a water aspirator and used for sample preparation. 8. Anti-AFP antibody (mouse monoclonal, Clone number: ME-101, affinity = 6 × 109; Abcam Ltd, Cambridge, UK) is dissolved in phosphate buffer at 3.2 mg/mL and stored in single-use aliquots at −80°C. A primary antibody solution is prepared by diluting the solution to 10 μg/mL in phosphate buffer (see Note 5). 9. Lyophilized powder of anti-AFP antibody with various salts (rabbit polyclonal; Monosan Ltd, Uden, Netherland) is dissolved in pure water at 1.0 mg/mL and stored at 4°C. A secondary antibody solution is prepared by diluting the solution to 10 μg/mL in phosphate buffer (see Note 5). 10. Anti-rabbit IgG antibody (goat polyclonal, 1.5 mg/mL; Zymed Laboratories, Inc., South San Francisco, CA) is dissolved at 10 μg/mL in phosphate buffer just before use (see Note 4). 11. A blocking solution is prepared by dissolving bovine serum albumin (BSA fraction V; Sigma-Aldrich, Inc., St. Louis, MO) at 10 mg/mL in phosphate buffer (see Note 6).
10
Arima et al.
3. Methods 3.1. Surface Plasmon Resonance-Based Immunoassay 3.1.1. Setup of SPR Apparatus
Various SPR instruments are commercially available from companies such as Biacore AB and Moritex Corp. The optical construction of an SPR instrument is simple, as shown schematically in Fig. 1b. We assembled an SPR instrument from optical parts as described (8). The glass plate is coupled to a hemicylindrical prism (custom-made, diameter: 25 mm, width: 10 mm; Sigma Koki Co., Ltd, Tokyo, Japan) with an index-matching fluid (n = 1.515; Cargille Laboratories, Ceder Grove, NJ), and the SPR flow cell is set on the glass plate. The flow cell is assembled with silicone rubber and a PVC plate and is connected to silicone tubes (inner diameter: 0.5 mm, outer diameter: 1 mm) (Fig. 4). The temperature of the flow cell is kept constant with a silicone rubber heater regulated by a digital controller (E5EK; Omron Corp., Kyoto, Japan). A peristaltic pump (MP-3 N; Tokyo Rikakikai Co., Ltd, Tokyo, Japan) delivers the liquid sample to the flow cell at the rate of 4 mL/min. A He–Ne laser beam (l = 633 nm, 05-LHP-151; Melles Griot, Carlsbad, CA) is separated into two by a beam splitter (NPCH10-6328; Sigma Koki). One beam is linearly p-polarized using a Gran-Thomson prism (GTPC-08-20AN; Sigma Koki), and the other is guided to a photodiode detector (S3590-01; Hamamatsu Photonics K.K., Shizuoka, Japan) to monitor the fluctuation of incident light intensity. The polarized light is focused by a lens (SLB-30-300PM, f = 300 mm; Sigma Koki) onto the backside of a gold thin film evaporated on a glass plate. The He–Ne laser, iris (IH-30; Sigma Koki), Gran-Thomson prism, beam splitter, and lens are placed on an optical rail (OBA-500SH; Sigma Koki) and fixed on an optical table by optical bases (MB-PH; Sigma Koki). The reflected light passes through a lens (SLB-30-60PM, f = 60 mm; Sigma Koki), and its intensity is measured by a photodiode detector. Reflectance is calculated from the intensities of the incident and reflected light (voltage), which are converted from the current of each photodiode. The sample stage and the detector for reflected light are rotated at intervals of 0.03° using a biaxial rotation stage (SGSP-120YAW-W; Sigma Koki), which is operated through an intelligent driver (CSG-522R; Sigma Koki) by homemade software, to obtain the relationship between incident angle and reflectance (Fig. 2a). To precisely determine a resonance angle, a quadratic function was fitted to an incident angle–reflectance profile ranging from 0.4° lower to 0.1° higher than an apparent minimum in reflectance, and the minimum of the quadratic function was regarded as the resonance angle. An SPR control and analysis program was developed on an integrated development environment using Object Pascal language (Boland Delphi 4 Pro. Jpn. ed.; Inprise Corp., Tokyo, Japan).
Surface Plasmon Resonance and Surface Plasmon Field-Enhanced
11
For real-time monitoring of binding processes, the change in reflected light intensity is monitored at a fixed angle (in our system, 0.5° lower than the resonance angle) during a measurement (Fig. 2a, arrows). Finally, the change in reflected light intensity is converted to SPR angle shift by homemade software. The amount of adsorbed protein is determined by the SPR angle shift using the following relationship: The amount of adsorbed protein (ng/cm2) = 500 × increase in resonance angle (degree), where the refractive index and density of protein are assumed to be 1.45 and 1.0 g/cm3, respectively. 3.1.2. Preparation of SPR Sensor
1. Glass plates for SPR (material: BK7, refractive index: 1.515, 25 mm × 25 mm × 1 mm) are purchased from Arteglass Associates (Kyoto, Japan). 2. Piranha solution, a 7:3 mixture of concentrated sulfuric acid and 30% hydrogen peroxide solution, is prepared (see Note 7). 3. BK7 glass plates are immersed in piranha solution for 5 min, rinsed twice with deionized water, rinsed with 2-propanol, and stored in 2-propanol until use. 4. Glass plates are dried with a stream of nitrogen gas. 5. Glass plates are placed in a thermal evaporation apparatus (V-KS200; Osaka Vacuum, Osaka, Japan) (see Note 8). 6. A gold wire (purity: 99.99%, f = 0.5 mm) and a chromium piece (purity: 99.99%) are placed in separate tungsten baskets. 7. The pressure in the evaporation chamber is decreased to less than 3 × 10−4 Pa. 8. A chromium layer of 1-nm thickness is deposited on the glass plate at 0.02 nm/s (see Note 9). 9. A gold layer is deposited on the glass plate at 0.05 nm/s for 4 nm, 0.3 nm/s for 38 nm, and 0.05 nm/s for 7 nm (total thickness: 49 nm) (see Note 10). 10. The gold-coated plates are immersed immediately in the TEG/ HEG mixture (see Subheading 1.2.3) for approximately 24 h at room temperature to form a mixed SAM (see Note 11). 11. The glass plates modified with TEG/HEG-mixed SAM are washed thoroughly with pure water and 2-propanol before use (see Note 12).
3.1.3. Determination of AFP Concentration in Human Plasma
1. A TEG/HEG-mixed SAM-coated glass plate is dried with a N2 gas stream and placed on a hemicylindrical prism with index-matching fluid (n = 1.515; Cargille Laboratories, Cedar Grove, NJ). 2. A flow cell chamber composed of a glass plate with prism, a 0.5-mm-thick silicone rubber spacer, and an upper
12
Arima et al.
plate is assembled and is placed in the SPR instrument (see Note 13). 3. The flow cell chamber and solutions are kept at 30°C (see Note 14). 4. For stabilization of baseline in the SPR instrument, degassed phosphate buffer solution is circulated by a peristaltic pump at 4.0 mL/min for 30 min before the start of measurements. 5. A mixture of 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC; Dojindo Laboratories, Kumamoto, Japan) and N-hydroxylsuccinimide (NHS; Nacalai Tesque, Kyoto, Japan) (stored as a mixed powder) is dissolved in degassed phosphate buffer at 0.1 and 0.05 M, respectively, just before use. 6. The mixture of 0.1 M EDC and 0.05 M NHS in phosphate buffer is flowed for 15 min to activate the COOH groups of the TEG/HEG-mixed SAM, and then the primary antibody solution (10 μg/mL) is immediately flowed for 25 min to achieve covalent immobilization. 7. A BSA blocking solution is flowed for 15 min to block nonspecific adsorption and to deactivate unreacted NHS ester groups on the surface (see Note 15). 8. Human plasma containing AFP (50–500 ng/mL) or a clinical sample is flowed into the SPF flow cell for 30 min. 9. PBS-Tween® solution is flowed for 15 min after flowing of plasma to remove nonspecifically adsorbed proteins (see Note 15). 10. A secondary antibody solution (10 μg/mL, polyclonal) is applied for 30 min, followed by a solution of anti-rabbit IgG antibody (10 μg/mL, polyclonal) for 30 min for the SPR signal enhancement (see Note 16). 11. As control experiments, the same procedures (steps 8–10) are performed in the absence of AFP in human plasma. 12. In this system, sample and buffer solutions (3 mL each) are circulated through the flow cell at 4.0 mL/min by a peristaltic pump. Between the different solution injections, the flow cell is washed with phosphate buffer for at least 5 min, except for between the two injections in step 6 (see Notes 17 and 18). 3.1.4. SPR Sensorgram
The SPR immunoassay is based on the formation of a sandwichtype immune complex with two kinds of antibody (Fig. 1.2b), which is also the basis of the widely used enzyme-linked immunosorbent assay (ELISA). Primary antibody is immobilized onto a SAM surface bearing carboxylic groups via covalent amide bonding by a NHS/EDC coupling method. After blocking treatment
Surface Plasmon Resonance and Surface Plasmon Field-Enhanced
13
with BSA or inert materials to prevent nonspecific adsorption of serum proteins, AFP in human plasma and secondary antibody (rabbit IgG) are applied sequentially. The SPR signal shift can be further enhanced by binding of anti-rabbit IgG antibody (polyclonal) (Fig. 1.5; Subheading 1.3.1.3). The concentration of a tumor marker in blood can be determined from the calibration curve, which is obtained from resonance angle shifts for solutions with various concentrations of tumor marker. Amplification of the SPR signal to detect low concentrations of AFP (25 pg/mL–1 ng/mL) can be accomplished using streptavidin-conjugated nanobeads and biotin-labeled antistreptavidin antibodies, according to the method reported for detection of brain natriuretic peptide (BNP) (6). Namely, after flowing of biotin-labeled secondary antibody, streptavidinnanobeads and biotin-labeled antistreptavidin antibody are flowed sequentially, followed by the flowing of 0.05% Tween 20 in PBS solution for the removal of nonspecific adsorbed proteins and beads. These procedures can be repeated to further amplify the SPR signal (see Note 19).
(a)
(b)
EDC / Primary NHS antibody
AFP in plasma
Secondary Tween20 antibody
700
3500
2500 2000
[AFP] = 500 ng / mL Secondary antibody
1500
1440
1000
1430 1420 1410
500
1400
[AFP] = 50 ng / mL
1390
0
1380 153
0
163
173
183
20 40 60 80 100 120 140 160 180
Time (min)
Angle shift value (mDA)
Angle shift value (mDA)
[AFP] = 500 ng / mL
BSA
3000
600 500 400 300
[AFP] = 50 ng / mL
200
Control (no AFP)
100 0
0
5
10
15
20
25
30
35
Time (min)
Fig. 5. SPR-based immunoassay for AFP in human plasma. a SPR profiles of sequential reactions during AFP detection in human plasma containing 50 or 500 ng/mL AFP (5). After immobilization of primary antibody by EDC/NHS and blocking treatment with BSA, plasma containing AFP was perfused. The sensor surface was washed with PBS-Tween®. Then, a solution of secondary antibody was flowed. An SPR signal shift (155 mDA) was clearly observed when the solution of secondary antibody was applied. The inset shows an SPR signal shift for the plasma containing 50 ng/mL AFP. b Amplification of SPR signal using goat polyclonal antibody against the secondary antibody (rabbit IgG). An SPR sensor surface was exposed to plasma supplemented with AFP ([AFP] = 50 or 500 ng/mL) or without AFP addition. The sensor surface was washed with PBS-Tween®, and then a solution of secondary antibody was flowed. The time course of the SPR profile during application of the solution of anti-rabbit IgG antibody was recorded.
14
Arima et al.
3.2. Surface Plasmon Field-Enhanced Fluorescence Spectroscopy-Based Immunoassay
The basal components of an SPFS apparatus are the same as for an SPR apparatus (Fig. 1.1b), except for the addition of a CCD camera on the sensor surface. When the incident angle approaches the SPR angle, the surface electric field intensity at the water/ metal interface strongly increases, as depicted by the dashed line in Fig. 1.6a (7). Peak intensities relative to the incoming intensity can reach an enhancement factor of ~16 for gold and incident light l = 635 nm. This enhanced field can be utilized to excite a fluorophore to detect fluorophore-labeled secondary antibodies in the SPFS-based immunoassay (Fig. 1.6b) (9). The intensity of fluorescence by SPFS is increased relative to that of the conventional total internal reflection due to the enhanced field in the SPR condition.
3.2.1. Setup of SPFS Apparatus
Figure 1.7a is a photo of our SPFS apparatus. The optical construction of an SPFS instrument is simple as shown schematically in Fig. 1.7b. A laser diode (LD, VHK laser diode module l. = 635 nm, 0.95 mW; Coherent, Santa Clara, CA), a lens (SLB-20-25P, f = 25 mm; Sigma Koki, Tokyo, Japan) for collimation of the laser light, a polarizing filter (TS0851-G; Sugitoh, Tokyo, Japan), and a neutral density filter (AND-20C-10, T = 10%; Sigma Koki) are placed on the same guide rail (OBS200G; Sigma Koki) with appropriate holders. The guide rail with these parts is placed on an arm of a rotational stage (MM-40θ; Chuo Seiki, Tokyo, Japan). Optical parts for detection of the reflected light, such as a lens (SLB-20-25P; Sigma Koki, Japan) and a photodiode detector (S2281-01; Hamamatsu Photonics, Hamamatsu, Japan), are placed on a guide rail and attached to an arm of another rotational stage. These rotational stages are operated by a two-axis stage controller (QT-CM2; Chuo Seiki). Reflectance is determined from the intensities of incident and
(a)
(b) 1
0.6 10 0.4 5
0.2 Incident angle, q
0
Field enhancement
Reflectance
15
Fluorophore-conjugated secondary antibody
Primary antibody
Tumor marker
Self-assembled monolayer
θ
0.8
0
Detector for fluorescence
20
Incident light
Gold thin film Reflected light
Fig. 6. Principle of surface plasmon field-enhanced fluorescence spectroscopy (SPFS) and its application to immunoassays. a Reflectance (solid line) and electric field enhancement (dashed line) relative to the incoming intensity as a function of incident angle q at an Au/water interface. b Schematic representation of SPFS-based sandwich-type immunoassay.
Surface Plasmon Resonance and Surface Plasmon Field-Enhanced
15
(a)
5 cm
(b)
SPFS PC
Stage controller
CCD camera Interference filter (670 nm) Objective (x20)
Glass plate (S-LAL10, Au:Cr = 49:1 nm)
Flow cell Prism (S-LAL10, n = 1.72)
θ
Polarizing filter LD (635 nm, 0.95 mW)
Photodiode
Lens
SPR Rotational stage
Fig. 7. a A surface plasmon field-enhanced fluorescence spectroscopy apparatus and b its schematic representation.
reflected light (voltage), which are converted from currents of the photodiode detector by the same algorithm used for the SPR measurement apparatus. Fluorescent light from fluorophores on a sensor surface is collected by an objective lens (SLWD Plan20×; Nikon, Tokyo, Japan), passed through an interference filter (l = 670 nm, transmittance 75%, FWHM = 7 nm; Optical Coatings Japan, Tokyo, Japan) fixed in a extension barrel (TS0155-H; Sugitoh, Tokyo, Japan), and then guided to a CCD camera with a charge multiplier (MC681-SPD; Texas Instruments, Dallas, TX) (see Note 1). A fluorescence image is acquired by an image capture board (MT-PCI2; Micro-Technica, Tokyo, Japan) and analyzed by homemade intensity scanning software (see Note 2). The light intensity in a fixed area (100 × 100 pixels) at the center of the laser light is monitored for a certain integration time, and the average intensity is recorded.
16
Arima et al.
The glass plate (S-LAL10, refractive index: 1.720; Sigma Koki) with a thin gold film (49 nm in thickness) is coupled to a triangular prism (25 × 25 × 25 mm, S-LAL10; Sigma Koki) with an index-matching fluid (n = 1.72, Cargille Laboratories). The flow cell illustrated in Fig. 1.4 is assembled on the glass plate. Instead of a hemicylindrical prism and a PVC plate, a triangular prism and a transparent PMMA plate are employed (see Note 20). 3.2.2. Preparation of Working Solutions
1. Phosphate-buffered saline (PBS) consists of 10 mM phosphate buffer with 140 mM NaCl and 3 mM KCl in pure water (pH 7.4). 2. Anti-AFP antibody used as a secondary antibody (1D5, mouse monoclonal; Japan Clinical Laboratories, Inc., Kyoto, Japan) is dissolved at 2.5 mg/mL in PBS and stored at 4°C. 3. Anti-AFP antibody used as a secondary antibody (6D2, mouse monoclonal; Japan Clinical Laboratories, Inc.) is dissolved at 2.5 mg/mL in PBS and stored at 4°C. 4. The secondary antibody is conjugated with Alexa Fluor 647 dye according to the standard protocol of a labeling kit from Molecular Probes (Eugene, OR), and the concentration of conjugate is 0) one can deduce that the increase of r2 at 1,499s to 1,500s may be the result of either the adlayer thickness increase or a bulk RI increase. But to exclude the latter interpretation we have decreased the RI of the biotin solution in respect to RI of the PBS by adding 20 μL of the pure water in 1 mL of PBS with biotin (before the biotin solution injection). Since nPBS − nH2O ≅ 0.0012, the change of the RI of the biotin solution (due to water injection) in respect of the RI of PBS is Δn ≅ −2 × 10−5, while the biotin itself in such a small concentration practically does not change the RI of the PBS (Δnbiotin < 10−7). This procedure was needed for us to be sure that the injection of the biotin solution will cause only the bulk RI to decrease. Therefore, the increase of r2 at 1,499s to 1,500s may be the result of only the adlayer thickness increase. The exact value of the corresponding calculated adlayer thickness increase Δd (but not its sign) in Fig. 4c depends on the exactness of the numerical values of the coefficients pointed
Optical Biosensors Based on Photonic Crystal Surface Waves
63
after Eq. 4. But, inasmuch as the expected bulk RI change Δn ≅ −2 × 10−5 corresponds well to the results from Fig. 4d (calculated with the same coefficients), we expect that the calculated value of the adlayer thickness increase Δd is also reasonably accurate. It may be noted here that both changes of buffer RI in Fig. 4d are in good agreement with the calculated values. During biotin injection1, 1 mL of biotin solution in PBS with 20 μL of pure water was added into the flow cell system (0.2 mL of pure PBS was at this time in the system) – the expected decrease of RI is equal to 2 × 10−5. After biotin injection 2, the same amount (1 mL of biotin solution in PBS with 20 μL of pure water) was added to 0.2 mL of the solution RI of which already was decreased. The calculated decrease of RI in this case equal to 3 × 10−6 is in good agreement with the data in Fig. 4d. 2. Measurement noises and mass detection limits. In compliance with (26), we suppose that the process of the biotin–streptavidin binding is a good candidate for comparison of the signal/ noise ratio of different label-free techniques. We believe that for comparison of the signal/noise ratio it is also very important always to point out the time of the measurement and the fact of posterior data averaging and/or smoothing (which increase the effective measurement time). In other words, the noise should be reduced to 1/√Hz value. In our experiments, the signal accumulation time was 1 second per point and no posterior data averaging or smoothing was done. The noise (i.e., standard deviation – std) of the thickness measurement was equal δd = std(da) ≅ 1.3 pm/√Hz. The noise of the measurement of the external medium RI was δn = std(ne) ≅ 5 × 10−7/√Hz. In Fig. 4c, one can see that we detected the streptavidin conformation process during free biotin binding with a (signal/noise)b ratio of about 15. The deposition of the streptavidin monolayer was detected with a (signal/noise)str ratio of about 5,000. Taking into account the (signal/noise) ratio, we obtain a minimal quantity of streptavidin str molecules that may be detected at the probed spot of our str N min = Nstr/(signal/noise)str ⯝ 200 000 streptavidin setup: molecules. This corresponds to a mass detection limit str str mmin = N min M str / N A @ 2 ´ 10 -14 g = 20 fg of the analyte on the probed spot of the surface (NA ≅ 6 × 1023 is Avogadro’s number). For biotin molecules we have a minimal detectable b str 8 quantity equal to N min = 2N / (signal/noise)b @ 1.3 ´ 10 b b biotin molecules or mmin = N min M b / N A @ 50 fg of the analyte on our probed spot. We believe that the noise of this technique could be further decreased by improving the quality of the dielectric multilayer coating and by decreasing the laser noise.
64
Konopsky and Alieva
Acknowledgement The authors thank S. Grachev for the kind donation of some biochemicals and for helpful advice about surface preparation. This work was partly supported by the European Network of Excellence, NMP3-CT- 2005-515703-2. References 1. Robinson, G. (1995) The commercial development of planar optical biosensors. Sens. Actuators B 29, 31–36 2. Cooper, M. A. (2003) Label-free screening of bio-molecular interactions. Anal. Bioanal. Chem. 377, 834–842 3. Homola, J., Yee, S. S. and Gauglitz, G. (1999) Surface plasmon resonance sensors: review. Sens. Actuators B 54, 3–15 4. Raether, H. (1988) Surface Plasmons. Springer, Berlin 5. Cush, R., Cronin, J., Stewart, W., Maule, C., Molloy, J. and Goddard, N. (1993) The resonant mirror – a novel optical biosensor for direct sensing of biomolecular interactions. I. Principle of operation and associated instrumentation. Biosens. Bioelectron. 8, 347–353 6. Alieva, E. V. and Konopsky, V. N. (2004) Biosensor based on surface plasmon interferometry independent on variations of liquid’s refraction index. Sens. Actuators B 99, 90–97 7. Slavík, R., Homola, J. and Vaisocherová, H. (2006) Advanced biosensing using simultaneous excitation of short and long range surface plasmons. Meas. Sci. Technol. 17, 932–938 8. Cross, G., Reeves, A., Brand, S., Swann, M., Peel, L., Freeman, N. and Lu, J. (2004) The metrics of surface adsorbed small molecules on the Young’s fringe dual-slab waveguide interferometer. J. Phys. D Appl. Phys. 37, 74–80 9. Yablonovitch, E. (1993) Photonic band-gap structures. J. Opt. Soc. Am. B 10, 283–295 10. Kossel, D. (1966) Analogies between thinfilm optics and electron band theory of solids. J. Opt. Soc. Am. 56, 1434–1434 11. Yeh, P., Yariv, A. and Hong, C.-S. (1977) Electromagnetic propagation in periodic stratified media. I. General theory. J. Opt. Soc. Am. 67, 423–438 12. Yeh, P., Yariv, A. and Cho, A. Y. (1978) Optical surface waves in periodic layered media. Appl. Phys. Lett. 32, 104–105 13. Robertson, W. M. and May, M. S. (1999) Surface electromagnetic waves on one-dimensional photonic band gap arrays. Appl. Phys. Lett. 74, 1800–1802
14. Villa, F., Regalado, L., Ramos-Mendieta, F., Gaspar-Armenta, J. and Lopez-Rios, T. (2002) Photonic crystal sensor based on surface waves for thin-film characterization. Opt. Lett. 27, 646–648 15. Li, J., Wang, H., Zhao, Y., Cheng, L., He, N. and Lu, Z. (2001) Assembly method fabricating linkers for covalently bonding DNA on glass surface. Sensors 1, 53–59 16. Palik, E. D. (1985) Handbook of Optical Constants of Solids. Academic, London 17. http://www.laser-export.com 18. http://www.ofr.com 19. http://www.hamamatsu.com 20. Konopsky, V. N. and Alieva, E. V. (2006) Long-range propagation of plasmon polaritons in a thin metal film on a one-dimensional photonic crystal surface. Phys. Rev. Lett. 97, 253904 21. http://www.cdpsystems.com 22. Elimelech, M. (1994) Particle deposition on ideal collectors from dilute flowing suspensions: Mathematical formulation, numerical solution, and simulations. Sep. Technol. 4, 186–212 23. Myszka, D. G., He, X., Dembo, M., Morton, T. A. and Goldstein, B. (1998) Extending the range of rate constants available from BIACORE: Interpreting mass transport-influenced binding data. Biophys. J. 75, 583–594 24. Hyre, D. E., Trong, I. L., Merritt, E. A., Eccleston, J. F., Green, N. M., Stenkamp, R. E. and Stayton, P. S. (2006) Cooperative hydrogen bond interactions in the streptavidin– biotin system. Protein Sci. 15, 459–467 25. Freitag, S., Trong, I. L., Klumb, L., Stayton, P. S. and Stenkamp, R. E. (1997) Structural studies of the streptavidin binding loop. Protein Sci. 6, 1157–1166 26. Zybin, A., Grunwald, C., Mirsky, V. M., Kuhlmann, J., Wolfbeis, O. S. and Niemax, K. (2005) Double-wavelength technique for surface plasmon resonance measurements: Basic concept and applications for single sensors and two-dimensional sensor arrays. Anal. Chem. 77, 2393–2399
Chapter 5 Surface Plasmon Resonance Biosensing Marek Piliarik, Hana Vaisocherová, and Jirˇí Homola Summary Surface plasmon resonance (SPR) biosensors belong to label-free optical biosensing technologies. The SPR method is based on optical measurement of refractive index changes associated with the binding of analyte molecules in a sample to biorecognize molecules immobilized on the SPR sensor. Since late 1990’s, SPR biosensors have become the main tool for the study of biomolecular interactions both in life science and pharmaceutical research. In addition, they have been increasingly applied in the detection of chemical and biological substances in important areas such as medical diagnostics, environmental monitoring, food safety and security. This chapter reviews the main principles of SPR biosensor technology and discusses applications of this technology for rapid, sensitive and specific detection of chemical and biological analytes. Key words: Optical biosensors, Affinity biosensing, Biorecognition elements, Detection of chemical and biological species, Bioassays.
1. Introduction Optical affinity biosensors based on surface plasmon resonance (SPR) present one of the most advanced label-free optical sensing technologies (1–3). Their ability to monitor the interaction between a molecule immobilized on the surface of the sensor and the interacting molecular partner in a solution have made SPR sensors a very powerful tool for biomolecular interaction analysis and biomolecular research in general (4). In recent years, SPR biosensors have been increasingly used also for the detection of chemical and biological substances related to medical diagnostics, environmental monitoring, food safety and security (5).
Avraham Rasooly and Keith E. Herold (eds.), Methods in Molecular Biology: Biosensors and Biodetection, Vol. 503 © Humana Press, a part of Springer Science + Business Media, LLC 2009 DOI: 10.1007/978-1-60327-567-5_5
65
66
Piliarik, Vaisocherová, and Homola
This chapter deals with SPR biosensor technology and its application in the quantification of chemical and biological species. 1.1. Principles of SPR Biosensing
SPR affinity biosensors are devices that consist of three main subsystems: sensor hardware (optical reader), biorecognition element, and sample preparation and delivery system (Fig. 1). In the optical reader of an SPR sensor, a light wave excites a special mode of electromagnetic field – surface plasmon (SP). Surface plasmon propagates along a thin metal film and its field probes the medium adjacent to the metal surface. Any change in the refractive index in the proximity of the metal surface results in a change in the velocity of the surface plasmon. This change in the propagation can be determined from the characteristics of the light wave coupled to the surface plasmon. Biorecognition elements specific to analyte molecules are immobilized on the surface of the metal. If a liquid sample is brought in contact with the sensor surface, molecules of analyte are captured by biorecognition molecules (Fig. 2). The binding gives rise to a refractive index change close to the sensor surface, which can be measured by the optical reader. The liquid sample is brought to the sensor surface using a sample preparation and a delivery system.
Optical reader change of refractive index Biorecognition element Interaction with analyte molecules Sample delivery system Fig. 1. Principal components of SPR affinity biosensor.
Surface plasmon
Sample with analyte
Biorecognition elements, n
Biorecognition elements, n+δn
Metal film
Metal film
Optical reader
ve
a tw
gh
Li
Fig. 2. Principle of operation of SPR affinity biosensors.
SP
Li
gh
tw
av
e
Surface Plasmon Resonance Biosensing
67
The following sections present the underlying principles and characteristics of the main parts of an SPR affinity biosensor. 1.2. Optical Reader
Nowadays, numerous SPR readers are available as commercial systems or research prototypes. Although their designs may differ considerably, the underlying physical principles of the SPR method remain the same. Therefore, the first part of this section is dedicated to fundamentals of surface plasmons and SPR sensing while the following part illuminates possible differences among the SPR sensing platforms and gives examples of the most frequent SPR sensing platforms.
1.2.1. Surface Plasmons and Their Optical Interrogation
A surface plasmon is an electromagnetic wave, which propagates along an interface between a dielectric and a metal and is characterized by the propagation constant and electromagnetic field distribution (6). The field of a surface plasmon is transversemagnetic (TM) polarized (the vector of magnetic intensity is perpendicular to the plane of incidence) and evanescently decays into both the media while the major part of the field is located in the dielectric. The propagation constant of SP bSP is determined by the optical constants of the surrounding media (permittivity eM of the metal and the refractive index of the dielectric nD) as follows: b SP =
e MnD2 w w nef = , c c e M + nD2
(1)
where w is the angular frequency, c is the speed of light in vacuum and nef denotes the effective refractive index of the surface plasmon. Although several metals can support surface plasmons at optical frequencies, all the main existing SPR (bio)sensors use gold due to its excellent chemical stability. The electro-magnetic field of the SP is strongly localized at the metal surface and its penetration depth into the dielectric Lpd (see Note 1) is typically 150–400 nm depending on the operating wavelength (Fig. 3). A change in the refractive index of the dielectric within a distance h from the metal surface produces a change in the effective refractive index of the surface plasmon nef. The magnitude of the change depends on the thickness of the layer h within which the refractive index change occurs, operating wavelength, and a refractive index distribution. If the thickness of the layer is much higher than that of the penetration depth Lpd of the SP field, the change in the effective refractive index of the surface plasmon can be calculated as Dnef =
nef3 (DnD )h > > L pd. nD3
(2)
Piliarik, Vaisocherová, and Homola 1.0
Magnetic field intensity [arb. u.]
68
Metal
Dielectric
Wavelength
0.8
630 nm 760 nm 850 nm
0.6
0.4 1/e
0.2
0.0 0.0
0.2 0.4 0.6 0.8 Perpendicular coordinate [μm]
1.0
Fig. 3. Distribution of electromagnetic field of a SP at the gold–water interface for three different wavelengths.
The change in the effective refractive index of a surface plasmon due to a refractive index change within a thin layer h 5. If the pH of the buffer is less than the isoelectric point (PI) of a protein, the protein will exhibit a net positive charge so it will promote its electrostatic pre-concentration into the dextran prior to covalent immobilization. The CMD enables simple coupling of amino-containing molecules to the carboxylate group of the matrix through succinimidyl ester chemistry resulting in the formation of a peptide bond between the amino-containing molecules and the matrix (Fig. 3a). The advantage of the CMD coated surface is that the 3-D nature of the hydrophilic matrix increases the ligand immobilization loading of the system compared to a planar layer. In addition, it keeps immobilized ligands away from direct contact with the surface, resulting in greater retention of the biological activity of biomolecules, reducing protein denaturation and non-specific adsorption. The disadvantage of this surface is that the orientation of immobilized recognition molecules will not be uniform due to the random nature of the immobilization chemistry. The second disadvantage is that this matrix can only be used for small analytes as the large analytes such as cells may be size-excluded. Thirdly, the molecule’s binding will not be sensed equally in the matrix due to the exponential nature of the evanescent field. In addition, from the kinetics point of view, molecules bound to sites farthest from the surface may sterically hinder the binding of other molecules to sites within the matrix (36, 37).
3.3.1. Procedure for Dextran Coating
1. Silanized chips are immersed overnight in 23% (w/w) dextran solution in 0.1 M NaOH (see Note 1). 2. The following day the chip is washed extensively with water to remove excess dextran. 3. The chip is treated with 1 M bromoacetic acid in 2 M NaOH for 6 h. 4. Rinse the chip extensively with water and dry; then it is ready to use.
3.3.2. Procedure for the Immobilization of AmineContaining Ligand to CMD
This protocol describes the immobilization of the protein on a dextran coated cuvette in the IAsys instrument. 1. Insert a CMD cuvette into the instrument and wash with PBST for 10 min for equilibration.
OH
(a) NHS
N
O
O NH2
+
EDC
CH3
Cl
N C N
O
+
O
OH O
+
NH
O O
OH OH
O
N
CH3
O
O
OH
OH OH
OH
NH
O
O
O
OH
O
OH OH
OH
OH
OH
CMD coated sensor surface O
(b)
O
N
S
N
O
O
Amino-modified CM-Dextran O
NH2
O
HN
HO
NH
O
O
O OH
HN
HO
OH
O
O
OH
O OH
NH
HO
OH
OH
O
Sulpho-SMCC
O OH
NH
O
O H
OH
SH2
CMD coated sensor surface
(c) (1)
H
Basic pH
OH
O
O
H
O
O
O
H
Glutaraldehyde
polymerized glutaraldehyde
(2) H
OH
O
O
O
H
(c)
O
H
+
H
O HN
H H
O
O
O
H NH2 NH2
NH2
NH2
H
NH2 NH
O
O
O
H
H NH2
NH2
H NH2
NH2 NH
NH2
NH2
Fig. 3. (a) Reaction scheme for carboxymethyl dextran with EDC/NHS for the immobilization of amine containing ligands. (b) Reaction scheme for Sulpho-SMCC modified immobilization of thiol containing ligand. (c) Reaction scheme (A) polymerization of glutaraldehyde and (B) the cross-linking of a ligand to an amino modified sensor surface.
106
Zourob et al. OH
(d) NHS
N
O
O
+ EDC
Cl
N C N
CH3
NH
+ O
HO OH
CH3
O N O
OH
O
O OH
OH
(e)
O
O
OH
S
O
N
O
O
HN
OH
OH
O
NH2
OH O
O
O
N
NH O
OH OH
O
O
NH
O
NH2
OH
O
HN
NH O
OH O
OH
Sulpho-SMCC
O
NH O
OH O
OH
SH2
Amino-modified carboxyle groups
(f)
Biotinylated ligand
Streptavidin
Biotinylated surface
Fig. 3. (continued) (d) Reaction scheme for carboxylate surface with EDC/NHS for the immobilization of amine containing ligands. (e) Reaction scheme for sulpho-SMCC mediated immobilization of thiol containing ligands. (f) Schematic diagram of capturing biotinylated ligand onto biotin functionalized surface via streptavidin.
Label-Free Detection with the Resonant Mirror Biosensor
(g)
107
O N O O N
OH
O
O
N S
S
O
N S S
+
S S
O
NH2 NH2 NH2
NH2
NH2 HN
NH2
O
NH2
HS
NH2 HN
NH2
NH2
Fig. 3. (continued) (g) Schematic reaction diagram of surface amines with SPDP, and subsequent reaction with free, accessible thiol from protein. (h) Formation of a lipid monolayer from solution onto hydrophobic sensor surface.
2. Start data acquisition and collect baseline data for 3 min. 3. Introduce a mixture of 200 μL of EDC and 200 μL of NHS for 7 min (see Notes 2 and 6). 4. Wash with PBST and collect baseline data for 3 min (see Note 7). 5. Change the PBST to 10 mM acetate buffer pH 5.5 to get a baseline (see Note 8). 6. Introduce IgG in acetate buffer to the cuvette for 5 min to begin protein electrostatic uptake prior covalent binding. 7. Wash with PBST and collect baseline data for 3 min. 8. Block the unreacted NHS-ester with 1 M ethanolamine pH 8.5 for 3 min (see Note 9). 9. Wash with PBST to get the baseline. 10. Wash the cuvette with 10 mM HCl for 3 min to remove the non-covalently bound ligand to the matrix. 11. Wash with PBST and collect baseline data for 3 min. 12. Calculate the amount of immobilized protein by subtracting the baseline level at step 4 from that at step 11.
108
Zourob et al.
3.3.3. Procedure for the Immobilization of Thiol-Containing Ligand to CMD
Protocol
This protocol describes a method for immobilization of a thiol containing ligand to CMD coated sensor surface as shown in Fig. 3b. This approach will include the modification of the CMD with a primary amine, followed by reaction with an amine–thiol reactive heterobifunctional cross-linking agent Sulfosuccinimide 4-(N-maleimidomethyl)-cyclohexane-1-carboxylate (Sulfo-SMCC). Covalent immobilization is achieved by the direct addition of thiol-bearing ligands to the modified sensor surface. Thiol groups offer a good approach to achieving orientation of a ligand, especially with cysteine incorporation into one terminal of a peptide. 1. Insert CMD cuvette into the instrument and wash with 200 μL of PBST and allow 10 min for equilibration. 2. Start data acquisition and collect 3 min of baseline data. 3. Incubate the surface with 50 μL of a mixture of 1:1 (v/v) EDC/ NHS for 10 min (see Notes 2 and 6). 4. Wash with 200 μL of PBST and collect baseline data for 3 min. 5. Incubate the surface again twice with 50 μL of a mixture of 1:1 (v/v) EDC/ NHS for 7 min. 6. Wash with 200 μL of PBST and collect baseline data for 3 min. 7. Incubate the surface with 50 μL of 1 M ethylene diamine pH 8.5 for 10 min to modify the carboxymethyl groups with amines. 8. Wash with 200 μL of PBST and collect baseline data for 2 min. 9. Activate the amine group with Sulfo-SMCC by washing and incubating the surface with 200 μL of 1 mg/mL SulfoSMCC for 10 min. 10. Wash with 200 μL of PBST and collect baseline data for 2 min. 11. Incubate the surface with 50 μL of Cys-GM-CSF to covalent binding for 8 min. 12. Wash with 200 μL of PBST and collect baseline data for 2 min. 13. Calculate the amount of immobilized protein by subtracting the baseline level after step 10 from that after step 12.
3.4. Aminosilane Functionalized Surface
The biosensor surface is coated with an aminosilane (AS) derivative, resulting in the introduction of amino groups to the surface via an alkyl chain linker. The amino derivatized surfaces are particularly useful for immobilization of ligands with PI < 4, where methodologies based on electrostatic concentration into a CMD matrix are not effective. The advantage of this chemistry is that
Label-Free Detection with the Resonant Mirror Biosensor
109
it allows large ligates such as membrane fragments, liposomes, viruses and cells which are unable to penetrate the 3D-hydrogel matrix to enter the most sensitive area of the evanescent field. Highly charged molecules can be used with this surface, which might be electrostatically repelled from the CMD matrix. This surface can be activated with cross-linkers such as glutaraldehyde or bis(sulfosuccinimidyl)suberate. The disadvantage of this method is that there is very little electrostatic concentration of ligand to the surface unlike the charged CMD matrix. This requires higher concentrations of ligand to give satisfactory surface coverage. Typically, concentrations in the range of 0.1–1 mg/mL are required. This protocol will describe immobilization of ligands via their amino groups to the surface amino-functionalized groups using polymerized glutaraldehyde (Fig. 3c). 3.4.1. Procedure for Polymerizing the Glutaraldehyde
1. Polymerized glutaraldehyde (PG) is prepared by adding 500 μL of 0.1 M NaOH to 5 mL of 5% (v/v) glutaraldehyde as shown in Fig. 3c(A). 2. Leave for 30 min to polymerize before neutralizing with 500 μL 0.1 M HCl. Presence of PG can be determined from a wavelength scan between 200 and 350 nm. An absorption peak is seen around 234 nm, which represents the polymer, whereas a peak at 280 nm is the monomer. The resulting polymerized glutaraldehyde solution must be stored at or below −18 °C.
3.4.2. Immobilization Procedure
1. Insert new AS cuvette into the instrument and wash with PBS for 10 min for equilibration (see Note 10). 2. Start data acquisition and collect 3 min of baseline data. 3. Introduce PG to the cuvette for 30 min. 4. Wash with PBS to get stable baseline. 5. Introduce the ligand solution 100 μg/mL and leave it for about 30 min (see Note 11). 6. Wash the cuvette with 1 M formic acid for 2 min to remove the non-covalently bound ligand to the surface. 7. Wash with PBS for about 5 min to get the baseline. 8. Wash the cuvette with 1 mg/mL BSA for 5 min to block unreacted PG (see Note 12). 9. Wash with PBS to get the baseline. 10. Calculate the amount of immobilized protein by subtracting the baseline level after step 4 from that after step 9.
3.5. Carboxylate Surface
The planar carboxylate surface allows the user to analyze biomolecular interactions in the absence of the 3-D CMD hydrogel layer. This surface is useful for large molecules (> 1,000 KDa),
110
Zourob et al.
particles or cells binding to the recognition molecules closer to the surface, where the evanescent field is more intense. Amino and thiol containing ligands can be immobilized on the carboxylic acid functionalized surface. 3.5.1. Immobilization of Amine-Containing Ligand to Carboxylate Surface
Protocol
The biosensor surface contains carboxylic acid functional groups that can react with the primary amine of the ligand via amide bond using EDC/NHS chemistry (Fig. 3d) as described in Subheading 3.13.1. 1. Insert carboxylate cuvette into the instrument and wash with PBS for 10 min for equilibration. 2. Start data acquisition and collect 3 min of baseline data. 3. Introduce a mixture of 200 μL of EDC and 200 μL of NHS for 7 min (see Notes 2 and 6). 4. Wash with PBST to get stable baseline. 5. Change the PBST to 10 mM acetate buffer pH 5.5 to get stable baseline (see Note 8). 6. Introduce RAMFc in acetate buffer to the cuvette for 10 min to begin electrostatic uptake of protein prior covalent coupling. 7. Wash with PBS for 3 min to get stable baseline (see Note 11). 8. Block un-reacted NHS-ester and uncovered surface with 2 mg/mL β-casein for 3 min. 9. Wash with PBS and collect 3 min of baseline data (see Note 12). 10. Wash residual non-covalently bound ligand with 1 M formic acid for 2 min. 11. Wash with PBS for 3 min to get the baseline. 12. Calculate the amount of immobilized protein by subtracting the baseline level after step 4 from that after step 11.
3.5.2. Immobilization of Thiol-Containing Ligand to Carboxylate Surface
Protocol
This protocol describes a method for immobilization of a thiol containing ligand to carboxylate functionalized sensor surface as shown in Fig. 3e. This approach includes modification of the carboxylate modified surface with a primary amine, followed by a reaction with an amine–thiol reactive heterbifunctional cross-linking agent Sulfosuccinimide 4-(N-maleimidomethyl)cyclohexane-1-carboxylate (Sulfo-SMCC). Covalent immobilization is achieved by the direct addition of thiol-bearing ligands to the modified sensor surface. Thiol groups offer a good approach to achieving orientation of a ligand, especially with cysteine incorporation into one terminal of a peptide. 1. Insert new carboxylate cuvette into the instrument and wash with 200 μL of PBS and allow 10 min for equilibration. 2. Start data acquisition and collect 3 min of baseline data.
Label-Free Detection with the Resonant Mirror Biosensor
111
3. Wash and incubate the surface with 50 μL of a mixture of 1:1 (v/v) EDC/ NHS for 10 min (see Notes 2 and 6). 4. Wash with 200 μL of PBS and collect baseline data for 3 min. 5. Wash twice again and incubate the surface with 50 μL of a mixture of 1:1 (v/v) EDC/ NHS for 7 min. 6. Wash with 200 μL of PBST and collect baseline data for 3 min. 7. The activated carboxyl group is then converted to amines by washing and the surface is incubated with 50 μL of 1 M ethylene diamine pH 8.5 for 10 min. 8. Wash with 200 μL of PBS and collect baseline data for 5 min. 9. Block any remaining activated carboxyl groups by washing twice and incubating the cuvette with 200 μL of 1 M Tris pH 8.5 for 5 min. 10. Wash with 200 μL of PBS and collect baseline data for 5 min. 11. Wash extensively with 200 μL of PBST and collect baseline data for 5 min. 12. Activate the amine group with Sulfo-SMCC by washing and incubating the surface with 200 μL of 1 mg/mL SulfoSMCC for 10 min. 13. Wash with 200 μL of PBST and collect baseline data for 2 min. 14. Incubate the surface with 50 μL of Cys-GM-CSF to covalent binding for 10 min. 15. Wash with 200 μL of PBST and collect baseline data for 2 min. 16. Calculate the amount of immobilized protein by subtracting the baseline level after step 13 from that after step 15. 3.6. Biotin Surface
3.6.1. Immobilization Procedure of Streptavidin
This method is widely used for immobilization of oligonucleotides, DNA, antibodies, peptides and other proteins. Usually, in this process biotinylated biomolecules are attached to the biotin functionalized sensor surface by depositing avidin or streptavidin or NeutrAvidin before introducing the biotinylated biomolecules (Fig. 3f). The advantage of this method is the convenient and rapid immobilization of biotinylated molecules. Since the interaction between the avidin and biotin is non-covalent, the surface can be regenerated by completely removing the avidin-captured molecules. Two examples will be described here, of immobilizing biotinylated protein ligands (Subheading 3.6.2), and biotinylated DNA (Subheading 3.15) onto a biotin-functionalized surface using streptavidin. 1. Insert new biotin cuvette into the instrument and wash with PBST for 10 min for equilibration (see Note 13).
112
Zourob et al.
2. Start data acquisition and collect baseline data for 3 min. 3. Introduce avidin in PBST and allow the binding to occur for 10 min (see Note 14). 4. Wash with PBST to get a baseline (see Note 15). 5. Calculate the amount of immobilized avidin by subtracting the baseline level after step 2 from that after step 4. 3.6.2. Immobilization Procedure for Biotinylated Ligand
The following steps begin after the streptavidin coating and washing steps as described in Subheading 3.6.1. 6. Introduce biotinylated protein G in PBST and allow the binding to occur for about 5 min. 7. Wash with PBST for 3 min to establish a baseline. 8. Introduce IgG in PBST and allow binding to occur for about 5 min. 9. Wash with PBST for 5 min to get a baseline. 10. Calculate the amount of immobilized protein G by subtracting the baseline level after step 5 from that after step 7.
3.6.3. Immobilization of Thiol Containing Ligands to Avidin
Protein Preparation
This protocol describes a method for the immobilization of proteins through available surface thiol groups using an amine-to-thiol heterobifunctional linker to the amine groups in an immobilized avidin protein layer (Fig. 3g). The remaining carboxylates on the avidin layer then serves to electrostatically attract a protein ligand. Covalent coupling through its free thiol groups then follows. The use of this protein layer in the immobilization of biomolecules allows ligands of interest to be coupled through a variety of crosslinking chemistries. The advantage of using this protein layer is that it is suitable for the immobilization of biomolecules of all sizes and reduces the non-specific binding. 1. Dissolve 500 μL of 0.5 mg/mL IgG in HBSTE. 2. Reduce the IgG protein to have free thiol by addition of DTT to a final concentration of 10 mM, and room temperature incubation for 30 min. 3. Desalt the product using a D-salt™ column (Pierce, 43230) using HBSTE as the pre-equilibrium and running buffer.
Desalting Protein
1. Equilibrate the column with 200 mL of HBSTE at room temperature, then add 500 μL of IgG mixture to the column head. 2. After loading, fill the funnel with HBSTE. Collect a number of fractions of drops (260 μL) each; the protein will be present in fraction numbers 6–9. 3. Wash the columns with 150 mL of HBSTE to remove lowmolecular weight compounds prior to re-use.
Label-Free Detection with the Resonant Mirror Biosensor Immobilization Procedure
113
1. nsert new biotin cuvette into the instrument and wash with 200 μL of HBSTE and allow 10 min for equilibration. 2. Start data acquisition and collect 3 min of baseline data. 3. Introduce 40 μL of avidin in HBSTE in the cuvette and allow binding to occur for about 3 min. 4. Wash with 200 μL of HBSTE and allow excess avidin to dissociate for about 3 min. 5. Wash extensively with 200 μL of 100 mM HCl for 2 min followed by washing with 150 μL of HBSTE for 3 min. 6. Add 60 μL of SPDP in HBSTE in the cuvette and allow the reaction to continue for 10 min. 7. Wash with 200 μL of HBSTE and collect baseline data for 3 min. 8. Wash with 200 μL of 10 mM acetate buffer pH 5.5, and establish baseline data for about 3 min. 9. Introduce 20 μL of protein containing free thiol (reduced IgG in this example) and incubate it for 15 min or desired level. 10. Wash with 200 μL of HBSE and collect baseline data for 3 min. 11. Wash with 200 μL of 100 mM HCl and incubate for 2 min to remove excess ligand. 12. Wash with 200 μL of HBSTE and collect baseline data for 3 min. 13. Add 600 ng/μL of a specific antibody to detect the level of binding to check if the immobilized protein is still active towards its cognate binding partner. 14. Then wash the free antibody from the cuvette and allow it to dissociate for 10 min. 15. Treat the surface with 100 mM HCl for 2 min to remove the bound antibody. 16. Wash and refill the cuvette with HBSTE to reveal a new baseline which matches that at step 14.
3.6.4. Immobilization of Amine Containing Ligands to Avidin
This protocol describes a method for the immobilization of proteins through available surface amino groups using an amine-to-thiol heterobifunctional linker to the amine groups in an immobilized avidin protein layer. The remaining carboxyls on the avidin layer then serve to electrostatically attract a protein ligand. Depending on the mode of activation, covalent coupling may then proceed through thiol groups as described in Subheading 3.6.3, or through the amine as described in this protocol. As previously mentioned, the use of a protein layer in the immobilization of biomolecules allows ligands of interest of all sizes to be coupled through a variety of cross-linking chemistries and reduces the non-specific binding.
114 Protocol
Zourob et al.
1. Insert new biotin cuvette into the instrument and wash with 200 μL of HBSTE and allow 10 min for equilibration. 2. Start data acquisition and collect 3 min of baseline data. 3. Introduce 40 μL of avidin in HBSTE in the cuvette and allow binding to occur for about 3 min. 4. Wash with 200 μL of HBSTE and incubate to allow excess avidin to dissociate for about 3 min. Wash again with 200 μL of HBSTE to establish a baseline. 5. Block biotin binding sites by addition of 20 μL of 5 mM biotin for 1 min. 6. Wash extensively with 200 μL of 100 mM HCl for 2 min followed by washing with 150 μL of HBSTE for 3 min and incubate for 3 min to allow excess avidin to dissociate. Wash again with 200 μL of HBSTE to establish a baseline. 7. Add 60 μL of SPDP in HBSTE in the cuvette and allow the reaction to continue for 10 min. 8. Wash with 200 μL of HBSTE and collect baseline data for 3 min. 9. Wash with 200 μL of 5 mM DTT, and allow thiol reduction to proceed for 7 min. 10. Wash with 200 μL of HBSTE and collect baseline data for 3 min. 11. Repeat the addition of 200 μL of SPDP and allow the activation to continue for 10 min. 12. Wash with 200 μL of HBSTE and collect baseline data for 3 min. This baseline will be used as starting place for the calculation of net immobilized protein. 13. Wash with 200 μL of 10 mM acetate buffer pH 5.5, and establish baseline data for about 3 min. 14. Introduce 20 μL of protein (1 μg IgG as example for electrostatic uptake), and allow reaction to occur for about 15 min or desired level. 15. Wash with 200 μL of HBSE and collect baseline data for 3 min. 16. Wash with 200 μL of 100 mM HCl and incubate for 2 min to remove excess ligand. (Optional step as might be the protein will not withstand this treatment). 17. Wash with 200 μL of HBSTE and collect baseline data for 3 min. 18. Wash with 200 μL of 1 M ethanolamine, pH 8.5 for 3 min to block residual NHS esters. 19. Wash with 200 μL of HBSTE and establish baseline data for about 3 min.
Label-Free Detection with the Resonant Mirror Biosensor
115
20. Add 100 ng/μL of a specific antibody to detect the level of binding to check if the immobilized protein still active towards its cognate binding partner. 21. Then wash the free antibody from the cuvette and allow to dissociate for 5 min. 22. Treat the surface with 100 mM HCl for 2 min to remove the bound antibody. 23. Wash and refill the cuvette with HBSTE to reveal a new baseline. 3.7. Immobilization of Phospholipids on Hydrophobic Surface
Protocol
The RM hydrophobic cuvette provides a substrate onto which hydrophobic molecules can be easily deposited. Lipids can be simply deposited from a solution avoiding the preparation of liposomes (Fig. 3h). This surface is used for the immobilization of lipids from solutions, obviating the requirement for liposomes (38). The surface is simply exposed to the required mixture of lipids in the solvent, the solvent is diluted out by the addition of buffer, and the lipids are forced to assemble on the surface. Before starting this protocol, it is essential to clean the IAsys instrument with ultra pure water, and aqueous solution containing 1 mM NaOH, 1% (w/w) SDS, 1% Tween 20, followed by extensive washing with water. To obtain reproducible results it is essential that the lipids used are not oxidized. It is recommended that stock lipid solutions contain butylated hydroxytoluene 0.1% (w/v) and are stored at less than −60 °C for no longer than 2 weeks. 1. Insert a new hydrophobic cuvette into the instrument and wash extensively with 250 μL of 2-propoanol. 2. Start data acquisition and collect 3 min of baseline data. 3. Wash extensively with 200 μL of PBS/AE and collect data for about 10 min (see Note 16). 4. Wash with 300 μL of 2-propanol and collect data for about 3 min. 5. Add 30 μL of DOPC and leave for 2 min (see Note 17). 6. Wash rapidly with 320 μL of PBS/AE and collect data for about 6 min (see Note 18). 7. Wash with 250 μL of 0.1 M HCl for 5 min. 8. Wash with 350 μL of PBS/AE and collect data for about 5 min. 9. Wash with 250 μL of NaOH and leave for 1 min. 10. Wash with 370 μL of PBS/AE and collect data for about 5 min. Measure the amount of lipid on the surface (see Note 19). 11. Add BSA solution to a final concentration of 1.5 mg/mL.
116
Zourob et al.
12. Wash with 350 μL with PBS/AE and leave for 3 min. 13. Steps 10–11 test for coverage of the sensor surface with lipid. In the presence of a complete lipid monolayer no BSA binding should be apparent after washing with PBS/AE. It is recommended that if saturated lipids (e.g. DPPC and DSPC) are used a rinse (wash three times with 200 μL) with 5% sucrose (w/w) in water is performed before washing with PBS/AE. Saturated lipid micelles will stick to the surface preventing monolayer formation. The increased density of the sucrose enables the micelles to float to the surface and be washed away. 3.8. Immobilization of Lipids on NonDerivatized Surface
Protocol
The non-derivatized cuvette offers an alternative surface to the hydrophobic cuvette for the immobilization of phospholipids (PLS). Single or mixed lipids can be immobilized from the solution in the detergent n-octyl b-D-glucopyranoside (OG). The lipid-coated surface can then be used for analysis of protein/lipid interactions, or of lipid-associated proteins with other soluble components. 1. Insert a new non-derivatized cuvette into the instrument and wash with 150 μL of PBS and allow 10 min for equilibration. 2. Start data acquisition and collect 3 min of baseline data. 3. Wash with 180 μL of PBS/OG, and collect data for about 3 min. 4. Sonicate the PLs for 10 s in a sonicating cleaning bath and introduce 100 μL to the instrument of the PLS solution till the response is about at a plateau. 5. Add 30 μL of PBS without removing the PLS solution. Leave until the binding response has approached a plateau. 6. Wash three times with 50 μL of PBS. 7. The amount of lipid immobilized can be determined by subtracting the initial baseline level in PBS from the final level after stage 6 (see Note 20).
3.9. Capture of His-Tagged Proteins
Protocol
Poly-histidine is a commonly used tag on recombinant proteins. This protocol describes a method for the chelation of His-tagged proteins on a nickel charged surface onto a carboxymethyl dextran coated surface. The immobilization process includes coupling of nitrotriacetic acid (NTA) to the dextran using EDC/NHS and charging the NTA with nickel ions. This complex will then be used to capture His-tagged protein from solutions. 1. Insert new carboxylate cuvette into the instrument and wash with 200 μL of HEPES and allow 10 min for equilibration. 2. Start data acquisition and collect 3 min of baseline data.
Label-Free Detection with the Resonant Mirror Biosensor
117
3. Wash and incubate the surface with 180 μL of a mixture of 1:1 (v/v) EDC/NHS for 15 min. 4. Wash with 200 μL of HEPES and collect baseline data for 3 min (see Note 2). 5. Incubate 50 μL of NTA solution for 8 min (see Note 21). 6. Wash with 200 μL of HEPES and collect baseline data for 3 min (see Note 22). 7. Incubate the cuvette with 1 M ethanolamine pH 8.5 for 8 min to block unreacted NHS-esters (see Notes 9 and 23). 8. Wash with 200 μL of HEPES and collect baseline data for 5 min. 9. Wash three times and incubate the surface with 200 μL of 500 μM of NiCl2 in HBSTE for 6 min. 10. Wash with 200 μL of HEPES and collect baseline data for 5 min. Optional
11. Wash three times with 200 μL of PGI wash buffer and leave for 5 min to remove loosely associated metal ions from the residual carboxyl’s on the CMD. 12. Wash with 200 μL of HEPES and collect baseline data for 3 min. 13. Spike in the His-tagged sample and allow binding 14. Wash with 200 μL of HEPES and collect baseline data for 3 min. If high concentration of His-tagged material has been used, it is possible that a very distinct dissociation phase will follow. This can take several minutes to settle down. It is possible to speed up the process by washing loosely associated material of the surface using PFI wash buffer (optional). 15. Wash and incubate the surface with 200 μL of PGI for 2 min to clear the his-tag from the surface (see Note 24). 16. Wash with 200 μL of HEPES and collect baseline data for 3 min.
3.10. Capture of FLAGTagged Proteins
Protocol
FLAG is a commonly used tag on recombinant proteins. The tag consists of eight amino acids in the sequence Asp-Tyr-Lys-Asp-AspAsp-Asp-Lys. This protocol describes a method for immobilization of anti-FLAG antibody to CMD for the capture of FLAG-tagged proteins. Captured recombinants can be easily eluted with competing FLAG peptides at nM concentrations (43). 1. Insert new carboxylate cuvette into the instrument and wash with 200 μL of PBS/T and allow 10 min for equilibration. 2. Start data acquisition and collect 3 min of baseline data. 3. Wash and incubate the surface with 180 μL of a mixture of 1:1 (v/v) EDC/ NHS for 10 min.
118
Zourob et al.
4. Wash with 200 μL of PBS/T and collect baseline data for 3 min. 5. Replace PBS/T with acetate buffer pH 4.5 and leave to equilibrate for 4 min. 6. Introduce Anti-FLAG antibody between 10 and 20 μg/ mL and leave for 5 min. An irrelevant antibody can be used immobilized at the same level as a control. 7. Wash with 200 μL of PBS/T and collect baseline data for 3 min. 8. Incubate the cuvette with 1 M ethanolamine pH 8.5 for 4 min to block unreacted NHS-esters. 9. Wash with 200 μL of PBS/T and collect baseline data for 5 min. 10. Wash extensively with 200 μL of 50 mM HCl and leave to equilibrate for 5 min to remove loosely bonded antibodies from the CMD matrix. 11. Wash with 200 μL of PBS/T and collect baseline data for 3 min. 12. Repeat steps 9 and 10 twice 13. Introduce 740 nM FLAG-BAP control peptide and incubate it for 5 min. 14. Wash with 200 μL of PBS/T and collect baseline data for 5 min. 15. Wash with 200 μL of 50 mM HCl and leave for 2 min to regenerate the surface. 16. Wash with 200 μL of PBS/T and collect baseline data for 5 min. 17. Repeat steps 13–16 to assess reproducibility of activity (see Note 25). 3.11. Sensitivity Enhancement Using Colloidal Gold Complexes
Protein-colloidal conjugate cause a higher refractive index and thickness change per binding event than native protein or other biomolecule. As a result, the colloidal gold nano-particles can be used to enhance the sensitivity of the assays. This protocol describes a general method for the preparation of colloidal gold-protein complexes and its use to improve the detection limit of an assay. Gold complexes can be prepared using the De Mey method. (33). The bioactivity of the colloidal gold complexes can be checked by immunoblotting with the appropriate antibody (Biocell gold conjugates technical information and guidelines for use in electron microscopy, Light Microscopy and Immunoblotting.). The average number of protein molecules per colloidal gold particle can be determined from radioactive studies (40).
Label-Free Detection with the Resonant Mirror Biosensor
119
Colloidal gold is susceptible to aggregation; therefore all glassware and plasticware should be thoroughly cleaned and if possible, silliconized. Protocol
1. Adjust the pH of the colloidal gold to 0.5 pH units above the PI of the protein with potassium carbonate. 2. Determine the number of gold particles by measuring the absorbance at λmax or 520 nm (A520 = 1 is 1.7 × 1011 particles/mL for 30 nm particles). 3. Add protein stepwise to colloidal gold to give a final molar ratio of 100:1 or 10:1, depending on how many protein molecules are required by gold particles. Typical number of particles will be 1 × 1012 particles/mL final concentration. 4. Add BSA to the solution to final concentration between 0.1 and 0.5% of BSA (w/v) to block any remaining exposed sites on the gold. 5. Leave on a rotary mixer for 15 min. 6. Spin the mixture in a microcentrifuge for 20 min. 7. Carefully remove the clear supernatant leaving loose red sediment. 8. Resuspend the gold complex in fresh BSA solution to remove the more free protein. 9. Repeat steps 6–8. A good final volume is 1/10th of the starting volume. 10. It is recommended at this stage to retest the concentration of the colloid as described in step 2. 11. Store the product in an air-tight siliconized glass vial at 4 °C until it used for assay enhancement (see Subheading 3.17).
3.12 Measuring Affinity and Rate Constants
Kinetic information indicates how fast the interactants come together, and how fast the resulting complex breaks down. Kinetic information is given by the rate constants of the forward and reverse reaction as shown in Eq. 1. The interaction of the ligate (L) with the immobilized ligand (G) on the sensor surface can be represented as: K on G + L ¾¾¾ ® GL.
G+L
kass kdiss
GL.
The rate of complex (GL) formation is = Kass (A)(B) The rate of the reverse reaction (dissociation) = Kdiss (AB) KD =
| G || L | kdiss 1 = = , GL kass KA
120
Zourob et al.
where KD and KA are the affinity constants (dissociation equilibrium constant in units (M) and association equilibrium constant in units (M−1) respectively). The parameters Kass (M−1 S−1) and Kdiss (S−1) refer to the association and dissociation rate constants respectively. By measuring (G), (L), and (GL) at equilibrium, the affinity can be determined. The biosensor gives rise to a signal that directly correlates to the amount of complex (GL) which is being formed. The dissociation equilibrium constant KD refers to the concentration at which 50% of the binding sites are occupied. KD and KA are opposite each other. If KD is low, the affinity of the interaction is high and in case the KA is high, the affinity is high. Upon exposure of the ligate to the ligand, initially the forward reaction predominates and GL is formed. With time, as the concentration of GL increases, the reverse reaction becomes significant. Eventually, at equilibrium, the time taken to reach this depends on the kinetics of the interaction and the concentration of the reactants, and the rate of the forward and reverse reaction, will be the same. GL continues to be formed from G and L, and GL continues to dissociate, and concentrations of GL, G and L remain constant. R eq =
Rmax [L] . K D + [L]
The classical route to determine equilibrium constants, is measuring the amount of complex GL at equilibrium from a fixed value of ligand. Rmax is the response when all possible ligand sites are occupied. At each concentration of ligate, a value of the response at equilibrium (Req) is obtained, equivalent to the amount of complex GL formed at equilibrium. At high concentrations of ligate L compared with the ligand G, the assumption that the concentration of L in solution is constant can be made. This is termed pseudo-first order condition. At high concentrations, Req is the same as Rmax. By plotting Req against the free ligate (L) concentration, the concentration required to saturate 50% of the available ligand sites can be determined; this value is the concentration which give rise to the response of Rmax/2. Rate of complex (GL) formation: d[GL] = kass [G][L] - kdiss [GL]. dt The concentration of free ligand sites decreases with ligate binding, and thus the concentration of free ligate sites at time t, is equal to the maximum number of sites (G)0 minus the sites bound (GL)t or ( (G)0−(GL)t); d[GL] = kass [G]0 - [GL]t [L] - kdiss [GL]t . dt
Label-Free Detection with the Resonant Mirror Biosensor
121
As complex (GL) formation results in a change in response (Rt), which is directly proportional to (GL), Rmax is equivalent to (L)0 and the above equation can be rewritten as: dRt = kass (Rmax - Rt )[L ] - kdiss Rt . dt Multiplying out the brackets and rearranging gives: dRt = kass (Rmax [L] - Rt (K ass [L] + K diss ). dt Originally this led to biosensor data being analyzed by linearization of the data for each ligate concentration (L) by plotting dR/dt against Rt. This should give a straight line of slope kass [L] + kdiss. Plotting slopes against concentrations of L will then give a straight line of slope equal to the association rate constant (Kass), and the intercept is equal to the dissociation rate constant (Kdiss). However, a problem with linearization is that the errors are compounded. By integration the equation can be transformed into a form suitable for application of non-linear regression:
(
)
Rt = Req 1 - e - (kass [L]kdiss )t . Using the above equation, binding data can be fitted at different concentrations to determine Kon (where Kon = Kass [L] + Kdiss. A plot of Kon against ligate concentration (L), will give a straight line with a slope of Kass and an intercept Kdiss. The determination of Kdiss from the intercept of Kon vs. (L) often has a high extrapolation error. A more robust approach is to initiate dissociation by removing the ligate with a buffer wash. Under these conditions (i.e. L = 0) dissociation of the ligate from the immobilized ligand is described by Eq. 3: Here the concentration of L is taken to be zero and the first order decay in response is given by: dR = - K diss Rt . dt On integration it gives: Rt = R0 e - (kdiss t ,
(3)
where R0 is the response at the initiation of dissociation. This simple equation is not applicable to all dissociation profiles due to incomplete dissociation which has be attributed to rebinding of the dissociated ligate. 3.13. Determination of the Concentration of Theophylline in Buffer and Serum
Theophylline has a molecular weight of 180 Da. The response of the RM for the direct detection of any analyte less than 5,000 Da will be too small to be measured reliably. Alternatively, indirect assay format must be employed. The assay is based on competition
122
Zourob et al.
between theophylline and a protein-theophylline conjugate for binding to a theophylline monoclonal antibody (Mab) captured on the sensor surface by an immobilized specific rabbit-mouse antibody (RAM-Fc) through the Fc region. The theophylline conjugate used in this assay is IgG (150,000 Da); it contains approximately 20 theophylline residues per protein molecule. 3.13.1 Assay Procedure in PBS/T
1. Insert new CMD cuvette into the instrument and wash with 200 μL of PBS/T and allow 10 min for equilibration. 2. Start data acquisition and collect 3 min of baseline data. 3. Wash and incubate the surface with 180 μL of a mixture of 1:1 (v/v) EDC/ NHS for 10 min. 4. Wash with 200 μL of PBS/T and collect baseline data for 3 min. 5. Introduce 10 μg/mL of RAM Fc for immobilization on CMD using the EDC/NHS chemistry. 6. Wash with 200 μL of PBS/T and collect baseline data for 3 min. 7. Incubate the cuvette with 1 M ethanolamine pH 8.5 for 4 min to block unreacted NHS-esters. 8. Wash with 200 μL of PBS/T and collect baseline data for 5 min. 9. Introduce 100 μg/mL of MAb for 10 min. 10. Wash with PBS/T and collect baseline data for 5 min. 11. Incubate a solution containing one of the theophylline/conjugate mixtures into the cuvette for 10 min. 12. Wash with PBS/T and collect baseline data for 5 min. 13. Regenerate the surface to remove the anti-theophylline and theophylline-protein-conjugate with 50 mM HCl for 3 min, followed by re-equilibration into PBS/T buffer. 14. Steps 5–13 were then repeated for further theophylline/ conjugate mixtures. 15. Construct the calibration curve.
3.13.2 Assay Procedure in Serum
Repeat the same experimental procedure as described for PBS/T. With the exception of theophylline, solutions were prepared in serum and were diluted 1:10 in PBS/T prior to mixing with the theophylline-protein-conjugate. In addition, a control for nonspecific binding was performed with a digoxin-protein-conjugate (see Note 26).
3.14 Protein– Carbohydrate Recognition
The importance of protein–carbohydrate recognition in modulating key cellular activities, such as receptor binding, adhesion, and mitogenesis, has become increasingly recognized in recent
Label-Free Detection with the Resonant Mirror Biosensor
123
years. Basic fibroblast growth factor (bFGF) is the archetype of the FGF family of growth factors and is involved in the regulation of growth and morphogenetic events as diverse as the induction of mesoderm and limb buds to the growth of skeleton and the mammary gland (41). To elicit their cellular effects, bFGF, like the other FGFs, must interact with a dual receptor system consisting of the receptors tyrosine kinases and heparan sulfate proteoglycans; bFGF binds to the heparin sulfate chains of the proteogylcans (42). Heparan sulphate is a sulphated glycosaminoglycan present on the surface of most mammalian cells. Heparin, a specialized form of heparan sulphate produced by mast cell, contains binding sequences for most heparan sulphate binding proteins, including bFGF (41). A number of peptides derived from the protein sequence of bFGF have been identified that are implicated in heparin binding (41). The RM used to study the interaction of heparin to bFGF was capable of quantifying the very low association rates and affinities of bFGF-derived peptides for heparin. Novel insights on structure and function were therefore obtained. A full report of this work can be found in (43). Protocol
1. Insert new CMD cuvette into the instrument and wash with 200 μL of PBS/T and allow 10 min for equilibration. 2. Start data acquisition and collect 3 min of baseline data. 3. Deposit streptavidin at a concentration of 200 μg/mL in 10 mM acetate buffer pH 5.0 for 15 min as described in Subheading 3.6.1. 4. Wash with 200 μL of 20 mM HCl for 3 min to remove noncovalently attached streptavidin. 5. Wash with 200 μL of PBS/T and collect baseline data for 3 min. 6. Introduce 100 μL of 1 mg/mL of biotinylated heparin and incubate for 30 min. 7. Regenerate the surface with Na2HPO4 buffer for 5 min. 8. Wash with PBS/T and collect baseline data for 5 min. 9. For association 100 μL of minimum five different concentrations per ligate (bFGF and peptides) in the PBS/T were analyzed for 5–10 min on at least two different surfaces CMD surfaces and one AS surface. 10. For dissociation run 200 μL of PBS/T for at least 5 min. 11. Regenerate the surface with Na2HPO4 buffer for 3 min, followed by re-equilibration in PBS/T. 12. Apply bFGF and bFGF-derived peptides as control analysis onto streptavidin surfaces in the absence of captured heparin (should not show significant interaction).
124
Zourob et al.
13. Calculate the kinetic association and dissociation constants (kass and kdiss) and the dissociation equilibrium constant (KD) as described in Subheading 3.12. Using a series of profiles for bFGF and bFF-peptide binding to heparin to generate the kinetics and equilibrium constants shown in Table 1 (see Note 27). 3.15 Nucleic Acid Hybridization
This protocol describes the use of RM for the rapid and unlabeled single stranded DNA–DNA hybridization using a biotinylated oligonucleotide, attached to CMD coated sensor surface via streptavidin (44). This approach is a convenient, generic approach and biotinylated oligonucleotides can be easily synthesized. The binding of biotin to streptavidin is virtually irreversible, requiring extreme conditions to separate the two molecules (45, 46). 1. Insert new CMD cuvette into the instrument and wash with 200 μL of PBS/T and allow 10 min for equilibration.
Protocol
2. Start data acquisition and collect 3 min of baseline data. 3. Streptavidin was immobilized at a concentration of 200 μg/ mL streptavidin in 10 mM acetate buffer pH 5.0 for 15 min as described in Subheading 3.6.1. 4. Wash with 200 μL of PBS/T and collect baseline data for 3 min. 5. Wash with 200 μL of 20 mM HCl for 3 min to remove noncovalently attached molecules. 6. Wash with 200 μL of PBS/T and collect baseline data for 3 min.
Table 1 kinetic and equilibrium constants for the interaction of bFGF with heparin Ligate Constant (units) a
a
BFGF (±SE)
BFGF (127–140) (±SE)
KD (M)
−9
84 (±55) × 10
30 (±4) × 10−6
KD (M)b
74 (±20) × 10−9
35 (±7) × 10−6
Kass (M−1S−1)c
9.3 (±2.5) × 10−4
4 (±0.16) × 10−2
Kdiss (S−1)d
6.8 (±0.4) × 10−3
1.4 (±0.38) × 10−2
KD calculated from five concentrations of ligate in three experiments (two carried out on CM-Dextran and one on AS) by binding curve analysis. KD was calculated from KD = (Kdiss/Kass) c Mean of three determinations (±SE) d Mean of nine determinations (±SE). (Reproduced from ref. 48 with permission from NeoSensors Ltd.)
Label-Free Detection with the Resonant Mirror Biosensor
125
7. Introduce 2 mg/mL Oilgo-1 in PBS/T and incubated for 15 min. 8. Monitor the binding response after washing with 200 μL of PBS/T and collect baseline data for 3 min. 9. Wash the surface with hybridization buffer (pH 8, 5× saline, sodium citrate, 5× Denhardts solution), followed by 5 mM sodium phosphate, 0.1% v/v Tween 20. 10. Introduce 4 μg/mL complementary oligonucleotides (Oilgo2) in hybridization buffer and incubate for 15 min. 11. Wash the surface with hybridization buffer and monitor the shift in resonance position (Fig. 4) (see Note 28). 12. Run control experiments on new streptavidin derivatized surfaces using (see Note 28): (a) Oligo-3 (5 μg/mL, 303 nM) with immobilized Oligo-1 at the sensor surface (Fig. 4). (b) Oligo-4 (2 μg/mL) with immobilized Oligo-2 (4 μg/ mL, 303 nM) at the sensor surface. Introduce Oligo-2 (4 μg/mL, 303 nM) in the absence of an immobilized specific oligonucelotide. 3.16. Fermentation Monitoring
The fermentation of micro-organisms and mammalian cells for the production of recombinant proteins is an area of increasing industrial interest. It could benefit greatly from the rapid assay to quantify the bioproducts of interest during the fermentation process. Here is one example reported for the use of the RM for
Fig. 4. Specific hybridization of Oligo-2 to captured complementary biotinylated Oligo-1 and the absence of hybridization of non-complementary Oligo-3 with Oligo-1. Arrows indicate: (a) addition of hybridization buffer, (b) addition of oligonucleotides. (Reproduced from ref. 46 with permission from NeoSensors Ltd.).
126
Zourob et al.
monitoring the recombinant antibody fragment, D1.3 Fv (specific for hen egg lysozyme, HEL) (47) produced by periplasmic secretion in E. coli fermentation broths (48). Protocol
1. Insert new CMD cuvette into the instrument and wash with 200 μL of PBS/T and allow 10 min for equilibration. 2. Start data acquisition and collect baseline data for 3 min. 3. Immobilize HEL (30 μg/mL) in acetate buffer pH 5 on CMD matrix for 5 min. 4. Wash with 200 μL of PBS/T and collect baseline data for 3 min. 5. Incubate the surface with 50 μL of 1 M ethylene diamine pH 8.5 for 10 min to modify the carboxymethyl groups with amines. 6. Wash with 200 μL of PBST and collect baseline data for 3 min. 7. Construct calibration curve using (negative) E. coli fermentation broth (one does not express D1.3 Fv that mimicked the process broth in terms of operation and final protein concentration (49) and by spiking different concentration of purified D1.3 into this broth. 8. Wash with 200 μL of PBS/T and collect baseline data for 3 min. 9. Incubate the sample from the fermenter at regular intervals for 5 min after removing the particulates using microcentrifuge 11,600 × g for 5 min. 10. Wash with 200 μL of PBS/T and collect baseline data for 3 min. 11. Regenerate the surface from the immobilized ligand with 100 mM HCl for 4 min. 12. Data can be analyzed using two methods either measuring the absolute change in response after incubating the sample with the ligand for 5 min or measuring the initial binding rate (using linear regression using FASTfit™ software) (50).
3.17. Bacterial Cell Detection
Protocol
The RM has been used to distinguish between bacterial strains on the basis of their surface proteins. Staphylococcus aureus (Cowan-1) cells, which express protein A on their surface, could be detected by binding to Human IgG (51, 52). Staphylococcus aureus strain (Wood-46) cells do not interact significantly with Human IgG as they do not express protein A on their surface (53) (see Note 29). 1. Insert AS cuvette into the instrument and wash with 200 μL of PBS and allow 10 min for equilibration.
Label-Free Detection with the Resonant Mirror Biosensor
127
2. Start data acquisition and collect 3 min of baseline data. 3. Activate the surface with 5% (v/v) glutaraldehyde in water for 30 min. 4. Wash with 200 μL of PBS and collect baseline data for 3 min. 5. Incubate the surface with 100 μg/mL IgG in PBS for 8 min. 6. Wash with 200 μL of PBS and collect baseline data for 3 min. 7. Block the surface with 3 mg/mL BSA for 5 min. 8. Wash with 200 μL of PBS and collect baseline data for 3 min. 9. Introduce S. aureus (Wood-46) and incubate for 20 min. 10. Wash with 200 μL of PBS and collect baseline data for 3 min (see Fig. 5a for the results). 11. Repeat the same procedure from steps 1 to 10 for Wood-46 strain (see Fig. 5a for the results). 12. Repeat the same procedure from steps 1 to 10 for S. aureus (Cowan-1) on BSA immobilized surface as control surface (see Fig. 5a for the results).
Fig. 5. (a) Calibration curve showing the response against cell concentration. The control data for bacterial cell strain Wood-46 and for the interaction between the Cowan-1 and immobilized BSA are also shown. (Reproduced from ref. 19 with permission from Analytical Chemistry). (b) Comparison of the direct binding assay with the colloidal gold sandwich assay. (Reproduced from ref. 19 with permission from Analytical Chemistry).
128
Zourob et al.
Fig. 5. (continued)
13. For sandwich assay using gold complex: repeat the same procedure as in direct assay from steps 1 to 10 followed by the addition of colloidal gold human IgG and incubate for 10 min. 14. Wash with 200 μL of PBS and collect baseline data for 3 min (see Fig. 5b for the results). 3.18. Receptor–Cell Interactions
Protocol
Carcinoembryonic antigen (CEA) (54) has been identified as immunocytochemical on tumor cell membrane from a variety of tissues, making it one of the most useful human tumor markers (55). Cell lines have been developed expressing CEA to study the etiology of carcinomas. In the present protocol the RM is used for the real-time detection of the binding of L cells bearing a cellexpressed CEA antigen, to anti-CEA antibody immobilized on AS sensor surface (see Note 29). 1. Insert new AS cuvette into the instrument and wash with PBS for 5 min for equilibration. 2. Start data acquisition and collect 3 min of baseline data. 3. Introduce PBS/BS3 to the cuvette for 10 min 4. Wash with PBS to get stable baseline. 5. Introduce the antibody solution 130 μg/mL and leave it for about 10 min (see Note 8). 6. Block remaining activated groups with 2 mg/mL BSA in PBS for 10 min.
Label-Free Detection with the Resonant Mirror Biosensor
129
7. Wash the cuvette with 20 mM HCl for 2 min 8. Wash with PBS/BSA for about 5 min to get the baseline. 9. Introduce cell solution. 10. Wash with PBS/BSA and collect 5 min of baseline data. 11. Regenerate the surface with 0 mM HCl for 2 min, PBS/T for 2 min, and finally for another 2 min with 20 mM HCl (see Fig. 6) (see Note 30). 12. Repeat steps 2–11 with non-expressing cells (Reference cells) (see Fig. 6). 3.19. Determination of the Kinetic Constants
Protocol
The use of RM for measuring equilibrium binding constants is attractive due to the label free detection unlike commonly used techniques such as ELISA, RIA, hapten inhibition, and fluorescence methods (56). This protocol describes the use of RM to determine the rate and affinity constants of the binding interaction between immobilized Staphylococcus aureus protein A and different concentration of human IgG in solution (57, 58). 1. Insert CMD cuvette into the instrument and wash with 200 μL of PBS and allow 5 min for equilibration.
Fig. 6. Binding of L cells to an AS surface derivatized with an anti-CEA antibody. (Reproduced from ref. 56 with permission from NeoSensors Ltd.).
130
Zourob et al.
2. Start data acquisition and collect 3 min of baseline data. 3. Introduce a mixture of 200 μL of EDC and 200 μL of NHS for 7 min. 4. Wash with PBST and establish a stable baseline. 5. Change the PBST to 10 mM acetate buffer pH 5.5 and collect 3 min baseline data. 6. Introduce 200 μL of 12.5 μg/mL Staphylococcus aureus protein in acetate buffer to the cuvette for 15 min. 7. Block the unreacted NHS-ester with 1 M ethanolamine pH 8.5 for 3 min. 8. Wash with PBST to get the baseline. 9. Wash the cuvette with 10 mM HCl for 3 min to remove the non-covalently bound ligand to the matrix. 10. Wash with PBST for 5 min to get the baseline and to determine the amount of protein A. 11. Incubate the first concentration of HIgG in PBS/T for 30 min. 12. Wash with PBS/T to get the baseline. 13. Regenerate the surface with 10 mM HCl for 3 min. 14. Re-establish baseline with PBS/T 15. Repeat steps 11–14 with different concentrations of HIgG (Fig. 7a) (see Note 31). 16. Plot dR/dt vs. R if the maximum binding response is known and the association between the two interactants is carried out under pseudo first order conditions for which the overall rate is: dR = kass (Rmax - R)[C] - kdiss R, dt dR = kass (Rmax [C] - R(K ass [C] + K diss ). dt However, if the maximum response is not known the ka can be obtained by plotting the slope, ks, of the plot dR/dt vs. R (Fig. 7b) for all concentrations of HIgG. The slope of this new plot is the ka and the intercept, in theory, is the dissociation rate constant (kd) (Fig. 7c). Ks = kaC + kd
.
Ka was calculated for HIgG and protein A 1.13 × 105M−1s−1 and the intercept value, kd, was 1.83 × 10−4s−1. Another way to determine the dissociate rate constant, which is independent of concentration, is to replace the sample HIgG with buffer and follow the dissociation phase. The dissociation
Label-Free Detection with the Resonant Mirror Biosensor
131
Fig. 7. (a) The biding curve of HIgG to immobilized SpA. (Reproduced from ref. 59 with permission from NeoSensors Ltd.). (b) Plots of dR/dt vs R for all HIgG concentrations. (Reproduced from ref. 59 with permission from NeoSensors Ltd.). (c) Slope Ks of dR/dt vs. R plotted against HIgG concentration. Determination of ka for SpA:HIgG. (Reproduced from ref. 52 with permission from NeoSensors Ltd.). (d) Plot of ln(R0/Rn) vs. tn − t0 to determine the dissociation rate constant for protein A:IgG. (Reproduced from ref. 59 with permission from NeoSensors Ltd.).
132
Zourob et al.
Fig. 7. (continued)
phase should follow the following equation, when rebinding is negligible due to complete removal of the dissociating analyte: dR = - K diss R. dt Integrating the above equation gives: dR = kd (tn - t0 ), dt
Label-Free Detection with the Resonant Mirror Biosensor
133
where R0 is the response at t0, and Rn is the response at tn. By constructing a plot of ln(R0/Rn) vs. tn − t0, kd is determined from the slope of the plot (Fig. 7d). The kd value was found to be 1.09 × 10−4. The overall affinity constant (kd/ka) of the HIgG and protein A was therefore calculated to be 1.62 × 10−9 M.
4. Notes 1. Silanization can be carried out by exposing the cleaned RM chip to 3-mainopropyletriethoxysilane vapour at 145 °C and about 3 mm Hg for 1 h (2). 2. NHS coupling kit is commercially available from Neosensors (Durham, UK, CUV999900) if you do not want to prepare it. EDC/NHS aliquots should be stored at or below −18 °C. 3. The use of coherent illumination proved troublesome, as dust particles within the instruments caused light scattering, producing a characteristic speckle pattern on the detector. 4. The finite bandwidth of these sources has the effect of broadening the apparent width of the resonance peaks in the angular scan mode, while the finite size of the sources reduces the degree of collimation, which also broadens the peak in the wavelength scan mode. 5. For the angular scan, an interference filter with 670 ± 10 nm wavelength, 25 mm in diameter (Comar instruments, Cambridge, UK), and in case of wavelength scan a band-pass interference filters of wavelengths 512, 531, 550, 570, 596, 605, 632, 645 and 670 nm (10 nm FWHM bandwidth, Ealing Electro-optics, Watford, UK) was used to provide light of different wavelengths. 6. Fresh dissolved EDC should be used. In case degraded EDC has been used the cuvette can be treated with 1 M Tris, 3 M NaCl, pH 8.0 to release any accumulated materials. 7. In case of highly charged ligand or ligates, it is desirable to reduce the charges on the matrix by activating the matrix with EDC and NHS, followed directly by blocking with ethanolamine prior to carrying out the protocol as written. 8. The pH of the acetate buffer should be below the PI of the protein ligand, but above a value of 3.5. 9. 1 M Tris–HCl, pH 8.0 may also be used to block the unreacted NHS ester.
134
Zourob et al.
10. Avoid exposing the AS surface to buffers containing detergent prior to immobilization as it may reduce or variable the levels of ligand immobilization. 11. High concentration of ligands are needed as there is very little electrostatic concentration of ligand to the As surface unlike CMD surface. 12. Smaller and less charged proteins such as β-casein or cytochrome c can be used. Be sure that the sensor surface is completely coated as failure to coat the surface with protein results in an increase in the non-specific binding when ligates are added. 13. It is advisable after the PBST wash to further treat the immobilized avidin layer with the same regenerants that may be used in subsequent ligate binding experiments. 14. Streptavidin and NeutrAvidin from Pierce can be used instead of avidin. 15. Biotin cuvette can be reused several times after removing the avidin layer by treating the surface for 2 min with saturated KOH outside the instrument. It is essential to wash the surface thoroughly and rapidly to remove all traces of KOH prior to reuse. 16. Prevent de-wetting of the sensor surface by extensive washing. 17. To obtain reproducible results it is essential that the lipids used are not oxidized. It is recommended that the stock lipid solution contain 0.1% (w/w) butylated hydroxytoluene and are stored at less than −60 °C for no longer than 2 weeks. 18. It is recommended in case of using saturated lipids to wash the surface with 50% sucrose (w/w) in water before washing back into the PBS/AE. The sucrose solution will float and washed away the formed sticky micelles instead of the formation of a monolayer. 19. The resonance scan should be a symmetrical peak, indicating uniformity immobilization throughout the sensed area. If the peak is not symmetrical, stages 6–9 can be repeated. 20. The PLs can be stripped off from the surface by incubating the cuvette with 2% OG in high quality water for 3 min followed by washing with PBS. 21. Incubation time can be extended up to 15 min to ensure that a reasonable quantity of NTA is immobilized. 22. Due to the expensive nature of the TNA, it is recommended to recover NTA solution as it can be re-used.
Label-Free Detection with the Resonant Mirror Biosensor
135
23. It is recommended to extend the blocking time due to the extension of the activation time. 24. Alternatively, the sample can be removed by washing with 10 mM HEPES, 0.15 M NaCl, 0.35 M EDTA, 0.005% Tween 20, pH 8.3. 25. This procedure can be used at any time during the experiment to test the activity of the surface. The protocol was also found to work with FLAG-Ubiquitin (Sigma, U5382). 26. No significant non-specific binding of a digoxin-protein conjugate was observed. 27. The dissociation equilibrium constant for the interaction of peptide bFGF (127–140) with heparin is a 1,000-fold higher than that observed for native bFGF, its association rate constant and dissociation equilibrium constants for the peptide were readily determined by the RM. The results showed how RM can be used to analyze protein and a very low affinity peptide–carbohydrate interactions and provides evidence that a part of the recognition site of bFGF for heparin lies within the amino-acid sequence of 127–140 of bFGF (43). 28. The results from Fig. 3 show the specific hybridization of complementary oligonucleotides (Oligo-2 to immobilized Oligo-1) under average stringency conditions, while the non-complementary DNA (Oligo-3) did not give any significant response. 29. The RM is unsuitable for cell detection due to the short penetration of the evanescent field at the sensor surface, which places the majority of the cells outside the field (59). However, the sensitivity of the RM technique can be increased 1,000-fold using a human immunoglobulin G (HIG)-colloidal gold complex (19). 30. Keep cells in ice between the bindings. 31. All experimental data shown were generated on a single cuvette.
Acknowledgements The authors thank W. Jones and S. Mian from Neosensors for their help and providing the material for this chapter. The views expressed here are those of the authors and do not represent those of Biophage Pharma Inc.
136
Zourob et al.
References 1. Cush, R., Cronin, J.M., Stewart, W.J., Maule, C.H., Molloy, J., and Goddard, N.J. (1993) The resonant mirror: a novel optical sensor for direct sensing of biomolecular interactions, Part I: Principle of operation and associated instrumentation. Biosens. Bioelectron. 8, 347–353 2. Buckle, P.E., Davies, R.J., Kinning, T., Yeung, D., Edwards, P.R., Pollard-Knight, D., and Lowe, C.R. (1993) The resonant mirror: a novel optical sensor for direct sensing of biomolecular interactions, Part II: Applications. Biosens. Bioelectron. 8, 355–363 3. Charalambous, B.M., Evans, J., Feavers, I.M., and Maiden, M.C. (1999) Comparative analysis of two meningococcal immunotyping monoclonal antibodies by resonant mirror biosensor and antibody gene sequencing. Clin. Diagn. Lab. Immunol. 6, 838–843 4. Kobayashi, K., Matsuura, E., Liu, Q., Furukawa, J., Kaihara, K., Inagaki, J., Atsumi, T., Sakairi, N., Yasuda, T., Voelker, D.R., and Koike, T., (2001) A specific ligand for β2-glycoprotein I mediates autoantibodydependent uptake of oxidized low density lipoprotein by macrophages. J. lipid Res. 42, 697–709 5. Zimmer, S., Stocker, A., Sarbolouki, M.N., Spycher, S.E., Sassoon, J., and Azzi, A. (2000) A novel human tocopherol-associated protein. Cloning, in vitro expression, and characterization. J. Biol. Chem. 275, 25672– 25680 6. Dmitriev, D.A., Massino, Y.S., and Segal, O.L. (2003) Kinetic analysis of interactions between bisepecific monoclonal antibodies and immobilized antigens using a resonant mirror biosensor. J. Immun. Meth. 280, 183–202 7. Rhodes, D.C. (2000) Binding of TammHorsfall protein to complement 1q measured by ELISA and resonant mirror biosensor techniques under various ionic-strength conditions. Immunol. Cell Biol. 78, 474–482 8. Nakra, P., Manivel, V., Vishwakarma, R.A., and Rao, K.V. (2000) Cell responses to a peptide epitope. X. epitope selection in a primary response is thermodynamically regulated. J. Immnol. 164, 5615–5625 9. Salek-Ardakani, S., Arrand, J.R., Shaw, D., and Mackett, M. (2000) Heparin and heparan sulfate bind interleukin-10 and modulate its activity. Blood 96, 1879–1888 10. Torigoe, H., Hari, Y., Sekiguchi, M., Obika, S., and Imanishi, T. (2001) 2′-O,4′-CMethylene bridged nucleic acid modification
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
promotes pyrimidine motif triplex DNA formation at physiological pH. Thermodynamic and kinetic studies. J. Biol. Chem. 276, 2354–2360 De Zutter, J., Forget, A., Logan, K., and Knight, K. (2001) Phe217 regulates the transfer of allosteric information across the subunit interface of the RecA protein filament. Structure 9, 47–56 Sassoon, J., Lilie, H., Baumann, U., and Kohli, J. (2001) Biochemical characterization of the structure-specific DNA-binding protein cmb1 from Schizosaccharomyces pombe. J. Mol. Biol. 309, 1101–1115 Despeyroux, D., Walker, N., Pearce, M., Fisher, M., Mcdonnell, M., Bailey, S.C., Griffiths, G.D., and Watts, P. (2000) Characterization of heterogeneity by electrospray mass spectrometry, capillary electrophoresis, and resonant mirror. Anal. Biochem. 279, 23–36 Hirmo, S., Artursson, E., Puu, G., Wadstorm, T., and Nilsson, B. (1998) Characterization of Helicobacter Pylori interactions with sialglycoconjugates using a resonant mirror biosensor. Anal. Biochem. 257, 63–66 Hirmo, S., Artursson, E., Puu, G., Wadstorm, T., and Nilsson, B. (1999) Helicobacter Pylori interactions with human gastric mucin studied with a resonant mirror biosensor. J. Microbiol. Meth. 37, 177–182 Sigurdsson, H.H., Loftsson, T., and Lehr, C.-M. (2006) Assessment of mucoadhesion by a resonant mirror biosensor. Int. J. Pharm. 325, 75–81 Woods, C.M., Neslund, G., Kornburst, E., and Flaim, S.F. (2000) Perflubron attenuates neutrophil adhesion to activated endothelial cells in vitro. Am. J. Physiol. Lung Cell Mol. Physiol 278, 1008–1017 Sheng, N., Fairbanks, M.B., Heinrikson, R.L., Canziani, G., Chaiken, I.M., Mosser, D.M., Zhang, H., and Colman, R.W. (2000) Cleaved high molecular weight kininogen binds directly to the integrin CD11b/CD18 (Mac-1) and blocks adhesion to fibrinogen and ICAM-1. Blood 95, 3788–3795 Watts, H.J., Lowe, C.R., and Pollard-Knight, D.V. (1994) Optical biosensor for monitoring microbial cells. Anal. Chem. 66, 2465– 2470 Nunomura, W., Takakuwa, Y., Parra, M., Conboy, C., and Mohandas, N. (2000) Regulation of protein 4.1R, p55, and glycophorin C ternary complex in human erythrocyte membrane. Biol. Chem. 275, 24540–24546
Label-Free Detection with the Resonant Mirror Biosensor 21. Nunomura, W., Takakuwa, Y., Parra, M., Conboy, C., and Mohandas, N. (2000) Ca2+dependent and Ca2+-independent Calmodulin Binding Sites in Erythrocyte Protein 4.1. Implications for regulation of protein 4.1 interactions with transmembrane proteins. J. Biol. Chem. 275, 6360–6367 22. Gill, A., Bracewell, D.G., Maule, C.H., Lowe, C.A., Hoare, M. (1998) Bioprocess. monitoring: An optical biosensor for rapid bioproduct analysis. J. Biotechnol. 65, 69–80 23. Rasooly, L., and Rasooly, A. (1999) Real time biosensor analysis of Staphylococcal enterotoxin A in food. Int. J. Food Microbiol. 49, 119–127 24. Skaldal, P. (1999) effect of methanol on the interaction of monoclonal antibody with free and immobilized atrazine studied using the resonant mirror-based biosensor. Biosens. Bioelectron. 14, 257–263 25. IAsys Technical note 3.1 BI (1994) The resonant mirror 26. Goddard, N.J., Pollard-Knight, D., and Maule, C.H. (1994) Real-time biomolecular interaction analysis using the resonant mirror sensors. Analyst, 119, 583–588 27. Goddard, N.J., Singh, K., Holmes, R.J., and Bastani, B. (1998) Resonant grating sensors using frustrated total-internal reflection. Sens. Actuators B 51, 131–136 28. Hulme, J., Malins, C., Singh, K., Fielden, P.R., and Goddard, N.J. (2002) Internally references resonant mirror for chemical and biochemical sensing. Analyst 127, 1233– 1236 29. Goddard, N.J., Singh, K., Hulme, J.P., Malins, C., and Holmes, R.J. (2002) Internally-referenced resonant mirror devices for dispersion compensation in chemical sensing and biosensing applications. Sens. Actuators A 100, 1–9 30. Ke, Y., Wilkinson, M.C., Fernig, D.G., Smith, J.A., Rudland, P.S., and Barraclough, R. (1992) A rapid procedure for production of human basic fibrobalst growth factor in E. coli. Biochim. Biophys. Acta 1131, 307–310 31. Kinsella, L., Chen, H.L., Smith, J.A., Rudland, P.S., and Fernig, D.G. (1998) Interactions of putative heparin-binding domains of basic fibrobalst growth factor and its receptor, FGFR-1, with heparin. Glycoconjugate J. 15, 419–422 32. Skerra, A., Pfitzinger, I., and Plückthun, A. (1991) The functional expression of antibody Fv fragemnts in Escherichia coli: Improved vectors and a generally applicable purification technique. Bio/Technology 9, 273–278
137
33. De Mey, J. (1986) The preparation and use of gold probes. Immunocytochemistry- Modern Methods and Applications, 2nd edition, Polak, J. M., Van Noorden, S., Wright and Sons, Bristol, UK 34. Richman, P.I., and Bodmer, W.F. (1987) Monoclonal antibodies to human colorectal epithelium: markers for differentiation and tumour characterization. Int J Cancer 39, 317–328 35. Willcocks, T.C., and Craig, I.W. (1990) Characterization of the genomic organization of human carcinoembryonic antigen (CEA): comparison with other family members and sequence analysis of 5′ controlling region. Genomics 8, 492–500 36. Edwards, P.R., Gill, A., Pollard-Knight, D., Hoare, M., Puckle, P.E., Lowe, C.R., and Leatherbarrow, R.J. (1995) Kinetics of protein–protein interactions at the surface of an optical biosensor. Anal. Biochem. 231, 210–217 37. Schuck, P. (1996) Kinetics of ligand binding to receptor immobilized in a polymer matrix, as detected with an evanescent wave biosensor. I. A computer simulation of the influence of mass transport. Biophys. J. 70, 1230–1249 38. Athanassopulou, N.R.J., Davies, P.R., Edwards, D.Yeung., Maule, C.H. (1998) Cholera toxin and GM1: a model membrane study with IAsys. Biochem. Soc. Trans. 27, 340–343 39. Brizard, B.L., Chubet, R.G., and Vizard, D.L. (1994) Immunoaffinity purification of FLAG epitope-tagged bacterial alkaline phosphatase using a novel monoclonal antibody and petide elution. BioTechniques 16, 730–735 40. IAsys Application note 1.3 (1994) Sensitivity enhancement using colloidal gold complexes 41. Fernig, D.G., and Gallagher, J.T. (1994) Fibroblast growth factors: an information network controlling tissue growth, morphogenesis and repair. Prog. Growth Factor Res. 5, 353–377 42. Turnbull, J.E., Fernig, D.G., Key, Y., Wilkinson, M.C., and Gallagher, J.T. (1992) Identification of the basic fibroblast growth factor binding sequence in fibroblast heparan sulphate. J. Biol Chem. 267, 10337–10341 43. IAsys application note 3.6 (1994) Molecular recognition, protein–carbohydrate interaction 44. Watts, H., Yeung, D., and Parkes, H. (1995) Real-time detection and quantification of DNA hybridization by an optical biosensor. Anal. Chem. 67, 4283–4289
138
Zourob et al.
45. Chaiet, I., and Wolf, F.J. (1964) The properties of streptavidin, a biotin-binding protein produced by Streptomycetes. Arch. Biochem. Biophys. 106, 1–5 46. IAsys application note 3.3 (1994) Molecular recognition Nucleic acid hybridization. 47. Ward, E.S., Gussow, D., Griffiths, A.D., Jones, P.T., Winter, G. (1989) Binding cativities of a reportoier of single immunoglobulin variable domains secreted from Escherichia coli. Nature 341, 544–546 48. IAsys application note 2.2 (1994), Kinetic analysis recombinant antibody fragment (D1.3 Fv) binding to immobilized hen egg lysozyme 49. Berry, M.J., Wattam, T.A., Willets, J., Lindner, N., de Graaf, T., Hunt, T., Gani, M., Davis, P.J., and Porter, P. (1994) Assay and purification of Fv fragments in fermenter cultures: design and evaluation of generic binding reagents. J. Immunol. Meth. 167, 173–182 50. Gill, A., Leatherbarrow, R.J., Hoare, M., Pollard-Knight, D.V., Lowe, P.A., Fortune, D.H. (1996) Analysis of kinetic data of antibody-antigen interaction from an optical biosensor by exponential curve fitting. J. Biotchnol. 18, 117–127 51. Grov, A., Myklestad, B., and Oeding, P. (1964) Immunochemical studies on antigen preparations from Staphaylococcus aureus.1. Isolation and characterization of antigen A. Acta Pathol. Microbiol. Scand 61, 588–596 52. Oeding, P., Grov, A., Myklestad, B. (1965) Antigenic properties of Staphylococci. Acta Pathol. Microbiol. Scand. 182, 183–190
53. Haukenes, G. (1974) Cellular antigen of Staphylococci. Acta Pathol. Microbiol. Scand. 236, 15–21 54. Gold, P., and Freedman, S.O. (1965) Demonstration of tumor-specific antigens in human colonic carcinomata by immunological tolerance and absorption techniques. J. Exp. Med. 121, 439–462 55. Goldenberg, D.M., Melville, M., and Carter, A.C. (1981) Carcinoembryonic antigen; its role as a marker in the management of cancer. A national Institute of Health Consensus Development Conference. Ann. Intern. Med. 94, 407–409 56. IAsys application note 5.2. (1994), Receptor–cell interactions, binding of L cells bearing the CEA antigen to an immobilized anti-CEA antibody 57. Absolom, D.R. and Van Oss, C.J. (1986) The nature of the antigen–antibody bound and the factors affecting its association and dissociation. CRC Crit. Rev. Immunol. 6, 1–46 58. Langone, J.J. (1982) Protein A of Staphylococcus aureus and related immunoglobulin receptors produced by Streptococci and Pneumococci. Adv. Immunol. 32, 157–252 59. IAsys application note 2.1 (1994), Kinetic analysis Protein A and Human IgG interaction 60. Zourob, M., Mohr, S., Treves-Brown, B.J., Fielden, P.R., McDonnell, M.B., and Nicholas, J.G. (2005) An integrated metal clad leaky waveguide sensor for detection of bacteria. Anal. Chem. 77, 232–242
Chapter 7 Label-Free Detection with the Liquid Core Optical Ring Resonator Sensing Platform Ian M. White, Hongying Zhu, Jonathan D. Suter, Xudong Fan, and Mohammed Zourob Summary Optical label-free detection prevents the cost and complexity of fluorescence and radio labeling while providing accurate quantitative and kinetic results. We have developed a new optical label-free sensor called the liquid core optical ring resonator (LCORR). The LCORR integrates optical ring resonator sensors into the microfluidic delivery system by using glass capillaries with a thin wall. The LCORR is capable of performing refractive index detection on liquid samples, as well as bio/chemical analyte detection down to detection limits on the scale of pg/mm2 on a sensing surface. Key words: Optical ring resonator, LCORR, Whispering gallery modes, Refractive index detection, Protease detection, DNA sequence detection.
1. Introduction The liquid core optical ring resonator (LCORR) sensing platform (1–8) integrates micro-capillary fluidics with label-free optical ring resonator sensing technology. Optical ring resonators have been studied for a decade for sensing applications (9–24). However, due to the use of the capillary as the ring resonator, the LCORR inherently integrates the sensor head with the sample fluidics, which can increase optical performance while simplifying the system design. The LCORR sensing platform is illustrated in Fig. 1. The ring resonator is formed in the circular cross section of the
Avraham Rasooly and Keith E. Herold (eds.), Methods in Molecular Biology: Biosensors and Biodetection, Vol. 503 © Humana Press, a part of Springer Science + Business Media, LLC 2009 DOI: 10.1007/978-1-60327-567-5_7
139
140
White et al.
Fig. 1. LCORR sensing platform. (a) The ring resonator is defined in the cross-section of the LCORR capillary, and is excited by evanescent coupling from a fiber taper or waveguide; (b) An evanescent field at the inner surface interacts with the sample; (c) Detection of analytes causes a spectral shift in the resonant wavelength; the spectral position of the resonant wavelength over time forms the sensing signal.
capillary. The capillary wall acts as a waveguide; light travels repeatedly in a circle around the circumference of the capillary as the ring guides the light through total internal reflection. Light is evanescently coupled into the ring resonator using a fiber taper or a pedestal waveguide. Light that forms an integer number of wavelengths around the circumference of the ring is resonant; the resonating modes are called whispering gallery modes (WGMs). As shown in Fig. 1b, the WGM has an evanescent field that extends beyond the inner wall of the LCORR capillary, where it interacts with the sample as it moves through the inside of the capillary. Analytes are detected by immobilizing biorecognition molecules (e.g. antibodies) that capture the analytes at the inner LCORR surface, where they interact with the evanescent field of the WGM. The presence of the analytes in the optical field changes the effective refractive index experienced by the WGM. Thus the effective optical path length around the ring changes, which causes the resonating wavelength to shift spectrally. This change in the WGM spectral location over a period of time as the sample passes through the LCORR capillary is the sensor signal, as illustrated in Fig. 1c. In this chapter, the method for creating LCORRs is described, the preparation of the LCORR sensor setup is presented, and the steps for performing sensing measurements are given in detail. Protocols for some sensing applications of the LCORR are
Label-Free Detection with the Liquid Core Optical Ring Resonator Sensing Platform
141
presented. These include a bulk refractive index measurement, an assay for detecting protein adsorption and protease activity, and an assay for detecting the presence of a specific DNA sequence in a sample.
2. Materials 2.1. Components and Equipment for Producing LCORRs and Fiber Tapers
1. CO2 laser. #F48-2(S)W, 25W (Synrad, Inc., Mukilteo, WA). 2. Glass tubes. Various tubes have been used as the LCORR preform, including aluminosilicate tubes (#A120-85-10, Sutter Instrument, Novato, CA), silica tubes, (#Q120-90-10, Sutter Instrument), and silica capillary tubing (#TSP530660, Polymicro Technologies, Phoenix, AZ). 3. Fiber optic cable. Single mode fiber, #SMF28 (Corning, Inc., Corning, NY). 4. Motion controller. 4-axis motion controller, #NI PCI-7390 (National Instrument, Austin, TX). 5. Mechanical slide type 1. Belt-drive slide #ZF1 (Techno, Inc., New Hyde Park, NY). 6. Mechanical slide type 2. Sherline linear slide (no part #) (Sherline, Inc., Vista, CA). 7. Data acquisition card. PCI-based 37-pin NI-DAQ, #NI PCI-6221 (National Instrument, Austin, TX). 8. UV-curable glue. Norland Optical Adhesive #8101 (Norland, Cranbury, NJ). 9. Plain microscope slides, Fisherbrand #12-550A (Fisher Scientific, Pittsburgh, PA). 10. Gas torch. The Little Torch (Smith Equipment, Watertown, SD). 11. Fiber optic clamp. adjustable force magnetic clamp, #T711250, (Thorlabs, Newton, NJ).
2.2. Components and Equipment for the Experimental Setup
1. Pumps. Various pumps have been used for moving the samples through the LCORRs, including a syringe pump (#55-1140, Harvard Apparatus, Holliston, MA) and a peristaltic pump (Masterflex #7562-10, Cole-Parmer, Vernon Hills, IL). 2. Tunable laser. Various tunable lasers have been used in the experimental setup, including a butterfly-packaged 1,550 nm distributed feedback (DFB) laser (JDSU #CQF935, JDS Uniphase Corp., Milpitas, CA), a 785 nm DFB laser (#DL100, Toptica Photonics), and a 980 nm external cavity laser (Velocity #6309, New Focus, San Jose, CA).
142
White et al.
3. Photodetector. Large area IR photoreceiver, #2033 (New Focus). 4. Data acquisition card. PCI-based 37-pin NI-DAQ, #NI PCI6221 (National Instrument, Austin, TX). 5. Thermo-electric cooler (TEC). #DT6-6-01 (Marlow Industries, Dallas, TX). 6. TEC controller. #LDT5910-B (ILX Lightwave, Bozemen, MT). 7. Thermister. #YSI 44036RC (YSI Temperature, Dayton, OH). 8. Tubing. Tygon microbore tubing, 0.01″ inner diameter, 0.03´ outer diameter, #EW-06418-01 (Cole-Parmer). 2.3. Buffers and Reagents
1. 18 MW water (referred to as water throughout). Produced by the EASYpure UV, #D7401 (Barnstead/Thermolyne Corp., Dubuque, IA). 2. Ethanol. Absolute, 200 proof, #E7023 (Sigma-Aldrich, St. Louis, MO). 3. Methanol. Absolute, acetone-free, #M1775 (Sigma-Aldrich). 4. Phosphate buffered saline (PBS) tablets. #P-4417 (SigmaAldrich). 5. Hydrochloric acid (HCl). Certified A. C. S. plus, #A144-212 (Fisher Scientific). 6. Hydrofluoric acid (HF). 48% A. C. S. reagent, #244279 (Sigma-Aldrich). 7. 3× SSC buffer. Prepare 0.45 M sodium chloride (NaCl, #S271, Fisher Scientific) and 0.045 M sodium citrate (#S1804, SigmaAldrich) in water.
2.4. Silanes, Cross-Linkers, and Biomolecules
1. 3-aminopropyl-trimethoxylilane (3-APS). 97%, #281778 (SigmaAldrich), store at 4 °C. 2. Dimethyl adipimidate dihydrochloride (DMA). 99%, #285625 (Sigma-Aldrich), store at 4 °C. 3. 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide HCl (EDC): #22980 (Pierce Biotechnology, Rockford, IL), store frozen. 4. N-Hydroxysulfosuccinimide (sulfo-NHS): #24310 (Pierce Biotechnology), store at 4 °C. 5. Bovine serum albumin (BSA). Minimum 96%, #A2153 (Sigma-Aldrich). 6. Oligonucleotide probes. Purchased through Sigma-Genosys (Sigma-Aldrich) with the custom-designed sequence: 5′-CCAACCAGAGAACCGCAGTCACAAT; the 5′ end is aminated, and has a 6-Carbon spacer. 7. DNA samples. All DNA samples are custom designed SigmaGenosys oligonucleotides of length 25mer, 50mer, and 100mer (Sigma-Aldrich).
Label-Free Detection with the Liquid Core Optical Ring Resonator Sensing Platform
143
3. Methods This chapter describes in detail the process for first setting up an LCORR measurement and then the process for performing a sample analysis with the LCORR. At this stage of the development of the LCORR for use in bio/chemical analysis, it is necessary to assemble many of the components in the lab, and to produce LCORRs on site. Figure 2 presents a flow chart of all the steps necessary to conduct a single sample analysis, from producing the LCORR to analyzing the data. This chapter further explains how to create the experimental setup, as well as how to assemble on-site LCORR and fiber taper production systems. 3.1. Producing LCORR Capillaries
The primary component of the LCORR sensing system is a glass capillary that acts simultaneously as a microfluidic channel for sample delivery and as a ring resonator for sample detection. This
Fig. 2. The process for conducting a bio/chemical molecule detection experiment using the LCORR sensing platform.
144
White et al.
capillary is unlike the typical capillaries utilized for liquid or gas sample movement because the wall of the LCORR must be very thin. In fact, for sensing purposes, the LCORR capillary wall must be less than 5 μm thick. As capillaries with this dimension are not available today, it is required to produce LCORR capillaries in the lab before conducting any detection assays. To produce a capillary with a thin wall, we stretch a commercially available glass capillary to thin its dimensions (see Note 1). This is analogous to the process of drawing fiber optic cable from a preform. The preform glass capillary is heated to the softening point while one end is pulled. As wall thickness is critical, attention must be given to the temperature and pulling speed, as these parameters dictate the change in the aspect ratio (diameter to wall thickness) during pulling. Higher speed and lower temperature will combat the effects of surface tension, which is pulling the softened glass radially inward and decreasing the aspect ratio. Preservation of nearly 100% of the aspect ratio during capillary drawing has been exhibited with our LCORR drawing technique, and has been demonstrated in similar work (25). The configuration utilized for drawing LCORR capillaries and a photo of the setup are shown in Fig. 3. The entire apparatus is contained within a clear acrylic enclosure to reduce air currents, which can cause fluctuations in the temperature of
Fig. 3. (a) Diagram of LCORR drawing setup (reproduced from ref. 8 with permission from SPIE). (b) Photo of a constructed setup to draw LCORRs.
Label-Free Detection with the Liquid Core Optical Ring Resonator Sensing Platform
145
the heating zone. CO2 lasers are used as the heat source to soften the glass capillary in the heating zone (see Note 2). Two CO2 lasers are used on opposite sides to provide more evenly distributed heating (see Note 3). The power of the lasers, which is computer-controlled using a data acquisition card, is dependent upon the type of glass used (see Notes 4 and 5). While the lasers soften the glass tube in one spot, the two stages holding each end of the glass tube are moved to stretch the glass. The tube can be taped onto each stage, although any temporary clamp will suffice. The stages are mechanical slides with stepper motors for movement. The movement of the stages is controlled with a PCIbased motion controller. One stage is pulled quickly away from the heating zone while the other is slowly pushed toward the heating zone, keeping constant the mass of glass in the heating zone. The ratio of pulling speed to feed-in speed controls the diameter of the pulled capillary (see Note 6). In this prototype implementation, a computer code controlling the stage movements and laser power are developed using LabView software (National Instruments). User inputs to the program include the desired laser power, pulling speed, and feed-in speed. The program operates through the DAQ card and the PCIbased motion controller to generate the appropriate voltage-based outputs to control the laser power and the pull/feed stages. The steps to be performed for pulling an LCORR capillary are listed below: 1. The preform glass capillary is bridged between the two stages, using clamps or tape to hold the glass in place. 2. The lasers are turned on to the power level used for softening the glass (see Note 7). 3. About 5 s are allowed to pass before activating the stages so that the glass has time to heat and soften. 4. The pulling and feed-in stages are moved at a constant speed until they reach the pre-set point, corresponding to the desired length. 5. The LCORR capillary is removed from the clamps by handling it with the remaining ends of the preform. With typical pulling conditions, the capillary is relatively robust, and can be handled easily. 6. The LCORR capillary can be cut from the final glass piece, although in some prototyping applications, it is desirable to keep at least one end of the remaining preform for handling purposes. The resonant region of the LCORR capillary extends from the thinned portion in the final heating zone to the portion of the capillary, where the equilibrium in the diameter is reached. Figure 4 illustrates the resonant region of a drawn LCORR capillary, and shows a photo of an LCORR drawn from a glass tube preform.
146
White et al.
3.2. Pulling Tapers
For prototyping and pre-clinical applications of the LCORR sensing system, it is sometimes practical to utilize a tapered fiber optic cable to excite WGMs in an LCORR (see Note 8). The operating principle of the fiber taper is that a fiber optic cable that is thinned to a few micrometers in diameter will have an evanescent field outside the cable. This evanescent field is capable of coupling light into the LCORR. In general, fiber tapers must be produced in the lab in which they will be used because of their fragility. Thus, taper production is presented here as a necessary component in the LCORR assay development. An illustration of the setup for pulling fiber tapers and a photo of the setup are presented in Fig. 5. The taper is produced by stretching a fiber optic cable under heat. Typical single mode fibers are utilized as the tapering fiber. The heat source is a clean flame provided by gas torch (see Note 9) (26). The fiber optic cable is stretched slowly from both sides of the heat zone while the heat zone is scanned back and forth by about 1 cm
Fig. 4. LCORR capillary after pulling from a preform.
Fig. 5. (a) Setup for pulling fiber tapers. (b) Photo of a constructed setup to pull tapers.
Label-Free Detection with the Liquid Core Optical Ring Resonator Sensing Platform
147
in order to provide a region of constant diameter. Just as with the LCORR pulling apparatus, the stages are mechanical slides controlled by a computer via a code developed in LabView and a PCI-based motion controller. The LabView program is very similar to the one used for pulling LCORR capillaries. It may be desirable to pass laser light through the fiber and to measure the loss during taper pulling to monitor the health of the taper while it is produced. The steps to be performed for pulling fiber taper are listed below: 1. Prepare a length of fiber cable that is long enough such that the leads on each side of the taper can be connected to the experimental setup (see Subheading 3.3). One meter is generally sufficient. 2. In the center of the fiber cable, strip away the polymer jacket along a 4 cm region and clean away dust and polymer remains. This will be the tapering region. 3. Using fiber optic clamps mounted on the stages, mount the fiber into the setup. 4. Position the stage holding the torch such that the tube is centered with respect to the tapering region. 5. Turn on the gas for the torch and light the flame. Position the torch so that the tapering region is being heated to the desired temperature. 6. Allow enough time before pulling for the oven to heat to the silica’s softening temperature. This may take 5 s. 7. When the glass is softened, the stages pull in opposite directions at a speed of around 0.007 cm/s. Meanwhile, the torch is scanned back and forth at a speed of 0.1 cm/s, with a travel distance of 5 mm. 8. When the desired taper diameter is reached, the stages immediately stop pulling and the torch is turned off (see Notes 10 and 11). 9. Mount the taper on a rigid holder such that the taper is exposed and not in contact with anything, while being anchored on both sides for support (see Note 12). Figure 6 shows a drawing and a photo of an anchored taper. 3.3. Creating the Experimental Setup
The purpose of the experimental setup is to pass the sample through the LCORR while monitoring the sample’s effect on the spectral position of the WGMs. The experimental setup is given in Fig. 7. The LCORR is connected to tubing so that the sample can be passed through, using a pump. The taper is connected at one end to a tunable laser (see Note 13) and at the other to a photodetector. The tunable laser scans across a spectral range
148
White et al.
Fig. 6. (a) Sketch and (b) photo of a fiber taper anchored onto a portable mount.
Fig. 7. LCORR sensing platform experimental setup (reproduced from ref. 7 with permission from SPIE).
wide enough to detect a WGM and track its movement. Typically, 10–15 GHz (i.e., about 100 picometers (pm) for a center wavelength of 1,550 nm) of tuning range is sufficient for this. When the scanning laser passes through a resonant wavelength, destructive interference occurs on the fiber taper at the coupling point with the LCORR, resulting in an observable decrease in the output power. Thus, scanning the laser across one WGM will
Label-Free Detection with the Liquid Core Optical Ring Resonator Sensing Platform
149
produce a measured waveform like the one shown in Fig. 7 (see Note 14). Tunable laser control and data collection are performed by a data acquisition (DAQ) card, under the direction of the code written in LabView. The DAQ voltage output scans the output wavelength of the laser while the input samples the voltage at the output of the photodetector. After each scan, the sample values are stored in respective files on the computer. Also, the LabView program displays the measured voltage in real time. Following the experiment, a simple program written in Matlab (Mathworks) is used to scan each file for the spectral location of the voltage minimum (indicative of the WGM), and then a data set is created to represent the WGM spectral position over time. This data set is the sensorgram. An exemple of a sensorgram is shown in Fig. 1c. As with many label-free optical sensors, the signal from the LCORR sensor is temperature dependent (4). Therefore, a temperature control system is required in the experimental setup to suppress temperature fluctuations in the LCORR sensing region that would be translated to noise in the sensing signal. Our design is based on a thermo-electric cooler (TEC) and TEC controller. The setup is illustrated in Fig. 8.
Fig. 8. (a) Temperature control setup for the LCORR sensor. (b) Photo of a taper and LCORR mounted on the temperature control setup. A white line is drawn along the taper so that it can be visualized.
150
White et al.
The steps to be performed for creating the experimental setup are listed below. 1. Create the temperature control setup. Place the TEC heat sink and TEC on the bench top (or on a platform elevated above the bench top). It is recommended to use a thermally conductive adhesive between materials in this setup. Place a machined (as shown in Fig. 8) conductive block on top of the TEC. This is the TEC controlled mount. 2. Place a thermister onto (or inside a machined hole of) the TEC controlled mount. Mounting the thermister closer to the LCORR sensing region results in a better temperature control performance. 3. Connect the leads of the TEC and the thermister to the TEC controller. 4. Next, prepare the LCORR. Attach tubing to each side of the LCORR while anchoring the LCORR onto the mount. The tubing diameter should be just large enough to fit around the diameter of the LCORR and should be compatible with either a peristaltic pump or a syringe pump. The mount must leave the sensing region of the LCORR accessible while providing support (see Notes 15 and 16). A drawing and a photo are shown in Fig. 9. 5. Mount the LCORR on a 3-dimensional optical stage, such that the sensing region is accessible to a fiber taper. 6. Mount the fiber taper on a 3-deminsional stage such that the tapered region can be brought into contact with the LCORR. 7. Connect one end of the taper to the tunable laser output and the other to the photodetector input (splicing or free-space coupling can be used). 8. Connect the DAQ voltage output to the laser wavelength control (see Note 17); connect the DAQ voltage input to the photodetector output. 9. Begin scanning the laser and monitoring the photodetector signal. 10. Bring the LCORR in contact with the tapered region of the fiber while monitoring the WGMs (see Note 18). Use a microscope camera to assist in the positioning process. The tapered region of the fiber should contact the LCORR resonant region. The two objects are at approximately 90° with respect to each other, as shown in Figs. 7 and 8. Attempt to excite the LCORR resonance with different positions along the taper to optimize the coupling strength. Generally, the best coupling will occur at the thinnest point of the taper.
Label-Free Detection with the Liquid Core Optical Ring Resonator Sensing Platform
151
Fig. 9. (a) Sketch and (b) photo of an LCORR with attached tubing, mounted on a portable mount.
11. When the preferred locations of the taper and LCORR are identified, lower the LCORR to be in contact with the LCORR heat sink as shown in Fig. 8. Then bring the taper into contact with the LCORR. 12. Connect one of the tubes that are connected to the LCORR to the pump and use the pump to move the sample through the LCORR. For liquid samples, a peristaltic pump or a syringe pump can be used for slow, constant flows (see Notes 19 and 20). 3.4. Preparing and Characterizing the LCORR for Sensing
After pulling the LCORR capillary, the wall thickness may not be as thin as desired for sensing purposes. Additionally, the sensitivity must be characterized before use. Sub-micron differences in wall thickness result in significant differences in the sensitivity of the LCORR, and thus it must be well-characterized. Therefore, once an LCORR is pulled, it is placed in the experimental setup for sensitivity optimization and characterization. To set the sensitivity of the LCORR, hydrofluoric acid (HF) is used to slowly etch away the inner capillary surface. The amount of glass to be removed depends on the preform and the capillary
152
White et al.
pulling process. For example, if the preform has an outer diameter of 1 mm and an inner diameter of 0.9 mm, and the LCORR’s outer diameter is 100 μm, then the thinnest that the wall can be is 10 μm. To achieve a good sensitivity, around 7 μm may need to be removed with HF. This is done by passing diluted concentrations of HF through the LCORR while intermittently characterizing the sensitivity. The sensitivity is characterized using solutions of known refractive indices, as the sensing mechanism of the LCORR is based on refractive index detection. We use solutions of ethanol in water (1), as the difference in refractive index of the varying concentrations is known (27). Determining the WGM spectral shift for a particular change in refractive index of the sample leads to the refractive index sensitivity of the LCORR sensor. The steps for optimizing and characterizing the sensitivity of the LCORR are listed here: 1. Prepare solutions of HF in water (see Note 21). The desired strength of HF depends on the type of glass used for the LCORR. Silica etches slowly, so stronger HF must be used. Aluminosilicate or other glasses etch much faster than silica. Typically, 5–10% HF is used for silica, while 1–2% may be used with aluminosilicate. 2. Prepare solutions of ethanol. The desired concentrations depend on the sensitivity goal for the LCORR. For example, if 10 nm/RIU (refractive index units) is desired, and considering that a 10 pm WGM spectral shift is practical to detect, the test samples should be prepared in increments of approximately 1 mRIU. This is equivalent to increments of about 2% (v/v) ethanol in water. 3. With the LCORR in contact with the taper, begin flowing HF through the LCORR (see Note 22). Monitor the WGM spectrum (as described in Subheading 3.3) while the LCORR is etching (see Note 23). When the LCORR becomes sensitive, the spectral position will drift to lower wavelengths. This is because the wall thickness is slowly decreasing, causing more light to interact with the sample, and thus changing the effective optical diameter of the LCORR. The speed of the WGM shift is an indicator of the wall thickness (see Note 24). 4. Once the spectral movement of the WGMs is easily visible, the sensitivity of the LCORR should be characterized. 5. Stop the etching by passing water through the LCORR. The WGM spectral movement should stop almost instantly after the water begins passing through the LCORR. 6. Once the WGM spectral position has stabilized, the refractive index sensitivity test can be conducted. While monitoring the WGM spectral position, change the sample solution in the LCORR from water to the lowest concentration of ethanol.
Label-Free Detection with the Liquid Core Optical Ring Resonator Sensing Platform
153
Record the WGM spectral shift (an exemplary WGM spectral shift is shown in Fig. 10). When the refractive index inside the LCORR increases, the WGM spectral position should move to a longer wavelength (red shift). Allow the mode to stabilize, and then change the solution in the LCORR back to water, while recording the WGM spectral shift. 7. Repeat this a few times. Record the WGM spectral shift in picometers (pm). If no spectral shift is visible, return to the HF etching step. If the WGM spectral shift is apparent at all, proceed with the characterization process. 8. Repeat step 6 (water–ethanol solution–water) with the other concentrations of ethanol while recording the WGM spectral shift that occurs for each solution change. 9. Determine the average WGM shift for each of the ethanol solutions. 10. Plot the spectral shift vs. change in the refractive index and perform a linear fit of the data. The slope of this fit is the refractive index sensitivity (see Note 25). An exemplary refractive index sensitivity plot is presented in Fig. 11 for an LCORR with a sensitivity of 315 pm/RIU. 11. If the refractive index sensitivity is sufficient, then the optimization and characterization process is complete. If not, then return to the HF etching step. If the sensitivity is close, then it may be prudent to reduce the concentration of the HF. 12. Clean the inside surface of the LCORR thoroughly with any desired glass treatment, such as 1:1 methanol:HCl. At least
Fig. 10. The observed WGM shift in response to a 10% ethanol solution replacing water inside an LCORR with a sensitivity of 315 pm/RIU.
154
White et al.
Fig. 11. The measured refractive index sensitivity for an exemplary LCORR. Data is recorded for ethanol solutions of 10%, 20%, 30%, 40%, and 50%. The values are the result of an average across at least six observed WGM shifts. The slope of the linear fit is the refractive index sensitivity.
2 h of treatment is recommended. Caution should be exercised with the reactivity of any of the glass treatments with tubing used in the experimental setup. 3.5. Bulk Refractive Index Detection
The characterization procedure outlined in Subheading 3.4 illustrates that the LCORR sensing platform can be used to identify the refractive index of sample liquids. Once the sensitivity has been characterized, liquids of unknown refractive index can be passed through the LCORR while the WGM spectral position is monitored. This can be used, for example, to identify if a sample has a small amount of contaminants. The following procedure is used to measure the bulk refractive index. 1. Prepare the experimental setup described in Subheading 3.3. 2. Prepare a baseline solution. Ideally, this solution will be relatively close in refractive index to the sample. For example, if looking for the amount of water contamination in ethanol, then pure ethanol should be used as the baseline solution. 3. Use the pump to drive the baseline solution through the LCORR. It is recommended to move the liquid slowly, e.g., 10 μL/min. 4. Monitor the WGM spectral position. 5. Quickly switch to the sample, while trying to minimize any air gaps between the baseline liquid and the sample (see Note 26).
Label-Free Detection with the Liquid Core Optical Ring Resonator Sensing Platform
155
6. Record the WGM spectral shift that occurs as the sample replaces the baseline liquid in the LCORR. 7. Switch back to the baseline liquid, and repeat the measurement several times. 8. Compute the average value of the WGM spectral shift. 9. The refractive index difference between the baseline and the sample is easily computed by dividing the measured WGM spectral shift by the LCORR sensitivity measured in the characterization procedure. 3.6. Detection of Protein Binding and Proteolytic Activity
The LCORR can utilize its refractive index sensing capabilities to detect biomolecule analytes that are captured at the inner surface of the capillary. Typically, biomolecules have refractive indices around 1.45–1.55, while buffers are typically close to the range of 1.3–1.35. Thus, when biomolecules bind to the surface, the local refractive index in the region of the WGM evanescent field increases. This RI increase is reflected in the sensor signal by a red shift of the WGM spectral position. Figure 12 illustrates an exemplary sensor signal reflecting the binding of BSA molecules at the inner surface of the capillary. Because the total WGM spectral shift is proportional to the number of molecules that bind to the surface, the concentration of analytes in the sample can be determined when using the LCORR as a sensor. Through the same refractive index sensing mechanism that causes a red-shift for binding analytes, a blue shift in the WGM spectral position occurs when analytes are removed from the capillary surface. Thus, the LCORR can also be used to detect
Fig. 12. Sensorgram showing the binding of BSA molecules at the inner surface of the LCORR capillary. The BSA concentration is 1 mg/mL.
156
White et al.
proteolytic activity. Figure 13 shows the sensor signal when trypsin is introduced into the same LCORR following the binding of BSA (shown in Fig. 12). Trypsin is known to cleave BSA molecules at a number of residues, which will cause a significant amount of the BSA mass to be removed from the LCORR. As an example of protein detection, the steps for detecting BSA molecules in a sample are listed below: 1. Prepare a solution of 10% ethanol in 90% water (v/v). 5 mL of solution is sufficient. 2. Prepare the surface functionalization silane. Use 1% (v/v) of 3-aminopropyltrimethoxysilane (3-APS) in the 10/90 ethanol/water. It is recommended to prepare about 2 mL of 3-APS solution. 3. Prepare the phosphate buffered silane (PBS). Dissolve a PBS tablet in water to prepare 0.01 M PBS as indicated in the manufacturer instructions. 4. Prepare the BSA solution. Dissolve the desired concentration of BSA in PBS. The sample volume required is very small. However, for practical purposes, it may be desirable to preparerepare at least 100 μL of BSA solution. If very low concentrations are used, a higher volume, such as 1 mL, should be prepared. A cross-linker will be added to this solution, but it should not be done until just before the sample is used. Otherwise, the protein molecules may be cross-linked together before the experiment begins. 5. Prepare the LCORR for surface functionalization. Begin by passing 1:1 HCl:methanol through the LCORR for at least
Fig. 13. Sensorgram showing the proteolytic activity of trypsin acting to remove BSA bound at the inner surface of the LCORR capillary. The trypsin concentration is 10 μg/mL.
Label-Free Detection with the Liquid Core Optical Ring Resonator Sensing Platform
157
30 min. Then rinse with methanol for at least 10 min. For all solutions throughout this measurement, the flow velocity should be kept below 10 μL/min. 6. Begin passing the 10/90 ethanol/water solution through the LCORR. Allow this solution to pass through the LCORR for approximately 15 min to establish a starting baseline for the WGM spectral position. Begin recording the WGM spectral position. 7. Quickly switch to 3-APS solution, minimizing the flow gap (without changing the pump speed) between the buffer solution and the 3-APS solution. The 3-APS should be passed through the LCORR for at least 15 min. The WGM spectral position should shift to a higher wavelength during the deposition, but should reach steady state after around 15 min (see Note 27). 8. Switch back to the 10/90 ethanol/water solution in the LCORR. This will cause some 3-APS to be released from the surface, resulting in a WGM blue shift. Once the blue shift is complete (10–15 min at most), allow another 10 min to establish the new signal baseline. The deposition of 3-APS should result in a red shift from the initial baseline WGM spectral position to the current position. 9. Begin passing PBS buffer through the LCORR and establish a new baseline position. 10. During this WGM stabilization, complete the preparation of the BSA solution. Add 1-Ethyl-3-[3-dimethylaminopropyl] carbodiimide hydrochloride (EDC) and N-hydroxysulfosuccinimide (sulfo-NHS) to the BSA/PBS solution to a concentration of 1 mg/mL each. EDC with sulfo-NHS is used to convert carboxyl groups to amine-reactive Sulfo-NHS esters (see Note 28). This solution should be prepared immediately before use. 11. Run the BSA solution through the LCORR for 30–45 min. A red shift with similar features to the one shown in Fig. 12 should occur. If the concentration of the BSA is high (greater than 0.1 mg/mL), then the total red shift may occur within as little as 5–10 min. 12. Rinse the LCORR with BSA. Again, the rinse should cause a blue-shift as some BSA will be removed from the surface of the LCORR. Once the blue shift is complete (10–15 min at most), allow another 10 min to establish the final signal baseline. The binding of BSA produces the red shift from the WGM spectral position after step 9 to the current WGM spectral position. 13. To determine the concentration of protein molecules in an unknown sample, the net WGM spectral shift should be
158
White et al.
compared to a calibration curve that plots the net WGM shift vs. protein concentration. The calibration curve is prepared using the same steps as above, but with samples that contain a known concentration of protein molecules. Naturally, the LCORR used to test the sample should have the same sensitivity as the LCORR used to produce the calibration curve. 14. Once the concentration in the sample is known, the reaction kinetics between the molecules in the sample and the capture molecules on the surface can be analyzed. The WGM spectral shift is related to the protein concentration in the sample by (14): dl WGM =
dlmax [protein] , K d + [protein]
where Kd is the dissociation constant and δlmax is the maximum WGM spectral shift in the calibration curve. Using this expression, the Kd can be determined for any protein sample. 15. If proteolytic activity is to be measured, the enzyme of interest should be dissolved in PBS to the desired concentration and passed through the LCORR. A blue shift in the WGM spectral position with features similar to the sensorgram shown in Fig. 13 should occur. 3.7. Detection of Specific DNA Sequences
One important application of the label-free biomolecule detection capability of the LCORR sensing platform is the identification of a specific DNA sequence in a sample. Similar to DNA microarray technology, an oligonucleotide probe that is designed to be the complement of the single-stranded target is immobilized on the sensor surface. If the target sequence exists in the sample, it will bind with an immobilized oligo probe at the LCORR inner surface. As shown in Subheading 3.6, the LCORR quantitatively detects biomolecules that bind to the surface. Surface chemistry for successful immobilization of the oligo probes is critical. First, the surface of the LCORR and the 5′ end of the oligo probe are functionalized with amine groups. Then, dimethyl adipimidate (DMA) is used to crosslink the oligo probes to the amino-functionalized LCORR surface. This surface chemistry is illustrated in Fig. 14. Upon immobilizing the probes, the LCORR is prepared to detect the presence of the complementary probe sequence in the sample. Figure 15 presents an exemplary sensorgram, which shows the WGM spectral shift during the functionalization, oligo probe immobilization, and sample analysis processes. The steps for conducting the DNA sequence detection measurement are outlined below: 1. Prepare a solution of 10% ethanol in 90% water (v/v). 5 mL of solution is sufficient.
Label-Free Detection with the Liquid Core Optical Ring Resonator Sensing Platform
159
Fig. 14. Surface chemistry for immobilizing oligo probes to the LCORR surface.
Fig. 15. Exemplary sensorgram for 3-APS functionalization, 25 base-pair oligo probe immobilization, and sample detection (25 base pair DNA sequence). In this case, all steps, including the 3-APS deposition, were performed in 3× SSC buffer. Therefore, there are no WGM spectral shifts due to buffer changes.
2. Prepare the surface functionalization silane. Use 1% (v/v) of 3-aminopropyltrimethoxysilane (3-APS) in the 10/90 ethanol/water. It is recommended to prepare about 2 mL of 3-APS solution. 3. Prepare 3× SSC buffer. 3× SSC is 0.45 M NaCl and 0.045 M sodium citrate. 10 mL of buffer is sufficient. 4. Prepare oligonucleotide probe solution. Dissolve the probe pellet into 3× SSC to produce a concentration of 10 μM. Prior to opening the tube of pelleted DNA, it may be advantageous to centrifuge it for 10–15 min to ensure that there is no aerosol material lost. Sonicate the solution to ensure that the oligos are dissolved. DMA will be added to this solution
160
White et al.
immediately before probe immobilization onto the LCORR. It should not be done in advance to avoid DMA hydrolysis and cross-linking of the oligo probes. 5. Prepare the DNA sample in 3× SSC (see Note 29). In many cases, the DNA sample will be the output of RT-PCR targeted at the sequence of interest. 6. Prepare the LCORR for surface functionalization and probe immobilization. Begin by passing 1:1 HCl:methanol through the LCORR for at least 30 min. Then rinse with methanol for at least 10 min. For all solutions, the flow velocity should be kept below 10 μL/min. 7. Begin passing 10/90 ethanol/water through the LCORR. If interested in using the LCORR sensing functionality to monitor the surface functionalization and probe immobilization processes, begin recording the WGM spectral position at this point. Let the solution run for at least 15 min while recording the spectral position in order to establish an initial baseline WGM position. 8. Quickly switch to 3-APS solution, minimizing the flow gap (without changing the pump speed) between the buffer solution and the 3-APS solution (see Note 26). The 3-APS should be passed through the LCORR for at least 15 min. The WGM spectral position should shift to a higher wavelength during the deposition, but should have reached steady state after around 15 min. 9. Switch back to the 10/90 ethanol/water solution in the LCORR. This will cause some 3-APS to be released from the surface, resulting in a WGM blue shift. Once the blue shift is complete (10–15 min at most), allow another 10 min to establish the new signal baseline. 10. Begin passing 3× SSC buffer through the LCORR and allow a new baseline position to be established. 11. During this WGM stabilization, complete the preparation of the oligo probe solution. Add DMA to the oligo solution to a concentration of 5 mg/mL. This solution should be prepared immediately (within 10 min) before use. 12. Run the DMA/oligo probe solution through the LCORR for 30–45 min. Because the DMA loses its reactivity after approximately one hour, there is no benefit in running the solution any longer. The WGM spectral position should undergo a red-shift as the oligo probes are immobilized onto the surface of the LCORR. 13. Rinse the LCORR with SSC. Again, the rinse should cause a blue-shift as some oligo will be removed from the surface of the LCORR. Once the blue shift is complete (10–15 min at most), allow another 10 min to establish the new signal baseline.
Label-Free Detection with the Liquid Core Optical Ring Resonator Sensing Platform
161
14. The LCORR is now prepared for the sample. Pass the sample through the LCORR while continuing to record the WGM spectral position. Allow the sample to pass through until the WGM has stabilized following the red shift. This may be as long as 1 h. Recycling the sample solution through the LCORR may enable more opportunities for the sample to bind at the LCORR surface. 15. Run a final SSC rinse through the LCORR. Once again, a blue shift will result as some target DNA is removed from the surface. After the blue shift is complete (10–15 min at most), allow another 10 min to establish the final WGM spectral position. 16. To determine the concentration of DNA molecules with the sequence of interest in an unknown sample, the net WGM spectral shift should be compared to a calibration curve that plots the net WGM shift vs. DNA concentration. The calibration curve is prepared using the same steps as above, but with samples that contain a known concentration of DNA molecules matching the sequence of interest. Naturally, the LCORR used to test the sample should have the same sensitivity as the LCORR used to produce the calibration curve. 17. Once the concentration in the sample is known, the reaction kinetics between the DNA molecules in the sample and the capture probe on the surface can be analyzed. The WGM spectral shift is related to the target DNA concentration in the sample by (14): dl WGM =
dlmax [DNA] , K d + [DNA]
where Kd is the dissociation constant and δlmax is the maximum WGM spectral shift in the calibration curve. Using this expression, the Kd can be determined for any sample.
4. Notes 1. Two different options have been used for the LCORR capillary preform: glass tubes with an outer diameter of 1.2 mm and an inner diameter of 0.9 mm, and silica capillaries with an outer diameter of 617 μm and an inner diameter of 535 μm. 2. A wire filament or heated ceramic oven can also be used as long as the temperature inside can be raised to the softening point of the glass without damaging the heating element.
162
White et al.
3. The lasers are slightly angled with respect to the perpendicular axis of the LCORR so that the beam of one laser is not hitting the opposite laser. 4. For quartz tubes with an outer diameter of 1.2 mm and an inner diameter of 0.9 mm (Suter Instruments), approximately 5 W from each laser has been used. For aluminosilicate tubes of the same size, approximately 2.5 W from each laser has been used. 5. It may be advantageous to have a feedback control design for the laser power, in which a laser micrometer monitors in real time the size and shape of the LCORR and invokes dynamic power adjustments of the lasers through computer control. 6. A typical pull speed is just over 1 cm/s while feeding at a speed of about 0.02 cm/s; this results in a diameter of approximately 115 μm from a glass tube with in initial diameter of 1.2 mm. 7. Extreme caution should be used with CO2 lasers. At these power levels, they can easily damage lab materials and are harmful to users. Moreover, the light is completely invisible. 8. Ultimately, in a clinical version of the LCORR, the capillaries would be mounted in a package on top of an array of pedestal waveguides, which would be used to excite WGMs (2–3). 9. A few options exist for pulling tapers (26, 28–30). 10. The desired diameter of the taper may depend on the wavelength of light utilized. Lower wavelengths will have a shorter evanescent field, and thus tapers for lower wavelength sources must be pulled thinner for sufficient evanescent exposure. 11. For SMF28, each stage pulls approximately 1.5 cm to produce a taper of approximately 3 μm. The difference in pulling length for a 1,500 nm source and a 600 nm source may be only about 1 mm. 12. A U-shaped holder made from glass microscope slides is assembled and UV-curable glue (Norland) is used to anchor the taper onto the arms of the mount. 13. Several options are available for tunable lasers. One common approach is to use distributed feedback (DFB) lasers, such as the Toptica LD100 or the JDSU CQF935. In this case, changing the laser gain current, tunes the output wavelength. Also, external cavity lasers, such as the New Focus Velocity 6300 series, can be used. For these lasers, applying a variable voltage signal onto a piezo-controlled grating or prism will tune the output wavelength. 14. Scanning the laser across 100 pm may result in the appearance of many WGMs.
Label-Free Detection with the Liquid Core Optical Ring Resonator Sensing Platform
163
15. UV-curable glue is used as the sealant. It is important to consider the sealant reactivity or solubility with any solutions that will pass through the LCORR. 16. The sealant tends to be pulled into the tubing by capillary action. It must be ensured that the sealant is not pulled far enough inside to block the sample’s path through the LCORR. 17. Two methods are commonly used for tuning lasers. In the case of distributed feedback (DFB) lasers, typically the gain current is modulated, which results in small wavelength shifts. In external cavity lasers, a piezo control for the external grating may be available for scanning the wavelength. 18. We use the DAQ and the LabView program to view the photodetector signal (and thus the WGMs) in real time. However, if this option is not available, a properly triggered oscilloscope will also work. 19. Some peristaltic pumps have been found to cause pressure variations inside the LCORR, which can add noise to the sensing signal (the WGM is extremely sensitive to the diameter of the LCORR, which can be altered slightly under varying pressure). 20. It is recommended to pull the sample instead of pushing it through the LCORR to reduce the possibility of breaking any of the seals along the fluidic path. 21. HF is a dangerous acid. Extreme caution should be exercised when using HF. Do not use glassware for containing the HF. 22. The LCORR etches more quickly if the HF solution is constantly moved through the LCORR, as opposed to sitting stagnant inside the LCORR. 23. A WGM spectral shift may occur very soon after the HF begins passing through the LCORR. However, this may be due to the temperature change caused by the reaction of the HF with the glass. 24. The spectral width of the WGM is expected to increase during the etching process. As the evanescent field in the liquid core increases, more light absorption occurs, which decreases the Q-factor and thus increases the spectral width of the mode. Furthermore, surface roughness of the LCORR due to etching can also reduce the Q-factor, especially as the evanescent field at the surface increases due to thinner walls. 25. First, convert the percentage of ethanol into a mole fraction of ethanol in water. Then use the expression from Ghoreyshi et al. (27) to find the difference in refractive index between water and the ethanol solution: 0.179258X - 0.380008X2 + 0.351867X3 – 0.124503X4
164
White et al.
where X is the mole fraction. This expression is validated for data taken at 632.8 nm at 25 °C. 26. Keeping the pump speed constant throughout the measurement is highly recommended, as pressure changes in the LCORR will cause shifts in the WGM spectral position. 27. Instead of pumping it through the LCORR, it can be made to sit still in the LCORR, allowing the 3-APS to diffuse to the surface. Changing the pump speed (or stopping the pump) corrupts the sensing signal, so this should only be done if the sensing signal is not desired for this step. 28. In the case of BSA deposition on an aminated surface, crosslinkers are not necessary, as the BSA will adsorb to the surface. Cross-linkers here are used for illustrative purposes, as in many typical measurements, they may be required to effectively bind the biomolecule to the surface. 29. If it is necessary to use a different buffer for the sample, then a baseline WGM spectral position must be established when switching buffers between the probe immobilization and sample deposition. This is similar to when the baseline is established for SSC buffer after the 3-APS deposition and 10/90 ethanol/water rinse.
References 1. White, I. M., Oveys, H., and Fan, X. (2006) Liquid core optical ring resonator sensors, Opt. Lett. 31, 1319–1321 2. White, I. M., Oveys, H., Fan, X., Smith, T. L., and Zhang, J. (2006) Integrated multiplexed biosensors based on liquid core optical ring resonators and antiresonant reflecting optical waveguides, Appl. Phys. Lett. 89, 191106191101–191106-1-3 3. White, I. M., Suter, J. D., Oveys, H., and Fan, X. (2007) Universal coupling between metal-clad waveguides and optical ring resonators, Opt. Express. 15, 646–651 4. Suter, J. D., White, I. M., Zhu, H., and Fan, X. (2007) Thermal characterization of liquid core optical ring resonator sensors, Appl. Opt. 46, 389–396 5. Zhu, H., White, I. M., Suter, J. D., Zourob, M., and Fan, X. (2007) An integrated refractive index optical ring resonator detector for capillary electrophoresis, Anal. Chem. 79, 930–937 6. Fan, X., White, I. M., Zhu, H., Suter, J. D., and Oveys, H. (2007) Overview of novel integrated optical ring resonator bio/chemical sensors, Proc. SPIE 6452, 64520M
7. White, I. M., Zhu, H., Suter, J. D., Oveys, H., and Fan, X. (2006) Liquid core optical ring resonator label-free biosensor array for lab-on-a-chip development, Proc. SPIE 6380, 63800F 8. White, I. M., Oveys, H., Fan, X., Smith, T. L., and Zhang, J. (2007) Demonstration of a liquid core optical ring resonator sensor coupled with an ARROW waveguide array, Proc. SPIE 6475, 647505 9. Laine, J. -P., Tapalian, H. C., Little, B. E., and Haus, H. A. (2001) Acceleration sensor based on high-Q optical microsphere resonator and pedestal antiresonant reflecting waveguide coupler, Sens. Actuators A 93, 1–7 10. Arnold, S., Khoshsima, M., Teraoka, I., Holler, S., and Vollmer, F. (2003) Shift of whispering-gallery modes in microspheres by protein adsorption, Opt. Lett. 28, 272–274 11. Vollmer, F., Arnold, S., Braun, D., Teraoka, I., and Libchaber, A. (2003) Multiplexed DNA quantification by spectroscopic shift of two microsphere cavities, Biophys. J. 85, 1974–1979 12. Hanumegowda, N. M., White, I. M., Oveys, H., and Fan, X. (2005) Label-free protease
Label-Free Detection with the Liquid Core Optical Ring Resonator Sensing Platform
13.
14.
15. 16.
17.
18.
19.
20.
21.
22.
sensors based on optical microsphere resonators, Sensor Lett. 3, 315–319 Hanumegowda, N. M., White, I. M., and Fan, X. (2006) Aqueous mercuric ion detection with microsphere optical ring resonator sensors, Sens. Actuators B 120, 207–212 Zhu, H., Suter, J. D., White, I. M., and Fan, X. (2006) Aptamer based microsphere biosensor for thrombin detection, Sensors 6, 785–795 Boyd, R. W. and Heebner, J. E. (2001) Sensitive disk resonator photonic biosensor, Appl. Opt. 40, 5742–5747 Blair, S. and Chen, Y. (2001) Resonantenhanced evanescent-wave fluorescence biosensing with cylindrical optical cavities, Appl. Opt. 40, 570–582 Armani, A. M. and Vahala, K. J. (2006) Heavy water detection using ultra-high-Q microcavities, Opt. Lett. 31, 1896–1898 Chao, C. -Y. and Guo, L. J. (2003) Biochemical sensors based on polymer microrings with sharp asymmetrical resonance, Appl. Phys. Lett. 83, 1527–1529 Krioukov, E., Greve, J., and Otto, C. (2003) Performance of integrated optical microcavities for refractive index and fluorescence sensing, Sens. Actuators B 90, 58–67 Ksendzov, A., Homer, M. L., and Manfreda, A. M. (2004) Integrated optics ring-resonator chemical sensor with polymer transduction layer, Electron. Lett. 40, 63–65 Ksendzov, A., and Lin, Y. (2005) Integrated optics ring-resonator sensors for protein detection, Opt. Lett. 30, 3344–3346 Chao, C. -Y., Fung, W., and Guo, L. J. (2006) Polymer microring resonators for biochemical sensing applications, IEEE J. Sel. Top. Quant. 12, 134–142
165
23. Yalcin, A., Popat, K. C., Aldridge, O. C., Desai, T. A., Hryniewicz, J., Chbouki, N., Little, B. E., King, O., Van, V., Chu, S., Gill, D., Anthes-Washburn, M., Unlu, M. S., and Goldberg, B. B. (2006) Optical sensing of biomolecules using microring resonators, IEEE J. Sel. Top. Quant. 12, 148–155 24. Yang, J. and Guo, L. J. (2006) Optical sensors based on active microcavities, IEEE J. Sel. Top. Quant. 12, 143–147 25. Fitt, A. D., Furusawa, K., Monro, T. M., and Please, C. P. (2001) Modeling the fabrication of hollow fibers: capillary drawing, J. Lightwave Technol. 19, 1924–1931 26. Sumetsky, M., Dulashko, Y., and Hale, A. (2004) Fabrication and study of bent and coiled free silica nanowires: self-coupling microloop optical interferometer, Opt. Express 12, 3521–3531 27. Ghoreyshi, A. A., Farhadpour, F. A., Soltanieh, M., and Bansal, A. (2003) Transport of small polar molecules across nonporous polymeric membranes, J. Membr. Sci. 211, 193–214 28. Laine, J. -P., Little, B. E., and Haus, H. A. (1999) Etch-eroded fiber coupler for whispering-gallery-mode excitation in high-Q silica microspheres, IEEE Photonic Tech. L. 11, 1429–1431 29. Knight, J. C., Cheung, G., Jacques, F., and Birks, T. A. (1997) Phased-matched excitation of whispering-gallery-mode resonances by a fiber taper, Opt. Lett. 22, 1129–1131 30. Tong, L., Gattass, R. R., Ashcom, J. B., He, S., Lou, J., Shen, M., Maxwell, I., and Mazur, E. (2003) Subwavelength-diameter silica wires for low-loss optical wave guiding, Nature 426, 816–819
Chapter 8 Reflectometric Interference Spectroscopy Guenther Proll, Goran Markovic, Lutz Steinle, and Guenter Gauglitz Summary Reflectometry is classified in comparison to the commercialized refractometric surface plasmon resonance (SPR). The advantages of direct optical detection depend on a sophisticated surface chemistry resulting in negligible nonspecific binding and high loading with recognition sites at the biopolymer sensitive layer of the transducer. Elaborate details on instrumental realization and surface chemistry are discussed for optimum application of reflectometric interference spectroscopy (RIfS). A standard protocol for a binding inhibition assay is given. It overcomes principal problems of any direct optical detection technique. Key words: Label-free optical biosensor, Reflectometric interference spectroscopy (RIfS).
1. Introduction The methods available for direct monitoring of biomolecular interaction can be divided into methods measuring changes in the refractive index of the interaction layer, and methods measuring changes in reflectometry at the layer (1). Regarding the first method, BiaCore (2) has opened the market by introducing surface plasmon resonance (SPR) (3) as a very promising tool in biomolecular interaction analysis (BIA). In contrast to refractometry, reflectometry concentrates on the measurement of changes in physical thickness. Reflectometry has been introduced many decades ago as ellipsometry using polarized light. Interference at the interfaces of the layer causes a change in the relative amount of amplitude of the two polarized radiation beams and in phase. This interferometric method has been applied to a very simple analytical method, called reflectometric interference
Avraham Rasooly and Keith E. Herold (eds.), Methods in Molecular Biology: Biosensors and Biodetection, Vol. 503 © Humana Press, a part of Springer Science + Business Media, LLC 2009 DOI: 10.1007/978-1-60327-567-5_8
167
168
Proll et al.
spectroscopy (RIfS) (Figs. 1 and 2) to be used as an example of a very robust, simple optical detection principle in chemical and biochemical sensing (4). This label-free optical detection method for surface interactions is based on white light interference at transparent thin layers. At each interface of thin layers of different materials with negligible absorption, radiation is partially reflected and transmitted. If the optical path length through these layers is less than the coherence wavelength, the different partial beams interfere, and form an interference pattern depending on the wavelength, the optical thickness, which is given by the product of the refractive index of the layer and its physical thickness, the incident angle, and the refractive index of the surrounding medium (5, 6). In case of perpendicular incidence, a nonabsorbing layer, and low reflectances, the reflec-tance R is given by:
Fig. 1. Scheme of the RIfS detection principle. The left part shows the superimposition of the reflected light beams and the change in optical thickness during a binding event on the sensor surface. The right part shows the corresponding change of the characteristic interference spectrum and the resulting binding curve.
Halogen lamp Photo diode array
PC
Transducer with Interference Layer
Fig. 2. RIfS setup.
Reflectometric Interference Spectroscopy
R = R1 + R2 + 2 R1R2 cos(4p nd / l ),
169
(1)
where R1 and R2 denote the Fresnel reflectance at the two interfaces, d is the physical thickness of the film, n its refractive index, and l the wavelength of incident light. A typical interference pattern showing the modulation of reflectance with cos(1/l) is given in Fig. 1. The optical thickness nd can be determined from the position of an extremum with a given order value m by nd =
ml . 2
(2).
RIfS uses the change in the optical properties in or at the top layer of a given layer system as detection principle (Figs. 1 and 2). The binding of an analyte molecule or particle to the sensor surface causes a shift of the interference pattern in the wavelength domain. To evaluate the binding signal, the locus of an extremum is tracked over time; thus, the change of the interference spectrum results in a time-resolved binding curve representing the binding of the analyte molecule to the sensor surface. A major advantage of RIfS is its resistance to changes in temperature (7). Refractometric methods such as SPR and ellipsometry, on the other hand, are very sensitive to temperature variations due to the high impact of temperature on the refractive index. Thus, temperature changes during a measurement cause negative effects with these methods, and quick changes of temp-erature between measurements are technically challenging. Since the refractive index n is dependent on the density given from the Clausius–Mossotti equation, and the density is dependent on the thermal expansion, the refractive index decreases with increasing temperature. Because of the thermal expansion of the biopolymer layer, the physical thickness d increases with increasing temperature. These two contrary temperature-dependent effects result in a rather low influence of temperature on the optical thickness – the product of the refractive index and the physical thickness.
2. Materials Common chemicals of analytical grade are purchased from Sigma or Merck. Milli-Q water is deionized water with a conductivity of 18.2 MΩ cm−1. 2.1.Transducer
The RIfS standard transducer consists of a glass substrate (D 263 glass, Schott AG, Germany) (~1-mm thick) coated with a 10-nm
170
Proll et al.
layer of a material with high refractive index (usually Ta2O5 or Nb2O5), and a top layer of SiO2 (330 nm). As a reference use a transducer without SiO2 layer. 2.1.1. Glass Substrates
1. BK 7 glass, n = 1.51, Schott, Mainz. 2. WG 345 glass, n = 1.699, Schott, Mainz. 3. Interference transducer: D 263, 10-nm Ta2O5, 500-nm SiO2, Schott, Mainz. 4. Goethe glass: multilayer system: 1-m D 23, 45-nm Ta2O5, 20-nm SiO2, Schott, Mainz (cut in small squares (10 mm × 10 mm) ).
2.1.2. Parallel Setup
1. White light source 100 W/12 V, Osram, Munich. 2. Lenses, mirror, and positioning optics, Spindler & Hoyer, Göttingen. 3. Polymer light guides (PMAA, n = 1.490, coupling element 1 × 2 (50:50), 1 mm, fiber diameter with SMA 905 fiber connectors, Microparts, Dortmund. 4. Optical 4 × 1 multiplexer DiCon VX 500-C, laser components, Olching. 5. MMS diode row spectrometer with Liliput-PC, ZEISS, Jena. 6. Nineteen-inches industrial standard housing by RS Components, Walldorf-Mörfelden. 7. 486 PC (Windows 95).
2.1.3. Single Setup
1. Modified simultaneous spectrometer SPECKOL 1100, Zeiss, Jena. 2. 486 PC (Windows 95).
2.1.4. Liquid Handling
1. Ten-position valve VICI, Valco Europa, Schenkon, Switzerland. 2. HPCL three-way valve VICI, Valco Europa, Schenkon, Switzerland. 3. Six-position valve, Bischoff, Leonberg. 4. Peristaltic pump Reglo-Digital MS2/8–160 ISM 832, Ismatec, Wertheim. 5. Peristaltic pump MS Fixo, Ismatec, Wertheim. 6. High-grade steel capillaries, screws, and fittings from Rheodyne, USA.
2.1.5. Software
1. MeasureCR for capturing spectras and controlling the systems. 2. IFZCR for evaluating the interferograms. 3. MS-Excel 7.0 and Microcal Origin 5.0 for further processing of measured data.
Reflectometric Interference Spectroscopy
2.2. Surface Chemistry
171
1. GOPTS (3-glycidyloxypropyltrimethoxysilane) purum: Toxic. Store at room temperature; keep under argon, sensitive to humidity (Fluka). 2. AMD (Aminodextran): MW 100 kD, level of amination 50%. Store at 4 °C under dry conditions (Innovent e.V. Technologieentwicklung, Jena, Germany). 3. Dicarboxypoly- and diaminopoly(ethylene glycol) (PEG): MW 2,000 Da. Store at −20 °C under dry conditions (Rapp Polymere, Tuebingen, Germany). 4. NHS (N-hydroxysuccinimide) purum: Store under room temperature (Fluka). 5. DIC (N,N´-diisopropylcarbodiimide) purum: Toxic. Store at room temperature, keep under argon, sensitive to humidity (Fluka). 6. EMCS (6-maleimidohexanoic acid N-succinimidyl ester) purum: Store at −20 °C under dry conditions (Sigma). 7. TBTU (2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium tetrafluoroborate). 8. HOBT (1-hydroxybenzotriazole) hydrate purum: Store at −20 °C at dry conditions (Fluka). 9. DIPEA (N,N-diisopropylethylamine): Highly inflammable. Store at room temperature (Sigma).
3. Methods 3.1. Setup for RIfS
In the standard laboratory setup for RIfS (see Note 1), the white light is guided via an optical fiber (1-mm PMMA fiber) to the back of the transducer mounted in a microfluidic flow cell, which in turn is attached to a liquid handling system. The reflected light is gathered in the same waveguide. A schematic of sample handling with flow injection analysis (FIA) for RIfS is shown in Fig. 3. The fiber optics used is bifurcated (50:50 ratio), with one tail leading to the light source and the other to the UV/Vis spectrometer. Possible light sources are halogen lamps (e.g., 10-W halogen lamp with fiber in-coupling optics consisting of front surface spherical mirror, collimating lens, and an infrared absorption filter) or LEDs. For the spectral detection of the reflected light, diode array spectrometers are used normally. A gap (approximately 100 μm) between transducer chip and the fiber output is filled with glycerol (80%) for refractive index matching. Samples are handled by a flow system (e.g., two peristaltic pumps, injectload valve, and six-way valve). Moreover, this flow system can be equipped with an autosampler. The various samples are injected
172
Proll et al.
Carrier Flow
D
P1 V1
Waste
P2
Waste
Loading: Injection:
V2 Sample Loop
Flow Direction while loading: while injection:
Fig. 3 Schematic of the FIA system for RIfS.
by the autosampler into a sample loop. From there, the sample is driven in continuous flow, passing the prepared transducer. The raw spectra are corrected for the dark current of the spectrometer by subtraction (if necessary), and normalized to the reflectance spectrum from a glass chip without the SiO2- and bio-layer. The position of an interference extremum at approximately 550 nm is determined by a parabolic fit to one half-wave of the interference spectrum. Optical thickness is calculated according to Eq. 2 (see Note 2). The costs of the setup can be reduced by sequentially illuminating the transducer with different suitable wavelengths coming from LEDs or a white light source with appropriate filters, detecting the intensity without spectral resolution using a photodiode, and reconstructing part of the interference pattern by fitting a parabola through the interpolation points. This detection principle can be used to realize a parallel screening system, which allows optical online detection of specific biomolecular interaction in 96- or 384-well microplate formats. Therefore, the whole area of the plate bottom consisting of a RIfS transducer is illuminated by a halogen light source combined with a filter wheel, which allows the subsequent passage of monochromatic light of seven different wavelengths. This is not only to reduce the costs, but is necessary because a CCD camera is used as a detector, which is able to detect the intensity of the whole area of interest in a single shot. 3.1.1. Silanization
1. Transducer chips are cleaned with freshly prepared Piranha solution (mixture of 30% hydrogen peroxide and concentrated sulfuric acid at a ratio of 2:3; caution: hot and aggressive!) for 30 min in an ultrasonic bath to clean the chip and to generate silanol groups on the transducer surface.
Reflectometric Interference Spectroscopy
173
2. After rinsing with double-distilled water and drying in a nitrogen stream, the surface is immediately activated by GOPTS at a surface concentration of approximately 10 μL cm−2 for 1 h in dryness (see Notes 3) (by assembling two slides face-to-face placed in a tray with ground joint) followed by cleaning with dry acetone and drying in a nitrogen stream (see Notes 4 and 5). 3.1.2. Biopolymer
When investigating interactions of biomolecules on surfaces, nonspecific binding to the sensor must be minimized. Therefore, the glass slide is coated with a shielding layer, which prevents nonspecific binding and additionally provides a large number of functional groups for the immobilization of the binding partner. The standard shielding chemicals used are dextrans, which form a 3D hydrogel loaded with a large number of binding sites, and polyethylene glycols (PEG) forming a two-dimensional brushlike monolayer with less binding sites but with a more defined surface. Amino and carboxy groups are normally functional groups for the immobilization process since the well-established peptide chemistry methods can be applied. Other shielding chemicals can be used depending on the application or the surface chemistry needed. 1. Aminodextran: The coupling of aminodextran as an aqueous solution (1:3) (approximately 15 μL cm−2) to the silanized surface is carried out by incubating over night (minimum 16 h) in a water-saturated atmosphere (see Note 6). After thoroughly rinsing with double-distilled water and drying, the prepared chips are stable for several months. 2. Polyethylenglycol: The coupling of diamino- or dicarboxypoly(ethylene glycol) (PEG) as a 1 mM solution in dichloromethane (DCM) (approximately 10 μL cm−2) to the silanized surface is achieved by incubating overnight at 70 °C, followed by thorough rinsing with Milli-Q water and drying in a nitrogen stream. 3. Change of functional groups: For some applications (e.g., immobilization of DNA oligomers with an aminolinker) the follow-up functionalization via peptide chemistry works better with diaminopoly(ethylene glycol) treated with 5 M glutaric anhydride (GA) in N,N-dimethylformamide (DMF) (10 μL cm−2) for 6 h to generate carboxylic groups on the surface. The same protocol can be used to modify AMD surfaces with carboxylic groups.
3.1.3. Immobilization of Amino-Terminated Ligands
1. The covalent coupling of amino-terminated molecules/ligands to biopolymers with carboxylic groups is done by standard peptide chemistry via activated esters (see Note 7). Therefore, the previously modified transducers are activated with a solution of NHS and DIC (1 M NHS, 1.2 M DIC in DMF,
174
Proll et al.
10 μL cm−2) for 2–4 h as sandwich pairs in a DMF saturated atmosphere (see Note 5). The transducers are rinsed first with DMF and then with dry Aceton, and dried in a nitrogen stream (see Notes 4 and 5). 2. To achieve a high density of covalently bound molecules, it is necessary to use an excess of reagents. Therefore, the ligand is applied as a DMF solution (10 μL cm−2) (if your ligand is insoluble in DMF, use another aprotic solvent of appropriate polarity) at a concentration of 1–3 μM to the activated surface for 4 h as sandwich pairs in a DMF saturated atmosphere (see Note 6). Clean the transducers by rinsing them with DMF and Milli-Q water and drying them in a nitrogen stream (see Note 4). 3.1.4. Immobilization of Carboxy-Terminated Ligands
1. For activation of the carboxy-terminated molecules, the ligand is dissolved in DMF at a concentration of 1–3 μM together with 1.5 M DIC. After quick mixing, this solution is immediately applied to a transducer modified with a biopolymer providing amino groups at a concentration of 10 μL cm−2. Place a second transducer on top to form a sandwich pair (see Note 6). 2. The reaction is finished after 4 h in a DMF-saturated atmosphere (see Note 6). Clean the transducers by rinsing them with DMF and Milli-Q water and drying them in a nitrogen stream (see Note 4).
3.1.5. Immobilization of Thiol-Terminated Ligands
1. Dissolve 1-mg EMCS per 10-μL DMF, and apply this solution to a transducer modified with a biopolymer offering amino groups at a surface concentration of 10 μL cm−2 for 6–12 h in a DMF-saturated atmosphere (see Note 6). After rinsing with DMF (see Note 4), dissolve the ligand at a concentration of 1–3 μM in DMF, and apply this solution to the transducers at a surface concentration of 10 μL cm−2 for 6–12 h as sandwich pairs in a DMF saturated atmosphere (see Note 6). 2. Clean the transducers by rinsing them with DMF and Milli-Q water and drying them in a nitrogen stream (see Note 4).
3.1.6. Immobilization of Biotin
1. Biotin is immobilized by TBTU activation: D-biotin (1 mg, 4 mmol), TBTU (1.4 mg, 4.4 mmol), and DIPEA (4 mL, 23.3 mmol) are mixed with DMF (50 mL) until the active ester is formed and the reaction mixture appears homogeneous. This solution is dripped on to a transducer at a surface concentration of 10 μL cm−2 pretreated with a biopolymer providing amino groups. Two of such transducers are put together to form a sandwich. 2. After a reaction time of 4 h in a saturated DMF atmosphere (see Note 6), the sandwich is separated and both transducers are rinsed with water and dried in a nitrogen stream.
Reflectometric Interference Spectroscopy
175
3.2. Measurement Protocol for Standard Binding Inhibition Assay for the Determination of the Affinity of an Antigen– Antibody Interaction
Binding inhibition assay to determine the affinity constant of antigen–antibody interaction. The transducer is modified according to protocols given before either with a carboxy or an amino-terminated antigen. Then, fixed amounts of antibody are preincubated with fixed volumes of differently concentrated antigens (see Note 8) to reach binding equilibrium before injection of the mixture with the FIA system (see Notes 9–12) on the transducer. Antigens with affinity to the antibody in the preincubated solution inhibit the binding of the antibody to the surface-immobilized antigen. In the case of diffusion (mass transport) limited binding, the diffusion rate obeys first Fick’s law. This results in a constant binding rate of the antibody to the surface and a linear binding curve. The slope of the binding curve is determined by the concentration of free antibody binding sites in solution. The slope of the binding curves decreases with increasing concentration of antigen in solution and reaches zero for high antigen concentrations. Ovalbumin (final concentration of 200 μg/mL) should be added to all antibody solutions to avoid loss of antibody by nonspecific binding in the fluidic system. The model function describing the concentration of antibodies with free binding sites is fitted to the titration curve using a Marquart–Levenberg nonlinear least-square algorithm (software ORIGIN from Microcal, Northampton/USA).
3.3. Data Evaluation
1. Determination of the concentration of active antibody Measuring the concentration of active antibody in a sample working at mass-transport limited conditions is essential. Accordingly, the rate of diffusion to the surface is much slower than the binding reaction to the surface. The reaction kinetics can therefore be neglected. Mass-transport limited conditions can be achieved by a high binding capacity of the surface and a low antibody concentration in solution. Under mass-transport limited conditions according to first Fick’s law, the resulting binding curve is linear, and its slope is proportional to the concentration of functional antibody. If all samples contain the same amount of protein, determined by UV spectroscopy or Bradford assay, it is possible to calculate the active antibody concentration in a.u. by the different slopes of the samples, where the sample with the highest slope is assigned the a.u. 1 per μg used protein. 2. Determination of affinity and kinetic constants To determine the affinity and the kinetics of an antibody (see Note 13), mass transport to the surface must be much faster than the rate of binding to the surface. If this is the case, the mass transport can be neglected. This can be achieved by a high concentration of bulk antibody and a low surface coverage of hapten derivative. The time dependence of the surface coverage can then be described by a pseudo first-order binding reaction as follows:
176
Proll et al.
dΓ (t ) = kac Ab (Γ max - Γ (t )) - kd Γ (t ), dt
(3)
with cAb: antibody concentration, G(t): surface coverage at time t, Gmax: maximal surface coverage, ka: association rate constant, and kd: dissociation rate constant. Solving this differential equation, the surface coverage that is equal to the resulting signal curve is given by Γ (t ) = Γ Eq (1 - e - kobst ),
(4)
with Γ Eq = Γ max
K aff c Ab , 1 + K aff c Ab
(5)
the equilibrium coverage and kobs = kac Ab + kd ,
(6)
the observed rate constant. The affinity constant Kaff is given by the fraction of the association rate constant, ka, and the dissociation rate constant, kd. To obtain accurate values for GEq and kobs, it is necessary that only the kinetics-controlled part of the signal curve is approximated, and that the surface and bulk are nearly in equilibrium at the end of the measurement. Now it is possible to determine ka as the slope of the straight line of the plot with kobs as ordinate and active antibody concentration as abscissa. In order to determine Kaff, Eq. 5 can be rewritten as 1 1 1 1 = + × . Γ Eq Γ max K aff × Γ max c Ab
(7)
With this formula, it is possible to obtain the value for Gmax with 1 1 plot. Inserting this value in Eq. 3, a Γ Eq c Ab nonlinear least square fit results in a value for Kaff. This method
a linear fit of the
æΓ ö gives nearly the same values as a Scatchard plot ç Eq Γ Eq ÷ . è c Ab ø
4. Notes 1. Classical RIfS setup (Figs. 1 and 2) 2. Very thick biolayers (more than approximately 100 nm) lead to measurements out of the linear correlation between
Reflectometric Interference Spectroscopy
177
change of optical thickness and shift of the interference spectrum. 3. Correct side of the RIfS transducers: Mark the uncoated side with a diamond pen to avoid mistakes during the surface chemistry steps. The coated side of the transducers appears colored because of the light interference. 4. In the case a transducer shows a gray shadow after rinsing and drying repeat this cleaning procedure. If this does not work, start the surface chemistry from the beginning (otherwise the modification will not be homogeneous and will show high nonspecific binding). 5. All surface chemistry protocols that produce highly reactive groups are sensitive to humidity (deactivation). This can be avoided by immediately applying the next step and working under dry ambient conditions. 6. Use a tray with ground joint. Place the transducer as sandwich pairs (by assembling two slides face-to-face) in the tray on a solid support and add the same solvent that is used in the reaction for a solvent-saturated atmosphere. 7. In the case of limited ligands it is possible to use commercially available piezo-based microdosing devices for printing a ligand solution in Milli-Q water. 8. Usually phosphate buffered saline (PBS, 150-mmol NaCl and 10-mmol dipotassium hydrogen phosphate in Milli-Q water at pH 7.4) is used for antigen–antibody interaction analysis. In general, all buffers can be used for RIfS measurements. Only restriction: do not use buffers with pH > 8.5 because of destruction of the surface modification. 9. Testing of the modified transducer for nonspecific binding: use ovalbumine (OVA) or bovine serum albumin (BSA) in excess (e.g., 1 mg mL−1) directly before the measurement. 10. Air disturbs the measurements: Use degassed buffers and avoid negative pressure (sucking) through the flow cell. In addition, it is helpful to keep the buffer under a slight positive Argon pressure. 11. If possible use the same buffers during a complete measurement to avoid artifacts because of changes in the refractive index. 12. After the interaction process, the transducer surface can be regenerated. To remove antibodies from their antigens, a solution of 0.5% SDS (sodium dodecyl sulphate) at pH 1.9 is applied via the FIA system. In the case of hybridization experiments, a solution of either 0.25% SDS at pH 2.5 or a solution containing 6 M guanidinium hydrochloride and 6 M urea at pH 2 can be used.
178
Proll et al.
13. RIfS is not restricted for biosensing; this technique can be used also for chemical sensing. Instead of biopolymers one can modify the transducer with, e.g., nanoporous polymer layers, rubber-like polymers, or molecular imprinted polymers (MIPs). The determination of kinetics is also possible. References 1. Gauglitz, G. (2005) Direct optical sensors: principles and selected applications. Anal Bioanal Chem. 381(1), 141–155 2. www.biacore.com 3. Homola, J., Yee, S.S., and Gauglitz, G. (1999) Surface plasmon resonance sensors: review. Sensor Actuator. B54, 3–15 4. Schmitt, H.M., Brecht, A., Piehler, J., and Gauglitz, G. (1997) An integrated system for optical biomolecular interaction analysis. Biosens Bioelectron. 12, 219–233
5. Brecht, A., Gauglitz, G., and Nahm, W. (1992) Interferometric measurements used in chemical and biochemical sensors. Analusis. 20, 135–140 6. Brecht, A., Gauglitz, G., Kraus, G., and Nahm, W. (1993) Chemical and biochemical sensors based on interferometry at thin layers. Sensor Actuator. 11B, 21–27 7. Proell, F., Moehrle, B., Kumpf, M., and Gauglitz, G. (2005) Label-free characterisation of oligonucleotide hybridisation using reflectometric interference spectroscopy. Anal Bioanal Chem. 382(2), 1889–1894
Chapter 9 Phase Sensitive Interferometry for Biosensing Applications Digant P. Davé Summary A simple yet highly sensitive implementation of an interferometric technique for a label-free molecular biosensing application is described. The intereferometric detection method is based on the phasesensitive detection of spectral interference fringes. The change in optical path length due to binding of biomolecules on functionalized optically clear substrates can be quantified by detecting the change in the phase of the spectral fringes. The common path interferometeric design permits measurement of sub- monolayer binding of biomolecules to the sensor surfaces. Key words: Interferometry, Biosensor, Phase-sensitive.
1. Introduction Optical techniques used for biosensing can be put in two distinct categories, namely labeled and label-free techniques. Labeling of biomolecules of interest using fluorescent tags is widely used in biosensing. Given the simplicity of detection and high sensitivity of labeled techniques, they are the preferred method of biosensing in a wide variety of biomolecular recognition applications. Despite the success of fluorescent tags for biosensing their limitations are well recognized. Ideally one would prefer to recognize the biomolecules of interest without labeling them since it is possible that the label may interfere with the activity of the biomolecule and its interaction with the recognition molecule. Moreover, robust chemistry needs to be developed to attach fluorescent tags to each biomolecule of interest. In the absence of a label, intrinsic properties of the biomolecule need to be exploited for detection. Optical properties that can Avraham Rasooly and Keith E. Herold (eds.), Methods in Molecular Biology: Biosensors and Biodetection, Vol. 503 © Humana Press, a part of Springer Science + Business Media, LLC 2009 DOI: 10.1007/978-1-60327-567-5_9
179
180
Davé
be potentially exploited for biosensing include, autofluorescnce, Raman scattering, refractive index, optical path length (product of refractive index and geometric size) and polarization. Autofluorescence and Raman scattering are typically weak signals for sensitive biosensing applications. Though Raman signal can be amplified using surface enhanced Raman spectroscopy (SERS), it requires nanostructured surfaces. A change in refractive index and optical path length (OPL) can be accurately measured with techniques such as surface plasmon resonance (SPR) and interferometry. SPR is now a well established method in biosensing, particularly to monitor and characterize the kinetics of biomolecular binding. The introduction of SPR instrumentation by Biacore enabled widespread use of SPR in biochemistry research and application. Using interferometry, a change in refractive index or OPL can be measured with a sensitivity that translates into biomolecule detection sensitivity that is comparable or even better than any other widely used biosensing technique. Despite the potential, the use of interferometry for biosensing has been limited until now. Sensitivity required for biomolecular sensing requires implementation of phase sensitive interferometry. In recent years a number of phase sensitive interferometeric techniques have been reported which are well suited for biosensing (1–19). 1.1. Phase Sensitive Interferometry
By measuring the phase on an interference signal, sub-wavelength changes in OPL can be measured. Although phase information is readily available in any interferometric setup, environmental noise corrupts the phase information, rendering it useless. For robust phase measurement, interferometer design should enable cancellation of common mode noise. A common path interferometric implementation thus enables common mode noise cancellation. In this chapter a simple spectral interferometric technique to measure OPL changes in the picometer range and its implementation for biosensing are described. In principle, the described technique is similar to reflectrometric interferometric spectroscopy implemented for biosensing. A significant advantage of the technique described is that it does not require the use of specially fabricated surfaces, and in fact can be easily interfaced with commonly used optically clear substrates. The technique is based on phase-sensitive implementation of low coherence interferometery in the spectral domain (19–26).
1.2. Spectral Domain Phase Sensitive Interferometry
Consider an optical setup in which light from a partially coherent single light source after traveling through two optical paths (reference and sample) is recombined and the spectrum is measured. The recorded spectrum is not only the superposition of the spectrum of light from the sample and the reference path but has an additional term that is present due to the
Phase Sensitive Interferometry for Biosensing Applications
181
non-zero cross-correlation of reference and sample spectrums. This non-zero cross-correlation term modulates the spectrum, giving rise to spectral interference. The modulation frequency is proportional to the optical path-length difference between reference and sample paths. Recorded spectral interference, due to mixing of light from the reference path and sample path can be written as S (k) = S ref (k) + S samp (k) + 2 S ref (k)S samp (k) × m (k) × cos(2p k Dp + j 0 ), (1) where k is the wave vector, m(k) is the spectral coherence function, and Δp is the optical path length difference between the reference path and sample path. Sref(k) and Ssamp(k) are the spectrum of light from the reference and sample path, respectively. Δp is a product of the geometrical path length difference (Δz) and refractive index of the medium separating reference and sample path which can be calculated by Fourier transformation of Eq. 1. For biosensing applications it is necessary to measure the change in optical path length with sub-wavelength resolution as the biomolecule of interest binds to the recognition molecule attached to a substrate. The change in the optical thickness of the biolayer is typically less than 10 nm depending on the size of the biomolecule. Sub-wavelength changes in Δp can me measured by detecting the change in phase of the frequency component of interest of the modulation term in Eq. 1 as follows æ Im(F (S (k)) ö 2p n Dz j |z =n Dz = tan -1 ç = lo è Re(F (S (k)) ÷ø 1.3. Implementation of SD-PSI for Biosensing
(2)
Fiber-based implementation of SD-PSI is shown in Fig. 1. SDPSI system can be constructed with commercially available components and minimal effort in optical alignment.
Fig. 1. Setup of a fiber-based spectral domain phase sensitive interferometer for biosensing applications.
182
Davé
2. Materials The part numbers listed below were used to build and test a prototype SD-PSI system operating at 830 nm for biosensing applications. Similar SD-PSI can be built at other operating wavelengths also. All the components can be purchased off-the-shelf from various manufactures. Given below is the list of components needed to set up a SD-PSI system for biosensing. 1. Fiber coupled SLD (SuperlumDiodes Inc. – SLD-37-MP will require a controller and a diode mount). 2. Single mode 2 × 2 fiber coupler (Thorlabs). 3. Fiber Isolator (OFR-IO-F-830). 4. Spectrometer (see Note 1). 5. Fiber Collimator (OFR-PAF-X-15-830). 6. Focusing Lens (10× microscope objective). 7. Cage System (Thorlabs). ●
Cage Rods (4) (SR6)
●
Cage Plate (2) (CP02)
●
Kinematic Mirror Mount (2) (KC1)
3. Methods 3.1. Interferometer Setup
Light from the broadband light source is coupled into a 2 × 2 (50:50 split ratio) single mode fiber (SMF) splitter. Any broadband light source can be used as long as the spectrum of the light source is stable and sufficient light can be coupled into a single mode fiber to maintain a signal-to-noise ratio, necessary to detect a desired phase change in the spectral interference fringes. The bandwidth of the light source affects the smallest thickness of the functionalized transparent substrate that can be used and also the precision with which the thickness of the biolayer can be measured. Fiber coupled superlumincsent diodes (SLD) are an excellent choice as broadband light sources for SD-PSI (see Note 2). These compact sources have a stable spectral and power output. Given the potential of damage to the SLD due to backreflection from the optical setup, a fiber isolator should be inserted between the SLD source and the input port of the 2 × 2 SMF coupler. In the common path configuration only one port of the coupler is used. The port that is not used should be angle polished or angle cleaved to avoid any back reflection from the fiber–air interface into the coupler. As an alternative the fiber can be dipped in 60–80% glyc-
Phase Sensitive Interferometry for Biosensing Applications
Functionalized glass slide
183
Microscope objective Kinematic mirror mount
Coupler output fiber
Inline collimator
FC/APC connector Fig. 2. Photograph of the SD-PSI sample path configuration showing various components integrated together using a cage system.
erol solution to avoid back reflection. This back reflection will reduce the signal-to-noise ratio of the interferometric signal and hence the sensitivity of measured phase (see Note 3). The other port of the fiber coupler should be terminated with an FC/APC connector. Light from the fiber port is first collimated and then focused onto the sample. A simple and robust setup for the sample path optics would be to mount all the opto-mechanical components on a cage system (Fig. 2). Optical components are mounted on cage plates and kinematic mirror mounts, which are in turn mounted on cage rods which ensure co-linearity of the optical train. Fiber connector with a snap-on collimator should be mounted on a kinematic mirror mount to enable a fine adjustment of collimated beam tip and tilt orientation with respect to the focusing lens. Choice of the focusing lens is not critical if it is assumed that the biomolecule binding per unit area remains constant. Under this assumption, phase measuring sensitivity remains constant, as the phase measured is the ratio of the area (footprint) of all attached biomolecules to the area of focused light on the biolayer being interrogated. A suitable sample geometry for biosensing is any transparent substrate (glass or plastic) of suitable thickness functionalized with chemistry for recognition of biomolecules of interest. The thickness of the transparent substrate that can be used is only limited by the resolution of the spectrometer used in the setup. The spectral modulation frequency is proportional
184
Davé
to the thickness of the substrate and that of the functionalized biolayer. In the sample optical configuration shown in Fig. 3, light reflects back onto the interferometer from the air–substrate (interface 1), substrate–functionalized biolayer (interface 2), functionalized biolayer–target biomolecule (interface 3) and target biomolecule–buffer (interface 4). Spectral modulation of various frequencies will occur due to interference between light reflection from various interfaces (1 and 2, 2 and 3, 3 and 4, 1 and 3, 1 and 4, 3 and 4). The modulation depth of the spectral interference signal at various frequencies is a function of the ratio of the reflecti-vities of the two interfaces involved. Inference will also occur due to multiple reflection of light from the various interfaces but the modulation depth is far smaller than the primary interference signal and the modulation frequency will be a multiple of the primary modulation frequency. Each interference signal occurs at a fixed spatial frequency band that is
Fig. 3. Depiction of change in spectral inference signal with the composite change in optical thickness of functional biosensing substrate. Example of bovine serum albumin binding to recognition molecule (EDC) is shown in the graph. EDC is a spacer that binds to functionalized –COOH glass coverslip.
Phase Sensitive Interferometry for Biosensing Applications
185
proportional to the optical thickness of the interfaces involved and can be measured by Fourier transformation of the recorded spectral interference. In the case of functionalized substrate for biosensing (Fig. 3), recording of kinetics and ultimately the detection of the target biomolecule concentration can be achieved by measuring the optical thickness changes that occur due to the formation of layer 4 with the binding of target biomolecule to the recognition layer 3. The spectral interference signal of interest is the interference signal that occurs between interface 1 and 4 or 2 and 4 or 3 and 4. Note that layer 3 is not a true monolayer unless all the recognition sites are filled with the target biomolecules. In principle, even if one biomolecule attaches to the recognition site a change in the phase of the interference signal will occur. Practically, the limit of the number of biomolecules detected will depend on the SNR and also on the spectral characteristics of the sources (wavelength, spectral stability and bandwidth). Phase noise of the reported SD-PSI systems is in the tens of picometer range, although the theoretical values calculated put the lower limit in the sub picometer range. Additional sources of noise will include mechanical vibration of the substrate resulting in probe beam scanning over the surface roughness, statistical binding and bonding variation of the target biomolecule and recognition molecule. Sensitivity with which biomolecules can be detected using SD-PSI can be written as Sensitivity (gm / mm 2 ) »
j N wm ´ , j T Am
(3)
where jT and jN are the total phase change due to a complete monolayer of biomolecule being interrogated and phase noise standard deviation of the SD-PSI system. wm is the weight of the biomolecule and Am is the footprint of the biomolecule. The sensitivity in Eq. 3 does not take into consideration the dynamics of target biomolecule diffusion and binding with recognition molecule, which can be affected by the size of the sensor area and methods to promote binding like agitation and circulation of analyte. Consider the example of IgG binding to a functionalized layer of anti-IgG. For a SD-PSI system with jN equivalent to 1 mrad (45 pm at 800 nm), approximating the IgG molecule to a sphere of 5 nm radius, a bound monolayer of IgG will give total phase change of jT (Δp = 7 nm at 800 nm, refractive index of 1.4), the sensitivity of detection is about 20 pg/mm2. Stepwise detection of biomolecule sensing on funtionalized optically clear substrate with a layer of recognition molecules using SD-PSI is diagrammatically shown in Fig. 4.
186
Davé
Fig. 4. Interference signal formed due to interference of partially reflected light from various interfaces. Binding of the target molecule to the receptor molecule will change the phase of the interference signal.
4. Notes 1. A scanning monochromator should not be used in place of an array based spectrometer as large spectral phase noise can be introduced due to mechanical scanning of the grating. 2. If the SD-PSI setup is used only to quantify the total concentration of biomolecules bound to the functionalized sensor surface, care must be taken to thoroughly clean the sensor chip after hybridization so that none of the cleaning solution or unbound target biomolecule residue remains on the chip. A reference surface with only the functionalized layer on the sensor chip is necessary to quantify the thickness of the bound target biomolecules. 3. Phase artifacts can be introduced in the measured binding kinetics sensogram due to the presence of gas bubbles at the sensor–liquid interface, changes in temperature of liquid in contact with the sensor surface, and mechanical movement of sensor chip. For binding kinetics measurements, a flow cell arrangement similar to the one used in SPR setups with degasser and proper temperature control of the liquids will minimize phase artifacts. References 1. G. H. Cross, A. A. Reeves, S. Brand, J. F. Popplewell, L. L. Peel, M. J. Swann, N. J. Freeman, “A new quantitative optical biosensor for protein characterization,” Biosensors and Bioelectronics 19, 383 (2003) 2. V. S. -Y. Lin, K. Motesharei, K. -P. S. Dancil, M. J. Sailor, M. R. Ghadiri, “A porous siliconbased optical interferometric biosensor,” Science 278, 840 (1997)
3. D. J. Bornhop, J. C. Latham, A. Kussrow, D. A. Markov, R. D. Jones, H. S. Sørensen, “Molecular interactions studied backscattering interferometry,” Science 317, 1732 (2007) 4. L. Peng, M. M. Varma, W. Cho, F. E. Regnier, D. D. Nolte, “Adaptive interferometry of protein on a BioCD,” Applied Optics. 46, 5384 (2007)
Phase Sensitive Interferometry for Biosensing Applications 5. M. M. Varma, H. D. Inerowicz, F. E. Regnier, D. D. Nolte, “High-speed label-free detection by spinning-disk micro-interferometry,” Biosensors and Bioelectronics 19, 1371–1376 (2004) 6. M. M. Varma, D. D. Nolte, H. D. Inerowicz, F. E. Regnier, “Spinning-disk self-referencing interferometry of antigen–antibody recognition,” Optics Letters 29, 950–952 (2004) 7. K. Haupt, A. -S. Belmont, S. Jaeger, D. Knopp, R. Niessner, G. Gauglitz, “Molecularly imprinted polymer films for reflectometric interference spectroscopic sensors,” Biosensors and Bioelectronics 22(12), 3267–3272 (2007) 8. K. AddedKroger, J. Bauer, B. Fleckenstein, J. Rademann, G. Jung, G. Gauglitz, “Epitopemapping of transglutaminase with parallel label-free optical detection,” Biosensors and Bioelectronics 17(11–12), 937–944 (2002) 9. A. Brecht, G. Gauglitz, G. Kraus, G. Lang, J. Piehler, J. Seemann, “Application of reflectometric interference spectroscopy to chemical and biochemical sensing,” In Sensor 95, 355–360 (1995) 10. K. Schmitt, B. Schirmer, A. Brandenburg, “Development of a highly sensitive interferometric biosensor,” Proceedings of SPIE 5461, 22 (2004) 11. O. Birkert, G. Gauglitz, “Development of an assay for label-free high-throughput screening of thrombin inhibitors by use of reflectometric interference spectroscopy,” Analytical Bioanalytical Chemistry 372, 141 (2002) 12. J. Hast, H. Heikkinen, L. Krehut, R. Myllyla, “Direct optical Biosensor based on optical feedback interferometry,” IEEE, 177 (2005) 13. W. B. Nowall, N. Dontha, W. G. Kuhr, “Electron transfer kinetics at a biotin/avidin patterned glassy carbon electrode,” Biosensors and Bioelectronics 13, 1237 (1998) 14. C. J. Easley, L. A. Legendre, M. G. Roper, T. A. Wavering, J. P. Ferrance and J. P. Landers, “Extrinsic fabry-perot interferometry for noncontact temperature control of nanolitervolume enzymatic reactions in glass microchips,” Analytical Chemistry 77, 1038 (2005) 15. B. H. Schneider, J. G. Edwards, N. F. Hartman, “Hartman interferometer: versatile integrated optic sensor for label-free, real-time quantification of nucleic acids, proteins, and pathogens,” Clinical Chemistry 43, 1757 (1997)
187
16. K. Schmitt, B. Schirmer, C. Hoffmann, A. Brandenburg, P. Meyrueis, “Interferometric biosensor based on planar optical waveguide sensor chips for label-free detection of surface bound bioreactions,” Biosensors and Bioelectronics 22, 2591 (2007) 17. N. Kinrot, M. Nathan, “Investigation of a periodically segmented waveguide Fabry– Pérot interferometer for use as a chemical/ biosensor,” Journal of Lightwave Technology 24, 2139 (2006) 18. D. A. Markov, K. Swinney, D. J. Bornhop, “Label-free molecular interaction determinations with nanoscale interferometry,” J. Am. Chem. Soc. 126, 16659 (2004) 19. J. Lu, C. M. Strohsahl, B. L. Miller, L. J. Rothberg, “Reflective interferometric detection of label-free oligonucleotides,” Analytical Chemistry 76, 4416 (2004) 20. R. Leitgeb, C. K. Hitzenberger, A. F. Fercher, “Performance of fourier domain vs. time domain optical coherence tomography,” Optics Express 11, 889 (2003) 21. J. F. de Boer, B. Cense, B. H. Park, M. C. Pierce, G. J. Tearney, B. Bouma, “Improved signal-to-noise ratio in spectral-domain compared with time-domain optical coherence tomography,” Optics Letters 28, 2067 (2003) 22. M. Choma, M. Sarunic, C. Yang, J. A. Izatt, “Sensitivity advantage of swept source and Fourier domain optical coherence tomography,” Optics Express 11, 2183 (2003) 23. M. A. Choma, A. K. Ellerbee, C. Yang, T. L. Creazzo, J. A. Izatt, “Spectral-domain phase microscopy,” Optics Letters 30, 1162 (2005) 24. C. Joo, T. Akkin, B. Cense, B. H. Park, J. F. de Boer, “Spectral-domain optical coherence phase microscopy for quantitative phase-contrast imaging,” Optics Letters 30, 2131 (2005) 25. N. Nassif, B. Cense, B. H. Park, M. Pierce, S. Yun, B. Bouma, G. Tearney, T. Chen, J. F. de Boer, “In vivo high-resolution video-rate spectral-domain optical coherence tomography of the human retina and optic nerve,” Optics Express 12, 367 (2004) 26. B. H. Park, M. C. Pierce, B. Cense, S.-H. Yun, M. Mujat, G. Tearney, B. Bouma, J. F. de Boer, “Real-time fiber-based multi-functional spectral-domain optical coherence tomography at 1.3 μm,” Optics Express 13, 3931 (2005)
Chapter 10 Label-Free Serodiagnosis on a Grating Coupler Thomas Nagel, Eva Ehrentreich-Förster, and Frank F. Bier Summary The unique feature of the label-free measurement techniques for screening specific binding molecules against a certain ligand is that knowledge about the analyte is not required. Due to the direct monitoring of the binding event, no further detection step, e.g., by a fluorescently labeled antibody, is necessary. This technique enables not only the analysis of binding properties, but also applications in serodiagnosis and in primary screening in drug discovery. Especially when complex biological solutions such as blood serum are used as sample fluids, the minimization of unspecific attachment is the crucial point of the assay. In this chapter, the basic handling of the grating coupler as an example of a label-free transducer is described, together with a simple protocol to minimize unspecific attachment when measuring undiluted blood serum. Key words: Label-free detection, Grating coupler, Blood serum, Serodiagnosis, Passivation of glass surfaces, Protein coupling.
1. Introduction During the last two decades, a variety of transducers in the field of optical label-free detection methods have been developed: surface plasmon resonance, resonant mirror, interferometric sensors and reflectometric interference spectroscopy (1), as also the grating coupler, another effective label-free transducer that is based on surface bound refractive index changes (2). All these methods allow the observation of binding events to the sensor surface in real-time. Due to the evanescent field of the guided wave in the grating coupler only refractive index changes at the near vicinity of the surface are detected. Such changes occur when proteins or other molecules bind to the surface. Interactions Avraham Rasooly and Keith E. Herold (eds.), Methods in Molecular Biology: Biosensors and Biodetection, Vol. 503 © Humana Press, a part of Springer Science + Business Media, LLC 2009 DOI: 10.1007/978-1-60327-567-5_10
189
190
Nagel, Ehrentreich-Förster, and Bier
between different types of biomolecules such as proteins, DNA and in some cases even small molecules (down to some hundred Daltons) are detectable. The avoidance of modifying the molecules of interest, e.g., with a fluorescence label, enables the work with biomolecules in their native structure. Furthermore, the real-time observation of the binding event allows the determination of association and dissociation of rate constants and consequently, more information can be obtained than from an end-point assay. Apart from these classical applications, efforts are made to use the label-free approach in the field of serodiagnosis and as a screening tool in drug discovery. For these applications, end-point determinations are sufficient in the majority of cases, which have a specific binder to the immobilized ligand, to identify these samples in a direct way. The crucial points for such measurements are the throughput of samples and the feasibility to use complex solutions like blood serum as sample fluids. The development of label-free devices that allow a throughput of thousands of samples within a day are far advanced (3, 4) and the use of complex sample solutions requires a coupling chemistry that minimizes the unspecific attachment to the surface (5, 6). To determine picomolar or even femtomolar concentrations of the analyte in blood as sample fluid, an indirect competitive assay can be performed (7). In this chapter, we describe a method for the passivation of glass surfaces for measurements with complex solutions, using a grating coupler as the transducer. In addition to these, the chemistry described here can also be used with similar systems based on sensor chips with a glass surface, such as interferometric sensors or reflectometric interference spectroscopy. 1.1. Sensor Principle
Three different grating coupler setups have been investigated: the input (8, 9), the output (10–12) and the reflection grating coupler (13). The advantage of the grating coupler in the reflection mode as compared to the other setups is that moving parts are not necessary. The optical arrangement is shown in Figs. 1 and 2. A convergent HeNe laser beam (λ = 633 nm) is irradiated onto the grating and the reflected light intensity is detected by a charge-coupled device (CCD) line sensor. Under the specific coupling angle α0, a part of the light is coupled into the waveguide and the electromagnetic field of the light wave reaches into the medium which has a lower refracting index (Fig. 3). This results in an exponentially decaying field, the evanescent wave. The position of the reduced intensity in the reflected light is observed and evaluated (Fig. 4). Each binding effect at the sensor surface results in a change of the effective refractive index Neff, leading to a shift of the coupling angle. This relation is given by the coupling condition: N eff = n0 sin a 0 +
kl . L
(1)
Label-Free Serodiagnosis on a Grating Coupler
191
Fig. 1. Optical arrangement for the grating coupler in the reflection mode. The distance between the grating coupler chip and the CCD sensor is about 100 cm. The wedge plate in front of the grating coupler chip and the positioning system is not pictured.
Fig. 2. A photograph of the optical arrangement of the grating coupler in the reflection mode. In the front is the flow cell which presses the grating coupler chip against the wedge plate. The cylinder lens, the neutral density filter and the laser are arranged in the background. In the right upper corner is the mirror which deflects the light to the here not shown CCD sensor.
where n0 denotes the refractive index in air, k the diffraction order, l the wavelength and L the grating period. The limit of detection of the grating coupler system is the minimum resolution of refractive index changes being in the range of 3 × 10−6. This corresponds to an approximate mass coverage of 10 pg/mm2 (14). Exemplary, biomolecules with a molecular weight of 150 kDa, like immunoglobulin G, can be detected down to a concentration
192
Nagel, Ehrentreich-Förster, and Bier
Fig. 3. Schematic illustration of the grating coupler chip with the in-coupled light under coupling conditions
Fig. 4. Intensity distribution recorded by the CCD sensor under coupling conditions (by courtesy of Jörg Henkel, FhG-IBMT). The shift of the left slope of the intensity minimum is determined.
of 0.3 nM (15). Since the grating coupler is sensitive only to changes of the refractive index very close to the surface, the use of complex sample solutions like blood serum affects the noise of the signal only in a minimal way. It is therefore possible to detect antibody concentrations in the low nM-range. The limiting factor is the discrimination between a specific binding event and unspecific attachment of sample compounds both happening at the surface. Consequently, the minimization of unspecific
Label-Free Serodiagnosis on a Grating Coupler
193
attachment is the crucial point for the detection of low analyte concentrations. 1.2. Surface Functionalization and Coupling Strategies
Coupling of biomolecules to surfaces can be performed with a considerable number of different strategies. The simplest approach is the physical adsorption of the ligand to the surface by electrostatic bonds, van der Waals interactions, hydrogen bonds, and/or hydrophobic interactions. However, this construction is thus stable only to a limited extent against changes in the ionic strength and the pH-value. Stable binding is obtained by using covalent coupling protocols which generate a linkage between reactive groups on the surface and the biomolecule (see Subheading 3.3 and 3.4). Reactive groups on the surface of the Ta2O5 waveguide can be generated for example by functionalization with an aminosilane layer in order to expose amino groups on the surface (Subheading 3.2.). Another method presents the affine coupling via the biotin-streptavidin system. Biotinlabeled biomolecules react under the formation of a very stable non-covalent interaction with streptavidin, avidin or neutravidin. The coupling procedure for different biomolecules always has to be optimized to establish a functional surface with the intact ligand in an epitope presenting orientation. A good collection of feasible coupling reactions and protocols can be found in (16–18). Two simple but effective covalent coupling procedures with a good surface passivation for measurements with complex solutions are described here in detail. More complex strategies for surface passivation have been described and evaluated in (5).
1.2.1. PDITC Coupling
1,4-phenylene diisothiocyanate (PDITC), a homobifunctional crosslinker, can be used to couple a biomolecule covalently to the sensor surface in a two-step process (18). In the first step, one of the two isothiocyanate groups of the crosslinker reacts with an amine group at the sensor surface under the formation of a thiourea linkage. The para-position of the isothiocyanate groups prevents the reaction with another amine group on the surface. In the second step, the free isothiocyanate group reacts with amine groups of the protein. This coupling chemistry results in a very dense surface coverage with biomolecules and, hence, minimizes unspecific binding to the surface. Consequently, the PDITC coupling procedure is ideal for measurements on complex solutions like blood serum (6). A typical sensorgram of a measurement on undiluted clinical samples (blood serum) for the direct detection of a tuberculosis specific antibody is shown in Fig. 5. In this case the label-free measurement with the grating coupler shows a similar sensitivity and specificity like an established ELISA test (6).
1.2.2. EDC/Sulfo-NHS Coupling
The covalent linkage between a carboxylate and an amine can be achieved with the reagents N-Ethyl-N´-(3-dimethylaminopropyl)
194
Nagel, Ehrentreich-Förster, and Bier 800 B 700 B
ΔNeff (*10−6)
600 500 400 R 300 200 R 100 1 0
2
B
0 2000
4000
6000
B 8000
time [s]
Fig. 5. Real-time measurement of clinical blood serum samples with the grating coupler. (1) tuberculosis negative blood serum; (2) tuberculosis positive blood serum; (R) regeneration (50 mM HCl); (B) PBS buffer.
carbodiimide hydrochloride (EDC) and N-hydroxysulfo succinimide (sulfo-NHS). EDC reacts with a carboxylate group to form an active ester group (O-acylisourea), but this intermediate is very unstable in aqueous solution. Sulfo-NHS increases the stability of the intermediate and rapidly reacts with amines forming a stable amid linkage (16). Beside the coupling of the carboxyl groups to the amines on the surface, self-polymerization occurs if amine groups are also present in the biomolecule. This self-polymerization yields in a dense surface coverage of the biomolecule.
2. Materials 2.1. Sensor Chip Modification 2.1.1. Cleaning and Activation of the Sensor Chips
1. Distilled H2O. 2. Ethanol 99%, denatured. 3. Piranha cleaning solution: hydrogen peroxide (30%) and sulfuric acid (95–97%) in a mixing ratio of 1:3. Always add the peroxide to the acid. An exothermic reaction occurs and the solution heats up above 100 °C. Mixing piranha solution with organic compounds may cause an explosion. Handle with care. Wear safety glasses and nitrile gloves and always work inside a fumehood. 4. Sodium hydroxide (5 M). Handle with care. Wear safety glasses and nitrile gloves and always work inside a fumehood.
Label-Free Serodiagnosis on a Grating Coupler
195
2.1.2. Silanization of the Surface
1. Aminopropyltriethoxysilane (APTES) solution: 10% (v/v) APTES in distilled H2O, adjust the pH with hydrochloric acid precisely to 3.45.
2.1.3. PDITC Coupling
1. Dimethylformamid, dry (DMF). 2. Phenylene diisothiocyanate, purum (PDITC): 300 mM in DMF. Handle with care. Wear safety glasses and nitrile gloves and always work inside a fumehood. 3. Phosphate-buffered saline (PBS): 137 mM sodium chloride, 10 mM di-sodium hydrogen phosphate, 2.7 mM potassium chloride, 2 mM potassium di-hydrogen phosphate, pH 7.4. 4. TRIS-HCL buffer: 1 M tris(hydroxymethyl)aminomethane, pH 7.4.
2.1.4. EDC/Sulfo-NHS Coupling
1. N-Ethyl-N´-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC). Protect from moisture. Store desiccated at −20 °C. 2. N-hydroxysulfosuccinimide (sulfo-NHS). 3. MES buffer: 50 mM 2-(N-morpholino)ethanesulfonic acid, pH 6. 4. Distilled H2O with 0.1% (v/v) Tween 20.
2.2. Measurement
1. Phosphate-buffered saline (PBS): 137 mM sodium chloride, 10 mM di-sodium hydrogen phosphate, 2.7 mM potassium chloride, 2 mM potassium di-hydrogen phosphate, pH 7.4. 2. Regeneration solution: 50 mM hydrochloric acid. 3. Human serum samples.
2.3. Instrumentation 2.3.1. Grating Coupler
The optical arrangement of the grating coupler is shown in Fig. 1 and Fig. 2. 1. Laser: 10 mW HeNe laser, λ = 633 nm, # 1137P (Uniphase, Manteca, CA, USA). 2. Laser power supply: #1202–2 (Uniphase, Manteca, CA, USA). 3. Neutral density filter: T = 10%, #063462 (Linos, Göttingen, Germany). 4. Cylinder lense: f = 40 mm, #063422 (Linos, Göttingen, Germany). 5. Chip holder: a. Wedge plate: #334482 (Linos, Göttingen, Germany). b. Positioning system consists of a X-Y stage, a Z-axis stage and a rotary stage (Linos, Göttingen, Germany). c. Grating coupler chips: Grating period = 0.75 μm, Ta2O5 waveguide, #ASI 3200 (Artificial Sensing Instruments, Zürich, Switzerland).
196
Nagel, Ehrentreich-Förster, and Bier
6. Mirror: 1/10 wave accuracy, Ø = 31,5 mm, #340008400 (Linos, Göttingen, Germany). 7. CCD line sensor: 1728 Pixels, #TH 7803ACC (Thomson, Saint-Egreve, France). 2.3.2. Flow System
1. Flow cell: PVC, Volume ∼ 9 μL (Fraunhofer IBMT, Potsdam, Germany). 2. Tubing: Teflon tube, Øi = 0.5 mm (ERC, Riemerling, Germany). 3. Peristaltic pump: Perimax 12 (Spetec, Erding, Germany).
2.3.3. Data Evaluation
1. Standard PC with a CCD sensor specific converter card (Thomson, Saint-Egreve, France). 2. The Linux based evaluation software (Fraunhofer IBMT, Potsdam, Germany) determined the shift of the minimum by the evaluation of the left slope of the measured signal. Fig. 4 shows a typical signal under coupling conditions. After the calculation of the inflection point of the slope, the minimum position of the inflection is determined and the shift of this position is recorded (5 Hz).
3. Methods 3.1. Cleaning and Activation of the Sensor Chips
All steps should be performed under gentle agitation. 1. Clean the chips with distilled H2O – ethanol – distilled H2O two times (each step 2 min). 2. Incubate in piranha cleaning solution for 5 min (see Note 1). 3. Wash intensively with distilled H2O. 4. Dip in 5 M sodium hydroxide for maximal 30 s (see Note 2). 5. Wash intensively with distilled H2O, avoid air contact and go on to the next step (see Note 3).
3.2. Silanization of the Surface
1. Incubate the chips in APTES-solution for 2 h at 80 °C (see Note 4). 2. Wash intensively with distilled H2O. 3. Dry at 110 °C for 1 h. 4. Store them dry and under vacuum.
3.3. PDITC Coupling
1. Dry the chips at 110 °C for 10 min. 2. Prepare a glass petri dish with a filter paper at the bottom. 3. Rinse the cooled chips with DMF and dry them in a nitrogen stream.
Label-Free Serodiagnosis on a Grating Coupler
197
4. Put the chips in a petri dish and pipette 150 μL PDITC solution onto the grating. Replace the air in the petri dish with a nitrogen stream and seal it with Parafilm. Incubation time: 2 h (see Note 5). 5. Wash the chips with DMF and dry them in a nitrogen stream. 6. Pipette the protein solution (in PBS) at a concentration in the range of 1 mg/mL onto the PDITC activated area (see Note 6). 7. Incubate overnight in a humidity chamber (sealed petri dish with 1 mL PBS buffer). 8. Wash with distilled H2O. 9. Deactivate the remaining isothiocyanate groups with TRISHCL buffer buffer for 2 h. 10. Wash with distilled H2O. 3.4. EDC/Sulfo-NHS Coupling
1. Mix the protein solution with MES buffer to a final concentration in the range of 0.1–1 mg/mL (see Note 6). 2. Add EDC (0.2 M) and sulfo-NHS (0.05 M) to the protein solution. 3. Pipette this mixture onto the sensor chip. 4. Incubate it for 3 h in an humidity chamber (sealed petri dish with 1 mL MES buffer). 5. Wash with distilled H2O containing 0.1% Tween 20. 6. Wash with distilled H2O.
3.5. Measurement
Avoid temperature variations and air bubbles in the flow system. Use always well degassed running buffer. The sample flow should be from the bottom to the top of the flow cell to ensure a complete filling with the sample (see Notes 7 and 8).
3.5.1. Chip Mounting and Basic Settings
1. Position the chip on the wedge plate (with a little droplet of refractive index oil between glass substrate and wedge plate) and fix it with the flow cell (see Fig. 2). 2. Fill the flow system with PBS buffer. 3. Irradiate the laser light on the grating and adjust the angle of the incident light to a good incoupling position (clear effacement in the reflected light) and guide the reflected light to the CCD sensor. The measured signal should be similar to Fig. 4. 4. Flush the system with PBS buffer until a stable baseline is achieved (see Note 9).
3.5.2. Serum Sample Measurement and Regeneration
See Fig. 5 for an exemplary serodiagnosis measurement for the detection of tuberculosis specific antibodies in blood serum.
198
Nagel, Ehrentreich-Förster, and Bier
1. Adjust flow rate to 25 μL/min (see Note 10). 2. Inject the sample into the flow system. To increase the contact time to the surface, stop pumping and incubate for some minutes. Serum samples can be used both undiluted and diluted (see Notes 11–14). 3. Start flow again and change back to PBS buffer and observe the amount of binding to the surface. 4. Regenerate the surface with 50 mM sodium hydroxide (1 min) (see Note 15). 5. Flush the system with PBS buffer and inject the next sample when the base line is stable. 6. Determine the changes of the refractive index (ΔNeff) caused by the different samples and evaluate these values (see Note 16).
4. Notes 1. Use only freshly prepared piranha solution. The chips should be hydrophilic after the treatment. 2. Do not incubate the chips longer than 30 s in the 5 M sodium hydroxide solution, otherwise the Ta2O5 waveguide will be corroded and destroyed. 3. To bridge a short period of time (in the range of minutes), transfer the chips into a container with water to avoid air contact. 4. Use only freshly prepared APTES solution with a well adjusted pH-value. Cover the reaction chamber to avoid evaporation. 5. The PDITC solution should not dry up during the two hours incubation time. Increase the volume if necessary. 6. Measurements with complex solutions like blood serum with optical label-free devices are only possible if the unspecific attachment to the surface can be minimized. In the literature, different approaches dealing with this topic are described (5, 6). High surface coverage with the ligand is the crucial point; either by the formation of multilayered constructions or just by coupling high concentrations of the ligand. Furthermore, the appropriate dilution of the sample or the addition of detergents and high salt concentrations in the running buffer can significantly minimize the unspecific attachment. 7. Perform the measurements always with well-tempered samples and buffers. Keep in mind that a change of one degree Celsius results in a refractive index change of 1 × 10−4.
Label-Free Serodiagnosis on a Grating Coupler
199
8. Air bubbles in the flow system have a negative influence on the measurement if they get into the measurement/flow cell. Degas the buffers in a vacuum chamber (pressure lower than 10 mbar) under gentle agitation. A set-up which sucks the buffer through the flow system is more susceptible for air bubble contamination than a set-up which applies positive pressure. A strong influence is also given by the flow rate and by the addition of detergents to the buffer. 9. Given that the measured effect is temperature dependent, use only solutions at room temperature. If the baseline is not stable, try to wash unbound material off with regeneration solution. 10. A flow rate between 20–100 μL/min is typical. Use only pumps with linear flow. For kinetic measurements it is essential to ensure that the reaction is mass transport-limited. 11. Always use standard solutions (e.g. monoclonal antibodies specific and unspecific to the coupled biomolecule) to check the success of the coupling procedure and the specificity of the binding. 12. Measurements of solutions with a large difference in their refractive index (e.g. buffer and serum) result in a huge signal shift (index jump). The determination of the amount of bound material on the surface is only possible after changing back to the running buffer. 13. Due to the fact that there is only one recognition area in the flow cell, it is necessary to check for unspecific binding to distinguish between a specific and an unspecific binding. The essential test is done with an unspecific antibody and the obtained signal is compared to the specific antibody binding. Furthermore, the amount of unspecific attachment of the sample fluid (e.g. blood serum, milk, etc.) is an important information about the passivation effect of the manufactured surface. In addition to that, the efficiency of the coupling procedure for each chip should be checked as quality control, using a standard solution of a specific monoclonal antibody. 14. Blood products are potentially extremely hazardous and all steps should be performed under adequate protection. In many countries authorizations are needed and specialized knowledge is a prerequisite for working with blood products. Blood serum should be stored in aliquots at −80 °C. Avoid repeated freezing and thawing. 15. To reuse the sensor chip, it is essential to empirically find the mildest regeneration condition to remove the analyte without impairment of the coupled ligand. A wide range of different regeneration solutions can be tested. Regeneration
200
Nagel, Ehrentreich-Förster, and Bier
with low pH solutions (e.g. 10–100 mM hydrochloric acid) is successful for protein surface regeneration. But also solutions with high pH values (10–100 mM sodium hydroxide), high ionic strength or with low concentrations of sodium dodecyl sulfate are potential regeneration solutions and have to be optimized for each assay. 16. The determined ΔNeff values of known tuberculosis negative serum samples are used to define a cut-off value (sum of the mean signal intensities plus twice the standard deviation). Consequently, samples with higher signals are positive while samples with smaller signals are negative.
References 1. Gauglitz, G. (2005) Direct optical sensors: principles and selected applications. Anal Bioanal Chem. 381, 141–155 2. Tiefenthaler, K. and Lukosz, W. (1989) Sensitivity of grating couplers as integratedoptical chemical sensors. J Opt Soc Am B. 6, 209–220 3. Comley, J. (2005) Label-free detection – new biosensors facilitate broader range of drug discovery applications. Drug Discovery World, Winter 2004/5, 63–74 4. Cooper, M. A. (2006) Current biosensor technologies in drug discovery. Drug Discovery World, Summer 2006, 68–82 5. Brynda, E., Houska, M., Brandenburg, A. and Wikerstal, A. (2002) Optical biosensors for real-time measurement of analytes in blood plasma. Biosens Bioelectron. 17, 665–675 6. Nagel, T., Ehrentreich-Förster, E., Singh, M., Schmitt, K., Brandenburg, A., Berka, A., Bier, F. F. (2008) Direct detection of tuberculosis infection in blood serum using three optical label-free approaches. Sens Actuators B: Chem. 129, 934–940 7. Ehrentreich-Förster, E., Scheller, F. W. and Bier, F. F. (2003) Detection of progesterone in whole blood samples. Biosens Bioelectron. 18, 375–380 8. Nellen, P. M. and Lukosz, W. (1990) Integrated optical input grating couplers as chemo- and immunosensors. Sens Actuators B. 1, 592–596 9. Nellen, P. M. and Lukosz, W. (1991) Model experiments with integrated optical input grating couplers as direct immunosensors. Biosens Bioelectron. 6, 517–525
10. Lukosz, W., Nellen, P. M., Stamm, C. and Weiss, P. (1990) Output grating couplers on planar waveguides as integrated optical chemical sensors. Sens Actuators B. 1, 585–588 11. Lukosz, W., Clerc, D., Nellen, P. M., Stamm, C. and Weiss, P. (1991) Output grating couplers on planar optical waveguides as direct immunosensors. Biosens Bioelectron. 6, 227–232 12. Clerc, D. and Lukosz, W. (1994) Integrated optical output grating coupler as biochemical sensor. Sens Actuators B. 19, 581–586 13. Brandenburg, A., Polzius, R., Bier, F. F., Bilitewski, U. and Wagner, E. (1996) Direct observation of affinity reactions by reflectedmode operation of integrated optical grating coupler. Sens Actuators B. 30, 55–59 14. Billitewski, U., Bier, F. F. and Brandenburg, A. (1998) Immunobiosensors based on grating couplers, in (Rogers, K. R. and Mulchandani, A., ed.) Methods in Biotechnology Vol. 7, Humana Press, Totowa, NJ, pp. 121–134 15. Clerc, D. and Lukosz, W. (1997) Direct immunosensing with an integrated-optical output grating coupler. Sens Actuators B. 40, 53–58 16. Hermanson, G. T. (ed.) (1996) Bioconjugate Techniques. Academic Press, San Diego, CA 17. Aslan, M. (ed.) (1998) Bioconjugation: Protein Coupling Techniques for the Biomedical Sciences. Stockton Press, London 18. Heise, C. and Bier, F. F. (2005) Immobilization of DNA on microarrays, in (Wittmann, C., ed.) Topics in Current Chemistry: Immobilisation of DNA on Chips II, Springer, Berlin Heidelberg, pp. 1–25
Chapter 11 CCD Camera Detection of HIV Infection1 John R. Day Summary Rapid and precise quantification of the infectivity of HIV is important for molecular virologic studies, as well as for measuring the activities of antiviral drugs and neutralizing antibodies. An indicator cell line, a CCD camera, and image-analysis software were used to quantify HIV infectivity. The cells of the P4R5 line, which express the receptors for HIV infection as well as b-galactosidase under the control of the HIV-1 long terminal repeat, were infected with HIV and then stained 2 days later with X-gal to turn the infected cells blue. Digital images of monolayers of the infected cells were captured using a high resolution CCD video camera and a macro video zoom lens. A software program was developed to process the images and to count the blue-stained foci of infection. The described method allows for the rapid quantification of the infected cells over a wide range of viral inocula with reproducibility, accuracy and at relatively low cost. Key words: Charge-coupled device, CCD; Imaging, HIV, Infectivity, Quantify, Rapid.
1. Introduction The quantification of a biological process can sometimes be cumbersome or labor-intensive. In the field of virology, several assays have been developed to measure the infectivity of the human immunodeficiency virus (HIV) during a single round of replication:
1 Portions reprinted from the Journal of Virological Methods, Vol. 137, J.R. Day, L.E. Martínez, R. Šášik, D.L. Hitchin, M.E. Dueck, D.D. Richman and J.C. Guatelli, A computer-based, image-analysis method to quantify HIV-1 infection in a single-cycle infectious center assay, Pages 125–133, Copyright 2006, with permission from Elsevier.
Avraham Rasooly and Keith E. Herold (eds.), Methods in Molecular Biology: Biosensors and Biodetection, Vol. 503 © Humana Press, a part of Springer Science + Business Media, LLC 2009 DOI: 10.1007/978-1-60327-567-5_11
203
204
Day
plaque assays (1), syncytium formation assays (2, 3), fluorescent and colorimetric focal immunoassays (4) and assays using indicator cell lines (5–10). The quantification of infection in these assays requires counting the number of plaques, syncytia, foci, or cells within an infected population of cells. The use of indicator cell lines, such as those containing an integrated b-galactosidase gene under the control of the HIV long terminal repeat (LTR), has enabled the straightforward determination of viral infectivity by visualizing infected cells that are stained blue with 5-bromo-4chloro-3-indolyl-b-D-galactopyranoside (X-gal). The Magi-CCR5 (6) and P4R5 (7) indicator cell lines are frequently used as targets of infection. These are HeLa-based adherent cells that have been engineered to express CD4, the HIV receptor, as well as CCR5, one of the two predominantly used coreceptors; HeLa cells naturally express CXCR4, the other major coreceptor used by HIV. To quantify the number of infected cells by using these b-galactosidase-based indicator cell lines, one approach is to manually count blue cells by eye through a microscope. This laborious method is subject to observer error and is difficult when the density of infected cells is high, potentially limiting the dynamic range of the assay. To facilitate the counting of infected cells with speed, accuracy and optimal dynamic range, a chargecoupled device (CCD) camera was used and software was developed to analyze digital images of infected cell monolayers (11). Accurate and reproducible cell counts over a wide range of viral inocula were obtained with this method. CCD camera technology has been available for years, but the recent explosion in the use of digital cameras for personal use has accelerated technology development and facilitated the incorporation of affordable CCD detection in the laboratory.
2. Materials 2.1. Imaging Apparatus
1. 5 megapixel CCD color camera with real-time viewing, C-mount optical lens interface, and FireWire IEEE 1394 digital interface: MicroPublisher 5.0 RTV (Model # MP5.0-RTV-CLR-10, QImaging, Burnaby, BC, Canada) (see Note 1). 2. Macro video zoom lens: Optem 18–108 mm, f/2.5, C-mount (Qioptiq Imaging Solutions, Rochester, NY, formerly ThalesOptem, Inc.) (see Note 2). 3. Copy stand with a 1/4″-20 threaded mounting screw (model # CS-3, Testrite Instrument Company, Inc.). Something similar would suffice.
CCD Camera Detection of HIV Infection
205
4. Fluorescent light box (typically used for viewing X-ray films) (see Note 3). 5. Camera mount adapter, custom made by our university machine shop, consisting of a tripod mounting screw (1/4″-20 thread) and a rectangular piece of aluminum (2.25 × 1 × 0.25 in.) with two drilled holes (see Note 4). Use of the adapter requires that the copy stand have a mounting rod that can extend away from the stand. 6. Personal computer for capturing and analyzing images: Windows 2000 or XP operating system, 1 available FireWire port (IEEE 1394) or a FireWire expansion card. 2.2. Image Acquisition and Analysis Software
1. Image acquisition software, QCapture Pro, v.5.0.0.16, with USB key (dongle) and drivers for PC (QImaging, Burnaby, BC, Canada). The Micropublisher 5.0 camera was bundled with QCapture Suite software, however, the upgrade to QCapture Pro was purchased to increase control over the image capture process. 2. Image analysis software: application (5MGL.exe) and parameter file (5MGL.par), collectively termed the “Romanizer.” The software was developed in-house in the Fortran 95 programming language for the specific purpose of counting HIV-infected cells. Use of the software for other purposes may not be suitable. Alternative image-analysis applications exist, both freeware and for purchase (see Note 5).
2.3. HIV Infectivity Assay
1. Indicator cells: P4R5 HeLa cells (P4.R5 MAGI cells, AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH contributed by Dr. Nathaniel Landau). The P4R5 cell line contains the b-galactosidase gene under the control of the HIV-1 promoter region (LTR). The derivation of the parental P4 line has been reported (7). P4R5 cells express on their surface the HIV entry receptors, CD4, CXCR4 and CCR5. 2. Complete media: Dulbecco’s Modified Eagle Medium (DMEM; Gibco, Grand Island, NY) supplemented with 10% fetal bovine serum (FBS; GemCell, Woodland, CA), 100 U/ mL penicillin, 100 μg/mL streptomycin (pen/strep; Gibco, Grand Island, NY), 2 mM L-glutamine (Gibco, Grand Island, NY) and 1 μg/mL puromycin. Store at 4°C. 3. 48-well tissue culture plate (Cat. # 353078, BD Falcon, San Jose, CA). 4. Phosphate buffered saline, 1× (PBS; Invitrogen, Carlsbad, CA). 5. Fix solution: 1% formaldehyde, 0.2% glutaraldehyde in PBS. Can be made in advance and stored in the dark at 4°C for 1–2 months.
206
Day
6. X-gal stain solution made fresh: For each 1.0 mL of solution, combine 949 μL PBS, 20 μL 0.2 M potassium ferrocyanide, 20 μL 0.2 M potassium ferricyanide, 1.0 μL 2.0 M Mg2CL, and 10 μL 40 mg/mL X-gal (5-bromo-4-chloro-3-indolyl-b-Dgalactopyranoside). Stock solutions of potassium ferrocyanide, potassium ferricyanide, Mg2Cl, and X-gal can be prepared and stored in aliquots at −20°C. X-gal is dissolved in dimethyl sulfoxide (DMSO) and should be stored in the dark. The X-gal solution will turn yellow over time, but this does not affect the activity. Discard the stock if it becomes green/brown. 7. The HIV virus is either isolated and cultured from blood, or produced in vitro by transient transfection of proviral plasmid DNA into cultured cells. Viruses do not need a complete HIV genome, but they must be competent for entry into P4R5 cells and they must be capable of expressing the HIV Tat protein. Tat transactivates the HIV LTR promoter region and will enable expression of b-galactosidase in the P4R5 indicator cells.
3. Methods In order to facilitate the counting of blue-stained HIV-infected cells, a CCD camera, macro lens, and a computer program were set up to capture and process images of plated cells. The setup does not require the exact equipment listed in Subheading 2. A camera, lens, light source, computer, and a method to mount the camera are the basic required elements. The ideas described here can be adapted to different configurations to suit the needs of the application. The method for analyzing the images will vary depending on the final goal of the analysis. Although the software we developed to count HIV-infected cells may not be suitable for other applications, alternative software programs may provide the necessary tools for your particular application. 3.1. Setup of Imaging Apparatus
1. Place the copy stand on a table and the light box (or light source of choice) on the base of the stand. 2. Assemble the camera by screwing the lens onto the camera. 3. Mount the camera to the stand using the 1/4″-20 threaded mount. The most straightforward way to mount the camera is illustrated in Fig. 1a. An alternative method to mount the camera requires a custom-built adapter (see Note 6). The alternative method is illustrated in Fig. 1b and a photograph of the setup is shown in Fig. 2. Screw the mount adapter onto the copy stand, then attach the camera to the adapter using the tripod mount screw (Fig. 3).
CCD Camera Detection of HIV Infection
207
Fig. 1. Schematic diagrams of the CCD camera setup. (a) The simplest setup using direct attachment of the camera to the copy stand. The FireWire ports on the camera are on the backside in the perspective shown. (b) Use of an aluminum mount adapter and tripod screw to mount the camera rotated at 180° (see Notes 4 and 6).
Fig. 2. Photograph of the actual camera setup showing the light box, copy stand, CCD camera, macro lens and computer.
208
Day
Fig. 3. Close-up photographs of the camera mounted to the copy stand using the mount adapter.
4. The height of the camera above the light box is adjustable and depends on the zoom setting of the lens. In our setup, the distance between the surface of the light box and the mounting screw hole of the camera was 14.25 in. (36.2 cm). The tip of the lens was 6.25 in. (15.9 cm) above the light box. Level the light box and camera using a standard bubble level. 5. Attach a FireWire cable to either one of the ports on the camera. Do not connect the cable to the computer until after loading the software and drivers. 3.2. Capture and Analysis of Images 3.2.1. Capturing Images
1. These instructions assume the use of the Micropublisher 5.0 RTV camera, Optem 18–108 mm macro video zoom lens, and QCapture Pro software. Other cameras, capture software, and analysis software may be used. Please follow the directions included with your specific components. 2. Follow the setup instructions from QImaging to load the camera drivers and QCapture Pro software. You will connect the FireWire cable to the computer during this process. 3. To define the capture settings, launch QCapture Pro and click the camera icon on the toolbar (or select Acquire/Video-Digital from the menu) to open the acquisition dialog box. Use the Basic Dialog (rather than the Advanced Dialog) by clicking the “Basic Dialog” button. If the button in the lower left says “Advanced” then you are already looking at the Basic Dialog. 4. Check the settings by clicking the “More >>” button at the bottom of the window. This expands the window. Using version 5.0.0.16 of QCapture Pro and with the specific light source we used, the settings were as follows (Fig. 4) (see Note 7 for possible variations): ● Exp Acq: 00.300.00, Adjust Exp for Binning (checked)
CCD Camera Detection of HIV Infection
209
Fig. 4. Screen shot of the capture settings within QCapture Pro capture software.
●
●
Binning: Pvw (2 × 2 for focusing, 4 × 4 for moving the plate); Acq (1 × 1) Capture area dimensions: Left (320), Top (0), Right (2239), Bottom (1919) [final resolution, 1920 × 1920 pixels
●
Capture depth: 24-bit color
●
Gain: 1100; Gamma: 1.90; Offset: 1120
●
White balance: R (2.327), G (1.0), B (1.375).
5. Click “Less 3 × 107 photons per ms at the photodetector) the signal to noise ratio of the photodiodes begins to deteriorate – above that, it is close to the theoretical limit (4). Below 107 photons per ms the SNR of a photodiode becomes worse than the one of a photomultiplier.A major reason for that is the dark noise of the device. Cooling down the detector can significantly improve the noise, but adds complexity to the sensor and increases the power budget. In general, the rule of the thumb is: if the intensity is above 107 photons per ms, use photodiode. However, only head-to-head comparison with the specific devices in place can show which detector is better. 4. Bandwidth vs. noise. The noise of an amplifier is an important feature when measuring low-level light. In this case, the low-noise level comes with a price – severe limitation of the bandwidth. The described photodetector acts as a band pass filter – its maximum amplification is between 1 and 10 kHz. The lower cut-off frequency is set by the RC chain R5C5 (R6C6 and R7C7, respectively). It is responsible for DC-blocking and attenuating the ever-present 60 and 120 Hz. The high cut-off frequency is set by R2C2 (R3C3, R4C4). and its variation has multiple consequences. First, the change of the resistor value varies the overall amplification. The feedback capacitor values influence both, the bandwidth of the amplifier (in other words, the maximum modulation frequency of the light source that will be amplified) and simultaneously the noise (Fig. 10). It could be seen that without feedback capacitors the network exhibits a peak in the amplification at 500 kHz (Fig. 10b) and noise levels of ∼0.8 V (Fig. 10a)! Introduction of the capacitor C1 (i.e. 0.2 pF) decreases the noise ∼20 times, but the bandwidth drops to ∼400 kHz. Further introduction of C2 (Fig. 10c, d) reduces the noise another ten times, but at the price of
322
Kostov
Fig. 10. Effect on the values of C1, C2 and C3 on the noise and bandwidth of the photodetector.
bandwidth being shrunk down to 10 kHz. Adding C3 finally reduces the noise to the target 3 mV without more changes in the bandwidth (Fig. 10e, f). Addition of C4 has very small effect, the effect on the values of C1, C2 and C3 on the noise and bandwidth of the photodetector is shown in Fig. 10. The 10 kHz bandwidth is sufficient for amplitude measurements of absorbance or fluorescence. However, if a higher bandwidth is
Photodiode-Based Detection System for Biosensors
323
needed, the capacitor values can be changed to accommodate that. It is a good idea to model the frequency response of the new network using a PSpice program (i.e. TINA-TI, available free of charge from Texas Instruments) when changing any of the passive elements values.
References 1. Gowar, J. (1993) Optical Communication Systems, 2 ed., Prentice-Hall, Hempstead, UK 2. Kostov Y., Rao G. (2000) Low-cost optical instrumentation for biomedical measurements. Rev. Sci. Instr., 70, 4361–4374 3. Graeme, J. (1995) Photodiode Amplifiers: Op Amp Solutions, McGraw-Hill, New York, NY, 4. Wu, J.-Y., Cohen, L.B. (1993) Fast Multisite Optical Measurement of Membrane Potential. In Fluorescent and Luminescent Probes for Biological activity. (Mason, W.T., ed.) 1st ed., Academic Press, London, UK
5. Kostov Y., Albano C.R., Rao G. (2000) All-solidstate GFP sensor. Biotech. Bioeng. 70, 473–477 6. Märtens, O. (2000) Precise synchronous detectors with improved dynamic reserve, IEEE Trans. Instr. Meas. 49, 1046–1049 7. Dalal, N.G., Cha, H.J., Kramer, S.F., Kostov, Y., Rao, G., Bentley, W.E. (2006) Rapid noninvasive monitoring of baculovirus infection for insect larvae using green fluorescent protein reporter under early-to-late promoter and a GFP-specific optical probe. Process Biochem. 41, 947–950
Chapter 18 Photodiode Array On-chip Biosensor for the Detection of E. coli O157:H7 Pathogenic Bacteria Joon Myong Song and Ho Taik Kwon Summary An integrated circuit (IC) of photodiode array (PDA) microchip system was used for the on-chip detection of E. coli O157:H7 based on an enzymatic bioassay and light absorption property of the reaction product. The PDA microchip consisting of an array of 12 × 12 photodiode detection elements served as a photosensor as well as a protein-immobilizing sample platform. As a result, E. coli O157:H7 could be detected directly on the surface of PDA detection elements. E. coli O157:H7 was detected by forming a “sandwich-type” enzymatic immunocomplex on the PDA detection elements using an on-chip bioassay. The quantitative analysis of E. coli O157:H7 immunocomplex was carried out based on the light absorption property of the enzymatic reaction products of E. coli O157:H7 immunocomplexes with respect to a red beam produced by light emitting diodes (LEDs) installed right above the PDA microchip. During the on-chip bioassay, the wet photodiode detection elements exposed to a lot of biological materials or buffer solutions were capable of maintaining their photosensing capabilities. The portable PDA on-chip biosensor permits direct optical detection of E. coli O157:H7 and eliminates the necessity of the conventional expensive microplate reader that is incompatible with the size of the protein microarray. Key words: Bipolar photodiode array microchip, On-chip bioassay, Biosensor, E. coli O157:H7 immunocomplex, Protein microarray, Light absorption.
1. Introduction Rapid and simultaneous monitoring of protein or DNA microarray has been performed using commercially available detectors such as microplate reader or microscope (1, 2). However, these detectors are bulky and not compatible with the biological microarray in size although their detections are highly sensitive, accurate, and reproducible. The construction of a miniature integrated biochip device containing a detector such as the complementary metal oxide semiconductor (CMOS) microchip can be Avraham Rasooly and Keith E. Herold (eds.), Methods in Molecular Biology: Biosensors and Biodetection, Vol. 503 © Humana Press, a part of Springer Science + Business Media, LLC 2009 DOI: 10.1007/978-1-60327-567-5_18
325
326
Song and Kwon
suggested as a solution to solve such incompatibility (3). In order to achieve further, a miniature biochip system, PDA on-chip detection system is an attractive tool (4). In this system the bipolar PDA microchip plays the dual role of a DNA/protein immobilizing sampling platform and of a phototransducer. Consequently, a complicated optical alignment in the detection system could be completely eliminated and further a miniature and simple integrated field-usable optical biochip system could be produced. In this work, the scheme of on-chip detection of the pathogenic bacteria E. coli O157:H7 involves three major steps: (1) immobilization of E. coli O157:H7 on the PDA surface as sandwich-type immunocomplexes, (2) carrying out an enzymatic bioassay involving the immunocomplexes that covers the PDA detection elements with enzymatic reaction products, and (3) employing the light absorption property of the enzymatic reaction products against the red beam irradiation from the LEDs. Immobilization of anti-E. coli O157:H7 antibody on the PDA detection elements is achieved by sequential treatment of the PDA surface with silanization reagent (3-aminopropyl) triethoxy silane or APTES), glutaraldehyde, and anti-E. coli O157:H7 antibody. The sandwich type of E. coli O157:H7 immunocomplexes on the surface of PDA are composed of anti-E. coli O157:H7 antibody, E. coli O157:H7 bacteria, and alkaline phosphatase-labeled anti- E. coli O157:H7 antibody. Reaction between alkaline phosphatase in the immunocomplex and nitroblue tetrazolium/5bromo-4-chloro-3-indolyl-phosphate (NBT/BCIP) formed the blue precipitates. The intensity of the irradiated red beam reaching the PDA detection elements is inversely proportional to the amount of blue precipitates produced by the enzymatic reaction on the surface of PDA. Based on this light absorption property, quantitative analysis of E. coli O157:H7 bacteria was achieved.
2. Materials 2.1. Surface Chemistry
1. 10% (v/v) 3-Aminopropyltriethoxysilane (APTES) (Catalog No. 440140, Sigma-Aldrich, St. Louis, MO) was prepared in dry toluene. 2. Phosphate buffer solution of pH 7.0 was prepared by dissolving 0.3402 g of KH2PO4 (Catalog No. P5379, Sigma-Aldrich, St. Louis, MO) and 0.4180 g of K2HPO4 (Catalog No. P2222) in 500 mL of water. 3. 5% (v/v) glutaraldehyde (Catalog No. 16310, Electron Microscopy Science, Fort Washington, PA) solution was made in phosphate buffer solution.
Photodiode Array On-chip Biosensor for the Detection of E. coli
2.2. Formation of E. coli Immunocomplex
327
1. Affinity purified antibody to E. coli O157:H7 (Catalog No. 01-95-90, Kirkegaard & Perry Labs., Gaithersburg, MD) stock solution: The lyophilized product was rehydrated per supplier’s protocol by dissolving the pellet in 1 mL of aqueous glycerol (50%) (antibody concentration = 1.0 mg/mL). The stock solution was stored at −20°C. Working solutions were prepared by diluting the stock solution with PBS immediately before use. 2. E. coli O157:H7 positive control (Catalog No. 50-95-90, Kirkegaard & Perry Labs., Gaithersburg, MD) stock solution (3 × 109 cfu/mL): Dextran stabilized product was rehydrated per supplier’s protocol by dissolving the lyophilized pellet in 1 mL of aqueous glycerol (50%). The stock solution was stored at −20°C. Working solutions were prepared by diluting the stock solution with PBS immediately before use. 3. Alkaline phosphatase-labeled affinity purified antibody to E. coli O157:H7 (Catalog No. 05-95-90, Kirkegaard & Perry Labs., Gaithersburg, MD) stock solution: The lyophilized product was rehydrated per supplier’s protocol by dissolving the pellet in 1 mL of aqueous glycerol (50%) (antibody concentration = 0.1 mg/mL). The stock solution was stored at −20°C. Working solutions were prepared by diluting the stock solution with PBS immediately before use. 4. BCIP/NBT enzymatic substrate solution: One part each of the BCIP concentrate and the NBT concentrate of the BCIP/ NBT phosphatase substrate system (Catalog No. 50-81-00, Kirkegaard & Perry Labs., Gaithersburg, MD) were mixed with the ten parts of the supplied 0.1 M Tris-HCl buffer solution before use. 5. Ethylenediamine tetraaceticacid (EDTA) solution: EDTA solution (9 mM) was prepared in 10 mM Tris-HCl medium.
3. Methods 3.1. PDA Microchip System
1. The PDA microchip system was fabricated using conventional bipolar semiconductor technology. As shown in Fig. 1a, the PDA microchip system consists of a PDA microchip containing a 12 × 12 array of photodiodes, LEDs, and a test board. The individual photodiode detection element is composed of a photodiode with 300 μm diameter, current amplifiers, and a current-to-voltage converter (Fig. 1b). The pixel-to-pixel distance is 100 μm. 2. The photodiode detection element is fabricated by the incorporation of the P-type implant layer on the N-type epitaxial
328
Test board
Song and Kwon
LED
PDA microchip
Fig. 1. (a) Photograph of the bipolar PDA microchip system. (b) Schematic diagram of the PDA microchip. The photocurrent of the individual photodiode is amplified by the current amplifier and converted to voltage by the current-to-voltage converter.
layer. A vertical structure of a photodiode detection element is shown in Fig. 2. Highly concentrated P-type silicon barriers were placed at both bottom ends of P-type silicon substrate for isolation of the individual photodiode detection element. The concentrated N-type buried layer was formed on the P-type substrate in order to reduce the resistance of the N-type epitaxial layer. Then, the N-type epitaxial layer was formed on the buried layer. The PN junction was achieved by the incorporation of the P-type implant into the N-type epitaxial layer. The N-type implant was incorporated to connect the N-epitaxial layer and Al metal. The surface of the fabricated PN junction layer was then oxidized. The contact layer was obtained by partial removal of the oxidized layer using a buffered oxide etchant, where the first metal wiring was performed. Al metal was deposited on the contact layer. The metal wiring process was completed on the deposited Al metal layer using photolithography. An Al striper was used as
Photodiode Array On-chip Biosensor for the Detection of E. coli
329
2’nd Metal 1’st Metal Oxide Layer (SiO2) Emitter Implant Layer (N-type)
Contact
Base Implant Layer (P-type)
Isolation (P-type) Epitaxial Layer (N-type) Bottom Isolation (P-type) Buried Layer (N-type) Substrate (P-type)
Fig. 2. A vertical structure of a photodiode detection element in the PDA microchip.
an etchant. The first metal wiring layer was coated with an insulation layer of silicon oxide to perform a second metal wiring on it. The second metal wiring was also completed using photolithography. Current amplifiers and current-to-voltage converters (4) in individual photodiode detection elements were fabricated by this metal wiring based on photolithography. As a result, the individual pixel consists of photodiode, current amplifier, and current-to-voltage converter. These components are integrated in an unit pixel and the PDA microchip contains these 144 pixels. The test board transfers photosignals from the PDA to a laptop computer. An array of LEDs was placed right above the PDA microchip to irradiate the PDA microchip with the red beam. 3. The individual photodiode can be addressed and read using digital I/O lines and an analog-to-digital conversion channel supplied by an RS232 interface to the laptop computer. Custom software written in the C language was used for data acquisition process. Figure 3 shows an algorithm of the custom software. There is an AT89S52 Micom (Micro-computer) in the test board. The Micom is operated by the custom software. Communication between the test board and the software in a laptop computer was achieved via the RS232 interface. The communication is initiated with a start signal in the software. When the Micom receives the start signal, the Micom sends a command to the LEDs installed right above the test board so that the LEDs beams are emitted. The produced LEDs beam is irradiated on the PDA microchip. After the emission of the LEDs beams, the Micom tests all the pixels in order to read output signals from the PDA. The Micom can access all the 12 column pixels in the first row one by one using row and column address decoders. Identical tests are performed with respect to all the 12 column pixels in
330
Song and Kwon
Test F/W Flow Chart Initialization
N Test Start?
Y PWM Pulse generation : -. 2 usec x PWM level timer running -. Interrupt Enable -. Delay 600 msec
Set Test Column and Row cell
Ready to read ADC0802
N
Does ADC interrupt generate?
Y Read ADC data Send ADC data to PC
Y Is any test column and Row cell ?
N Fig. 3. An algorithm of the custom software that operates the entire PDA microchip system. The software was written in the C language.
the remaining 11 rows. The output signals from the PDA are converted to digital values by the 8-bit ADC (analog to digital converter) in the test board (5). Then Micom reads output signals and transfers those signals to the laptop computer via the RS232 interface. Figure 4 shows a block diagram to represent an operation process of the test board. 4. As shown in Fig. 5, E. coli O157:H7 immunocomplex was immobilized directly on the surface of photodiode detection elements.
Photodiode Array On-chip Biosensor for the Detection of E. coli
Row[1..12] Col[1..12] Analog S / W_1
Analog S / W_13
OUT
331
8bit_ADC
Col[1..12] Analog S / W_2 Light Source Col[1..12] Analog S / W_3
Row_1 Row_2 Row_3 12 by 12 Row_4 Photosensor Row_5 Array Row_6
Row Decoder
Col[1..12] Analog S / W_4
Col[1..12]
Analog S / W_5 u-COM
PC
… Row_12
Col[1..12] Analog S / W_6
PWM
… Col[1..12] Analog S / W_12 Column Decoder
LPF
To Light Source
Fig. 4. A block diagram that shows an operation process of the test board containing the PDA microchip. The analog-todigital converter and column and row address decoders are contained in the test board.
Fig. 5. A schematic diagram of the protein microarray directly immobilized on the surface of PDA detection elements. The E. coli O157:H7 immunocomplex was formed by sequential on-chip antibody–antigen reactions. Blue precipitates were produced as a result of enzymatic reaction between alkaline phosphatase-labeled anti-E. coli O157:H7 antibodies and NBT/BCIP.
332
Song and Kwon
3.2. Test of PDA Microchip System
1. The blank signal of the PDA microchip system was measured by irradiating a bare (virgin) chip with the red beam from LEDs in a dark room. The PDA microchip system was manipulated so that the signal becomes 2.5 when the red beam is irradiated onto the bare PDA microchip.
3.3. Surface Chemistry
1. First a set of six PDA microchips was treated with 10% 3-APTES solution at 115°C in an oven for 2 h. After 2 h, each chip was rinsed thoroughly with toluene and dried at room temperature. 2. The silanized chips were then treated with 2.5% glutaraldehyde solution for 1 h at room temperature. Each chip was washed thoroughly with potassium phosphate buffer (pH 7.0) to remove any unbound glutaraldehyde. 3. For immobilization of anti-E. coli O157:H7 antibody on the PDA microchip, the anti-E. coli O157:H7 antibody stock solution (1 mg/mL) was diluted with PBS to a final concentration of 20 μg/mL. Eighty microliter of the diluted solution was spotted onto the photodiode detection elements of each chip, and the chips were incubated overnight at 4°C. The unbound anti-E. coli O157:H7 antibodies were rinsed off with PBS.
3.4. Formation of E. coli Immunocomplex
1. Anti-E. coli O157:H7 antibody-immobilized chips were then subjected to 10% BSA blocking solution (Catalog No. 50-61-01, Kirkegaard & Perry Labs., Gaithersburg, MD) for 1 h at room temperature. 2. The E. coli O157:H7 stock solution (3 × 109 cfu/mL) was serially diluted with PBS to get six E. coli O157:H7 sample solutions of concentrations 5.0 × 104, 2.0 × 105, 7.5 × 105, 2.0 × 106, 5.0 × 106, 1.0 × 107 cfu/mL. 3. Eighty microliter of each of E. coli O157:H7 sample solutions were then spotted separately on six BSA-treated PDA microchips. The chips were incubated for 1 h at room temperature. Unbound E. coli O157:H7 was rinsed off with PBS. 4. The alkaline phosphatase-labeled anti-E. coli O157:H7 stock solution was diluted with PBS (1:50). Eighty microliter of the diluted solution was spotted onto the PDA microchip to react with E. coli O157:H7 captured by the anti-E. coli O157:H7 antibody immobilized on the photodiode detection elements. 5. After 1 h incubation at room temperature, the unbound alkaline phosphatase-labeled anti-E. coli O157:H7 antibodies were rinsed off with PBS.
3.5. Enzymatic Reaction for Producing Blue Precipitate
1. 40 μL of the BCIP/NBT working solution was added onto the PDA microchip containing the E. coli immunocomplexes. The optimal enzymatic reaction between the BCIP/NBT
Photodiode Array On-chip Biosensor for the Detection of E. coli
333
and the alkaline phosphatase of the E. coli immunocomplex was achieved by incubating the chips at 37°C for 1 h. The amount of blue precipitates produced by the enzymatic reaction is proportional to the number of E. coli O157:H7 attached to the surface. 2. The enzymatic reaction was ceased by adding 40 μL of Tris-EDTA buffer solution (final EDTA concentration 1 mM) onto the PDA microchip. 3.6. Detection on the PDA Microchip
1. The quantitative detection of E. coli O157:H7 was performed by the irradiation of the red beam onto the PDA microchip covered with blue precipitates. 2. Control signal is the average digital value (Sc) obtained from all the pixels when the red beam is irradiated onto an entire PDA microchip whose on-chip bioassay is performed without E. coli O157:H7. 3. Each PDA chip was irradiated with the red beam and the average signal was obtained as a function of the number of E. coli O157:H7. The average digital value obtained from 12 photodiode detection elements at this time is Ss and the magnitude of Ss is dependent on the number of E. coli O157:H7. The magnitude of Ss is reduced compared to Sc because the intensity of red beam reaching the photodiode detection elements decreases due to absorption by blue precipitates. 4. The final signal at a given E. coli concentration is Sc–Ss. Unknown concentration of bacteria can be detected from a plot of (Sc–Ss) values as a function of known E. coli concentration. 5. The background noise level is the standard deviation of control signals obtained from control experiments. The noise level was determined to be 35 mV. Based on the background noise level of 35 mV, a detection limit of 4.5 × 104 E. coli. O157:H7 (at S/N = 3) can be obtained. The entire output signals from 144 photodiode detection elements are displayed as the digital value in the custom software.
4. Notes 1. On-chip detection of E. coli O157:H7 pathogenic bacteria using the PDA microchip may lead to erroneous result if the pHs of reaction solutions deviate from the neutral range (pH 7.0 to 7.5). The PDA detection elements may get damaged, e.g., corroded when the pH of the reaction solution is out of neutral range.
334
Song and Kwon
4.500 4.000 3.500
Signal (V)
3.000 2.500 2.000 1.500 1.000 0.500 0.000 After APTES
After glutaraldehyde After Goat antiE.coli O157:H7
After BSA
After E.coli O157:H7
After phosphataselabeled goat antiE.coli O157:H7
Fig. 6. Photosignals as measured at every reaction step during the on-chip bioassay. The pHs of solutions for BSA treatment and E. coli O157:H7 capturing were 8.0 and 8.2, respectively. As a result, large deviation of photosignals from the average value of 2.37 V was observed. When the pH of the reaction solution for on-chip bioassay was maintained at 7.0, photosignals were stabilized around the average value of 2.37 V, as measured after APTES, glutaraldehydes, anti- E. coli O157:H7 antibody, and alkaline phosphatase-labeled anti-E. coli O157:H7 antibody treatments.
2. To confirm normal operation of the PDA microchip during the on-chip reactions, the signal at every on-chip reaction step has to be monitored. As shown in Fig. 3, a large deviation of the signal (from 2.37 (average) ± 0.053 V) reveals abnormality of the photodiode detection element as a photosensor. 3. Figure 6 is a typical example that represents the variation of photosignals at every reaction step due to the variation of pH of reaction solutions. The pHs of BSA and E. coli O157:H7 solution were 8.0 and 8.2, respectively. As a result, large deviations of the signals could be observed. 4. A normal operation of the PDA microchip is assured as long as pH range of the on-chip reaction solution is maintained between 7.0 and 7.5.
Acknowledgments This research is sponsored by the Korean Ministry of Commerce, Industry and Energy under contract 10023590-2006-02 program with Seoul National University.
Photodiode Array On-chip Biosensor for the Detection of E. coli
335
References 1. Nakauchi, G., Inaki, Y., Kitaoka, S., Yokoyama, C., and Tanabe, T. (2002) Application of L-cystein derivative to DNA microarray. Nucleic Acids Res. Suppl. 2, 257–258 2. Allain, L. R., Askari, M., Stokes, D. L., and VoDinh, T. (2001) Microarray sampling-platform fabrication using bubble-jet technology for a biochip system. Freenius J. Anal. Chem. 371, 146–150 3. Song, J. M., Culha, M., Kasili, P. M., Griffin, G. D., and Vo-Dinh, T. (2005) Detection
of single bacteria using a compact CMOSbased immunosensor. Biosens. Bioelectron. 20, 2203–2209 4. Song, J. M., Yang, M., and Kwan, H. T. (2007) Development of a novel DNA chip based on a bipolar semiconductor microchip system. Biosens. Bioelectron. 22, 1447–1453 5. Kwon, H. T., Yang, M. S. (2003) Apparatus for analyzing for a disease using photodetector cell circuit. KR 0400202
Chapter 19 DNA Analysis with a Photo-Diode Array Sensor Hideki Kambara and Guohua Zhou Summary A simple instrument or device for easy DNA analysis is required. A combination of bioluminometric assay and photo-detection with an inexpensive photo-diode array provides a simple instrument for various DNA analyses. Its characteristics and applications for DNA analyses are described. Key words: DNA analysis, DNA sequencing, Pyrosequencing, SNPs, Photodiode, Photo-diode array, Luminescence, Step-by-step sequencing.
1. Introduction Gel electrophoresis coupled with laser induced fluorescence detection is generally used for analyzing DNA (1). A capillary array DNA sequencer has been successfully used to sequence the human genome (2, 3). With the completion of the human genome project in 2003, massive amounts of DNA data are available to be used in solving medical, environmental, and food problems. In the post genome-sequencing era, simple and inexpensive DNA analysis devices or instruments are required in addition to a high throughput instrument. Simple and inexpensive DNA analyzers can be made by using inexpensive photo-sensors. Here the characteristics of a small DNA analyzer with a photo-diode array and its application to DNA analysis are described. 1.1. Pyrosequencing
The small DNA analyzer is based on bioluminometric detection with a photo-diode and step-by-step nucleotide incorporation reactions (pyrosequencing) (4, 5). Figure 1 shows the principle of
Avraham Rasooly and Keith E. Herold (eds.), Methods in Molecular Biology: Biosensors and Biodetection, Vol. 503 © Humana Press, a part of Springer Science + Business Media, LLC 2009 DOI: 10.1007/978-1-60327-567-5_19
337
Kambara and Zhou primer 5’ A C G ssDNA
3
+
T G
T A C G
C A T G
dATP
degradation
dCTP
degradation
dTTP
degradation
dGTP
PPi
+ + +
5’
A C G
3’
T G +
5’ 3’
ATP T A
PPi
luminescence
C G G
ATP
C C T T
A A
luminescence
A C G T A C G G A A T G C A T G C C T T A A 1 min
CCC AA TTGG TT G
C C T T A A
luciferase reaction
C A T G
dATP
fluorescence
338
GG CC AA TT A T G
AA C
3 bases 2 bases 1 base noise
time
Fig. 1. Principle of pyrosequencing technology and pyrogram: Four enzymatic reactions are carried out simultaneously. As the reaction solution is diluted by the dNTP additions, peak heights decrease gradually as indicated by the dashed lines. It is partly because the reaction speed slows down with the step while the peak areas stay almost constant.
pyrosequencing. The target DNA is amplified by PCR to prepare a single stranded template DNA. The reaction chamber contains reagents for ATP production as well as luminescence reaction. After the template DNA is purified, it is put into the reaction chamber together with the DNA polymerase and a primer. The primer hybridizes to the template DNA, and this is followed by nucleotide incorporation reactions obtained by sequentially adding four deoxynucleotide species (dNTPs) one-by-one. When an added nucleotide species is complementary to the target, nucleotide incorporation occurs to extend the complementary DNA strand and to produce an inorganic pyrophosphate (PPi) as a by-product. The pyrophosphate reacts with adenosine 5´ phosphosulfate (APS) to produce ATP by the enzymatic reaction of ATP sulfurylase. ATP reacts with luciferin to emit luminescence through a luciferase reaction. The luminescence is detected with an array sensor or a photo-diode. The base sequence is determined by observing the luminescence and the injected nucleotide species. The enzymatic reactions are summarized in Fig. 2. The original method of pyrosequencing uses APS as a substrate for the reaction to produce ATP. This is the substrate for luciferase reaction as well. Although the reaction speed of
DNA Analysis with a Photo-Diode Array Sensor
339
dNTP DNA polymerase
Primer
ssDNA
Extension
PPi AMP + oxyluciferin + CO2
APS
ATP sulfurylase
PPi
Luciferase
ATP
SO2 2-
hv Luciferin + O2
Fig. 2. Enzymatic reactions used in pyrosequencing. Nucleotide incorporation reaction, ATP production reaction, nucleotide degradation reaction and luminometric reaction occur simultaneously in one chamber.
luciferase-APS reaction is slow, the reaction gives a large background because the amount of APS is large compared to that of ATP. Another enzymatic reaction for producing ATP from PPi can be used. This reaction uses another enzyme, pyruvate orthophosphate dikinase (PPDK), to produce ATP with the substrates of AMP and phosphoenolpyruvate (PEP). DNA polymerase (ssDNA - primer)n + dNTP ¾¾¾¾¾¾ ®(ssDNA - primer)n + 1 + PPi 2+
PPDK, Mg AMP + PPi + PEP ¾¾¾¾¾ ® ATP + pyruvate + Pi 2+
luciferase, Mg ATP + luciferin + O2 ¾¾¾¾¾¾ ® AMP + Oxyluciferin + CO2 + PPi + light
As neither AMP nor PEP can be a substrate for the luciferase reaction, almost no background signal is produced even if a large amount of AMP or luciferase is used in the reaction (6). 1.2. Instrumentation for Pyrosequencing 1.2.1. Outline of the System
1.2.2. Device
A schematic diagram of a pyrosequencing system is shown in Fig. 3. It consists of eight sets of four reagent reservoirs and fluidics for injecting small amounts of reagents into eight reaction chambers, respectively. Each reaction chamber has a photo detector attached at the bottom. The system was made in house. Photographs of the prototype system are shown in Figs. 4 and 5. All movements are controlled with a microcomputer (Renesas Technology (Tokyo), HD64F3052BF25) according to the time chart as shown in Fig. 6. The required photo-detection sensitivity is estimated as follows. 1. Estimation of luminescence emitted from a reaction chamber: Usually 0.5–1 pmol of DNA sample is used for DNA sequencing.
340
Kambara and Zhou
dNTP reagent reservoirs dispensing nozzles
gas cylinder PC
valves loading
up and down valve control motor dispenser Position control
H8 microcomputer
reaction chamber vibration control
reaction chambers vibrator
reaction chamber holder
AD converter photo diode
Amp1
x1 x10 x100
Amp8
Amp2
multiplexer
Amp
Fig. 3. Schematic diagram of the instrument.
air pressure dispensers reaction chamber
computer small DNA analyzer
Fig. 4. Photograph of a small DNA analyzer.
DNA Analysis with a Photo-Diode Array Sensor
341
reagent reservoirs with dispensers
reaction chamber
Fig. 5. Photograph of the inside of the instrument.
reagent reservoir
air pressure
up
down dispensers reaction chamber reaction mixture injected dNTP
mixing by vibration
0.1s
0.2s
Dispenser position 2s Nucleotide pressure valve
0.2s 0.1s
Negative pressure valve 4.5s Reaction chamber vibration
115.5s
Fig. 6. Time chart for operating the system.
When one base extension reaction is carried out with 1 pmol of template DNA, one pmol of PPi is produced. If all of them are consumed for producing ATP and then luminescence, the total number of photons caused by the extension reaction is
342
Kambara and Zhou
about 4 × 1011 assuming that the quantum yield of luciferase reaction is 0.7. 2. Effect of ATP degradation by apyrase and ATP reproduction by a cycle reaction: The by-product of luciferase reaction, PPi, reacts with AMP and PEP again to produce ATP, and this causes a cycle reaction that can last until all AMP and PEP are consumed to produce ATP provided apyrase, which degrades ATP, does not exist in the reaction chamber. As ATP degradation and luciferase reactions are carried out simultaneously in one chamber, the average photon number caused by a nucleotide incorporation reaction is estimated to be as large as the number of incorporated nucleotides. 3. Estimation of photo-electron current detected with a photo-diode: Assuming that the luminescence lasts for 40 s, the photon flux emitted from the reaction chamber is estimated to be 1010/s. Assuming that 1% of luminescence is collected with the detector, the collected photo-flux is 108photons/s.(Generally, a large % of luminescence can be collected with a photo-detector.) The photoelectron current is determined by the characteristics of the detector device as shown in Fig. 7. The detection sensitivity of 0.3 A/W is obtained with a photodiode S1133 (from Hamamatsu Photonics, Shizuoka, Japan) for 560 nm light. As the energy of 560 nm light is 2.23 eV/photon, a photon flux of 108/s corresponds to an energy flux of 3.5 × 10−11 W, which gives a photo-current of 10−11A on a S1133 photo-diode. As
Si photodiode S1133
Package
ceramics
detection area
2.8X2.4 mm
detectable wavelength
320-730nm
most efficient wavelength (lp)
560nm
photosensitivity at lp
0.3A / W
dark current max.
0.01nA
Fig. 7. Photosensitivity curves of photo-diode S1133.
DNA Analysis with a Photo-Diode Array Sensor
343
the photo-current corresponds to a flux of 0.6 × 108 photoelectrons/s, the quantum yield is about 0.6. 4. Electric circuit for a photo-diode array: An example of the electric circuit used for S1133 is shown in Fig. 8. An output voltage is determined by the resistor Rf (1 GΩ) and the amplification factor of the buffer amplifier (×2, ×20, and ×200 scales can be selected). A signal of 2 V is obtained by receiving 10−11 A of photocurrent (×200 scale). This is easily detected with a photo-diode S1133. 5. Detection limit of the system: The detection limit is determined by the fluctuation of dark current of a photo-diode. A photodiode S1133 can detect 10−14A of photo-current at S/N = 3, which corresponds to 2 mV signals (×200 scale). It is possible to detect DNAs at a f-mole level with the photo-diode. The photo-diode is covered with a conductive film coated with ITO (Indium Tin Oxide) to reduce the electric noise. R1 C
R2
Rf
R3 − 5VA
− 5VA
C0
C0
Rs
_ PD1
+
_
A/D +
opA129UB
op07CS C0
C0 +5VA
+5VA
+5VA C0 C0
C − 5VA
Rf
1
8
Multiplexer Max4051ACSE
− 5VA C0
PD: photodiode S1130-1 opA129UB : Operational Amplifier op07CS : Operational Amplifier Rf: 1GΩ R1: 180kΩ R2: 18 kΩ R3: 1.8kΩ Rs: 1 kΩ C : 100pF C0: 100,000pF
_ PD8
+
opA129UB C0
+5VA
Multiplexer (MAX40551ACSE, Maxim lntegrated Products, Sunnyvale, CA) Operational Amplifier Operational Amplifier AD converter ( ADS1271PW, Texas Instruments, Dallas, TX)
Fig. 8. Electric circuit used for luminescence detection with S1133 photodiode.
344
Kambara and Zhou
6. Signal to noise ratio: In the typical pyrosequencing condition, a half picomole of template DNA is used. It gives a signal of about 0.5 V. The background luminescence due to a side reaction of the luciferase with the APS in the conventional condition is in a range of 0.5–1 V. The background becomes negligible in the new enzymatic reaction using PPDK and AMP. In the latter case, the background signal as well as the noise level is determined by the luminescence and its fluctuation due to the luciferase reaction of dNTP as well as residual ATP or PPi in the reagents. Usually it is in a range of 5–10 mV. Therefore the signal-to-noise ratio for sequencing 0.5 pmol of DNA in the latter condition is about 50 or more. 1.2.3. Reaction Chamber
A part of an immuno-plate (eight micro-wells of 350 μL) from NUNC (Nunc-Immuno™ BreakApart™ Modules; CAT No473539) is used as reaction chamber. A micro-well plate is fixed on a holder. The holder is attached with a small vibrator to mix reagents. The bottoms of the micro-wells are flat and placed 0.5 mm above the photodiodes. Usually the reaction chambers are filled with 30 μL or more of reaction solution for luminometric assays. If a reduced amount of reaction solution is required, handmade small chambers (as shown in Fig. 9) can be used.
1.2.4. Dispenser
A plastic dispenser chip that is made up of four reagent reservoirs with four dispensing nozzles (capillary tubes) attached to the reservoirs (Fig. 10) is used. The capacity of each reservoir is 40 μL. Usually 20 μL of reagent is supplied to each reservoir, which is enough for 50 such injections. The capillary tubes are 20 mm long and their inner diameter is 50 mm. The reservoirs contain four different dNTPs dissolved in buffer for pyrosequencing. Four different ddNTPs in buffer are used instead of dNTPs for SNPs
Reaction chambers (NUNC)
Reaction chambers in a holder
Small reaction chambers (homemade type)
Fig. 9. Photograph of reaction chambers. A part of micro-titer plate from NUNC is used routinely.
DNA Analysis with a Photo-Diode Array Sensor
345
capillary nozzle
reagent reservoir
Fig. 10. Photograph of a dispenser with four reservoirs each containing one of four nucleotide species. Each nucleotide is injected into a reaction chamber from the capillary by applying air pressure on the reservoir.
detection. Dispenser chips are filled with dNTPs or ddNTPs just before the experiment to avoid the decomposition of the reagents.
2. Material 2.1. Material for PCR Amplification
1. High Fidelity PCR Buffer Invitorogen, (Carlsbad, CA), Cat. No. 11304-011. 2. dNTP mixture (10 mM), Invitorogen, (Carlsbad, CA), Cat. No. 18427-013. 3. MgSO4 (50 mM), Invitorogen, (Carlsbad, CA), Cat. No. 11304-011. 4. Autoclaved, distilled water, Invitorogen, (Carlsbad, CA), Cat. No. 10977-023. 5. PCR Primers, Sigma Genosys (Hokkaido, Japan). 6. Platinum Taq High Fidelity, AB: Applied Biosystems (Foster City, CA) Cat. No.11304-01.
2.2. Reagents for Pyrosequencing
1. PPDK-E, Kikkoman (Chiba, Japan). 2. Luciferase(LUC-H), Kikkoman (Chiba, Japan). 3. Apyrase, grade VI, Sigma (ST. Louis, MO). 4. Luciferin, Kikkoman (Chiba, Japan).
346
Kambara and Zhou
5. PEP, Kikkoman (Chiba, Japan). 6. BSA, Sigma (St Louis, MO). 7. DTT, Sigma (St Louis, MO). 8. Nucleotides Sp-dATP-α-S Biolog Life Science Institute (Bremen, Germany). 9. dTTP, dCTP, and dGTP, GE Healthcare (Piscataway, NJ). 10. Tris (hydroxymethyl) aminomethane, GIBCO Industries (Langley, OK). 11. EDTA and Mg(Ac)2, Sigma (St. Louis, MO). 2.3. Preparation of PCR Reaction Reagent
The components of the PCR reaction mixture are listed in Table 1. 1. The primer sequences used to amplify the 181-bp fragments of TPMT gene from human genomic DNA (NCBI accession No AB045146) are as follows. Forward primer: 5´-TGTTGAAGTACCAGCATGCAC-3´ Reverse primers: 5´-biotin-AAATTACTTACCATTTGCGATCA-3´.
Table 1 PCR reaction mixture Volume (mL)
Final concentration
Supplier
Cat. No.
Store at
10× High Fidelity PCR Buffer
5
1×
I
11304–011
−20°C
dNTP mixture (10 mM)
1
0.2 mM each
I
18427–013
−20°C
MgSO4 (50 mM)
2
2 mM
I
11304–011
−20°C
Primer forward, 5´-biotin* (original concentration: 10 μM)
1
0.2 μM
SG
–
−20°C
Primer reverse (original concentration:10 μM)
1
0.2 μM
SG
–
−20°C
Template DNA 50 ng/μL (extracted from genomic DNA)
1
1 ng/μL
–
–
4°C
Platinum Taq high fidelity
0.2
1.0 U
AB
11304–011
−20°C
Autoclaved, distilled water
39
I
10977–023
Room temperature
Components
I Invitorogen(Carlsbad, CA); SG Sigma Genosys (Hokkaido, Japan); AB Applied Biosystems (Foster City, CA)
DNA Analysis with a Photo-Diode Array Sensor
347
2. Buffer solution for primers: 10 mM Tris–HCl (pH 7.5). 3. Buffer solution for template DNA: TE buffer (10 mM Tris– HCl (pH 7.5), 1 mM EDTA). 2.4. Reagents Used for Preparing SingleStranded DNA Template
1. Beads for magnetic separation: Dynabeads® M-280 Streptavidin (2.8 μm ID) (Dynal A.C., Oslo, Norway).
2.5. Reagents and Buffer Preparation for Pyrosequencing
The reagents used in pyrosequencing are listed in Table 2 together with the suppliers. 1. Preparation of buffer for luminometric assay: A half liter of 2 × Buffer solution (pH 7.75) is prepared by mixing 200 mM Tris-HCl, 4 mM EDTA and 40 mM Mg(Ac)2 in ultra pure water. The solution is stored at 4°C overnight followed by filtering with a filter (450 Filter Units-500 mL Capacity, MF75-series; Nalgene labware, Cat No 450-0020). The filtered solution is fractioned in falcon tubes (15 mL each) to be stored at 4°C. This is to reduce the risk of contamination.
2. Binding & Wash buffer: 10 mM Tris–HCl (pH 7.5), 1 mM EDTA, 2 M NaCl.
2. Preparation of reaction mixture for luminometric assay with PPDK-AMP system: The reaction mixture for luminometric assay and pyrosequencing consists of the reagents listed in Table 2. The reaction mixture is prepared in a 2 mL tube at first. Then
Table 2 Luminometric assay mixture (2 mL) Components
Volume (µL)
Final concentration
Original concentration
Supplier (state)
2 × Buffer
1,000
–
–
–
Ultra pure water
385.6
–
–
G
PPDK-E
62.8
33.8 U/mL
1076 U/mL
K (liquid)
Luciferase (LUC-H)
245.6
523.0 GLU/mL
4258.1 GLU/ mL
K (liquid)
Apyrase, grade VI
36
1.8 U/mL
100 U/mLl
S (solid)
Luciferin
200
0.4 mM
4 mM
K (liquid)
PEP
2
0.04 mM
40 mM
K (liquid)
AMP
8
0.2 mM
50 mM
S (solid)
BSA
20
0.10%
10%
S (liquid)
DTT
40
2 mM
100 mM
S (solid)
K Kikkoman (Chiba, Japan); S Sigma (St Louis, MO); G GIBCO Industries(Langley, OK)
348
Kambara and Zhou
it is subdivided into four 500-μL tubes to store at −20°C. For preventing the decomposition of the reagents, repetitive freezing and thawing should be avoided. Therefore, the reaction mixture should be divided into small bottles. The frozen reagent should be kept at room temperature for 30 min before use. 3. Preparation of nucleotide reagents: The concentrations of dNTPs are listed in Table 3. Each dNTP is dissolved in the buffer (pH 7.75) containing the components in Table 4. Each dNTP is stored at −20°C. As dNTP solution frequently contains PPi which produces a background signal, inorganic pyrophosphatase (PPase; USB Corp. (Cleaveland, OH) ) is added to the dNTP solution, which is incubated at room temperature for 15 min. The concentration of PPase in the dNTP solution is 20 mU/μL and 3.2 U of PPase should be added to 160 μL of each dNTP solution. The enzymatic reaction is stopped by placing the dNTP tube on ice. It is recommended to use the dNTP solutions within 24 h. 4. Primer: The primer (sequence: 5´-TGTTGAAGTACCAGCATGCAC-3´) of concentration 10 μM is produced by Sigma Genosys (Hokkaido, Japan). It is stored at −20°C before use. 5. Template DNA: The fragment of TPMT gene (181 bp) from human genomic DNA is used. 6. DNA polymerase: The original concentration of DNA polymerase I, Klenow Fragment, Exo-cloned from Ambion (Austin, TX) is 5 U/μL. It is stored at −20°C before use. In the case of 8-chamber pyrosequencing, the total amount of reaction mixture for analyzing eight samples in parallel is 240 μL (30 μL × 8). Add 1.5 μL of DNA polymerase to 240 μL of luminometric assay reaction mixture. 7. Preparation of terminator (ddNTPs) reagents: Four ddNTPs are from GE Healthcare. Each of them is dissolved in the buffer listed in Table 4 with the concentration of 125 μM.
Table 3 Preparation of dNTPs Nucleotides
Supplier (state)
Final concentration (µM)
Sp-dATP-α-S
BLSI (liquid)
250
dCTP
GEH (solid)
125
dGTP
GEH (solid)
125
dTTP
GEH (solid)
125
BLSI Biolog Life Science Institute (Bremen, Germany); GEH GE Healthcare (Piscataway, NJ)
DNA Analysis with a Photo-Diode Array Sensor
349
Table 4 Buffer composition Components
Supplier (state)
Concentration (mM)
Tris (hydroxymethyl) aminomethane
G (solid)
100
EDTA
W (solid)
0.5
Mg(Ac)2
S (solid)
5
DTT
S (solid)
1
G GIBCO Industries (Langley, OK); W WAKO Pure Chemical Industries (Osaka, Japan); S SIGMA (St.Louis, MO)
3. Methods 3.1. PCR Amplification and Amplicon Purification
The principle of PCR is shown in Fig. 11. The reverse primer is labeled with biotin at 5′-end for separating single-stranded template DNA. 1. Take 39 μL of autoclaved distilled water into a PCR tube (0.2 μL, Applied Biosystems Cat. No 8010540). 2. Add 5 μL of 10× High Fidelity PCR Buffer to the solution. 3. Add 2 μL of MgSO4 (50 mM) to the solution. 4. Add 1 μL of dNTP mixture (10 mM) to the solution. 5. Add 1 μL of forward primer and 1 μL of reverse primer (5′-biotin) to the solution. 6. Add 1 μL of template DNA (50 ng/μL: extracted from genome DNA) into the solution. 7. Add 0.2 μL of Platinum Taq High Fidelity 8. Mix the reagent gently by pipetting the solution several times. 9. Set the tube on a Gene Amp PCR System 9700 (Applied Biosystems, Foster City, Ca). 10. Denature the reagent at 94°C for 2 min and follow that with 40 cycle reactions (94°C for 15 s; 55°C for 30 s; 70°C for 60 s). 11. Remove the tube and place it on ice. 12. Take 1 μL of the product for confirming the length of the product by the 2100 Bioanalyzer (Agilent Technologies, Santa Clara, Ca) (This step can be skipped.).
350
Kambara and Zhou 3’ template DNA 5’
5’ 3’
denaturing and primer annealing 5’ reverse primer 3’
3’ forward primer 5’
1st cycle
extension 5’ 3’
3’ 5’
denaturing and primer annealing 5’
3’
3’
5’
2nd cycle
extension 5’
3’
3’
5’
~ 40cycles 5’
3’
amplified DNA fragments
3’
5’
Fig. 11. Outline of the PCR amplification process.
13. Purify PCR products with QIAquick PCR Purification Kit (Qiagen, Cat. No. 28104) according to the instructions attached to the product. Next, dissolve the purified product in 25 μL of TE buffer. 3.2. The Preparation of Single-Stranded Template DNA
Figure 12 is a schematic of the single-stranded DNA production process. 1. Take 25 μL of Dynabeads (from Dynabeads® M-280, Cat 112-062, 10 mL) solution into a 1.5 mL tube. 2. Separate the Dynabeads from the original buffer with a magnet (Dynal, Cat No 120.20). 3. Add 50 μL of binding & wash buffer (10 mM Tris–HCl (pH7.5), 1 mM EDTA, 2 M NaCl into the solution. Wash the beads dissolved in the solution by vibrating the tube with a Voltex mixer. After washing the beads, remove the buffer solution. Repeat the washing process three times.
DNA Analysis with a Photo-Diode Array Sensor
double stranded DNA
biotin
351
bead streptavidin
+
Dynabead M-280 PCR solution
immobilization DNA on the bead
alkaline denaturation 0.1N NaOH
separation bead with a magnet
magnet
neutralization
washing
binding / wash buffer
single stranded DNA
DNA samples immobilized on bead
DNA samples in a solution
Fig. 12. Outline of the DNA purification process.
4. Add 25 μL of PCR product together with 25 μL of binding & wash buffer to the tube and mix with the beads by stirring at room temperature for 30 min (PCR product, double-stranded DNA, is immobilized onto the bead surfaces.). 5. Separate the beads from the solution with a magnet. Add 50 μL of NaOH (0.1 M) to the beads and keep the mixture at room temperature for 5 min. The complementary DNA strands are separated in the solution and the target DNA stays on the beads. 6. Separate the beads from the solution with a magnet and the target DNA is on the bead surface while the complementary DNA strands are in the solution. 7. Neutralize the solution containing the complementary DNA strands to pH7.5–8.0 by adding 50 μL of 0.1 M HCl to obtain free template DNA in 100 μL of solution.
352
Kambara and Zhou
8. Wash the beads in the tube using the same procedure described in step 3. Then wash them with 125 μL of ultra pure water three times. 9. After removing the water, add 50 μL of binding & wash buffer and vibrate the tube with a Vortex mixer. 10. Now you have single-stranded DNA samples immobilized on bead surfaces as well as in a solution. 3.3. DNA Sequencing by Pyrosequencing
Figures 4 and 5 are photographs of a small DNA analyzer based on luminometric assay. The measurement is carried out according to the following process. 1. Open the cover of the instrument. 2. Set eight reagent reservoirs in the upper holders. 3. Put reaction chambers in the reaction chamber holders. 4. Close the cover of the instrument. 5. Turn on the switch of the instrument and the computer. 6. Various operations are carried out according to the time chart as shown in Fig. 6. (a) Move down the reagent reservoirs with four capillary nozzles to dip into the reaction solutions. (b) Inject one nucleotide species at a time from a nozzle by applying a small pressure onto one of the four reservoirs for 0.2 s. The injection of four nucleotide species is carried out by turns. The amount of reagent dispensed per shot is 0.4 μL. Then the reservoirs with capillary nozzles are lifted up to be placed in the air. As the four capillary nozzles are dipped in the reaction solution together, small amounts of reagents besides the injected reagent species are frequently leaked out from the nozzles. To prevent the leakage of reagents, a small negative pressure is applied to the reservoirs to make air gaps in the capillary nozzles which are useful to prevent the undesirable leakage of reagent during the injection. (c) The reaction chamber holder is vibrated with a small motor (Tsukasa Electric Co. (Tokyo), TG-87A-GU) during the measurement after dispensing dNTP into the solution. (d) The detection of photo signals starts with the injection of dNTP and lasts for 120 s. The detection is carried out every 0.2 s. Then the measurement cycle is repeated 100 times for sequencing about 50 bases. 7. All the data obtained in a cycle are transferred to a computer at a time.
DNA Analysis with a Photo-Diode Array Sensor
353
8. A set of data per cycle contains 600 data points which make one peak in a pyrogram. The 600 data points are summed up to give the signal strength per nucleotide incorporation reaction which is used for sequence analysis. The time chart of the operation is shown in Fig. 6. The interval of every injection is 120 s, which is selected so that the degradation of dNTP by Apyrase is completed before the next dNTP injection. The main steps needed for measurements: 1. Turn on the main switch and confirm that the background signal is less than 0.005 V. The background signal may be caused by stray lights or an induction current that can be removed by shielding the photo-diodes with a conductive plastic film. 2. Take 30 μL of the pyrosequencing mixture into a reaction chamber. 3. Put the reaction chamber onto the instrument to check the background signal again. If the background signal is less than 0.01 V, you can proceed. If you have a large background signal, your sample may be contaminated with PPi or ATP. If so, wait for a while until apyrase degrades most of the ATP. It is also recommended to add a small amount of PPase into the reaction mixture prior to the addition of APS for degrading any residual PPi (When the ATP sulfurylase-APS system is used to produce ATP from PPi, the background due to APS becomes around 0.5 V due to the APS. (see Notes 1-3). 4. After confirming that the background signals are negligibly small, add 1 μL of DNA polymerase to the mixture in a reaction chamber. It is recommended to add a ddNTP mixture containing four ddNTPs of 3 pmol each to terminate the strand extension of contaminated DNA in the mixture that may be introduced in the preparation process. Keep the mixture at 30°C for 2 min so that the added ddNTPs are degraded by apyrase. 5. Take 1 μL of primer and 5 μL of template DNA into one tube (0.5 mL). Put the tube in a water bath to increase the temperature to 96°C for 5 s and cool it down to room temperature for primer hybridization to the template DNA. 6. Take 1 μL of the template DNA hybridized with the primer into the pyrosequencing mixture in a reaction chamber. (Now all the extendable DNA termini except for the primers are terminated and signals due to the primer extension are obtained.) 7. Put dispensers, filled with 20 μL of dNTPs in each of the four reservoirs, on the instrument. The dispenser units move down to dispense reagents. The capillary nozzle ends of the dispensers are placed 0.5 mm below the reaction mixture surfaces in reaction chambers while dNTPs are being dispensed (see Notes 1-3).
354
Kambara and Zhou
8. Start the sequencing measurement. At first, the dispenser unit moves down to dip the capillary nozzles into the reaction solution and then air pressure is applied to the dispensers for 0.2 s to inject 0.4 μL of dNTP into the reaction chambers. The reaction chambers are vibrated with a vibrating motor attached to the chamber holder for 30 s after the injection. Luminescence measurement starts from 5 s after the injection. 9. The order of dNTP injection is dATPαS, dTTP, dGTP, and dCTP. Observed signals together with the injected nucleotide species from a program are shown in Fig. 13. The DNA sequence can be determined by observing the signal for each nucleotide injection. Although no signal appears for the injections of dATP, dTTP, and dGTP, a signal obtained by injecting dCTP can be observed in the figure. This indicates the addition of base “C” to the complementary DNA strand. Here, for convenience, we call the signal intensity corresponding to one base extension a signal-unit here for the convenience. As two “C”s were incorporated, the signal intensity was two signal-units in Fig. 13. By injecting dTTP, a signal as large as one signal-unit appeared, and this indicated that the extended DNA strand had the sequence of 5´-primer-CCT-3´. Then, we observed a photo-signal as large as two signal-units that appeared after dGTP injection indicating that two G were added to the complementary DNA strand. The sequence was determined to be 5´-primer-CCTGGATT-3´. The next signal for G was a half signal-unit and the following “A” signal was about one and a half signal-units indicating the sample included a hetero-sequence of G/A at the indicated region. The sequence was determined as 5´-primer-CCTGGATT(G/A)ATGGCAACT-3´. As described here, it becomes frequently difficult to accurately determine the sequence of genomic samples when SNPs especially heterozygous sequences appear because it requires accurate and quantitative signal intensities for sequencing analysis. 3.4. SNPs Detection by One Base Extension with Ddntps
The procedure for SNP typing with ddNTPs is basically the same as that for pyrosequencing except that four terminators (ddNTPs) are used instead of dNTPs for nucleotide incorporation reactions (7). Each terminator is injected once. Obtaining pyrograms for hetero-samples is complex as has been demonstrated previously, however, the new method gives a simple spectra even when a sample has a plural of SNP sites. A schematic of the method is shown in Fig. 12a. A primer is constructed to hybridize to the target DNA as the nucleotide incorporation occurs at the SNPs site. Nucleotide incorporation reactions are conducted by injecting four ddNTPs one at a time. Only one base extension occurs when the injected ddNTP
Signal intensity (V)
0
CC
T
GG
500
A
TT
G
0.5
A
1.5
T C
1000
GG AA C T
TGC
1500
AA T
2000
T
C
Time (sec)
CC T
2500
A T
3000
T
CCC AAA
C A
3500
T G
T
C
4000
AAA
Fig. 13. Pyrogram obtained with a fragment of human TPMT gene. The dNTP dispensing order is A → T → G → C → A. The base species on the peaks indicate the sequence of DNA complementary to tvhe template.
0.0
0.4
0.8
1.2
DNA Analysis with a Photo-Diode Array Sensor 355
356
Kambara and Zhou
is complementary to the target DNA, which makes the spectrum much simpler than that for pyrosequencing. Obtaining a spectrum from one-base-extension reaction with ddNTPs is simpler than that with dNTPs as is shown in Fig. 14. As the use of ddNTPs makes spectral patterns very simple, plural SNPs can be determined simultaneously. Figure 15 shows the nine possible spectral patterns for dual SNPs. The dual SNPs can be determined by observing the spectral pattern. Even triple SNPs can be determined simultaneously with the method. The new method uses ddATP although dATPαS is usually used in pyrosequencing to reduce background luminescence. Although dATP can be a substrate for a luciferase reaction, ddATP is not a substrate for a luciferase reaction, but is a good substrate for nucleotide incorporation reaction. This is an advantage of the new SNP typing method.
AAATC
+
GAATC
ddCTP AAATC
+
AAATC
+
dCTP C
AAATC
+
C GAATC
T
TTT AAATC
C GAATC
ddATP
No reaction
T
+
CTT GAATC
C
dTTP decomposition C
T
dATPαS
T A
No reaction
ddGTP
C
dTTP C
+
C GAATC
dCTP decomposition
ddTTP T AAATC
GAATC
TTTA AAATC
+
CTTA GAATC
A C
dATP αS decomposition T C
dGTP
T A G
TTTAG AAATC
+
A
CTTAG GAATC
G
C
dGTP decomposition
(a)
(b)
Fig. 14. Principle of SNP detection by ddNTP addition (a) and pyrosequencing (b): Four ddNTPs or dNTPs are injected into a reaction chamber in turn. Signals appear only when the nucleotide incorporations occur at the SNP site in case (a), which makes the spectrum simple. However, a pyrogram for a SNP sample becomes complex because nucleotide incorporation reactions can occur at a position other than the SNP site.
DNA Analysis with a Photo-Diode Array Sensor 4
4
(AA/TT)
3
4
(AC/TT)
3 2
2
1
1
1
0 A
T
G
C
4
0 A
T
G
C
4
(AA/TC)
3
A
(AC/TC)
3
2
1
1
1
T
G
C
4
0
A
T
G
C
4
(AA/CC)
3
0
(AC/CC)
2
2
1
1
1
0 T
G
C
A
T
G
C
(CC/CC)
3
2
A
C
4
3
0
G
(CC/TC)
3
2
A
T
4
2
0
(CC/TT)
3
2
0
357
0 A
T
G
C
A
T
G
C
Fig. 15. Spectral patterns for two SNPs obtained with ddNTP addition. As the patterns are limited to nine for two SNP sites, two SNPs are easily determined by observing the pattern.
4. Notes 1. ATP production system using ATP sulfurylase and APS: The pyrosequencing reaction mixture used for ATP sulfurylaseAPS system is shown in Table 5. As APS is a substrate of luciferase reaction, the background signal due to APS is as large as 0.5 V. The background signal decreases with time because APS is consumed for the reactions. 2. The four nozzles of a dispensing unit are dipped into the reaction solution at a time. The pressure is applied on one of the nozzles (working nozzle) to inject the corresponding dNTP. The other three nozzles (resting nozzles) are also dipped in the solution, however, the leakage from the other dNTPs has to be avoided. To prevent dNTP leakage from non-working nozzles, an air gap is formed in the nozzles by applying a small negative pressure on all the nozzles when they are at the waiting position in the air. 3. Phase shift in pyrosequencing.
358
Kambara and Zhou
Table 5 Luminometric assay mixture based on APS and ATP sulfurylase (2 mL) Components
Volume (µL)
Final concentration
Original concentration
Supplier (state)
2 × Buffer
1,000
–
–
–
Ultra pure water
338
–
–
G
ATP sulfurylase
20
200 mU/mL
0.02 U/μL
S (solid)
Luciferase (LUC-H)
246
523 GLU/mL
4,258 GLU/mL
K (liquid)
Apyrase, Grade VI
36
1.8 U/mL
100 U/mL
S (solid)
Luciferin
200
0.4 mM
4 mM
K (liquid)
Adenosine 5′ phosphosulfate (APS)
100
5 μM
0.1 mM
S (solid)
Bovine serum albumin (BSA)
20
0.10%
10%
S (liquid)
DTT
40
2 mM
100 mM
S (solid)
K Kikkoman (Chiba, Japan); S Sigma (St Louis, MO); G GIBCO Industries (Langley, OK)
In pyrosequencing, the uniformity of signals is important for accurate base reading. The signal intensity is generally proportional to the number of extended DNA strands and influenced by the reaction yield for nucleotide incorporation reactions. Sometimes the nucleotide incorporation reactions do not occur uniformly in a reaction chamber which causes positive and negative phase shifts in the reaction (6, 8). A positive phase shift occurs when the degradation of injected nucleotide species has not been completed before the injection of the next nucleotide species and two nucleotide species are incorporated at the same time for some DNA strands. This produces DNA strands in different lengths and in advanced reaction steps. The minus phase shift occurs when the incorporation reactions for all DNA strands are not completed because the degradation of nucleotides is too fast or the mixing of the injected nucleotides in a chamber is insufficient. Then the not-extended DNA strands coexist together with the extended DNA strands in a reaction chamber. Examples of the phase shifts are shown in Fig. 16. The accumulation of both phase shifts prevents an accurate base calling because signals are always observed for any nucleotide species injection. DNA of up to 40 bases can be rather easily sequenced and the readable length can be extended to more than 70 bases by optimizing the conditions.
DNA Analysis with a Photo-Diode Array Sensor dNTP injection order : A
Sequence of complementary DNA strand : AGTTGTCCTA
A
A
T
A
G
AG
C
AG
A
AG A
G C
G
C A
T
100
0.0
G
C
400
A
0
T
G C
100
A
T
G
200
C
A
3.3mU/μl (b)
300
T
G
C
400
500
T
G
G
A
2.2mU/μl (c)
c
0.0
A
0
T
G
C A
100
T
G
200
1.2
C
A
300
T
G
C
400
500
TT
0.8
A
0.4
G
A
T G
1.1mU/μl (d)
T
G
0
C
A
200
T
G
C
400
1.6
600
CC
c
0.0
600
CC
TT
0.3
500
T
G
0.6
AGTTG AGTTGT AGTTG AGTTGTCC
T
300
CC
G
A
0.2
AGTT
G C A
200 TT
0.4
relative intensity
T
A T
0
0.6
C A
6.7mU/μl (a)
g
A
0.0
insufficient dNTP degradation
T
G
G
A
0.1
G
CC
TT
0.2
imperfect strand extension
T
359
A
T
600
G C
800
1000
TT A
0.8 0.0
A
0
G
G T
G
200
C
A
400
T
G
C C
600
CC
T A
800
T
G
0.6mU/μl (e)
C
1000 1200
time, s
Fig. 16. Pyrograms showing plus and minus phase shift. (a) apyrase:6.7 mU/µL, (b) apyrase:3.3 mU/µL, (c) apyrase:2.2 mU/µL, (d) apyrase:1.1 mU/µL, (e) apyrase:0.6 mU/µL. The spectral patterns change with the amount of apyrase in a reaction chamber. When the amount of injected dNTP is not sufficient or the apyrase amount is large, dNTP is degraded before the DNA strand extension is completed. The extended DNA strands together with non extended DNA strands are produced in the chamber. The inhomogeneous reaction produces minus phase shift in a spectrum as shown in (a). When the amount of apyrase is very small, insufficient dNTP degradation occurs where injected dNTP stays for a long period in the chamber. They are incorporated together with the newly injected dNTP species and positive phase shift is thus produced as shown in (e).
Acknowledgments The author would like to thank Dr Tomoharu Kajiyama and Dr Akihiko Kishimoto for the development of technologies and instruments, and Ms Sumiyo Takiguchi and Ms Mari Goto for their help in experiments.
References 1. Kheterpal, I. and Mathies, R.A. (1999) Capillary array electrophoresis DNA sequencing. Anal Chem, 71, 31A–37A 2. Zubritsky, E. (2002) How analytical chemists saved the human genome project. Anal Chem, 74, 23A–26A 3. Kambara, H. and Takahashi, S. (1993) Multiple-sheathflow capillary array DNA analyser. Nature, 361, 565–566
4. Ronaghi, M., Uhlen, M. and Nyren, P. (1998) A sequencing method based on real-time pyrophosphate. Science, 281, 363–365 5. Zhou, G.-H., Kamahori, M., Okano, K., Harada, K. and Kambara, H. (2001) Miniaturized pyrosequencer for DNA analysis with capillaries to deliver deoxynucleotides. Electrophoresis, 22, 3497–3504 6. Zhou, G.-H., Kajiyama, T., Gotou, M., Kishimoto, A., Suzuki, S. and Kambara, H. (2006)
360
Kambara and Zhou
Enzyme system for improving the detection limit in pyrosequencing. Anal Chem, 78, 4482–4489 7. Zhou, G.-H., Gotou, M., Kajiyama, T. and Kambara, H. (2005) Multiplex SNP typing by bioluminometric assay coupled with terminator incorporation (BATI). Nucleic Acids Res, 33, e133
8. Gharizadeh, B., Nordstrom, T., Ahmadian, A., Ronaghi, M. and Nyren, P. (2002) Long-read pyrosequencing using pure 2´-deoxyadenosine-5´-O’- (1-thiotriphosphate) Sp-isomer. Anal Biochem, 301, 82–90
Chapter 20 Miniaturized and Integrated Fluorescence Detectors for Microfluidic Capillary Electrophoresis Devices Toshihiro Kamei Summary Microfluidic devices are revolutionary in their ability to use very small quantities of liquid samples and to perform biochemical analyses with unprecedented speed. Toward the goal of a lab-on-a-chip that integrates a series of analysis steps and analytical components into a single microfluidic device, one of the most critical aspects of size reduction is miniaturizing and integrating the fluorescence detection system. We present here details of a new integrated fluorescence detection system. A microfluidic biochemical analysis device is mounted on a compact detection platform that comprises a fluorescence-collecting microlens and micromachined fluorescence detector in which a multilayer optical interference filter is monolithically integrated and patterned on a hydrogenated amorphous silicon (a-Si:H) photodiode. A central aperture in the micromachined a-Si:H fluorescence detector allows semiconductor laser light to pass up through the detector and to irradiate a microchannel of the microfluidic analysis device. Such an optical configuration enables a detachable, reusable, compact module to be constructed for the excitation source and detector. The micromachined a-Si:H fluorescence detector exhibits high sensitivity for practical fluorescent labeling dye, making it ideal for application to portable point-of-care microfluidic biochemical analysis devices. Key words: Microfluidic, Lab-on-a-chip, Hydrogenated amorphous silicon, Photodiode, Fluorescence, Electrophoresis.
1. Introduction The manipulation of a minute quantity of fluid (pl–nl) in a microchannel, termed microfluidics, has emerged in the last decade as an interdisciplinary field between molecular biology and electronics. Similar to the scaling law for a metal-oxide semiconductor field-effect transistor (MOSFET) in an integrated circuit that, as transistors gets smaller, they can switch faster Avraham Rasooly and Keith E. Herold (eds.), Methods in Molecular Biology: Biosensors and Biodetection, Vol. 503 © Humana Press, a part of Springer Science + Business Media, LLC 2009 DOI: 10.1007/978-1-60327-567-5_20
361
362
Kamei
and use less power, the polymerase chain reaction (PCR) in a nanolitter reactor has been dramatically speeded up from the conventional microlitter scale. In addition, an unprecedentedly small sample plug of approximately 100 μm that can be electrokinetically formed in a microchannel network has achieved capillary electrophoresis (CE) with high-speed and high-separation efficiency (1). Microfluidic plumbing technology based on a silicone elastomeric membrane has been used to perform cell sorting and combinatorial screening of protein crystallization conditions (2–4). Despite this rapid progress in microfluidic biochemical assays, a high-sensitivity microfluidic analysis still typically requires a bulky fluorescence detection system comprising a laser, optical device, and detector. In order to realize the potential portability of a microfluidic analysis system, the fluorescence detection system must also be miniaturized and integrated. Hydrogenated amorphous silicon (a-Si:H) is an ideal choice of material for integrating a fluorescence detector for a variety of reasons (a) it exhibits high sensitivity at the emission wavelength of most practical labeling dyes such as green fluorescence protein, DNA intercalators, ethidium bromide, and fluorescein; (b) it exhibits a low-dark current suited for low-noise measurement; (c) it can be monolithically integrated on a laser diode due to its disordered structure and low-temperature fabrication process; and (d) its manufacture is inexpensive (5). Although an avalanche Si photodiode (APD) shows much higher sensitivity, the limit of detection (LOD) of an integrated fluorescence detector is determined by the efficacy of the integrated optics rather than by the detector sensitivity itself; there has been no need to use a higher sensitivity detector than a photodiode at this point. Our systematic study has shown that an a-Si:H PIN photodiode exhibited LOD of 99.9%) for bottom electrode sputtering. 4. A solution (stripper 106, TOK, Japan) for stripping photoresist. 5. A solution containing 6–16% of cerium ammonium nitrate and 4–11% of nitric acid for Cr etching. 6. A spin coater (e.g., Mikasa 1H-DX2) to spin photoresist on a wafer. We typically use OFPR-8600LB 33cp photoresist (TOK, Japan), except for the Al lift-off process for which SIPR-96843.0 photoresist (Shinetsu Chemical Co., Japan) is used. We use a stepper (e.g., Ultratech) for wafer exposure, but a mask aligner (e.g., MA6, Suss Microtech) works as well. 7. High-purity SiH4 gas (>99.999%) for a-Si:H deposition. 8. A high-purity and high-density indium tin oxide (ITO) target (purity >99.99%; relative density >95%) for top electrode sputtering. 9. A high-purity Al target (>99.999%) for Al electrode sputtering.
2.2. Integration and Patterning of the Optical Interference Filter
1. A mixed solution of phosphoric acid (78.9%), nitric acid (2.8%), and acetic acid (3%) for Al etching.
2.3. Integration Detection Platform
1. A 2-mm diameter half-ball lens made of BK7 (Edmund Optics, USA) for the fluorescence-collecting microlens.
2. The SiO2/Ta2O5 multilayer filter is designed and deposited by such companies as Optoquest Co. (Japan) and Barr associates (USA), using ion-assisted deposition (IAD) or ion beam sputtering (IBS).
2. A semiconductor laser emitting light at 488 nm (e.g., Sapphire, Coherent, USA). 3. Black-anodize an Al platform to suppress light reflection on the surface. 2.4. Microfluidic Electrophoresis Device Fabrication
1. Borofloat glass wafers (76-mm diameter, 1.1-mm thick, Schott, NY) for microfluidic electrophoresis devices. 2. A mask aligner (e.g., MA6, Suss Microtech) for a wafer exposure. 3. HF solution (50%, Semiconductor grade, Morita Chemical, Japan) for channel etching of the glass wafer.
364
Kamei
4. Piranha is heated mixture of 80% concentrated sulfuric acid and 20% hydrogen peroxide. 5. A furnace (KDF-90, Denken, Japan) for thermally boding glass wafers. 2.5. Microfluidic Electrophoresis Analysis
1. Microfluidic electrophoresis devices can be fabricated by such companies as Institute of Microchemical Technology (IMT, Japan) or Micronit Microfludics (Netherlands). 2. A high-voltage power supply designed for microfluidic electrophoresis (e.g., HVS448, LabSmith, USA).
3. Methods 3.1. a-Si:H Photodiode Fabrication
The details of fabrication procedures of a-Si:H PIN photodiode are described in this section. Brief fabrication steps are as follows: Cr deposition on a glass substrate/resist patterning/etching; deposition of nip a-Si:H and then ITO/resist patterning/etching; SiN deposition/resist patterning/etching; resist patterning/Al deposition/lift-off. An a-Si:H PIN photodiode typically exhibits excellent sensitivity to visible light and its quantum efficiency exceeds 80% at a wavelength between 500 and 600 nm. Dark current of an a-Si:H photodiode should also be very low: several pA or less at room temperature when its outer diameter is 2 mm. These features are suited for high sensitivity visible fluorescence detection: 1. These instructions assume the use of multichamber plasma-enhanced chemical vapor deposition apparatus (e.g., PD-2203LS, SAMCO, Japan) such that each layer of the a-Si:H PIN photodiode is deposited in a separate chamber, suppressing cross contamination by impurities such as B and P. 2. Rinse glass wafers with water (see Note 1), before ultrasonically cleaning for 5 min. Immerse them in acetone, before ultrasonically cleaning again for 10 min. Rinse them with water and immerse them in the cleaning solution (Semico Clean 56), before ultrasonically cleaning for a further 20 min. Thoroughly rinse them with water and dry. 3. Sputter chromium on a glass wafer to a thickness of approximately 200 nm. Pattern the photoresist on the Cr-deposited glass wafer by photolithography. Wet etching is performed by dipping it in a Cr etchant until the prescribed region becomes transparent. Remove the photoresist by immersing in the resist-stripping solution at 80°C, dip the wafer in isopropyl alcohol for 1 min, rinse thoroughly with water and dry.
Miniaturized and Integrated Fluorescence Detectors
365
4. Deposit P-doped a-Si:H (n layer) at a temperature of 250°C to a thickness of approximately 20 nm by plasma decomposition of a source gas mixture of SiH4 and PH3 (e.g., PH3/SiH4 = 1,000 ppm). 5. Deposit nondoped a-Si:H (i layer) at a temperature of 250°C to a thickness of approximately 500 nm by plasma decomposition of an SiH4 source gas. 6. Deposit B-doped a-Si:H (p layer) at a temperature of 250°C to a thickness of approximately 20 nm by plasma decomposition of a source gas mixture of SiH4 and B2H6 (e.g., B2H6/SiH4 = 5,000 ppm). 7. Sputter ITO to a thickness of approximately 70 nm. Pattern photoresist on the wafer by photolithography. Etch ITO with a mixture of CH4 and H2 (e.g., CH4/H2 = 1/5) and then a-Si:H with SF6 in reactive ion etching (RIE) apparatus. Remove the photoresist by immersing in the resist stripping solution at 80°C, dip the wafer in isopropyl alcohol for 1 min, rinse thoroughly with water and dry. 8. Deposit SiN by plasma decomposition of a mixture of SiH4, NH3, and N2 at 300°C. The thickness should be approximately 300 nm. Pattern photoresist on the wafer by photolithography. Etch with SF6 in RIE. Remove the photoresist by immersing in the resist stripping solution at 80°C, dip the wafer in isopropyl alcohol for 1 min, rinse thoroughly with water and dry. 9. Pattern photoresist on the wafer by photolithography. Then sputter aluminum on it, lifting off the unwanted Al by dipping it in the resist stripping solution at 80°C. Dip the wafer in isopropyl alcohol for 1 min, rinse thoroughly with water and dry. 3.2. Integration and Patterning of the Optical Interference Filter
The details of SiO2/Ta2O5 multilayer optical interference filter patterning are described in this section. Both materials such as SiO2 and Ta2O5 are difficult to etch by RIE, so it is virtually impossible to directly etch such a thick SiO2/Ta2O5 optical filter. We have also found no photoresist appropriate to lift off such a thick optical filter and to be tolerant to filter coating temperature (230°C). Therefore, a lift off process using an Al/Si bilayer as a sacrificial layer has been adopted to pattern the optical filter. 1. Sputter aluminum to a thickness of approximately 10 μm, and then deposit Si to a thickness of 200 nm. Pattern photoresist on the wafer by photolithography. Anisotropically etch the Si film by SF6 in RIE, before isotropically wet etching with Al to form an overhang structure. 2. Send the sample to the company to deposit the SiO2/Ta2O5 multilayer optical interference filter on it (see Note 2). The spectroscopic properties of the filter we typically use are shown in Fig. 1.
366
Kamei
3. Immerse the wafer in an aluminum etchant at approximately 60°C, lifting off the unwanted region of the optical interference filter (see Note 3). Rinse thoroughly with water and dry. 4. Deposit SiO2 by plasma decomposition of a mixture of tetraethoxysilane (TEOS) and O2 at 300°C. The thickness should be approximately 400 nm. 5. Pattern photoresist on the wafer by photolithography. Etch SiO2 layer to make contact with Al as well as the Cr electrode. A plan view of the micromachined a-Si:H fluorescence detector is shown in Fig. 2a together with a schematic illustration of its cross-sectional view in Fig. 2b. The integrated a-Si:H fluorescence detector that can be coupled with a microfluidic electrophoresis device is shown in Fig. 3, and a more extended system diagram is shown in Fig. 4. A half-ball lens (2-mm diameter) and the micromachined a-Si:H fluores-
(a) 100
T (%)
80 60 40 20 0
(b) 102 101
Log T (%)
3.3. Integrated Detection System
100 10−1 10−2 10−3 10−4 400
450
500
550
600
650
700
Wavelength (nm) Fig. 1. Simulated spectroscopic properties of the sio2/Ta2O5 multilayer optical interference filter we typically use. a Transmission (%) vs. wavelength (nm); b Log (transmission (%) ) vs. wavelength (nm).
Miniaturized and Integrated Fluorescence Detectors
367
(a) μFD
Cr
Al
Aperture
500μm
(b) SiO2 / Ta2O5 Optical filter
SiN
ITO p i a-Si:H n
μFD~7μm
SiO Al Glass substrate
Cr
Fig. 2 a. Plan view of the optical micrograph of a micromachined a-Si:H fluorescence detector (μFD). b Schematic cross-sectional view of the detector. (c) 1 mm Microchannel 2 mm-dia. Half-Ball Lens
Fluorescence
Semiconductor Laser (488 nm)
Fig. 3 a. A microfluidic electrophoresis device mounted on a black-anodized Al platform that comprises a half-ball lens and a micromachined a-Si:H fluorescence detector. b A micromachined fluorescence detector placed underneath the Al platform. c A schematic cross-sectional view of the microfluidic electrophoresis device, half-ball lens, and fluorescence detector. The excitation light from a semiconductor laser passes up through the fluorescence detector and half-ball lens to irradiate a microchannel of the microfluidic electrophoresis device.
368
Kamei
HV
Computer
microfluidic electrophoresis devices
micromachined fluorescence detector
LIA f
R
Θ
LASER Chopper
Lens
Mirror
Fig. 4. A system diagram of the integrated detection system for a microfluidic electrophoresis device. The excitation light passes through an optical chopper, lens, mirror, and a micromachined a-Si:H fluorescence detector, and is loosely focused in the microchannel of a microfluidic electrophoresis device. The photocurrent under a reverse-bias voltage of 1 V synchronized to the chopped laser light is detected by a lock-in amplifier (LIA). Data acquisition and high-voltage (HV) control for microfluidic separation are accomplished by a personal computer.
cence detector were assembled and supported by black-anodized aluminum, forming a compact platform for attaching to a microfluidic electrophoresis device. The annular micromachined a-Si:H fluorescence detector and transparent glass substrate allow vertical laser excitation while avoiding direct incidence of the excitation light on the detector. Fluorescence is collected by the half-ball lens and transmitted by the optical interference filter, which simultaneously eliminates the excitation light. Ray trace simulation (ZEMAX, Focus Software, CA, USA) indicated that a 2-mm-diameter half-ball lens would approximately collimate the fluorescence emitted from a microchannel located behind the 1-mm-thick Borofloat glass substrate. The procedure to construct the integrated detection system is as follows: 1. Mount the half-ball lens on the Al platform by using epoxy resin to make their surface flat. 2. Loosely focus laser light (approximately 30 μm) through a convex lens (NA ~0.01). The focal point may be close to the microchannel. 3. Adjust a mirror underneath the detector (see Fig. 4) so that laser light passes upward normal to the half-ball lens.
Miniaturized and Integrated Fluorescence Detectors
369
4. Place an iris above the center of the Al platform (half-ball lens) and then adjust the position of the Al platform (half-ball lens) by two orthogonal micrometers so that the laser light goes through the iris, making sure the laser light passes through the center of the half-ball lens. 5. Place and fix the micromachined fluorescence detector on the Al platform to coaxially align the detector with the half ball lens. 6. Apply a reverse-bias voltage of 1 V to the a-Si:H photodiode and connect its output to a lock-in amplifier (LIA; e.g., SR840, Stanford Research Systems, CA, USA). Direct the laser light through an optical chopper (e.g., SR540, Stanford Research Systems, CA, USA) and detect the photocurrent synchronized with the chopped laser light to reduce background interference due to environmental light. 7. Digitize and acquire an LIA output at 20 Hz by using LabView (e.g., NI-DAQ E6024, National Instruments, TX, USA). 3.4. Microfluidic Electrophoresis Device Fabrication
The procedure to fabricate microfluidic electrophoretic devices is described in this section. Plasma-deposited Si film is used as a hard mask for HF wet etching and glass wafers are thermally bonded. The mask design of the microfluidic CE, which we typically use is shown in Fig. 5. The width of microchannel in the mask is 20 μm. 1. Rinse glass wafers with water, before ultrasonically cleaning for 5 min. Immerse them in acetone, before ultrasonically cleaning again for 10 min. Rinse them with water and immerse them in the cleaning solution, before ultrasonically cleaning for a further 20 min. Thoroughly rinse them with water and dry. 2. Deposit a-Si:H at a temperature of 250°C to a thickness of approximately 250 nm by plasma decomposition of an SiH4 source gas. 3. Pattern the photoresist on the Si-deposited glass wafer by photolithography. 4. Etch a-Si:H with SF6 in RIE. Remove the photoresist by immersing in the resist stripping solution at 80°C, dip the wafer in isopropyl alcohol for 1 min, rinse thoroughly with water, and dry. 5. Dip the glass wafer in HF solution for 7 min, producing a 120-μm wide, 50-μm deep microchannel. Rinse thoroughly with water and dry. 6. Drill access holes with diamond bits for sample, waste, cathode, and anode reservoirs. Immerse the drilled wafers in acetone, before ultrasonically cleaning for 15 min. Thoroughly rinse them with water and dry.
370
Kamei
5
4
3
1
2
S C
A
W
5
4
5
4
3 3
1
2 2
1
2
3
4
5
1
1
2
3
4
5
1
2
3
4
5
Fig. 5. Mask design of the microfluidic electrophoresis device, which we typically use. S, W, C, and A, respectively, stand for the sample, waste, cathode, and anode reservoirs.
7. Strip off both sides of a-Si:H layers with SF6 in RIE. 8. Immerse them in the cleaning solution (Semico Clean 56), before ultrasonically cleaning for 15 min. Thoroughly rinse them with water. 9. Immerse them in piranha at 120°C for 15 min. Thoroughly rinse them with water and dry. 10. Align the drilled and etched wafer with a blank wafer by hand. Sandwich them by macor blocks and insert them into a furnace. 11. Raise temperature in the furnace until 660°C at a rate of 5–10°C/min. Keep temperature at 660°C for 12 h and naturally lower temperature. 3.5. Microfluidic Electrophoresis Analysis
The procedure to perform microfluidic electrophoretic separation of DNA restriction fragment digests is described in this section. A so-called cross-injection method has been adopted. A slight modification of the mask design or sample preparation enables detection and identification of pathogens, enantiomer detection of amino acids, and detection of glucose (5, 7). A further reduction of background photocurrent due to laser light scattering
Miniaturized and Integrated Fluorescence Detectors
371
should improve LOD, making the micromachined fluorescence detector more applicable to biochemical assays including DNA sequencing. 1. Dissolve hydroxyethylcellulose (HEC, Hercules, DE, USA) in a 1×Tris-acetate/EDTA (TAE) buffer to 1% w/v for the sieving matrix. Dilute oxyazole yellow (YO, Invitrogen, OR, USA), an intercalating dye used for fluorescent labeling of DNA fragments, to 1 μM in the sieving matrix. Degas this mixture under vacuum while stirring for 5 min and then centrifuge it to remove any bubbles. 2. Apply labeled HEC to the anode with a syringe and force it through the channel. 3. Dilute a HaeIII digest of φX174 bacteriophage DNA (Takara, Japan) to 100 ng/μL in water (see Note 4) and then put the diluted DNA sample (5 μL) in the sample reservoir. Put 10 μL of the 1×TAE buffer in each of the other reservoirs. 4. Drop index-matching fluid (series A, n = 1.474, Cargille Laboratory, NJ, USA) on the half-ball lens to optically couple with the microfluidic CE. 5. Adjust the position of the microfluidic CE chip with the two orthogonal micrometers so that the laser irradiates the microchannel of the microfluidic electrophoresis device. This irradiated point corresponds to a detection point several cm away (approximately 4 cm in our case) from the injection cross point of the channels. 6. Insert Pt electrodes into all reservoirs to make electrical contact. 7. Inject the DNA sample at 830 V/cm for 20–40 s from the sample reservoir to waste reservoir while the potentials of the other reservoirs are close to the floating potential. 8. Switch the electric field to electrophorese the sample to the separation channel at 230 V/cm while applying a back-biased electric field of 160 V/cm between the injection cross point and both the sample and waste reservoirs to avoid any bleeding of the sample. 9. Start to measure the LIA output after the electrophoresis starts. A plot of the LIA output as a function of time gives an electropherogram. Figure 6 presents results obtained according to the above procedure. Despite neither data processing nor baseline subtraction, all 11 peaks of the DNA fragments could be detected with a good S/N ratio and well resolved, including 271 and 281 bp peaks, in approximately 2-min separation. Provided that the S/N ratio for the peak of each fragment is determined as the ratio of the peak height to noise level of the baseline, the LOD values
Kamei
872 1078 1353
372
C
271
281
194 234
310
W
S
118
603
Fluorescence
~3.5 cm
72
A
0
20
40
60
80
100
120
140
160
Time (s) Fig. 6. Electropherogram of the Hae III digest of φX174 bacteriophage DNA (100 ng/μL) using the micromachined fluorescence detector. The DNA fragment length is indicated in bp for each peak. The channel geometry for the microfluidic CE is schematically shown top left, where S, W, C, and A, respectively, stand for the sample, waste, cathode, and anode reservoirs.
(S/N = 3) for each fragment vary from 150 to 430 pg/μL with an average figure of 250 pg/μL. Theoretical plates for each fragment range from 26,000 to 128,000, with an average figure of 68,000. A conventional confocal fluorescence detection system typically exhibits an LOD value of 10–100 pM in fluorescein concentration, still much lower than the micromachined fluorescence detector, but such high sensitivity is not needed for many applications. When high-sensitivity detection is required, it may be more effective and easier to compensate the detector sensitivity by coupling with the DNA preamplification and/or cell preconcentration method (8), because the optical interference filter we adopted in this work is already superior and suited for visible fluorescence detection from fluorescein and YO.
4. Notes 1. Unless stated otherwise, all the solutions should be prepared in water that has a resistivity of more than 18 MΩ cm and total organic content of less than ten parts per billion. This is referred to as “water” in this text. The glass wafers should also be rinsed with this water. 2. Design of an optical interference filter is proprietary but its structure we have adopted here is essentially multiple Fabry–Perot cavity. For example, in the case of design: Air
Miniaturized and Integrated Fluorescence Detectors
373
100
T (%)
80 60 40 20 0 400
450
500
550
600
650
700
Wavelength (nm) Fig. 7. Computed transmittance of the multiple cavity filter against wavelength (nm). Design: Air|LHLL( (HLH)4L)10HLHHLH|glass. L = SiO2 (n = 1.45980), H = Ta2O5 (n = 2.14064) and reference wavelength is 553 nm.
|LHLL( (HLH)4L)10HLHHLH|glass where L and H stand for a quarter-wave layer of high and low refractive index, respectively; L = SiO2, H = Ta2O5, and reference wavelength is 553 nm, computed transmittance of such a filter is shown in Fig. 7. Although more ripples appear in Fig. 7 compared to commercial filter shown in Fig. 1a, similar features could be reproduced. 3. The lift-off process to pattern the optical interference filter is the most difficult; it might take 1 week and the Al etchant often damages the Al electrode of the a-Si:H photodiode. To mitigate against this, we make many scratches on the unwanted region (optical filter/Al/Si) with a pitch of approximately 2 mm, prior to the lift-off process, reducing the lift-off time to 4–5 h and preventing the Al electrode from being damaged. 4. The DNA sample is normally diluted with the running buffer, in this case, the 1×TAE buffer. The DNA sample is diluted with water, so that stacking takes place to concentrate the sample plug in the separation channel and enhance the peak intensity.
Acknowledgments This study was carried out partly in collaboration with Prof. Richard Mathies’ group at University of California in Berkeley. I would like to particularly thank Dr. James Scherrer for his contribution to construct the optical setup and Dr. Brian Paegel for his consultation concerning microfluidic electrophoresis. This work was supported in part by New Energy and Industrial Technology Development Organization (NEDO) of Japan.
374
Kamei
References 1. Harrison, D. J., Fluri, K., Seiler, K., Fan, Z. H., Effenhauser, C. S., and Manz, A. (1993) Micromachining a miniaturized capillary electrophoresis-based chemical-analysis system on a chip. Science 261, 895–97 2. Unger,M.A.,Chou,H.P.,Thorsen,T.,Scherer,A., and Quake, S. R. (2000) Monolithic microfabricated valves and pumps by multilayer soft lithography. Science 288, 113–16 3. Fu, A. Y., Chou, H. P., Spence, C., Arnold, F. H., and Quake, S. R. (2002) An integrated microfabricated cell sorter. Analytical Chemistry 74, 2451–57 4. Hansen, C. L., Skordalakes, E., Berger, J. M., and Quake, S. R. (2002) A robust and scalable microfluidic metering method that allows protein crystal growth by free interface diffusion. Proceedings of the National Academy of Sciences of the United States of America 99, 16531–36
5. Kamei, T., Paegel, B. M., Scherer, J. R., Skelley, A. M., Street, R. A., and Mathies, R. A. (2003) Integrated hydrogenated amorphous Si photodiode detector for microfluidic bioanalytical devices. Analytical Chemistry 75, 5300–05 6. Kamei, T., and Wada, T. (2006) Contact-lens type of micromachined hydrogenated amorphous Si fluorescence detector coupled with microfluidic electrophoresis devices. Applied Physics Letters 89, 114101 7. Kamei, T., Toriello, N. M., Lagally, E. T., Blazej, R. G., Scherer, J. R., Street, R. A., and Mathies, R. A. (2005) Microfluidic genetic analysis with an integrated a-Si: H detector. Biomedical Microdevices 7, 147–52 8. Lagally, E. T., Lee, S. H., and Soh, H. T. (2005) Integrated microsystem for dielectrophoretic cell concentration and genetic detection. Lab on a Chip 5, 1053–58
Chapter 21 Photomultiplier Tubes in Biosensors Yafeng Guan Summary Photomultiplier tubes (PMT) are widely used for the weak light detection in some types of biosensors. A light detection system for biosensors based on PMT generally contains optic fibers, PMT, and filters. Basic principles of those accessories were provided in this chapter. The guides to selecting fibers, filters, PMT, and power suppliers in practical applications were presented. Major problems that may occur with the instruments were listed and discussed. Key words: Fluorescence detection, LIF, LED-IF, PMT.
1. Introduction Light detection technology is a powerful tool that provides deeper understanding of more sophisticated phenomena. Measurement using light detectors offers unique advantages such as nondestructive analysis of a substance, high-speed properties, and extremely high detectability. By coupling optical principles and methods with the biological specificity, optical biosensors have been a particular field of biosensors. In these cases, photomultiplier tubes (PMT) is often the optimal optical detector for biosensors due to its superior properties of high sensitivity to low-light-level, remarkable stability, fast response to wide wavelength spectral, and wide linearity. PMT in biosensors is generally used to measure the light intensity of fluorescence, absorption light, and chemiluminescence, produced during the process of the specific combination between the biological recognition element and substrate. Because of the wide wavelength response
Avraham Rasooly and Keith E. Herold (eds.), Methods in Molecular Biology: Biosensors and Biodetection, Vol. 503 © Humana Press, a part of Springer Science + Business Media, LLC 2009 DOI: 10.1007/978-1-60327-567-5_21
375
376
Guan
of PMT, filters or monochromators should be used to enhance the wavelength resolution of sensors when applied to single wavelength detection. In this chapter, basic principles and structures of PMT will not be introduced in detail. We will briefly describe an optical detection system using PMT to detect the slight changes of the weak light signal produced in biosensors. In addition, some important aspects related to the correct use of PMT, such as optic fibers, filters, and power suppliers, will be mentioned in proper place.
2. Materials A light diode emitting at 470 nm (50 mW, Shifeng Optics, China) was chosen as the light source. The lens of f10 and f15 (GCL010131, GCL-010132, Daheng Optics, China) was used to focus the excitation light. Filter for the light source was BP470 (Huibo Optics, China). The sample flowed in a fused-silica capillary (100 μm i. d., Yongnian, Inc., China). The optical fiber (core 0.5 mm, cladding 0.6 mm, Chunhui, Inc., China) was used to collect the fluorescence. The emitting filter was chosen to be BP530 (Huibo Optics, China). PMT (H5784) was obtained from Hamamatsu, Japan. A weak current amplifier TJ-110 was from Taiji Computer cooperation limited, China. The data acquisition station KF-98 (Taiji co., China) was used for signal acquisition and data processing.
3. Methods 3.1. Filter
Filters are used to pass a band of wavelengths (bandpass filters) or to block wavelengths longer or shorter than some desired value (cutoff filters). The use of filter before the PMT can reduce the background light and enhance the sensitivity (1). The characteristics of the filters are illustrated by plots of their spectral transmittance (T) vs. wavelength, as shown in Fig. 1. A filter with higher spectral transmittance can be beneficial to the sensitivity of the biosensor constructed. The choice of the filters should be made by considering wavelength of the excitation light and the emission light of the system used in the biosensors. Figure 2 shows a typical fluorescence system and the corresponding filter used before the PMT. To get high sensitivity, the lm should be consistent with the wavelength
Photomultiplier Tubes in Biosensors
377
(a) 100
T (%)
80 60 40 20 0
(b) 102
Log T (%)
101 100 10−1 10−2 10−3 10−4 400
450
500
550
600
650
700
Wavelength (nm) Fig. 1. Transmittance of bandpass (a) and short-wavelength cutoff (b) filters. a lm, the wavelength of maximum transmittance; Tm, the maximum transmittance; and Δl, FWHM. b lm, the wavelength at which the transmittance over 90% of its maximum value; lh, the wavelength at which T = 0.5Tm; and lc, the cutoff wavelength.
of the emission light, and kept away from the wavelength of the excitation light used in the system. 3.2. Optical Fiber
Optical fibers are experiencing greater use in biosensors for several reasons. Because they are mechanically flexible, light can be transmitted over curved paths. Thus the optical fiber can replace several mirrors in directing light between two points in a biosensor. A single optical fiber cannot transmit an image; bundles of fine glass, quartz, or plastic fibers can be used for image transmission if the fibers are small enough in diameter that each fiber transmits rays from a small area of the object. The user can choose the optical fiber according to the biosensor constructed.
3.3. Photomultiplier Tube
When light enters a PMT, it is detected and it produces a signal through the following processes:
378
Guan
(a) μFD
Cr
Al
500μm
Aperture
(b) SiO2 / Ta2O5 Optical filter
SiN
SiO Al Glass substrate
ITO p i a-Si:H n
μFD~7μm
Cr
Fig. 2. The fluorescence system and the corresponding filter used before the PMT. a A typical fluorescence system. ex is the excitation light and em is the emission light. b The bandpass filter chosen for the system.
1. Passes through the input window. 2. Electrons in the photocathode are excited by the light and emitted into the vacuum. 3. Photoelectrons are accelerated and focused by the focusing electrode onto the first dynode where they are multiplied by means of secondary electron emission. The secondary emission is repeated at the successive dynodes. 4. The multiplied secondary electrons emitted from the last dynode are finally collected by the anode. It is important to know the conditions of the light to be detected before we choose a proper PMT. The parameters listed in Table 1 should be taken into account when making a selection. Spectral response characteristics of PMT are mainly related to photocathodes and window materials. Here we introduce these materials and the corresponding detection wavelength range. 3.3.1. Photocathodes
Most photocathodes (2–4) are made of compound semiconductors, which consist of alkali metals with a low work function. There are approximately ten kinds of photocathodes currently
Photomultiplier Tubes in Biosensors
379
Table 1 The parameters should be taken into account when selecting a PMT Incident light conditions
Selection reference
Light wavelength
Window material Photocathode spectral response
Light intensity
Number of dynodes Dynode type Voltage applied to dynodes
Light beam size
Effective diameter Viewing configuration (side-on of head-on)
Speed of the optical phenomenon
Response time
employed in practical applications. Each photocathode is available as a transmission (semitransparent) type or a reflection (opaque) type, with different device characteristics. The photocathode materials commonly used in PMT are as follows: 1. Ag–O–Cs. Transmission type photocathodes using this material are sensitive from the visible light through the near infrared region, i.e., from 300 to 1,200 nm, which the reflection type exhibits a slightly narrower spectral response region from 300 to 1,100 nm. Compared with other photocathodes, this photocathode has lower sensitivity in the visible region, but good sensitivity at longer wavelengths in the near infrared region. So both transmission and reflection type Ag–O–Cs photocathodes are mainly used for near infrared detection. 2. GaAs (Cs). A GaAs crystal activated with cesium is used for both reflection type and transmission type photocathodes. The reflection type GaAs (Cs) photocathode has sensitivity across a wide range from the ultraviolet through the near infrared region around 900 nm. It demonstrates a nearly flat, high-sensitive spectral response curve from 300 to 850 nm. The transmission type has a narrower spectral response range because shorter wavelengths are absorbed. It should be noted that if exposed to incident light with high intensity, these photocathodes tend to suffer sensitivity degradation when compared with other photocathodes primarily composed of alkali metals.
380
Guan
3. InGaAs. This photocathode provides a spectral response extending further into the infrared region than the GaAs photocathode. Additionally, it offers a superior signal-to-noise ratio in the neighborhood of 900–1,000 nm in comparison with the Ag–O–Cs photocathode. 4. Cs–I and Cs–Te. Cs–I and Cs–Te are not sensitive to solar radiation therefore often called “solar blind.” The sensitivity of Cs–I sharply falls off at wavelengths longer than 200 and it is exclusively used for vacuum ultraviolet detection. Cs–Te is not sensitive to wavelengths longer than 300 nm. 5. Multialkali (Sb–Na–K–Cs). This photocathode uses three or more kinds of alkali metals. Due to high sensitivity over a wide spectral response range from the ultraviolet through the near infrared region around 850 nm, this photocathode is widely used in broadband spectrophotometers. 6. High-temperature, low-noise bialkali (Sb–Na–K). As with bialkali photocathodes, two kinds of alkali metals are used in this photocathode type. The spectral response range is almost identical to that of bialkali photocathodes, but the sensitivity is somewhat lower. This photocathode can withstand operating temperatures up to 175°C while other normal photocathodes are guaranteed to no higher than 50°C. For this reason, it is ideally suited for use in oil well logging where PMT are often subjected to high temperatures. In addition, when used at room temperatures, this photocathode exhibits very low dark, which makes it very useful in low-level light measurement such as photon counting applications where low noise is a prerequisite. 3.3.2. Window Materials
Most photocathodes have high sensitivity down to the ultraviolet region. However, because ultraviolet radiation tends to be absorbed by the window material, the short wavelength limit is determined by the ultraviolet transmittance of the window material (5, 6). The window materials commonly used in PMT are as follows: 1. Borosilicate glass. This is the most commonly used window material. Because the borosilicate glass has a thermal expansion coefficient very close to that of Kovar alloy, which is used for the leads of PMT, it is often called “Kovar glass.” The borosilicate glass does not transmit ultraviolet radiation shorter than 300 nm. It is not suited for ultraviolet detection shorter than this wavelength. Moreover, some types of head-on PMT using a bialkali photocathode employ a special borosilicate glass (so-called “K-free glass”) containing a very small amount of potassium (K40), which may cause unwanted background counts. The K-free glass is mainly used for PMT designed for scintillation counting where low background counts are desirable.
Photomultiplier Tubes in Biosensors
381
2. UV glass (UV-transmitting glass). As the name implies, this transmits ultraviolet radiation well. The short wavelength cutoff of the UV glass extends to 185 nm. 3. Synthetic silica. Synthetic silica transmits ultraviolet radiation down to 160 nm and in comparison with fused silica, offers lower levels of absorption in the ultraviolet region. Since silica has a thermal expansion coefficient greatly different from that of the Kovar alloy used for the stem pins (leads) of PMT, it is not suited for used as the bulb stem. As a result, a borosilicate glass is used for the bulb stem and then a graded seal, using glasses with gradually changing thermal silica bulb. Because of this structure, the graded seal is very fragile and proper care should be taken when handing the tube. In addition, helium gas may permeate through the silica bulb and cause an increase in noise. Avoid operating or storing such tubes in environments where helium is present.
3.3.3. Dynode Types and Features
4. MgF2 crystal. The crystals of alkali halide are superior in transmitting ultraviolet radiation, but have disadvantage of deliquescence. A magnesium fluoride (MgF2) crystal is used as a practical window material because it offers very low deliquescence and allows transmission of vacuum ultraviolet radiation down to 115 nm. There are a variety of dynode types available and each type exhibits different gain, time response, uniformity, and secondaryelectron collection efficiency depending upon the structure and the number of stages. The optimum dynode type must be selected according to application: 1. Circular-cage type. The circular-cage type has an advantage of compactness and is used in all side-on PMT and in some head-on PMT. The circular-cage type also features fast time response. 2. Box-and-grid type. This type, widely used in head-on PMT, is superior in photoelectron collection efficiency. Accordingly, PMT using this dynode offer high detection efficiency and good uniformity. 3. Linear-focused type. As with the box-and-grid type, the linearfocused type is widely used in head-on PMT. Its prime features include fast time response, good time resolution, and excellent pulse linearity. 4. Venetian blind type. The Venetian blind type creates an electric field that easily collects electrons, and is mainly used for head-on PMT with a large photocathode diameter. 5. Mesh type. This type of dynode uses mesh electrodes stacked in close proximity to each other. There are two types: coarse mesh type and fine mesh type. Both are excellent in output linearity and have high immunity to magnetic fields. When used with a cross wire anode or multianode, the position of
382
Guan
incident light can be detected. Fine mesh types are developed primarily for PMT, which are used in high magnetic fields. 6. MCP (microchannel plate). A microchannel plate (MCP) with 1-mm thickness is used as the base for this dynode structure. This structure exhibits dramatically improved time resolution as compared with other discrete dynode structure. It also assures stable gain in high magnetic field and provides position-sensitive capabilities when combined with a special anode. 7. Metal channel dynode. This dynode structure consists of extremely thin electrodes fabricated by advanced micromachining technology and precisely stacked according to computer simulation of electron trajectories. Since each dynode is in close proximity to one another, the electron path length is very short ensuring excellent time characteristics and stable gain even in magnetic fields. 8. Electron bombardment type. In this type, photoelectrons are accelerated by a high voltage and strike a semiconductor so that the photoelectron energy is transferred to the semiconductor, producing a gain. This simple structure features a small noise figure, excellent uniformity, and high linearity. Figure 3 shows a typical application example in which a PMT is used in a biosensor. A blue LED driven by a 5 V constant voltage source through a 100 Ω current-limiting resistor was used as the excitation source. LED light was collimated and focused with two quartz achromatic lenses into the capillary. To
7
6
3
5
4
1 2
2
Fig. 3. The structure of a biosensor (1) light source; (2) lens; (3) filter for the light source; (4) sample; (5) optical fiber; (6) filter; and (7) PMT.
Photomultiplier Tubes in Biosensors
383
reduce the scattering light from the capillary wall, an aperture of 0.2 mm was used to restrict the beam size. An interference filter was inserted between two lenses to eliminate the interference of long wavelength from LED. A detection window on the capillary was formed by burning off the polyimide coating (5 mm in length) with an electrical coiled resistance. Fluorescence was collected with a right-angle geometry by an optical fiber (see Note 1) and passed through two blocks of interference filters. The distance between the fiber and the capillary, as well as the distance between the fiber and the filter (see Notes 2 & 3), was set to be 0.5 mm. The fluorescence signal was then detected by a metal package miniaturized PMT (see Notes 4–8). The signal from the PMT was acquired by chromatographic workstation. Figure 4 illustrated a solid schematic view of the sensor. The application of the biosensor to detect the riboflavin in watermelon sample was demonstrated. The limit of detection for riboflavin was 10 μg/L. The sensor exhibited an excellent linear behavior over the concentration range of 10–1,000 μg/L (R = 0.9996). The riboflavin in the watermelon sample was determined to be 100 μg/L (S/N = 3). Figure 5 showed the chromatogram obtained from an analysis of riboflavin solution with several concentrations. The sensor can also be applied to electrophoresis to detect several biomolecules. Figure 6 demonstrated the analysis of three amino acids labeled by FITC. The
Fig. 4. The solid schematic view of the biosensor (1) light source; (2) lens; (3) filter for the light source; (4) sample; (5) optical fiber; (6) filter; and (7) PMT.
384
Guan
Fig. 5. Chromatogram of riboflavin solutions with several concentrations.
Fig. 6. Electropherogram of FITC-labeled amino acids (1) FITC-labeled lysine; (2) excess FITC; (3) FITC-labeled Tryptophan; and (4) FITC-labeled phenylalanine.
detection limits of 10, 9, and 4.8 nM (S/N = 3) were achieved for lysine (Lys), tryptophan (Trp), and phenylalanine (Phe), respectively. Potential applications of the sensor are routine analysis of protein, peptide, amino acids, and others compounds. 3.4. Selecting a Power Supply For PMT
The operation stability of a PMT depends on the total stability of the power supply characteristics including drift, ripple, temperature dependence, input regulation, and load regulation. The power supply must provide high stability that is at least ten times as stable as the output stability required for the PMT. Series-regulator type high-voltage power supplies have been widely used with PMT. Recently, a variety of switching-regulator types have been put on the market and are becoming widely
Photomultiplier Tubes in Biosensors
385
used. Most of the switching-regulator type power supplies offer compactness and light weight, yet provide high voltage and high current. However, with some models, the switching noise is superimposed on the AC input and high-voltage output or the noise is radiated. Thus, sufficient care is required when selecting this type of power supply, especially in low-light-level detection, measurement involving fast signal processing, and photon-counting applications. The high-voltage power supply should have sufficient capacity to supply a maximum output current which is at least two times the current actually flowing through the voltage-divider circuit used with the PMT. Table 2 shows the guide for selecting the correct high-voltage power supply. Recently, commercial PMT modules comprised of a PMT, a high-voltage power supply (HV) circuit, and a voltage-divider circuit are available on the market. Using this type of PMT modules eliminate an external HV from an external power supply. All that need is simple wiring and low-voltage input. Supply approximately 15 V to the low-voltage input, ground the GND terminal, and connect the control voltage and reference voltage input according to the gain adjustment method as described by manufactures (see Notes 9 & 10). 3.5. Connecting the PMT Output to Data Acquisition System
The output of a PMT is a current, while the expected signal of signal conditioning circuit is a voltage. Therefore, the current output should be converted into a voltage signal by some means. One simple method for the current output of a PMT into a voltage output is to use a load resistance. Since the PMT may be thought of as an ideal constant current source, the voltage output of I × RL (I is output current and RL is the load resistance) can be obtained. Another method is to use a current-to-voltage conversion circuit using an operational amplifier. A basic circuit using an operational amplifier is shown in Fig. 7. TJ-110 we used is a
Table 2 The characteristics of the high-voltage power supply for PMT Line regulation
±0.01% or less
Load regulation
±0.2% or less
Ripple noise
0.05% or less
Temperature coefficient
±0.05%/K or less
386
Guan
Fig. 7. Current-to-voltage conversion circuit using an operational amplifier.
typical current amplifier module using an operational amplifier current-to-voltage conversion circuit. With this circuit, the output voltage Vo is given by Vo = –Ip· zRf · Some PMT modules contain an integrated current-to-voltage conversion circuit and the output of PMT is a voltage. It is not necessary to use an additional conversion circuit. After conversion, the voltage signal is measured with a high sensitive ammeter or connected to the input terminals of the data acquisition system KF-98.
4. Notes 1. For optical fibers, make sure that the end should be flattened and mirror polished. Pay attention to the minimum bending radius of the fiber used; do not exceed the value. 2. Do not touch the filter with bare hands or expose it to the dust. If not used, store it in dry environments. 3. Make sure that the optical fiber placed to the filter as near as possible, but had no physical contact with the filter, since the fiber may cause damage to the film of the filter. 4. PMT is a very high sensitive photodetector; the users have to read-through the guide for users before use, and handle/operate the module carefully. 5. The PMT is fragile by shock and vibration, so handle it carefully not to drop or add excess shock. 6. Dust and fingerprints in the window will cause loss of signal light transmittance, so do not touch the window portion of the PMT with bare hands or expose it to the dust. Should it be dirty, wipe it with alcohol.
Photomultiplier Tubes in Biosensors
387
7. Do not expose the PMT to strong light, even when it is not operated. Direct sunlight and other strong light illumination may cause damage to it. When the PMT is not used, keep it in dark storage. 8. The PMT should not be stored or operated in an environment of high-pressure Helium gas over partial pressure in air. Helium gas can penetrate through the window of the PMT, and increase the noise level or damage the performances. 9. Check the wiring before turning on the power supply. If the voltage is turned on at incorrect cable wiring, for example, when the voltage polarity is supplied incorrectly, the PMT can be damaged. 10. Do not supply the voltage over the maximum of the guide for users said. It will cause the damage to the PMT.
References 1. D. James, J.R. Ingel, R.C. Stanley, Spectrochemical Analysis, Prentice-Hall, Englewood Cliffs, NJ (1988) 2. Hamamatsu Photonics Catalog: Photomultiplier Tubes 3. T. Hiruma, SAMPE Journal 24, 35 (1988)
4. A.H. Sommer, Photoemissive Materials, Robert E. Krieger Publishing Company, Huntington, NY (1980) 5. Handbook of Optics, McGraw-Hill, New York, NY (1987) 6. J.A.R. Samson: “Techniques of Vacuum Ultraviolet Spectroscopy”: Wiley, New York, NY (1967)
Chapter 22 Integrating Waveguide Biosensor Shuhong Li, Platte Amstutz III, Cha-Mei Tang, Jun Hang, Peixuan Zhu, Yunqi Zhang, Daniel R. Shelton, Jeffrey S. Karns Summary The Integrating Waveguide Biosensor was developed for rapid and sensitive detection of bacterial cells, spores, and toxins. A sandwich format of immunoassay was employed using Salmonella as model. The analyte was immunocaptured on the inner surface of the waveguide and then detected by the antibody conjugated with fluorescent dye. The waveguide was illuminated by an excitation light at a 90° angle. The emitted light from fluorescent labels on the surface of the waveguide was efficiently collected and channeled to a detector at the end of the waveguide, while minimizing interference from the excitation light. Utilizing fluorescent dye Cy5, a 635-nm diode laser for excitation, and a photomultiplier tube detector, the Integrating Waveguide Sensor System was able to detect approximately ten captured cells of Salmonella. Key words: Biosensor, Integrating waveguide sensor, Integrating waveguide biosensor, Fluorescence detection, Salmonella.
1. Introduction When fluorescent labels emit light, the emission is typically in all directions such that only a small fraction of the light is collected by the detector as signal. Simultaneously, light from the excitation source, auto fluorescence from the sample and the sample container, Raman emission from water, and other light collected by the detector contribute to background noise. The ability to maximize the signal while minimizing the background noise, i.e., high signal-to-noise ratio results in a lower limit of detection and improved instrument sensitivity.
Avraham Rasooly and Keith E. Herold (eds.), Methods in Molecular Biology: Biosensors and Biodetection, Vol. 503 © Humana Press, a part of Springer Science + Business Media, LLC 2009 DOI: 10.1007/978-1-60327-567-5_22
389
390
Li et al.
A platform to detect analyte on the surface of a waveguide is shown in Fig. 1, demonstrating a captured analyte on the inside surface of the capillary waveguide binding a fluorescent detection antibody. The signal is obtained by illuminating the capillary tube at a 90° angle relative to its length and subsequent collection of the emitted fluorescence from one end of the waveguide. The emitted fluorescent light is efficiently gathered and guided to the end of the capillary tube waveguide and goes through a set of lenses and optical filters to the optical detector. The signal is maximized by integrating the emitted light from all of the fluorescent labels on the capillary tube surface. At the same time, the background noise is reduced as a consequence of the excitation light being directed at a 90° angle to the waveguide surface. The Integrating Waveguide Biosensor (IWB) has an inherently high signal-to-noise ratio, because it gathers a high percentage of the fluorescence signal, and because the noise contributed by the excitation is reduced. The IWB was originally developed by Ligler et al. at the Naval Research Laboratory (NRL) in Washington, DC (1, 2). Ligler et al. (2) reported a detection limit of 40 pg/mL for mouse IgG and 30 pg/mL for staphylococcal enterotoxin B (SEB) in a sandwich assay format, which is about 100-fold more sensitive than the evanescent-wave fiber-optic and array biosensor technologies previously developed (3–8). We have constructed a compact and portable instrument using a low-cost capillary tube as the waveguide. When the IWB was applied to the detection of Salmonella, the limit of detection was ten captured Salmonella cells. The IWB utilizing solid phase assay of analyte on the waveguide surface belongs to a family of integrating waveguide sensors detection principles. The implementation of the liquid-phase
Fig. 1. Basic principle of the integrating waveguide sensor where a capillary is used as the waveguide. The analyte is captured on the inner surface of the capillary waveguide. Fig. 2. Schematic of Creatv’s Integrating Waveguide Sensor.
Integrating Waveguide Biosensor
391
detection of the Integrating Waveguide Sensor is described in Chap. 24 (9).
2. Materials 2.1. Instrument Components
1. Illumination light source. Diode laser of 635 nm, 30 mW (LaserMax, Rochester, NY). 2. Signal cleanup filters. A RG665 color glass filter (Edmund Industrial Optics, Barrington, NJ) and an interference band-pass 670-nm filter (Omega Optical, Brattleboro, VT). 3. Detector. Photomultiplier tube (PMT) H7827-11 (Hamamatsu, Bridgewater, NJ). 4. Signal digitalizing. DT9802 Data acquisition card (Data Translation, Inc., Marlboro, MA). 5. Laser line generator. A rod lens from LaserMax, Rochester, NY, combined with a cylinder lens from Edmund Industrial Optics, Barrington, NJ. 6. Signal collimating lens. An aspheric lens combined with a double convex lens. 7. Signal focusing lens. Double convex lens from Edmund Industrial Optics, Barrington, NJ. 8. User interface. LabVIEW™ software (National Instruments Corp., Austin, TX).
2.2. Simulation and Analysis
1. Optical ray tracing. TracePro software (Lambda Research Corp., Littleton, MA).
2.3. Capillary Waveguide
1. Capillary waveguides. 50 mm long borosilicate glass capillary tube with 1.66 mm outer diameter and 1.23 mm inner diameter (Drummond Scientific Company, Broomall, PA).
2.4. Chemicals and Reagents for Assay
1. NeutrAvidin™ (Pierce Biotechnology, Rockford, IL). 2. GMBS. 4-maleimidobutyric acid N-hydroxysuccinimide ester (Sigma-Aldrich, St. Louis, MO). 3. Buffers. PBST (10 μg/mL in PBS containing 0.05% Tween) and PBSTB (PBST containing 2% BSA). 4. Analyte. Salmonella typhimurium (ATCC 53648) (ATCC, Manassas, VA). 5. Capture antibody. Biotinylated rabbit polyclonal antibody (Biodesign International, Saco, Maine). 6. Detector antibody. Cy5-conjugated goat polyclonal antibody (KPL, Gaithersburg, MD).
392
Li et al.
3. Methods The IWB consists of an illumination optical system, capillary tube waveguide, detection optical system, and data acquisition system (Fig. 2); the details of each component part will be discussed in the following sections. One of the design changes from NRL’s experimental instrument was the elimination of the laser chopper and lock-in amplifier, which were necessary to minimize ambient background light. In contrast, the light-tight construction of Creatv’s instrument eliminates interference from ambient light, obviating the chopper and lock-in amplifier, thereby reducing both the size and the cost. Other design changes are automation of data acquisition and packaging into a portable bench-top format. Laser beam is expanded and collimated to fit the dimension of the capillary tube, so that the capillary tube can be evenly illuminated. Light exiting from the end of the capillary tube is collimated before passing through optical filters, in order to block excitation light and pass emission light. After passing the optical filters, the signal is focused down to the detection area of the PMT. 3.1. Illumination Optical System
The illumination system provides a collimated beam to the analytesensing surface. The system consists of a laser, a cylinder lens to generate a line pattern, and a collimating cylinder lens, as shown in Fig. 2. The features of the system include: ● Stable laser illumination ●
Collimated beam
●
Low noise in the fluorescence emission range
●
Illumination of the whole length of capillary waveguide
Fig. 3. Results of tracepro simulation of NRL’s fused-silica capillary tube (0.70 mm inner
Integrating Waveguide Biosensor
393
The optical system was optimized using Cy5 as the fluorescent dye. The peak absorption wavelength of Cy5 is 647 nm, while the peak emission wavelength is 670 nm, with a range of 650– 750 nm. A diode laser (LaserMax, Rochester, NY) at 635 nm was chosen to provide maximum separation of the excitation and emission wavelengths. Capillary tube is 1.66 mm in diameter and 50 mm long. In order to illuminate the whole tube, the laser beam needs to be about 2 mm × 50 mm in size. The dimensions of the laser beam was expanded to 2 mm × 50 mm to cover the capillary tube waveguide and collimated to provide illumination at a 90° angle to the capillary surface. 3.2. Waveguide
Borosilicate capillary tube is chosen as waveguide and they have the following properties: ● Good transmission for both laser excitation and fluorescent emission wavelengths ●
Low autofluorescence
●
Convenient and inexpensive
Compatible with bioassays In NRL’s initial development, the capillary tubes they used is fused-silica capillary pieces 38 mm long (0.70 mm inner diameter and 0.85 mm outer diameter) with optically polished ends and coated on the outer surface with PTFE were obtained from Polymicro Technologies (Phoenix, AZ). These capillary tubes provided a low limit of detection, but their high cost and fragile nature made them unsuitable for routine use. In order to select the optimum capillary tube waveguide geometry, simulations of the optical properties of various geometries and materials were conducted using the optical ray tracing software TracePro (Lambda Research Corp., Littleton, MA) to perform analysis of illumination of the fluorescent labels on the surface of the capillaries. TracePro uses Monte Carlo simulations to set up the rays for ray tracing and to compute optical flux. TracePro accounts for absorption, reflection, refraction, and scattering of light according to the material and surface properties of the capillary tube. Simulations were performed using incident rays of collimated 635-nm laser light. Initial simulations of the NRL capillary tube waveguide indicated that the NRL’s capillary geometry did not provide uniform illumination of the analyte sensing surface on the inner surface of the capillary. TracePro ray tracing results for the NRL capillaries is shown in Fig. 3a. The trajectories of the incident rays from the excitation laser light (entering from the top of the figure), are bent in the silica capillary (index of refraction = 1.46), and are reflected at the capillary tube surfaces due to change of the refraction index. The capillary is filled with buffer, because ●
394
Li et al.
diameter and 0.85 mm outer diameter) with liquid inside ray trajectories (a) and intensity plot (b).
NRL’s experiments obtained better results with fluid inside the capillary. The rays shown in Fig. 3a are displayed in shades of gray: black rays denote the rays with flux equal or greater than 67% of the incident laser light; medium gray rays denote the rays with flux between 33 and 67% of the incident laser light; while
Integrating Waveguide Biosensor
395
light rays denote the rays with flux less than 33% of the incident laser light. Figure 3a shows that bottom portions of the capillary surface received concentrated excitation light, while some areas to either sides of the capillary received no excitation light. TracePro can also display the results of the excitation laser intensity for a cross section of the capillary tube, where the grayscale from black to white indicates no laser light to brightest laser light, respectively. The dotted white circle in Fig. 3b denotes the location of the inner surface of the capillary tube used by NRL and shows that this capillary geometry does not provide uniform illumination of the analyte sensing surface. We subsequently analyzed several borosilicate glass capillary tubes of various sizes (index of refraction n = 1.52) commercially available from Drummond Scientific. Representative results are shown for the capillary tube with outer diameter of 1.661 mm and inner diameter of 1.226 mm, which was easy to handle and had superior performance (Fig. 4a, b). The TracePro simulation was performed with empty capillaries. Ray tracing trajectories (Fig. 4a) and grayscale cross-sectional view of laser excitation intensity (Fig. 4b) show nearly uniform illumination of the analyte sensing surface, indicated by the dotted white circle. Based on these simulations, this capillary geometry was chosen for the experiments.
Fig. 4. Results of tracepro simulation of Drummond borosilicate capillary tube (1.226 mm inner diameter and 1.661 mm outer diameter) with air inside ray trajectories (a) and intensity plot (b).
396
Li et al.
Fig. 4. (continued)
In order to design the detector optical system, we needed to estimate the properties of the signal that exits from the end of the capillary tube. A different TracePro simulation of the propagation of fluorescent emission signal from Cy5 dyes on the inner surface of the capillary was gathered and guided to the end of tube. The result is presented in Fig. 5 as a candela plot of luminous intensity plot or emission flux distribution vs. angle at the exit of the capillary tube from one dye with 10,000 emission rays. Darker gray indicates more flux. Simulation with large number of dyes would provide uniform angular distribution. The software TracePro, however, can only simulate one dye at a time. Figure 5 indicates that the emission signal exits the tube end in a cone shape with wide angles ranging from 30 to 60°. Thus, it is very important for the detecting optics to collect all the signals efficiently from the large emission angles from the end of the capillary tube. 3.3. Detection Optical System
The purpose of the detector optical system is to: Efficiently collect the emission light from the end of the waveguide
●
●
Filter out any laser light
Integrating Waveguide Biosensor
397
Fig. 5. Candela plot of emission flux distribution from Drummond capillary tube vs. angle at the exit of the capillary tube. Darker gray indicates more flux.
●
Select the relevant portion of the emission light for detection
Focus the transmitted signal to the detector The light from the end of the waveguide includes both excitation and emission light. The excitation light has to be separated out before testing the signal. A 670-nm band-pass filter combined with a 665-nm long-pass filter are used to block excitation light. In order for band-pass filter to work properly, the light has to be collimated. The light from the end of the capillary has large angles (about 60° half angle). An aspheric lens plus a double convex lens are incorporated to collimate the light, as shown in Fig. 2. After passing through the collimating lenses and filters, the excitation light is removed, and only the emission light passes through. Finally, a focusing lens focuses the signal onto the PMT. It is important that the capillary tube should be aligned with the illumination light and the detection system. The alignment of optics is guaranteed by mounting the illumination optics, signal detection optics, and optics holder in a precision machined chamber. ●
398
Li et al.
3.4. Data Acquisition and User Interface
The output form PMT is voltage signal ranging from 0 to 10 V. A data acquisition card is used to digitalize the voltage signal and send it to a computer. The gain of PMT is adjustable. If the signal is too low, the PMT gain can be raised; and if the signal is too high and saturates the PMT, the gain can be lowered. The signal has to be normalized accordingly. LabVIEW software is used to control the IWB and collect the data. The user interface, shown in Fig. 6, consists of three panels: SET-UP PANEL, TEST PANEL, and RESULT PANEL. User fills in the information on the SET-UP PANEL. User starts the test by clicking the TEST button. For each testing, the chart at the right side of the front panel will show all the readings collected by PMT as a function of time. An average value will be calculated and displayed on top of the chart. The test data can be saved in EXCEL format. The file name and path will be shown on the RESULT PANEL. Figure 7 shows a schematic of the interior of the instrument. This model uses a H7827-11 Hamamatsu PMT, a DT9802 data acquisition card (Data Translation, Inc.), and a computercontrolled user interface using LabVIEW™ software (National Instruments Corp.).
3.5. Testing Results
IWB was applied to the detection of Salmonella using sandwich immunoassay. Glass capillary tubes were prepared as previously
Fig. 6. Computer interface of Creatv’s prototype Integrating Waveguide Biosensor.
Integrating Waveguide Biosensor
399
Fig. 7. Schematic of Creatv’s portable prototype IWB using capillary tubes as waveguides for Cy5 fluorescent dye.
described (2). Briefly, capillary tubes were cleaned with methanol/ HCl and sulfuric acid, dried with nitrogen, silanized with 3-mercaptopropyl trimethoxysilane in anhydrous toluene under nitrogen atmosphere, incubated with 4-maleimidobutyric acid N-hydroxysuccinimide ester (GMBS) (Sigma-Aldrich, St. Louis, MO), and then treated with NeutrAvidin (Pierce Biotechnology, Rockford, IL). After conjugation of NeutrAvidin inside the capillary tubes, they were washed with PBST, and stored at 4°C until use. The capillary tubes were incubated with PBSTB containing 2% BSA, and then a biotinylated capture antibody in PBSTB was immobilized inside the capillary. After rinsing with PBST, sample was introduced into the capillary tubes and incubated for 1 h at room temperature. The capillary tubes were subsequently washed with PBSTB, and then filled with Cy5 conjugated detector antibody in PBSTB for 1 h. After removal of the unbound detector antibody and thorough washing with PBST, the capillary tubes were tested using the IWB instrument. A serially diluted culture of S. typhimurium (ATCC strain 53648) in PBST buffer (phosphate buffered saline + Tween + Triton) was tested using the Creatv’s IWB. Biotinylated rabbit polyclonal antibody from Biodesign International (Saco, Maine) was used for capture and Cy5 conjugated goat polyclonal antibody from KPL (Gaithersburg, MD) was used for detection. The fluorescence signals obtained using 40–4 × 106 Salmonella cells per capillary are shown in Fig. 8a. The detection threshold (Mean + 3 SD (Standard Deviation) of zero concentration) was 23.4 mV. These data indicate that as few as 40 Salmonella cells per capillary resulted in a detectable signal of 26.9 mV. Based on an estimated capture efficiency of 18–29% (data not shown), the absolute detection sensitivity of the IWB is approximately ten cells. This is consistent with our previous assay to detect E.
400
Li et al.
Fig. 8. Sandwich immunoassay results for detection of Salmonella using the IWB using Cy5 fluorescent label. a Fluorescence signal response to total cell number of Salmonella in a capillary. Threshold detection limit (mean value plus three times of standard deviation) is 23.4 mV, as indicated by the dashed line. b Linear correlation between log10 of net fluorescence signal and log10 of total cells per capillary.
coli (10). For high cell concentrations, the R2 value is 0.998 (Fig. 8b). When the cells per sample are low, the signal readings are accurate, but cells per sample and percentage of cells captured by the waveguide are susceptible to sampling error. We do not have means to provide accurate count of the number of cells captured on the capillary waveguide. The linear relationship between log10 of fluorescence signal and log10 of input cell concentration (Fig. 8b) allows for quantitative detection for high concentrations and only semiquantitative detection for lower concentrations.
Integrating Waveguide Biosensor
401
4. Notes 1. Collimating the laser light or signal through the interference filter is important to reduce background noise. 2. In capillary tube preparation, all the steps from cleaning until up to the immobilization of NeutrAvidin should be done on the same day. After conjugation of NeutrAvidin inside capillary tubes, the capillary tubes were washed with PBST, and may be stored at 4°C until use. 3. The instrument requires a warm up time of about 20 min. This is for the PMT to stabilize.
Acknowledgments We would like to thank Dr. Francis Ligler (Naval Research Laboratory, Washington DC) for helpful advice and comments. This work was supported by NIH Small Business Innovation Research grants R43 CA094430 and R43 AI052684 and a grant from the Maryland Technology Development Corporation (TEDCO).
References 1. Feldstein, M. J., MacCraith, B. D., Ligler, F. S. (2000) Integrating multi-waveguide sensor, US Patent No. 6,137,117 issued on October 24, 2000 2. Ligler, F. S., Breimer, M., Golden, J. P., Nivens, D. A., Dodson, J. P., Green, T. M., Haders, D. P., Sadik, O. A. (2002) Integrating waveguide biosensor. Anal. Chem. 74, 713–719 3. Anderson, G. P., Golden, J. P., Ligler, F. S. (1994) An evanescent wave biosensor – Part I: Fluorescent signal acquisition from step-etched fiber optic probes. IEEE Trans. Biomed. Eng. 41, 578–584 4. Feldstein, M. J. et al. (1999) J. Biomed. Microdevices 1, 139–153 5. Golden, J. P., Anderson, G. P., Ogert, R. A., Breslin, K. A., Liger, F. S. (1993) Evanescent-wave fiber optic biosensor: challenges for real-world sensing, SPIE Proceedings Series, Vol. 1976 (Meeting 8–9 Sep. 1992, in Boston, MA), pp. 2–8
6. Golden, J. P., Anderson, G. P., Rabbany, S. Y., Ligler, F. S. (1994) An evanescent wave biosensor – Part II: Fluorescent signal acquisition from tapered fiber optic probes. IEEE Trans. Biomed. Eng. 41, 585–591 7. Rowe, C. A., Scruggs, S. B., Feldstein, M. J., Golden, J. P., Ligler, F. S. (1999) An array immunosensor for simultaneous detection of clinical analytes. Anal. Chem. 71, 433–439 8. Rowe-Taitt, C. A., Golden, J. P., Feldstein, M. J., Cras, J. J., Hoffman, K. E., Ligler, F. S. (2000) Array biosensor for detection of biohazards. Biosens. Bioelectron. 14, 785–794 9. Li, S., Zhang, Y., Amstutz, P., Tang, C.-M., Multiplex integrating Waveguide Sensor – Signalyte™-II, Chapter 24 in this book 10. Zhu, P., Shelton, D. R., Karns, J. S., Sundaram, A., Li, S., Amstutz, P., and Tang, C.-M. (2005) Detection of water-borne E. coli O157 using the integrated waveguide biosensor. Biosens. Bioelectron. 21, 678–683
Chapter 23 Detection of Fluorescence Generated in Microfluidic Channel Using In-Fiber Grooves and In-Fiber Microchannel Sensors Rudi Irawan and Swee Chuan Tjin Summary In life sciences, the problem of very small volume of sample, analytes, and reagents is often faced. Microfluidic technology is ideal for handling costly and difficult-to-obtain samples, analytes, and reagents, because it requires very small volume of samples, in order of μL or even nL. Among many types of optical techniques commonly used for biosensing in microfluidic chip, fluorescence detection technique is the most common. The standard free-space detection techniques used to detect fluorescence emission from microfluidic channel often suffer issues like scattering noise, crosstalks, misalignment, autofluorescence of substrate, and low collection efficiency. This chapter describes two fluorescence detection methods, based on in-fiber microchannels and in-fiber grooves, which can solve those problems, as the techniques integrate the excitation and emission light paths, and the sensing part. Utilizing an optical fiber as a sensing component makes these detection techniques suitable for lab-on-a-chip or μTAS applications. Key words: Optical fiber sensor, In-fiber microchannel, In-fiber grooves, Fluorescence, Microfluidic chip.
1. Introduction Among the optical techniques commonly used for biosensing applications, fluorescence-based sensing is the most common and the most highly developed due to their high sensitivity, versatility, accuracy, and fairly good selectivity. Currently, the fluorescence sensors incorporated with microfluidic devices mainly use free space configurations (1, 2), which can experience issues like noise and crosstalks due to back scattering, as well as low fluorescence collection efficiency (3–5) that can degrade the sensitivity of the Avraham Rasooly and Keith E. Herold (eds.), Methods in Molecular Biology: Biosensors and Biodetection, Vol. 503 © Humana Press, a part of Springer Science + Business Media, LLC 2009 DOI: 10.1007/978-1-60327-567-5_23
403
404
Irawan and Tjin
system. The detection system in free space configurations may also collect the autofluorescence signals from microfluidic substrates, which may be relatively high and time dependent (6, 7). The integration of the excitation and emission waveguide, and the sensing part may be able to solve these matters (8, 9). In this chapter, we propose a fiber optic as a waveguide and sensor. A fiber optic sensor can be embedded into a microfluidic chip easily due to its size and flexibility (9, 10), and also be used to transport the excitation light to the fluorophores immobilized inside the microchannel, or to collect the generated fluorescence emission, or both (3, 8, 9, 11). Because the optical fiber can transmit the excitation light to the sample and transmit the emitted fluorescence to the detector, it makes the system simpler and more compact for sample excitation and emission collection than a free space configuration method where the light source and the detector may be situated next to each other (3, 8–10). Optical fiber sensors are usually etched (to thin the cladding), tapered, or side polished to improve light coupling from within the fiber to the external environment, measurand. In the experiments explained in this chapter, we construct microstructures on fibers using a direct write CO2 laser machine instead of etching or polishing method, so that the optical fiber sensor can be constructed cheaply and rapidly. Two types of optical fiber sensor microstructures are explored here, i.e., in-fiber microchannel and in-fiber grooves. The in-fiber microchannel and in-fiber grooves enhance the interaction between the transmitted light through the fiber and the environment, increase the collection efficiency of fluorescence emission to a great degree, and enables the sensor to be able to detect very low light intensities emitted by low concentrations of fluorophores, so that the sensitivity of a fluorescence sensor is improved significantly. The investigations are conducted to find out how the length of the channels and the number of the grooves affect the sensitivity of the sensor as well as the smallest concentration of fluorescein solutions that can be detected. A comparison is also made into how a PMMA optical fiber sensor performs against a sensor fabricated in a silica fiber. Another advantage of in-fiber microchannel structure is that it also simplifies the fabrication of a microfluidic channel for fluid transport, because it can also function as a microfluidic channel. This chapter describes an optical fiber fluorescence sensor based on in-fiber microchannel or in-fiber grooves that can be cheaply and rapidly fabricated using a direct write CO2 laser system, and embedded in a microfluidic card suitable for lab-on-a-chip and μTAS applications. The general schematic diagram of the system, including fluid delivery, microfluidic card, optical fiber sensor, and detection system, is illustrated in Fig. 1.
Detection of Fluorescence Generated in Microfluidic Channel Using In-Fiber Grooves
405
2. Materials 2.1. Fabrication of In-Fiber-Microchannel Sensor
1. PMMA optical fiber (PMMA fiber), 1 mm core diameter with 10 μm fluorinated polymer cladding, or Plastic clad silica fiber (PCS fiber), 600 μm/750 μm (core/clad diameter) from Ceram Optec Industries, Inc. 2. Computer-controlled continuous wave CO2 laser direct writing machine, from Epilog Laser (Legend 24TT) as shown in Fig. 2, with two-dimensional robotic arms and z-stage for
Injection valve and sample loop with disposable syringe Buffer (PBS)
Waste
Voltmeter
550 nm Highpass filter PMT
470nm LED Syringe pump
Microfluidic card with in-fiber microstructures
Lens system to focus light and 470nm bandpass filter
Power Supply for PMT
Fig. 1. General schematic diagram of the system.
Fig. 2. Direct write CO2 laser machine used to fabricate microfluidic cards and microstructures in optical fibers.
406
Irawan and Tjin
laser focusing adjustment as sketched in Fig. 3. This CO2 laser machine is originally designed for engraving polymeric materials. 3. Computer-aided design software, such as CorelDRAW. 4. Ultrasonic cleaner (Bransonics, B1510). 5. Flame source, propane flame. 2.2. Fabrication of InFiber-Grooves Sensor
1. PMMA optical fiber (PMMA fiber), 500-μm core diameter with 10-μm fluorinated polymer cladding. 2. Computer-controlled continuous wave CO2 laser direct writing machine (Epilog Laser, Legend 24TT) with two-dimensional robotic arms and z-stage for laser focusing adjustment (see Figs. 2 and 3). 3. Computer-aided design software, such as CorelDRAW. 4. Ultrasonic cleaner (Bransonics, B1510).
2.3. Fabrication of Microfluidic Card with Microchannel
1. PMMA sheet, 1 mm thick. 2. Mylar sheet from Duponts, 0.002˝ thick, manufactured with double-layer 3M501 adhesives. 3. Mylar sheet from Duponts, 0.002˝ thick, with no adhesive. 4. Computer-controlled continuous wave CO2 laser direct writing machine (Epilog Laser, Legend 24TT) with two-dimensional robotic arms (see Figs. 2 and 3). 5. Computer-aided design software, such as CorelDRAW. 6. Nanoport assemblies (PN: N-126H) and drug delivery tubes for inlets and outlets tubing (Capillary PEEK Tubing, PN: 1569) from Upchurch Scientific. 7. Pressure device, a modified mechanical vice grip/clamp (to press and clamp Mylar and PMMA sheets together), and heater or oven. 8. Epoxy Adhesive (PN: N-008) from Upchurch Scientific.
Y-stage Mirror
Y-stage Mirror
Focusing Lens Z-stage Mirror
CO2 laser
Fig. 3. Layout of the direct write CO2 laser machine.
Working piece
X-stage
Detection of Fluorescence Generated in Microfluidic Channel Using In-Fiber Grooves
1. Light source, blue LED (NSPB300A, Nichia Chemical Industries) with wavelength centered at 470 nm (FWHM: 28 nm), that covers the excitation wavelength of fluorescein (see Fig. 4). 2. Power supply, current source (Phihong Dual Tracking DC Power Supply, PP-30-6-2B) for the LED. 3. Band-pass filter (SP Corion Filters, 470 ± 5 nm) to limit the spectral width of the excitation light. 4. Microfluidic chip with microchannel and optical fiber sensor embedded into it. 5. High-pass filter (Andover Corp, 550 nm) to filter out the excitation light entering the detector. 6. One-millimeter diameter pin hole made from a black anodized thin aluminium plate, installed at the front side of the detector to prevent the stray light entering the detector. 7. Two 40× objective lenses (infinity corrected objectives) from Olympus to focus the excitation light into the fiber and to collect the fluorescence emission at the other end of the fiber. 8. Convex lens (25 mm focal length) to focus the fluorescence light into the detector. 9. X–Y–Z micropositioners (Newport Corporation) to position the focusing and collection lens systems accordingly. 10. Optical detector, a minicompact module of photon multiplier tube (PMT, Hamamatsu H5784-02), which has spectral response of 300–880 nm. This PMT module, which includes a high-gain built-in amplifier and a built-in DC–DC highvoltage converter, is compact (2 cm × 2 cm × 6 cm) and
Intensity (a.u.)
2.4. Fluoresence Detection System
407
300
350
400
450
500
550
Wavelength (nm)
Fig. 4. Spectrum of the light source.
600
650
700
408
Irawan and Tjin
requires only 15 V DC (see Fig. 5), so that it is very convenient to develop a compact and sensitive detection system. 11. Power supply (Hamamatsu, C7169) for the minicompact module of PMT H5784-02 (see Fig. 6). This power supply is specially designed for the minicompact module of PMT. 12. A digital multimeter, Agilent 34401A Multimeter equipped with automatic data acquisition system and computer. 13. A graphical software Origin for calculating data statistical analysis and plotting graphs. 2.5. Fluorophore Solutions Used for Sensor Tests
1. Fluorescein powders purchased from Sigma-Aldrich. 2. Phosphate-buffer saline (PBS) at pH 7.4.
Fig. 5. PMT module used as the photodetector.
Fig. 6. Power supply for the PMT module.
Detection of Fluorescence Generated in Microfluidic Channel Using In-Fiber Grooves
409
3. Magnetic stirrer, micropipette, and beakers. 4. Microfluidic pump system (Precision Syringe Pump from Kloehn Co. Ltd, Las Vegas, Versa Pump 6 P/N 54022, see Fig. 7, and injection valve P/N V-451 and sample loop P/N 54022 from Upchurch Scientific, Oak Harbor, USA, see Figs. 8 and 9). 5. Syringe for the microfluidic pump, purchased from Kloehn Co. Ltd, Las Vegas. 6. Disposable syringe for the injection valve and sample loop.
(A) Buffer (PBS) Injection valve and sample loop Syringe pump Microfluidic card
Waste
(B) Fig. 7. a Precision syringe pump. b Connection of syringe pump, injection valve, sample loop, microfluidic card, and waste.
410
Irawan and Tjin
Injection valve
Sample loop Syringe
Fig. 8. Injection valve and sample loop.
1
1
Pump
6
2
Sample in
Pump
6
2
Sample in
Column
5
3
Sample waste
Column
5
3
Sample waste
4
4
(A)
(B)
Fig. 9. Schematic diagram of injection valve and sample loop as viewed from rotary knob. a Connections when samples are loaded into the sample loop. b Connections when samples are pumped through the microfluidic channel.
3. Methods The methods described below outline (a) fabrications of both types of optical fiber sensors, in-fiber-microchannel type and in-fibergrooves type, (b) fabrication of microfluidic card and embedding the optical fiber sensor in the microfluidic card, and (c) testing the sensors. 3.1. Fabrications of In-Fiber-Microchannel and In-Fiber-Grooves Sensors
Because the PMMA fiber has a very thin cladding, a few microns, removing the cladding before constructing the microchannels or grooves in PMMA fibers is not required. In contrast, removing the jacket and cladding of silica fibers is necessary, so that it is recommended to use type silica fibers that have jackets and claddings that can be removed easily, like the silica fibers suggested here, PCS fibers.
Detection of Fluorescence Generated in Microfluidic Channel Using In-Fiber Grooves
411
1. The microchannels and grooves that are going to be fabricated in the fibers are drawn precisely using CorelDRAW, then the drawings are transferred to the CO2 laser direct writing machine. In our tests, we made various lengths of microchannels, 5, 10, 15, 20, and 25 mm, and the width of all chanels is 100 μm. The numbers of grooves were made 5, 10, 20, 30, 40, and 50. Examples of CorelDRAW drawings for constructions of in-fiber microchannel and in-fiber grooves are shown in Fig. 10. 2. The cladding and jacket of silica fibers are removed using flame, then they are cleaned using Kimwipes and ethanol. 3. Before the fabrications of microstructures, the fibers are cleaned thorougly to get rid of from dust and finger prints that may stick on the surfaces of fibers. PMMA fibers should not be cleaned using acetone or ethanol. Wiping the PMMA fibers using wet Kimwipes should be enough. 4. The CO2 laser machine stage has no readily built optical fiber holder. We made the optical fiber holder from 2-mm thick of Acrylic sheet. A V-groove was constructed in the Acrylic sheet, and clamps were placed at two ends of the V-groove. Hence, for fabrications of in-fiber-microchannel and in-fibergrooves sensors, the optical fibers were placed and clamped inside the V-groove of Acrylic sheet to hold the optical fibers securely and firmly. 5. The setting of the CO2 laser machine, particularly the laser power and scanning speed, must be adjusted before printing microchannels or grooves on the fibers. The laser setting needs to be optimized and it depends on the characteristics of material and the quality of the CO2 laser system used. Ideally, low laser power should be used, but the laser beam should be often multiply scanned over the same area of the optical fiber to create well-defined structures. The microstructure depth increases linearly with the laser power set and the number of laser passes (12). It was found that because the glass transition temperature of silica is higher than PMMA, to obtain the same depth of structures, the silica core fibers require higher laser
Direction of laser scan
Written pattern
(A)
Direction of laser scan Written pattern
(B)
Fig. 10. Examples of CorelDRAW drawings for constructions of microstructures in optical fibers. a For in-fiber microchannel. b For in-fiber grooves.
412
Irawan and Tjin
power and/or a greater number of laser passes than the PMMA fibers do. In our experiments, to get approximately 200-μm depth microchannels in the PCS fiber, the power of CO2 laser machine was set at 32 W, 15 laser passes, and 3.25 mm/s scanning speed. On the other hand, to get the same depth in the PMMA fibers, the power of CO2 laser machine was set at 20 W and single laser pass. However, the laser setting may vary from machine to machine and from time to time. There are many factors affecting the laser power hitting the optical fiber, such as the conditions and the qualities of the lenses and the mirrors installed in the laser machine. Before the microstructure fabrications are started, the vertical position of optical fiber is adjusted to the position of the focal point of laser beam by adjusting the X-stage of CO2 laser machine. 6. The ablation process on the PMMA fibers forms a thin layer of white residue in the microstructures. This white layer can be partly eliminated by annealing the microstructures of fibers at 80°C for an hour. On the other hand, the ablation process on the silica core fibers still maintains the microstructures with clear surfaces. 7. Any debris or particles left by microfabrication processes in the fibers are cleaned by sonicating the microstructures in water. 8. The sizes of in-fiber-microchannel and in-fiber-grooves were estimated using Scanning Electron Microscope and/or surface profiler, actual SEM images of microstructures constructed in optical fibers using a direct write CO2 laser machine shown in Fig. 11. 3.2. Fabrications of Microfluidic Cards with Microchannels and Embedding the Optical Fiber Sensors in the Cards
1. Thin Mylars manufactured with double-layer adhesive are placed at the top and bottom of 1 mm PMMA (see Fig. 12a). 2. The drawings of microfluidic card and microchannel that are going to be fabricated on PMMA and Mylar sheets are precisely drawn using CorelDRAW, then the drawings are transferred to the CO2 laser direct writing machine. The width of microchannels should not be bigger than the diameters of the optical fibers used for sensor and the length of the microchannels is at least the same as that of the sensing area of the optical fibers. Examples of CorelDRAW drawings for fabrication of microfluidic card are sketched in Fig. 13. 3. The setting of the CO2 laser machine, particularly the laser power and scanning speed, must be adjusted before printing or cutting microfluidic cards and microchannels on PMMA and Mylar sheets. The laser setting needs to be optimized depending on the characteristics of material and laser machine. Since the fabrications of microfluidic cards and microchannels are based on cutting the polymers, their qualities are less sensitive to the setting of the CO2 laser machine as compared to
Detection of Fluorescence Generated in Microfluidic Channel Using In-Fiber Grooves
413
Fig. 11. SEM images of microstructures constructed in optical fibers using a direct write CO2 laser machine. a 600-μm core diameter PCS fiber with an in-fiber microchannel. b 1-mm core diameter PMMA fiber with an in-fiber microchannel. c Cross section of microchannel in 1 mm diameter PMMA fiber. d 500-μm core diameter PMMA fiber with in-fiber grooves.
Microfluidic Channel
PMMA
(A)
Mylar with double adhesives
(B)
Microfluidic channel Optical Fiber Sensor embedded into microfluidic channel
Mylar
(C) Fig. 12. Cross section of microfluidic cards. a Microfluidic card substrate before writing microfluidic channel. b After writing microfluidic channel. c After the optical fiber sensor imbedded in the microfluidic card.
the fabrications of the microstructures of optical fiber sensors. However, low scanning speed results in smoother microchannel wall, and too much laser power can cause the microchannel size bigger than the specified size due to the burning effect. After cutting, the cross section of microfluidic card and microchannel are as depicted in Fig. 12b.
414
Irawan and Tjin
(A)
(B)
(C) Fig. 13. Examples of CorelDRAW drawings for microfluidic card fabrications. a Middle layer. b Top layer. c Bottom layer.
4. The optical fiber sensor is placed at the bottom of the microchannel. The gaps between the optical fiber sensor and the wall of microchannel, if any, must be eliminated to avoid leak. It can be achieved by fitting the optial fiber sensor against the microchannel walls tightly (see Figs. 12c and 14a). 5. Then, the microchannels are closed using nonadhesive Mylar sheets pressed against the adhesive layer of Mylar attached to PMMA. To get good seal, they are pressed between two plates at temperature 40°C for few hours (see Figs. 12c and 15). 6. The inlet and outlet of microchannels are made at the top Mylar sheet (see Figs. 15 and 16). The inlet port is connected to a syringe pump through a Nanoport assembly and drug delivery tube, from Upchurch Scientific, Inc., Washington and the outlet is connected to a waste container. To prevent any leak, the nanoport assemblies are glued to the inlet and the outlet of microchannels using Epoxy Adhesive and the drug delivery tubes are tightened to the nanoport assemblies by finger tightening. 3.3. Testing the Sensors
1. The optical fiber sensor under the test, installed inside a microchannel of a microfluidic card, is placed in the experimental setup as shown in Fig. 14b. 2. The excitation beam, from a blue LED, is filtered by a bandpass interference filter (470 ± 5 nm), then it is focused into the one end of the optical fiber sensor by an objective lens. The circuit of the LED is shown in Fig. 17, the power supply, Phihong Dual DC Power Supply, is operated at constant current mode.
Detection of Fluorescence Generated in Microfluidic Channel Using In-Fiber Grooves
415
Fluorescence Light
Microchannel Fiber in microfluidic card Sensing part of fiber
Excitation Light
(A)
LSFF
Microfluidic card with in-fiber microstructures
550nm high pass filter PMT
470nm LED
LSCF 5.0V ?
V V
OFF
A A V?
A
COM
(B) Fig. 14. a Illustration of microfluidic card, microchannel, and embedded optical fiber sensor. b Experimental setup. LSFF: lens system to focus the excitation light into the fiber. It consists of a 470-nm LED, an objective lens, and 470-nm band-pass filter. The 470-nm LED is attached to the objective lens and the 470-nm band-pass filter is placed inside the objective lens. The distance between the objective lens and the fiber is adjusted using a micropositioner, so that the excitation light is coupled into the fiber efficiently. LSCF: lens system to couple the fluorescence emission from the fiber into the detector. It consists of an objective lens, a convex lens, and 550-nm high-pass filter. High-pass filter is attached to the PMT detector, and the convex lens is 25 mm from the PMT. The distance between the objective lens and the fiber is adjusted using a micropositioner, so that it reaches maximum fluorescence light coupling.
Inlet
A
Microchannel
Outlet
Groove
Optical fiber and grooves A (A) Microchannel
In-fiber channel
A
Optical fiber and channel A
(B)
Fig. 15. Optical fiber sensor embedded in a microfluidic channel. a Side-view and crosssection in-fiber grooves embedded in a microfluidic channel. b Side-view and crosssection in-fiber channel embedded in a microfluidic channel.
416
Irawan and Tjin
(A)
Drug delivery tubes
(B)
Nanoport assemblies Optical fiber Microfluidic channel and embedded optical fiber sensor
Fig. 16. Microfluidic card devices. a The substrate is clear PMMA. b The substrate is black PMMA.
Fig. 17. Circuit of the LED. The power supply, Phihong Dual DC Power Supply, is operated at constant current mode.
20 mA
LED
3. A collection lens system, an objective lens, to collect the fluorescence emission transmitted through the optical fiber sensor is installed at the other end of the fiber. The output of the collection lens system is focused into the detector, a minicompact PMT module, by a convex lens. Between the detector and the convex lens, a high-pass filter (550 nm) is installed to avoid the excitation beam from entering the photodetector that can cause noise of detection. A 1-mm diameter pin hole is also placed at the detector to prevent the stray light entering the detector. Since the photodetector used is a compact PMT module, which includes a high-gain built-in amplifier, the outputs of the photodetector are readily measured using a digital multimeter connected to a computer through a data acquisition system. This compact PMT module is also accompanied by a specifically dedicated power supply, so that the electrical connections of the detection system are simple (see Fig. 18). The digital multimeter used is equipped with built-in and convenient data acquisition interface and software. What we need to do is just to connect the multimeter to the one of the external ports of a computer and run the provided software. The data will be downloaded and stored automatically into a designated file. The time interval and frequency of data downloading can be determined accordingly. The details of the computer interface and the algorithms of the digital multimeter data acquisition system can be found in the manual of Agilent 34401A multimeter. 4. Various concentrations of fluorescein solutions are prepared using fluorescein powder and phosphate-buffer saline (PBS)
Detection of Fluorescence Generated in Microfluidic Channel Using In-Fiber Grooves
417
Cables to connect the PMT and power supply PMT
Cable to connect the output of PMT and a digital multimeter
Power Supply
Fig. 18. Electrical connections of the PMT module and power supply.
to stabilize its pH at 7.4. The solutions are stored inside syringes and protected from light exposure using a black cover to avoid photobleaching before measurements. The syringes are placed into the injection valve and the solutions are loaded into the sample loop (see Figs. 8 and 9). Then, the microfluidic syringe pump pumps PBS to drive different concentrations of the fluorescein solutions previously loaded inside the sample loop into the microchannels (see Fig. 7). This method is useful to avoid cross contaminations of sample. The connection of microfluidic syringe pump, injection valve, sample loop, and microfluidic card can be illustrated as shown in Figs. 1 and 7b. 5. The sensitivity of the optical fiber sensors is evaluated by filling the microchannels with known concentrations of fluorescein solutions. The tests are started from the lowest concentration of prepared fluorescein solutions, and then progress to consecutively higher concentrations. During the tests, the samples are continuously pumped through the microchannel at a constant speed, so that the photobleaching effect is minimized. 6. Between the measurements, the microchannel and the optical fiber sensors are cleaned thoroughly by flushing with buffer. 7. The fluorescence intensities of various fluorescein concentrations are corrected with respect to the background fluorescence, the fluorescence signals when the microfluidic channel is loaded by buffer only. Therefore, all the fluorescence intensities shown in Figs. 19–23 are corrected signals, which are
Irawan and Tjin 10000
Fluorescence Intensity (a.u)
9500 9000 8500 8000 7500 7000 6500 6000 5500 5000 0
5
10
15
20 25 30 35 40 Number of grooves
45
50
55
60
Fig. 19. Effect of the numbers of grooves to the sensitivity of sensor.
65000 60000
Fluorescence intensity (a.u)
418
55000 50000 45000 40000 35000 30000 25000 20000 15000 10000 5000 0 0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Fluorescein concentration (μg / l)
Fig. 20. Fluorescein concentrations vs. fluorescence intensities. The fiber is made from PMMA and the number of grooves is 20.
the background fluorescence signals when the buffer only is inside the microchannels are subtracted from the fluorescence signals measured when the fluorescein solutions are inside the microchannels. 8. The data are taken at least ten times. The means, the errors, and the graphs are calculated and plotted using a graphical
Detection of Fluorescence Generated in Microfluidic Channel Using In-Fiber Grooves
419
4400
Fluorescence Intensity (a.u)
4200 4000 3800 3600 3400 3200 3000 2800 2600 0.0
0.5
1.0 1.5 Channel length (cm)
2.0
2.5
Fluorescence intensity (a.u.)
Fig. 21. Effect of microchannel lengths to the sensitivity of sensor.
15000 14000 13000 12000 11000 10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 0 0.0
0.1
0.2
0.3 0.4 0.5 0.6 0.7 0.8 Fluorescein concentration (μg / l)
0.9
1.0
Fig. 22. Fluorescein concentrations vs. fluorescence intensities for in-fiber microchannel sensor made from PCS fiber.
software, Origin. The results show the sensors have good signal to noise ratio. For example, the optical fiber fluorescence sensor based on in-fiber microchannel is able to detect 0.005 μg/L of fluorescein in PBS solution with the signal to noise ratio better than 5.
Irawan and Tjin
Fluorescence intensity (a.u.)
420
15000 14000 13000 12000 11000 10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 0 0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Fluorescein concentration (μg /l)
Fig. 23. Fluorescein concentrations vs. fluorescence intensities for in-fiber microchannel sensor made from PMMA fiber.
4. Notes 1. The length of in-fiber microchannel or number of in-fiber grooves in the optical fiber sensor determines the sensitivity of the sensor. Therefore, it is important to check what is the optimum length of in-fiber microchannel or optimum number of in-fiber grooves for each designed optical fiber sensor. 2. The procedures of sensor tests are the same for both optical fiber sensors based on in-fiber microchannel and in-fiber grooves. 3. Fluorescein solutions diluted in PBS are used for sensor sensitivity tests and Microfluidic card substrates, PMMA and Mylar, may have auto background fluorescence. Therefore, all the fluorescence intensities shown in Figs. 19–23 are previously corrected by the background fluorescence, the fluorescence signal when only buffer is inside the channel. The readings are taken few times, at least ten times, to get statistical average data and error bars. 4. Fluorescein solutions are very sensitive to strong light exposure. Exposing fluorescein solutions to strong intensity of light can cause photobleaching. Therefore, before measurements, fluorescein solutions must be protected against light exposure. Fluorescence intensity and spectrum of fluorescein are also affected by pH. Hence, fluorescein powder must be
Detection of Fluorescence Generated in Microfluidic Channel Using In-Fiber Grooves
421
diluted using buffer with consistent pH. Fluorescein solutions are slightly hazardous in case of skin contact (irritant), eye contact, and ingestion. Therefore, it is recommended to use glove, safety glass, and laboratory coat while handling fluorescein solutions, and waste must be disposed according to local environmental control regulations. 5. The microfluidic pump causes vibrations, and the alignment of optical system, light source, detector, and microfluidic card is rather sensitive to vibrations. Therefore, it is recommended that the microfluidic pump and the sensing system are placed on different tables or placed on vibration absorbers. 6. The PMT used is a compact PMT module, which includes a high-gain built-in amplifier and a built-in DC–DC highvoltage converter, and is equipped with a special power supply. The outputs of PMT are readily measured using a digital multimeter. Cable connections from the PMT to the power supply and a digital multimeter can be refered to the user manual of H5784-02 PMT from Hamamatsu. 7. Mylar sheets from Duponts can be purchased with adhesive layers or without adhesive layer depending on our needs or orders.
Acknowledgments We thank to the Biomedical Research Council of Singapore for the financial support under the Singapore-University of Washington Alliance Programme, and Republic Polytechnic, Singapore for providing us facilities and supports to write the manuscript. We also thank to Chia Meng Tay for his helps to fabricate microstructures in the fibers.
References 1 Irawan, R., Tjin, S. C., and Fu, C. Y. (2005) Integration of a fluorescence detection system and a laminate-based disposable microfluidic chip. Microwave and Optical Technology Letters 45(5), 456–460 2 Yao, B., Luo, G., Wang, L., Gao, Y., Lei, G., Ren, K., Chen, L., Wang, Y., Hu, Y., and Qiu, Y. (2005) A microfluidic device using a green organic light emitting diode as an integrated excitation source. Lab on a chip 5, 1041–1047 3 Irawan, R., Tjin, S. C., Zhang, D., and Fang, X.-Q. (2005) Fluorescence detection system
and laminate-based disposable microfluidic chip. Chinese Optics Letters 3, S173–175 4 Irawan, R., Tjin, S. C., Yager, P., and Zhang, D. (2005) Cross-talk problem on a fluorescence multi-channel microfluidic chip system. Biomedical Microdevices 7(3), 205–211 5 Van Orden, A., Machara, N. P., Goodwin, P. M., and Keller, R. A. (1998) Single-molecule identification in flowing sample streams by fluorescence burst size and intraburst fluorescence decay rate. Analytical Chemistry 70, 1444–1451
422
Irawan and Tjin
6. Piruska, A., Nikcevic, I., Lee, S. H., Ahn, C., Heineman, W. R., Limbach, P. A., and Seliskar, C. J. (2005) The autofluorescence of plastic materials and chips measured under laser irradiation. Lab on a Chip 5, 1348–1354 7. Hawkins, K. R. and Yager, P. (2003) Nonlinear decrease of background fluorescence in polymer thin-films – a survey of materials and how they can complicate fluorescence detection in μTAS. Lab on a Chip 3, 248–252 8. Hubner, J., Mogensen, K. B., Jorgensen, A. M., Friis, P., Telleman, P., and Kutter, J. P. (2001) Integrated optical measurement system for fluorescence spectroscopy in microfluidic channels. Review of Scientific Instruments 72(1), 229–233
9. Irawan, R., Tay, C. M., Tjin, S. W., and Fu, C. Y. (2006) Compact fluorescence detection using in-fiber microchannels – its potential for lab-on-a-chip applications. Lab on a Chip 6, 1095–1098 10. Polynkin, P., Polynkin, A., Peyghambarian, N., and Mansuripur, M. (2005) Evanescent fieldbased optical fiber sensing device for measuring the refractive index of liquids in microfluidic channels. Optics Letters 30(11), 1273–1275 11. Tjin, S. C. and Irawan, R. (2006) Microfluidic immunoassay device. Patent Reference No. PCT/SG2006/000044. Pending patent 12. Klank, H., Kutter, J. P., and Geschke, O. (2002) CO2-laser micromachining and backend processing for rapid production of PMMA-based microfluidic systems. Lab on a Chip 2, 242–246
Chapter 24 Multiplex Integrating Waveguide Sensor: Signalyte™-II Shuhong Li, Yunqi Zhang, Platte Amstutz III, and Cha-Mei Tang Summary A platform to detect multiplex fluorescent labels was developed based on liquid phase implementation of the Integrating Waveguide Sensor detection principles. The liquid sample is held in a capillary cuvette with a lens at one end. The excitation light incident on the cuvette at 90° angle. The emitted fluorescence is efficiently gathered and propagated to the end of the waveguide cuvette, exiting via the lens to the detector. The capillary cuvette acts as a waveguide to efficiently gather the emission signal, providing high detection sensitivity for small sample sizes. Excitation sources ranging from 470 to 635 nm are four high-powered LEDs, allowing for multiplex fluorescence assays and a spectrometer is used to collect the signal from 390 to 790 nm. The cuvette can hold 1–35 μL samples. This technology can be used for a wide variety of assays and detection needs, such as FRET, end point PCR reading, immunoassays, chemiluminescence detection, multiplex quantum dots assays, polarization assays, etc. Key words: Biosensor, Spectrofluorometer, Fluorometer, Integrating waveguide sensor.
1. Introduction The Integrating Waveguide Biosensor (IWB) technology, which detects bound analytes on the inner surface of a capillary tube, was shown to be a very sensitive detection platform in another chapter in this book (1–3). Many assays, however, are more naturally performed in solution and this chapter describes an implementation of the Integrating Waveguide Sensor for liquid-phase assays, called Signalyte™-II. The basic principle of the liquid-phase integrating waveguide sensor (4) is shown in Fig. 1. Detection and quantitation are achieved by illuminating the cuvette (i.e., optical waveguide) at a 90° angle relative to the length of the waveguide and subsequent Avraham Rasooly and Keith E. Herold (eds.), Methods in Molecular Biology: Biosensors and Biodetection, Vol. 503 © Humana Press, a part of Springer Science + Business Media, LLC 2009 DOI: 10.1007/978-1-60327-567-5_24
423
424
Li et al.
Fig. 1. Schematic showing of the principle of liquid-phase integrating waveguide sensor detection, where the fluorescent labels are in the sample solution inside a cuvette. The excitation light impinges perpendicular to the cuvette. The cuvette and the sample together acts as a waveguide sending the signal out of the end of the cuvette. The lens at the end of the cuvette focuses the emission signal.
collection of emitted fluorescence from the end of the waveguide. There is a half-ball lens at the end of the cuvette, which focuses the signal down to the detection optical system. The principle of the detection is similar to the solid-phase integrating waveguide sensor described by Li et al. in Chap. 22 (3). Signalyte™-II provides high sensitivity for small sample sizes. Noise from excitation light is reduced because the excitation light is perpendicular to the detector. Signal from fluorescence is increased because the cuvette and the sample inside together act as a waveguide for the fluorescent signal from the whole sample to the end and exiting via the half-ball lens. Signalyte™-II provides four high-power LEDs for illumination. Wavelength of the LED can be varied. In this paper, they are 470, 530, 590 and
Multiplex Integrating Waveguide Sensor: Signalyte™-II
425
635 nm. Four excitation sources enable multiplex assays utilizing a wide range of fluorescent dyes. Creatv also have a single excitation source model called Signalyte™. Sensitivity results for Cy5 and FITC fluorescent labels using Signalyte™-II are described below. Signalyte™-II enables assays that the fluorescent labels are in the liquid such as FRET-based assays, chemiluminescence detection, end point PCR, and DNA and RNA detections without using thermal cycling amplification, immunoassays, multiplex quantum dots assays, polarization assays, etc.
2. Materials 2.1. Instrument Components
1. Excitation light source: LEDs of 470, 530, 590, and 635 nm (Lumileds, http://www.lumiledsfuture.com). 2. Band-pass filters for cleaning up LED light sources: 3. Dichroic filters: Chroma, Rockingham, VT. 4. LED collimation lens: aspheric lens. 5. Lens for illuminating cuvette: cylinder lens. 6. Lens for signal detecting: aspheric lens. 7. Long-pass filters: color glass filters. 8. Filter wheel (Thorlabs, Newton, NJ). 9. BTC111 Spectrometer (B&W Tek, Newark, DE). 10. SmartMotor: SM2315D (ONExia, West Chester, PA)
2.2. Software
1. Optical ray tracing: TracePro software (Lambda Research Corp., Littleton, MA). 2. Instrument control and user interface: LabVIEW (National Instruments, Austin, TX).
2.3. Cuvette, Chemicals, and Reagents
1. Cuvette (Roche Diagnostics, Indianapolis, IN). 2. Cy5 labeling kit PA35000 (GE Healthcare). 3. FITC dye: AC11925 from (Fisher, Pittsburgh, PA).
3. Methods Signalyte™-II consists of LED illumination system, cuvette, spectrometer detection system, sample movement system, and computer control system (Fig. 2); the details of each component part will be discussed in the following sessions.
426
Li et al.
Fig. 2. Schematic of a liquid-phase Integrating Waveguide Sensor, Signalyte™-II, with four excitation light sources, spectrometer as detector and computer control.
3.1. LED Illumination System
Currently, the illumination system uses four LEDs with central wavelength of 470 nm, 530 nm, 590 nm, and 635 nm respectively. The four LEDs provide the excitation for the most commonly used fluorescence dyes in market. Different choices of LED wavelength can be easily implemented to suite the user need. LED has more wavelength options and costs less than laser. High-power 3-W LED was chosen to compensate the fact that LED is less powerful in a narrow band of wavelength than laser. A band-pass filter is needed to clean up the spectrum in order to reduce noise in the signal wavelength regime. The beam from LED was collimated and expanded to about 2 mm × 30 mm using a collimating lens and a cylindrical lens in order to illuminate the cuvette. Compare to laser, LED has large emission angle. In order to collimate as much light as possible, a collimating lens with short focal distance and large NA is used. Each LED has its own collimating lens and band-path filter. All four LEDs and their optics are mounted onto a common housing so that the four LEDs share one light path to illuminate the cuvette. There are three dichroic filters in the housing. The dichroic filters are arranged in a way so that the light from each LED can go to the cuvette. A cylinder lens is attached at the end of the housing in order to focus the beam into 2 mm thick in one direction.
3.2. Simulation of Cuvette
To help selecting the cuvette, we conducted simulations of various geometries of cuvette using the optical ray tracing software TracePro (Lambda Research Corp., Littleton, MA).
Multiplex Integrating Waveguide Sensor: Signalyte™-II
427
The cuvette is equivalent of a regular capillary tube and a lens at the end, which focuses the light from the capillary waveguide and seals the cuvette. A variety of lens shapes was simulated, including: 1. Flat end 2. Round end 3. A half-ball lens at the end 4. A lens that is thinner than half-ball lens 5. A lens that is thicker than half-ball lens TracePro is used to simulate the cuvette. TracePro uses Monte Carlo simulations to set up the rays for ray tracing and to compute optical flux. TracePro accounts for absorption, reflection, refraction, and scattering of light according to the material and the surface properties of the capillary tube. To analysis the illumination of the flurescent labels in the liquid sample inside the cuvette, simulations were performed using incident rays of collimated 635-nm light and the result is shown in Fig. 3. In Fig. 3a, the trajectories of the incident rays from the excitation light (entering from the top of the figure), are bent by the borosilicate glass cuvette (index of refraction of n = 1.52), and reflected at the capillary tube surfaces due to change of index of refraction. The cuvette is filled with buffer. The rays shown in Fig. 3a are displayed in shades of gray: black rays denote the rays with flux equal or greater than 67% of the incident light; medium gray rays denote the rays with flux between 33 and 67% of the incident light; while light rays denote the rays with flux less than 33% of the incident light. Figure 3b is the irradiance map on the inner surface of the glass cuvette, which shows that the sample inside the cuvette is fairly evenly illuminated. We simulate the transmission of fluorescent signal in cuvette. The cuvette is filled with solution with fluorescent dyes inside. The fluorescent dyes are considered point source in TracePro simulation. Light emits from the fluorescent dyes is guided by the cuvette to the lens at its end, and then focused. The size of the focal point and angular distribution are compared. Among the different lens shapes, the cuvette with a half-ball lens provides the best result with the smallest focusing point as well as a moderate angular distribution. Figure 4 shows the power density at the focal plane. The cuvette has a small focal point of less than 2 mm in diameter. The cuvette with a half-ball lens is commercially available. A 35-μL cuvette made with borosilicate glass with a 1.2 mm ID (inner diameter) and 1.55 mm OD (outer diameter) is available through Roche Diagnostics (Indianapolis, IN) and the borosilicate glass cuvette is produced by Drummond Scientific Company (Broomall, PA).
428
Li et al.
Fig. 3. a Results of TracePro ray tracing simulation of excitation light trajectories on borosilicate cuvettes (1.2 mm inner diameter and 1.55 mm outer diameter) with water inside. The liquid sample is well illuminated. b Irradiance map of cuvette inner surface generated from Fig. 3a.
Multiplex Integrating Waveguide Sensor: Signalyte™-II
429
Fig. 4. Results of TracePro simulation intensity plot at the focal spot (0.3 mm from the end of the half-ball lens) from the 35-μL borosilicate cuvettes with water inside.
3.3. Signal Detection System
Signal detection system consists of a set of lenses that focus the signal from the cuvette, through a long-pass color glass filter to the spectrometer. A long-pass filter for each excitation LED wavelength is needed to block the excitation LED light, so that only the emission signal is detected by the spectrometer. All four long-pass filters are incorporated into a filter wheel and controlled by computer. The input to the spectrometer is a 0.29 NA slit. The spectrometer is sensitive from 390 to 790 nm with a cooled CCD inside. To achieve high sensitivity in Signalyte™-II, noise has to be reduced. Noise can be classified into two sources: ● Background noise from the instrument Nonspecific binding of fluorescent dyes in sample The background noise exists even when there is no fluorescence in cuvette. The background noise is from the instrument itself. It is caused by LED, lenses and filter, and electronic noise from spectrometer. In illumination system, LED has wide spectrum that extends to signal range, although band-pass filter is used to cut off the unwanted spectrum, because of the imperfection of optics, there are still some noise coming through the optics and goes to the detection system. Long-pass filter in detection
●
430
Li et al.
system blocks most of the excitation light, and the part that leaks into spectrometer forms the background noise. Electronic noise of spectrometer also contributes to background noise. In order to minimize background noise, collimation of LED source, choice of filters, and optimizations of lenses are all important. Electronic noise is reduced by choosing spectrometer with cooled CCD. Nonspecific binding comes from fluorescent dyes that are not captured by purpose. This can be reduced by optimizing the assay. 3.4. Sample Movement
Samples are loaded into a holder that can hold nine cuvettes. For the eight cuvettes on the right side of the holder, the distance between the adjacent cuvettes is compatible with a multichannel pipetter. After finishing the assay in a 96 well plate, samples can be transferred to the holder using a multichannel pipetter. The cuvette at the “0” position on the left is for reference sample (for example, negative control). In order to differentiate the reference sample from other testing samples, the distance between the eighth cuvette and the ninth cuvette is larger than the distance between other adjacent cuvettes. The illumination and signal detection systems are at fixed positions. A Smartmotor™ is incorporated to move each cuvette to the testing position. After a sample is moved to the testing position, a LED is turned on to illuminate the cuvette, and the signal is detected by a spectrometer. It is important that the cuvette should be aligned with the illumination light, and the lens at the bottom of the cuvette should be aligned with the detection system. The alignment of optics is guaranteed by mounting the illumination optics, signal detection optics, and tube holder in a main mounting piece machined with precision. Sample holder movements are programmed and calibrated so that every cuvette is moved to the same testing position.
3.5. Computer Control System
●
LED control
●
Motion control
●
Spectrometer control
Filter wheel control When Signalyte™-II runs a test, one tube is moved to the testing position, stopped; one LED is turned on, spectrometer takes the data, and the data are transferred to computer, processed, and displayed on the screen. A computer control system programmed in LabVIEW handles LED on/off, sample motion control, spectrometer data acquisition, filter wheel motion control, data processing and display, etc. All four LEDs are connected to a digital I/O control device. The digital I/O control device, motor, spectrometer, and filter wheel are connected to computer through a USB HUB. They are converted to USB port first if they are not originally USB. ●
Multiplex Integrating Waveguide Sensor: Signalyte™-II
431
Fig. 5. User interface print screen. The full spectrum is the plot at the bottom. The bar chart at the upper left shows the peak value chosen by the user and the spectrum on the top right shows the user-selected section.
A user interface is also needed for the user to interact with the instrument. Figure 5 shows the user interface. The user can input a note for each sample, choose the exposure time, select the spectrum range to display, and pick the peak emission wavelength to generate the result. Up to eight samples can be tested with one load. A zero concentration sample cuvette will also be tested to provide the background reference. The user can enter notes for each tube. A motion control system is incorporated to move each cuvette to the testing position. A progress bar shows the number of the sample that is under testing. After all the eight samples are tested, a full spectrum of all the samples is shown on the bottom of panel. A bar chart that shows the signal at the emission peak designated by the user, and a subset of the spectra selected by the user is shown on the top right panel. Both the bar chart and the subset of the spectrum can be saved as EXCEL and/or JPEG files. 3.6. Results
A model of the exterior of the Signalyte™-II is shown in Fig. 6. On the upper right side, there is a door that can be opened when user
432
Li et al.
Fig. 6. Photography of a Signalyte™-II instrument.
loads samples. Eight samples, plus one reference, can be loaded in the holder and can be tested in about 1 min. In order to test the limit of detection for different LED light source, we selected two typical fluorescence dyes: Cy5 was tested by using red LED (635 nm) and FITC was tested by using blue LED (470 nm). Cy5 dye was diluted to serial dilutions using PBS buffer. Background noise is obtained by testing PBS buffer when illuminated by red LED. This background noise is subtracted from the signals. The results of the spectrum are shown in Fig. 7. X-axis is wavelength in nanometer unit. Y-axis is relative intensity without unit. The limit of detection of Cy5 is 2.5 pM (0.088 fmol in 35 μL). FITC dye was diluted to serial dilutions using PBS buffer. Background noise is obtained by testing PBS buffer when illuminated by blue LED. This background noise is subtracted from the signals. The results of FITC testing are shown in Fig. 8. The limit of detection of FITC is 25 pM (0.88 fmol in 35 μL). The difference in limit of detection for Cy5 and FITC is mainly because the difference in excitation efficiency and difference of dye’s photon efficiency. The excitation efficiency for Cy5 excited at 635 nm is only 65%, while the excitation efficiency for FITC excited at 470 nm is only 45%. On the other hand, Cy5 has higher photon efficiency than FITC that means with the same excitation power, Cy5 convert more energy into emission.
Multiplex Integrating Waveguide Sensor: Signalyte™-II
433
Fig. 7. Testing result with the background removed for Cy5 using 635-nm LED showing the limit of detection of 2.5 pM for a 35-μL sample size.
Fig. 8. Testing result with the background removed for FITC using 470-nm LED showing the limit of detection of 25 pM for a 35-μL sample size.
If a user wants to test a sample with unknown concentration, a standard curve needs to be established first. A standard curve is a plot of the signal vs. concentration. First, a serial dilution of assay is prepared with the same protocol of the testing sample, except with known concentrations. Second, the serial dilution samples are tested on Signalyte™-II. Then, after subtracting background from all the signals, a plot of signal vs. concentration is generated. That is usually a linear curve. The relationship between concentration and signal can be expressed in a formula. Then, the unknown sample is tested. The signal for the unknown sample is obtained by subtracting background. Concentration of the
434
Li et al.
unknown sample can then be calculated according to the standard curve. The standard curve only needs to be tested once for the same assay. The generation of the curve and the calculation of concentration can also be programmed in the instrument.
4. Notes 1. IWB was originally developed by Ligler et al. at the Naval Research Laboratory (NRL) in Washington, DC. Creatv Microtech licensed the technology and developed Signalyte™-II to use the technology on liquid phase platform. 2. High-power LED generates heat during operation. Temperature rise will cause color shift, which means change in emission wavelength. To avoid this, a heat sink is needed to be mounted on the back of LED. 3. System alignment is important for obtaining high sensitivity. Cuvette is only 1.55 mm outer diameter, and it needs to be aligned with both the excitation and detection optics. When user loads the sample into the holder, make sure every cuvette is in proper position, not tilted. 4. Signalyte™-II requires a warm up time of about 20 min. This is for the CCD in spectrometer to be fully cooled in order to get a low and stable background.
References 1. Feldstein, M. J., MacCraith, B. D., Ligler, F. S., US Patent No. 6,137,117 issued on 10/24/2000 2. Ligler, F. S., Breimer, M., Golden, J. P., Nivens, D. A., Dodson, J. P., Green, T. M., Haders, D. P., Sadik, O. A., Integrating waveguide sensor. Anal. Chem. 74, 713–719, 2002
3. Li, S., Amstutz, P., Tang, C. M., Hang, J., Zhu, P., Zhang, Y., Shelton, D. R., and Karns, J. S., Integrating Waveguide Sensor, Chapter 22 in this book 4. Tang, C.-M. and Amstutz, P., III, Sensitive Emission Light Gathering and Detection System, US Patent Application, filed on October 3, 2006
Chapter 25 CCD Based Fiber-Optic Spectrometer Detection Rakesh Kapoor Summary Highly sensitive and cost effective measurement tools are required in biotechnology research and applications. Fluorescence provides very simple, cost effective, and sensitive methods in most of the biosensor techniques. Spectrometer is an essential tool for any kind of spectroscopic measurements. A charged coupled device (CCD)-based fiber optic spectrometer is highly compact, light weight, and an extremely easy to use tool. In this chapter, we have described the use of CCD-based fiber-optic spectrometers in detection of fluorescence signal from a fiber-optic-based sensor. The method can easily be extended to fluorescence detection in any other application. Key words: Spectrometer, CCD, Charge coupled device, Fiber-optic, Fluorescence, Fiber-optic sensor.
1. Introduction Biotechnology research and applications require highly sensitive and cost effective measurement tools, thus fluorescence provides very simple, cost effective, and sensitive methods in most of the biosensor techniques. These biosensors can be used to detect various kinds of bioactive compounds like an enzyme, an antibody, or a nucleic acid (1). The fluorescence is generated either by the attached fluorescence molecules to the analytes (2, 3) or it could be autofluorescence generated by the analyte. To achieve better results and faster measurements, new technologies are emerging. A charged coupled device (CCD)-based fiber-optic spectrometer is one such device. These spectrometers are highly compact, light weight, and extremely easy to use. Here we have described the use of CCD-based fiber-optic spectrometers
Avraham Rasooly and Keith E. Herold (eds.), Methods in Molecular Biology: Biosensors and Biodetection, Vol. 503 © Humana Press, a part of Springer Science + Business Media, LLC 2009 DOI: 10.1007/978-1-60327-567-5_25
435
436
Kapoor
in detection of fluorescence. To illustrate various aspects of its use we have chosen its application in a fiber-optic-based sensor (4).
2. Materials 2.1. CCD-Based Fiber-Optic Spectrometer
1. High-resolution Spectrometer (HR2000+, Ocean Optics, Inc., Dunedin, FL). 2. HR Grating 400–800 nm range; Installed (H9, Ocean Optics, Inc., Dunedin, FL). 3. Order-sorting detector filter; Installed (OFLV-400800, Ocean Optics, Inc., Dunedin, FL). 4. Detector Collection lens; Installed (L2, Ocean Optics, Inc., Dunedin, FL). 5. Optical bench entrance aperture, 200 μm width; Installed (SLIT-200, Ocean Optics, Inc., Dunedin, FL). 6. Laptop computer with USB2 port and Window XP operating system (Latitude D620, Dell, Inc.). 7. 600-μm core silica clad, collection fiber with SMA 905 terminators on both ends (P6002-VIS/NIR, Ocean Optics, Inc., Dunedin, FL).
2.2. Laser Diode
1. Laser diode 405 nm and 4 mW Power (LDM405, Thorlabs, Inc., Newton NJ). 2. 405-nm narrow band filter (NT43104, Edmund Optics, Inc, Barrington, NJ). 3. Lens tube (SM1L05, Thorlabs, Inc., Newton NJ). 4. Lens tube Spanner Wrench (SPW602, Thorlabs, Inc., Newton NJ).
2.3. Collection/ Excitation Chamber
1. 30-mm Cage Cube (C4W, Thorlabs, Inc., Newton NJ). 2. Two packaged collimation lenses of f = 11.0 mm and NA = 0.25 (F220SMA-A, Thorlabs, Inc., Newton NJ). 3. Two collimation lens Mounting Adapters (AD11F, Thorlabs, Inc., Newton NJ). 4. Kinematic Mounting Plate (B4C, Thorlabs, Inc., Newton NJ). 5. Cage Cover Plate (B1C, Thorlabs, Inc., Newton NJ). 6. Short-pass filter (SP500R/25, Maier Photonics, Inc., Manchester Center, VT). 7. 1˝ Optic Mount for mounting Beam splitter (B5C, Thorlabs, Inc., Newton, NJ). 8. 1˝ End cap (SM1CP1, Thorlabs, Inc., Newton NJ).
CCD Based Fiber-Optic Spectrometer Detection
437
9. Laser Line Long-pass Filter 457.9NM (NT47501, Edmund Optics, Inc, Barrington, NJ). 10. Lens tube (SM1L10, Thorlabs, Inc., Newton NJ). 11. Fiber adapter (SM1SMA, Thorlabs, Inc., Newton NJ). 2.4. Fiber Probe
1. 600-μm core silica clad fiber (Fiber-600-VIS/NIR, Ocean Optics, Inc., Dunedin, FL). 2. Bare Fiber Terminator (BFTU, Thorlabs, Inc., Newton NJ). 3. SMA 905 Fiber Connector (10640A, Thorlabs, Inc., Newton NJ). 4. Diamond Wedge Scribe (S90W, Thorlabs, Inc., Newton NJ). 5. Bunsen Burner (17928027, VWR Scientific Products). 6. Ultrasonic Cleaner (1533722C, Fisher Scientific International). 7. Centrifuge (13100510, Fisher Scientific International). 8. Hydrofluoric acid 50% (A1461LB, Fisher Scientific International). 9. PBS (BP24384, Fisher Scientific International). 10. Dry Acetone (AC32680, Fisher Scientific International). 11. Sodium Hydrochloride (Fisher Scientific International). 12. Sodium Bicarbonate (Fisher Scientific International). 13. Aminosilane Reagent (3-Aminopropyltriethoxysilane) (80370, Pierce Biotechnology, Inc., Rockford, IL). 14. Cysteamine hydrochloride (MEA) (30078, Sigma-Aldrich Co., St. Louis, MO). 15. EDTA (17892, Pierce Biotechnology, Inc., Rockford, IL). 16. Desalting Column (CS-800, Princeton Separations, Adelphia NJ). 17. Antibodies to be immobilized. 18. Sulfo-SMCC cross linker (22322, Pierce Biotechnology, Inc., Rockford, IL).
3. Methods 3.1. Selection of Spectrometer and Its Accessories
1. For the biosensor applications it is assumed that real time signal monitoring is the preferred method. This requires that the spectrometer should have a high speed of spectra recording. HR2000+ from Ocean Optics has the capability of transferring spectra continuously at a rate of 1 ms per spectra. Such a high speed is ideal for real time recording of the signal.
438
Kapoor
2. Grating is the main component of any spectrometer. Choice of grating depends on the required spectral range. For most of biosensor experiments, the spectral range is generally from 400 to 800 nm (4). For this kind of spectral range, a preinstalled H9 grating should be ordered with the spectrometer. 3. Most of the time higher order spectra interfere with required first order spectrum therefore it is always safe to get a spectrometer with preinstalled long-pass filter. HR2000+ should be obtained with a preinstalled OFLV (Variable Long-pass Order-sorting Filters) filter to the detector’s window to eliminate second- and third-order effects. This will give a clean first-order spectrum. 4. Another parameter to decide while getting a spectrometer is the slit width as most of the CCD-based miniature spectrometers come with a preinstalled slits. The width of the slit determines the amount of light entering the spectrometer. Higher the width more light will enter the instrument but smaller slit width gives better optical resolution. If no slit is used, the diameter of the fiber connected to the spectrometer determines the size of the entrance aperture. In most of the biosensor experiments, the recorded spectra are either from a molecular dye or from the autofluorescence of the sample. The spectral width of these spectra is typically from 30 to 100 nm (4). A highest available slit width of 200 μm is reasonable to get maximum light collection efficiency and optimum resolution of 5 nm. 5. Generally slit height is larger than the CCD array height, to utilize maximum input intensity a cylindrical lens is needed to focus the light from the tall slit onto the shorter detector elements. The spectrometer should be ordered with a preinstalled L4 lens. This cylindrical lens is fixed to the detector’s window and it increases light-collection efficiency and reduces stray light. It is also useful in a configuration with a largediameter fiber for low light-level applications. 3.2. Assembling Collection/Excitation Chamber
A photograph of assembled collection chamber is shown in Fig. 1: 1. Laser diodes produce significant emission in the red spectral range besides lasing at the laser wavelength. This red light can interfere with the fluorescence signal. To avoid this unnecessary emission (see Note 1) a laser line filter NT43104 (Edmund Optics, Inc.) should be installed in front of the Laser diode LDM405 (Thorlabs, Inc.). Filter should be first mounted into the lens tube SM1L05 (Thorlabs, Inc.) as this lens tube fits well on the laser diode head. 2. Mount the laser assembly (see Note 2) on one of the four side holes of the cage cube C4W (Thorlabs, Inc.).
CCD Based Fiber-Optic Spectrometer Detection
439
Fig. 1. a Photograph of the collection chamber along with a fiber probe. b Photograph showing interior of the collection chamber.
3. Intensity of small amount of scattered or back-reflected laser light is generally comparable with the fluorescence signal and thus can interfere with its detection. To avoid this interference, direct entry of the scattered and back-reflected laser light should be prevented from entering the spectrometer. This can be achieved by using laser line long-pass filter NT47501 (Edmund Optics, Inc.). Transmission of this filter for fluorescence signal wavelengths (457–630 nm) is around 98% but for laser wavelength (405 nm) its transmission is only 10−6%. First mount the filter NT47-501 (see Note 3) in a SM1L10 lens tube and then mount one end of this tube on one of the side holes (right angle to the laser mounted hole) of the cage cube C4W. On the other end of the lens tube, mount the collimation lens F220SMA-A (Thorlabs, Inc.) with the help of lens Mounting Adapters AD11F (Thorlabs, Inc.). Collimation lens should face toward the cage cube. This is the fluorescence collection port (Fig. 1) of our experimental setup. Connect one end of a 2-m fiber P600-2-VIS/NIR (Ocean Optics, Inc.) to this port (SMA connector side of the collimation lens assembly) and other end to the spectrometer input port. 4. Third hole of the cage cube opposite (180°) to the laser mounted hole is used as the output port or fiber probe port.
440
Kapoor
Second collimation lens F220SMA-A (Thorlabs, Inc.) should be mounted with the help of lens Mounting Adapters AD11F (Thorlabs, Inc.) on this hole. Collimation lens should face toward the cage cube. 5. Block the fourth hole with the help of 1″ end cap SM1CP1 (Thorlabs, Inc.). 6. Short-pass filter SP500R/25 (Maier Photonics, Inc.) should be mounted on a 1˝ Optic Mount B5C (Thorlabs, Inc). This filter has about 98% reflectivity from 530 to 689 nm and about 90% transmission for 405 nm. The mounted short-pass filter should further be mounted on the kinematic mounting plate B4C (Thorlabs, Inc.). 7. Kinematic mounting plate with short-pass filter SP500R/25 should be fitted on bottom of the cage cube in such a way (see Note 4) that filter is inside the cage and makes 45° angle with the incoming laser light (Fig. 1). The angle should be such that the reflected fraction of the laser should hit the blind port with end cap SM1CP1. For optimum alignment of the short-pass filter replace the SM1CP1 end cap (see Note 5) with fiber adapter SM1SMA (Thorlabs, Inc.). Connect one end of the P600-025-VIS/NIR fiber to the fiber adapter and point other end toward a white paper. Turn the laser switch on and make sure the shutter in front of laser is on. Now slowly rotate kinematic mounting plate for maximum laser output intensity through the P600-025-VIS/NIR fiber. At the maximum output intensity position tighten the four screws of the kinematic mounting plate. Once the alignment is over, replace the fiber adapter with end cap. 8. Now the setup is ready to record fluorescence signal. A photograph of the experimental setup is shown in Fig. 2 and a schematic of the setup is shown in Fig. 3. 3.3. Preparation of Fluorescence Fiber Probe
1. Use a diamond wedge scribe S90W (Thorlabs, Inc.) to cut a 10-cm long fiber piece from Fiber-600-VIS/NIR (Ocean Optics, Inc.)
3.3.1. Fabrication of Silica Fiber Probe
2. With the help of a Bunsen Burner, burn off 2 cm buffer coating from one end of the 10-cm fiber probe. 3. Rinse the probe with distilled water in ultrasonic cleaner for 4 min. 4. Prepare a 15% Hydrofluoric acid solution by mixing about 2.2 parts of distilled water with one part of 50% hydrofluoric acid. 5. Dip buffer less part (2 cm) of probe in to the 15% hydrofluoric acid until etch to the desired diameter of 270 μm (for 15% HF, the etching rate is about 18.5 μm/h) (see Note 6). 6. Again rinse the etched probe with distilled water in ultrasonic cleaner for 4 min.
CCD Based Fiber-Optic Spectrometer Detection
441
Fig. 2. Photograph of the experimental setup. The laser used in this setup has a separate laser diode driver but LDM405 from Thorlabs, Inc. has a built-in driver.
Spectrometer
LD
Collection fiber
ND
BF
CL
Computer Collection Chamber
LF
CL SF
Fiber Probe
Fig. 3. Schematic of evanescent wave excited fluorescence-based fiber-optic sensor. LD Laser diode, CL Collimation lens, BF Band-pass filter, ND Neutral density filter, LF Long-pass filter, and SF Short-pass filter.
7. After etching burn off the buffer coating from the whole 10 cm long probe on a Bunsen burner. 8. Rinse the probe with distilled water in ultrasonic cleaner for 4 min.
442
Kapoor
9. Rinse the probe with 1 N NaOH in ultrasonic cleaner for 4 min. 10. Rinse the probe with distilled water in ultrasonic cleaner for 10 min. 11. Rinse the probe with acetone in ultrasonic cleaner for 10 min. 3.3.2. Immobilization of Antibodies on Silica Fiber Probe Aminosilylate the Fiber Probe Surface
1. Thoroughly wash and dry the glass or quartz fiber probe surface to be coated (see Note 7). 2. Prepare a 2% solution of the Aminosilane Reagent (3-Aminopropyltriethoxysilane) in acetone. For example, mix 1 part Aminosilane Reagent with 49 parts dry (i.e., water-free) acetone. Prepare a volume sufficient to immerse required probe length. (2 cm probe length will require 400 μl of solution in 1.0 mL without-cap plastic tube or 400 μl in 0.6 mL centrifuge tube.) 3. Immerse probes in the diluted reagent for 60 s. 4. Rinse surface with dry acetone.
Partially Reduce Antibody to Produce Sulfhydryls for Coupling
1. Prepare 0.5 M MEA stock solution by dissolving 6 mg Cysteamine hydrochloride (MEA) in 100 μl Coupling Buffer. (For future use, store this solution at 4°C). 2. Prepare 0.1 M EDTA stock solution by dissolving 3.72 g EDTA power in 100 mL PBS buffer. (This solution could be stored at room temperature.) 3. Prepare Reducing Agent by mixing 50 mM MEA and 10 mM EDTA into Antibody solution. (For total 20 μL of Antibody before dilution, mix 2.5 μL 0.5 M MEA, 2.5 μL 0.1 M EDTA, and 20 μL 0.5 mg/mL Antibody in a 0.2 mL tube.) 4. Incubate the Reducing Agent for 90 min at 37°C (after this step start working on Subheading “Maleimide-Activate the Amino-Modified Surface,” after completing Subheading “Maleimide-Activate the Amino-Modified Surface,” come back to step 5 of this section). 5. Prepare a Desalting Column CS-800 (Princeton Separations) by adding 650 μL EDTA buffer in Desalting Column powder and incubate at least 30 min. Then spin it for 2 min (the maximum amount of reagent for one column could desalt is 100 μL). 6. Purify the reduced antibody from the Reducing Agent using the Desalting Column (proceed to Subheading “Cross-Link Sulfhydryl-Containing Antibody to Activated Surface” as you must have already completed section 3).
Maleimide-Activate the Amino-Modified Surface
1. Prepare a 4.2 mM Sulfo-SMCC cross linker solution in PBS. (For total 400 μL of cross linker solution, add 0.8 mg Sulfo-SMCC
CCD Based Fiber-Optic Spectrometer Detection
443
powder to 400 μL PBS buffer). This solution must be used immediately to avoid hydrolysis (see Notes 8 and 9). 2. Incubate probes in cross linker solution for 1 h at room temperature. 3. Rinse the modified surface with Coupling Buffer (see Note 10). Cross-Link SulfhydrylContaining Antibody to Activated Surface
1. Dilute the reduced antibody solution to about 5–10 μg/mL. 2. Submerge the maleimide-activated fiber probe with the antibody solution. The antibody solution may be diluted in Coupling Buffer to a volume sufficient to submerge the required probe length (generally 2 cm). 3. Incubate it for 4 h at room temperature. 4. Thoroughly rinse the surface with Coupling Buffer to ensure that only covalently attached antibody molecules remain. 5. The surface is now ready to use for detection assays and other applications. Depending on stability of the particular antibody, the surface material may be dried for storage or kept covered in buffer containing 0.02% sodium azide.
3.4. Recording Spectrum/Signal
When immobilized antibodies come in contact with the specific analyte, the analyte get attached to the probe. Now either the analyte itself produce some kind of autofluorescence, or fluorescence can be generated by attaching another labeled antibody to this analyte (sandwich method). 1. Mount the prepared fiber probe on a bare fiber terminator BFTU (Thorlabs, Inc.) by inserting its unetched side into SMA 905 fiber connector attached to BFTU. A photograph of a mounted fiber is shown in Fig. 4. 2. Connect mounted fiber-optic sensor probe to the output port (see Note 11) of the collection chamber. 3. First install the spectrometer software (see Note 12) on the laptop Latitude D620 (Dell, Inc.) and then connect the
Fig. 4. A photograph of a mounted fiber probe.
444
Kapoor
spectrometer HR2000+ to the USB2 port of the laptop with the provided USB cable. Once the HR2000+ is installed, it must be configured by operating software’s Configure Spectrometer options so that operating software (OOIBase32) recognizes the HR2000+ spectrometer. Details can be obtained from the operation manual of the software. 4. For power stabilization, it is always a good idea to start the laser few minutes before starting the experiment. 5. Start the spectrometer operating software and open the laser shutter. Record the fluorescence signal by clicking the appropriate button. Try to adjust the integration time of the spectrometer to a value, such that the maximum spectral amplitude is less than 90% of the full scale. This will ensure that recorded signal is not saturated. 6. For maximizing the signal adjust the three hex adjuster fine alignment screws (see Note 13) on the back (Fig. 1) of kinematic mounting plate B4C (Thorlabs, Inc.) 7. Once the maximum signal value is known and integration time is adjusted, close the laser shutter and record the dark back ground signal (see Note 14) and store it. 8. Save this signal and turn the background deduction button on. (Details about these various operations can be read in the software manual of the spectrometer). 9. Now record the spectrum of the collected (see Note 15) fluorescence and save it to the file.
4. Notes 1. Red tail emission from the laser diode can be further reduced if the laser is operated at much higher power than the required power and neutral density filters are placed in the laser path along with the band-pass filter to reduce the power level to the required value. 2. All the SM1 series lens tubes and cage cube (Thorlabs, Inc.) holes have compatible threading therefore mounting of these tubes on the cage cube is straightforward. 3. Make sure that the coated part of the laser line filter NT43-104 faces the laser side. Generally there is a mark on the ring of the filter which indicates the coated side. 4. Make sure that the coated part of short-pass filter faces the spectrometer port or collection port.
CCD Based Fiber-Optic Spectrometer Detection
445
5. Replacement of end cap SM1CP1 with fiber adapter SM1SMA and its connection with the fiber P600-025-VIS/ NIR is only done for alignment purpose. After completing the short-pass filter alignment, fiber adapter should be replaced back with end cap. 6. Make sure the probe stand vertically in the acid solution. 7. Perform steps 2 and 3 in a fume hood. 8. The Reducing Agent (Subheading “Partially Reduce Antibody to Produce Sulfhydryls for Coupling”) should be made first because it needs 90 min incubating time. 9. The better reaction of NHS-ester is in the environment of pH = 8.3. However, maleimide group will slowly hydrolyze and loses its reaction specificity for sulfhydryls at pH values >7.5. For these reasons, conjugations with these cross linkers are usually performed at pH 7.2–7.5. 10. The maleimide-activated surface may be dried and stored desiccated at 4°C for later use. 11. This connector is for 630-μm outer diameter fiber. For other diameter fibers appropriate SMA 905 fiber connector should be used. 12. This port is called output port as the excitation laser comes out of this port although it also acts as an input port for collected fluorescence. 13. Never connect the spectrometer to the computer, without first installing the required software and drivers. This is a onetime step if you are using the same computer for future experiments. 14. This is one time adjustment for improving the signal to noise ratio. 15. Always keep the integration time of the back ground signal and actual signal same.
References 1. Mehrvar, M., Bis, C., Scharer, J.M., Young, M.M., Luong, J.H., 2000. Fiber-optic biosensors – trends and advances. Anal. Sci. 16, 677–692 2. Hirschfeld, T.E., Block, M.J., 1984. Fluorescent immunoassay employing optical fiber in capillary tube, US Patent No. 4447546
3. Andrade, J.D., VanWagenen, R.A., Gregonis, D.E., Newby, K., Lin, J.N., 1985. Remote fiberoptic biosensors based on evanescent-excited fluoroimmunoassay: Concept and progress. IEEE Trans. Electron. Devices ED-32(7), 1175–1179 4. Kapoor, R., Kaur, N., Nishanth, E.T., Bergey, E.J. Halvorsen, S.W., Prasad, P.N., 2004. Biosens. Bioelectron. 20, 345–349
INDEX A Absorption sensor.......................................................... 311 Achromatic lenses.........................................38, 42, 46, 382 Acousto-optic tunable filter ................................... 293–304 A/D converter ............................ 38, 43, 222, 313, 317, 318 Adherent cell ......................................................... 204, 212 Affinity selection ............................................................. 73 Alkaline denaturation .................................................... 351 Alkaline-phosphatase .............................326, 327, 331–334 Alkanethiols ...........................................................5, 18, 75 Alkylthiols ................................................................. 74, 75 Amine groups ......................... 108, 111–113, 158, 193, 194 Amine-modified surfaces ...............................110, 422–443 Amine-thiol........................................................... 108, 110 Aminodextran........................................................ 171, 173 Amorphous silicon ................................................ 361, 362 Amplification .........................3, 5, 6, 13, 21, 247, 307, 308, 311–314, 317, 321, 343, 345, 349–350, 372, 425 Analog-to-digital conversion ......................................... 329 Angular scan ................................. 71, 90–92, 102, 103, 133 Antibody................................ 3–7, 9, 12–14, 16–19, 21–24, 28, 31, 33, 34, 39, 44–46, 73, 77, 81, 82, 84, 85 Antibody purification ............................................ 277, 327 Anti-streptavidin antibody .......................................... 6, 13 ATP................................................ 338, 339, 341, 342, 344, 346, 348, 353, 354, 356–358 Autofluorescence ........................... 180, 239, 242, 248–250, 252, 256, 257, 264, 393, 403, 404, 435, 438, 443
B Bacteriophage .................................................275, 371, 372 Bacteriophage DNA .............................................. 371, 372 Bandpass filter ............................... 222, 248, 252, 255, 275, 276, 294, 311, 376, 378, 405, 414 Bandwidth ......................133, 182, 185, 245, 247, 262, 263, 294–296, 301, 307, 309, 312–314, 318, 321, 322 Base implant layer ......................................................... 329 Basic fibroblast growth factor (bFGF) ...100, 123, 124, 135 BCIP/NBT ........................................................... 327, 332 Beads ....................................................6, 13, 347, 350–352 Beam directing .......................................241, 251, 254–255 Beam splitter ....................................................4, 7, 10, 436 Beamsteering ................................................................... 38 Bending .................................. 241, 252, 255, 311, 318, 386 Binding inhibition assay .............................84, 85, 167, 175
Binding peptide ............................................................. 124 Biotin........... 6, 13, 49, 52, 54, 57–63, 74, 75, 78–81, 95, 97 Bipolar photodiode .........................................325, 326, 328 Bipolar semiconductor ................................................... 327 Blocking .............................9, 12, 13, 16–18, 78, 81, 82, 96, 101, 133, 135, 278, 285, 295, 300, 304, 321, 332 Blood 3, 5, 9, 13, 189, 190, 192–194, 197–199, 206, 242, 252, 283 Board assembly .............................................................. 316 Board preparation .................................................. 314–316 Board testing ......................................................... 316–317 Bonding ............................................ 12, 185, 265–266, 270 BoNT-A .........................................................259, 267–269 Botulinum neurotoxin ............................259, 261, 267–269 BSA blocking .......................................................... 12, 332 Buried layer ........................................................... 328, 329
C Campylobacter jejuni.................................................275, 288 Capillary ........................................ 139–141, 143–146, 151, 155, 156, 161, 163 Capillary electrophoresis.........................221–235, 361–373 Capillary waveguides ..............................390–392, 400, 427 Carcinoembryonic antigen (CEA) .............5, 102, 128, 129 Casein 76, 78, 82, 96, 110, 134 CCD.........................................8, 14, 15, 17, 25, 26, 28–31, 34, 73, 85, 103, 172, 190–192, 196, 197, 203–218, 221, 223–228, 234 Cell culture .................................................................... 212 Cell-targeting ................................................................ 217 Chopper .........................................................368, 369, 392 Chromium ....................................................11, 18, 23, 364 Cleaning .................................................. 39, 43, 44, 46, 86, 116, 173, 177, 186, 194, 196, 363, 364, 369, 370, 401, 425 CMOS ...................................................246, 313, 314, 325 CO2 laser ....................................................................... 406 Collimated beam ................................. 28, 29, 46, 103, 183, 244, 254, 302, 392, 426 Collimated light .......................8, 14, 19, 28, 40, 42, 43, 46, 244, 257, 302, 304, 382, 393, 397, 401, 426, 427 Collimating lens ..............................................92, 102, 171, 240, 241, 244, 251, 253, 254, 257, 304, 391, 392, 397, 425, 426, 436, 439–441 Colloidal gold .............................. 94, 95, 99, 101, 118–119, 127, 128, 135
447
BIOSENSORS AND BIODETECTION 448 Index Colorimetric .......................................................... 204, 259 Combinatorial ............................................................... 362 Competitive ....................................... 18, 84, 190, 230, 273, 281, 285–286, 290 Competitive immunoassay.............. 273, 281, 285–286, 290 Conjugated ................................. 3, 6, 14, 16, 389, 391, 399 Continuous flow immobilization Convex lens ................................. 42, 46, 95, 103, 278, 296, 298, 300–303, 368, 391, 397, 415, 416 Cooled CCD ......................................... 242, 257, 259, 264, 275, 276, 279, 287, 429, 430 Cooling.......................................... 214, 239, 243, 249–250, 253, 254, 260, 262, 264, 269, 308, 321 CorelDRAW ..........................................406, 411, 412, 414 Coupling optics ..................................................72, 73, 171 Critical angle ..........................................................4, 40, 90 Crosslinking .................................................................. 112 Current amplifiers ..................................327–329, 376, 386 Current-to-voltage conversion .............................. 385, 386 Current-to-voltage converters ............................... 327–329 Cy5...... .......................................... 277, 282, 285, 286, 290, 389, 391, 393, 396, 399, 400, 425, 432, 433
D Dark current .................................. 172, 242, 247–250, 255, 308, 309, 342, 343, 362, 364 Data acquisition............................................................. 386 Deoxynivalenol ...............................................274, 281, 285 Detection cell .........................................223–227, 233–235 Dextran..........................74, 95–97, 104–108, 116, 124, 327 Dielectric layer...................................... 50, 54, 90, 245, 257 DMAP ...................................................................... 76–78 DNA...........................................35, 74, 111, 124, 135, 139, 141, 158, 159, 161, 173, 190 analysis............................................................. 337–359 sequencing ................................ 337, 339, 352–354, 371 dNTPs... ................................. 338, 344, 345, 348, 353–357 Double-layer adhesive ................................................... 412 Double-sided adhesive................................................... 270 Double sided adhesive tape ........................................... 270 Dual-waveguide interferometric ...................................... 50 Dynode ...................................................378, 379, 381–384
E E.coli..... ................................... 101, 126, 307, 318, 325–334 EDC..... ...................... 12, 13, 17, 19, 23, 33, 76, 77, 79, 96, 97, 99, 100, 102, 105–108, 110, 111, 116, 117, 122, 130, 133, 142, 157, 184, 193–195, 197 EDC-NHS.................................. 12, 13, 19, 105, 106, 108, 110, 111, 116, 117, 122, 133 EL-CCD ................................................259, 261, 263–269 Electrode .........................363, 366, 371, 373, 378, 381, 382 Electroluminescence .............................................. 259–271 Electrophoresis ....................... 221–235, 337, 361–373, 383
ELISA... ..................................... 12, 18, 129, 193, 287, 288 Ellipsometry .....................................................22, 167, 169 Emission filter ....................................... 261, 264, 265, 275, 276, 279, 294, 312, 319 Epitaxial layer ........................................................ 328, 329 Epoxy.... ..........................................................268, 406, 414 Etching ...........................152, 153, 163, 315, 363–365, 369, 404, 440, 441 Evanescent wave ..................4, 37, 53, 61, 90, 190, 390, 441 Excitation filter............................... 261, 264, 265, 294, 312
F α-Fetoprotein ................................................................ 5, 9 Fiber optic ..................................... 37, 38, 40–43, 141, 144, 146, 147, 171, 311, 390, 404, 435–445 Fiber probe ..................................... 226, 234, 237, 439–443 Flow cell ........................................ 4, 10–12, 15–18, 24, 25, 30, 31, 50, 52, 54, 57, 60, 62 Flow through ................................................77, 79–81, 309 Fluidics ............................................ 76, 77, 79, 81, 85, 139, 262, 264, 279–281, 287, 339, 361 Fluorescein .................................... 223–224, 297, 302–304, 362, 372, 404, 407, 408, 416–421 Fluorescein isothiocyanate isomer I (FITC)................. 221, 223–224, 228, 230, 235, 259, 261–263, 265, 267–268, 383–384, 425, 432–433 Fluorescence ...........................3–19, 46, 129, 139, 180, 190, 221–226, 230–233, 235, 239–242, 244–246, 248–250, 252, 255–257, 259–261, 263–264, 273, 279, 293–304, 307, 309–313, 318, 322, 337, Fluorescence activated cell sorting................................. 362 Fluorescence detector ..................... 223, 232, 260, 361–373 Fluorogenic peptide substrate .................259, 261, 267–269 Fluoroimmunoassays ............................................. 273–290 Fluorometer .....................259, 262, 265, 267, 269, 319, 423 Fluorophore ....... 3, 6, 7, 14–15, 17, 278, 282, 311, 404, 408 Flurocence ..................................................................... 262 Folding .............. 94, 135, 252, 255, 284–285, 298, 301, 390 Food..........................................65, 74, 76, 78–79, 273, 275, 277–278, 283–284, 288, 316, 337 Frame grabber.........................................296, 298–299, 302 Frequency variation ....................................................... 321 FTIC.... ..........................................................221, 231, 235 Fumonisin.............................................................. 281, 285 Functionalization ...................... 76, 156, 158–160, 173, 193
G Gelatin............................................................278, 285, 290 Glucose .................................................................. 104, 370 Gold thin film ..............................................4, 5, 10, 14, 19 Gold wire......................................................................... 11 Grating ........................... 69, 91, 92, 95, 103, 162, 163, 186, 189–200, 227, 229, 233, 294–295, 300, 436, 438 Grating coupler ............................................... 69, 189–200
BIOSENSORS AND BIODETECTION 449 Index H
L
Half-ball lens ...................363, 366–369, 371, 424, 427, 429 Hapten .................................................................. 129, 175 Hemicylindrical prism ......................................7, 10–11, 16 His-tagged..................................................98, 99, 116–117 Human immunodeficiency virus (HIV) .................................................203–218, 318 Hybridization .................................... 74, 89, 100, 124, 125, 135, 177, 186, 353 Hybridoma ...................................................................... 39 Hydrogel.....................................................74, 95, 109, 173 Hydrophobic ......................................... 43, 74, 95, 98, 107, 115, 116, 193
Label-free ...............................4, 40, 49, 58, 63, 65, 89–135, 139–164, 167, 168, 179, 189–200 Labeling .............................................. 16, 37, 49, 139, 179, 251–253, 256, 277, 281, 282, 361, 362, 371, 425 LabVIEW ................................... 40, 42, 46, 145, 147, 149, 163, 222, 318, 319, 369, 391, 398, 425, 430 Laminating ............................................................ 266, 270 Laminator...................................................................... 262 Laser.... ............................. 4, 7, 8, 10, 14, 15, 24, 28, 53–55, 61, 63, 72, 91, 92, 94, 95, 102, 141, 145, 147–150, 162, 163, 170, 190–191, 195, 197, 222, 230, 239, 241–246, 251–255, 257, 258, 260, 261, 264, 265, 269, 276, 278 Laser-induced-fluorescence (LIF) ..........221, 222, 229, 375 Laser line generator ....................................................... 391 Leaky waveguide ............................................................. 89 Lens..... .............................................. 4, 7, 8, 10, 14–15, 24, 25, 28, 29, 46, 92, 95, 102, 103, 171, 182, 183, 191, 203, 204, 206–209, 212, 214, 216, 239–242, 244–245, 251–257, 259, 261–262, 264 Light.................................................. 4–6, 8, 10, 11, 14–16, 18, 19, 21, 22, 25–29, 31, 34, 37, 40–43, 46, 50, 53, 54, 59, 66, 67, 69, 70, 72, 73, 85, 90–92, 94, 95, 102, 103, 118, 133, 140, 142, 146, 147, 152, 162, 163, 167–168 Light box ........................................205–208, 210, 215–217 Light-emitting diodes (LED) .........................92, 102, 121, 221–226, 228–235, 260, 294, 297, 309–311, 316–320, 328, 375, 382–383, 405, 407, 414–416, 424–426, 429, 430, 432,–434 Light scatter ........................... 133, 270, 281, 303, 362, 370 Line scanning ...........................................37, 38, 40, 43–45 Liposome ............................................................... 109, 115 Liquid core optical ring resonator (LCORR) .................................................. 139–164 Liquid crystal variable retarder (LCVR) ..............................................293, 296–303 Luciferase ...................................... 338–339, 342, 344, 345, 347, 356–358 Luminescence ........................ 221–235, 259–271, 337–339, 341–344, 354, 356, 375, 423, 425 Luminescence detector .......................................... 221–235 Luminometric .................337, 339, 344, 347, 348, 352, 358 Lysozyme .............................................................. 101, 126
I IC clock ......................................................................... 318 Illumination...................... 4, 19, 28, 67, 72, 85, 91, 92, 133, 172, 214, 215, 217, 224–226, 235, 259–271, 273, 304, 309, 311, 317, 387, 390–393, 395, 397, 423–430, 432 Immobilization .................................... 6, 12–13, 16, 17, 58, 73–81, 89, 90, 95–97, 104–106, 108–113, 115–117, 122, 134, 158–160, 164, 173–174, 326, 332, 351, 401, 442 Immunoassay ..................................... 3, 5, 7, 10, 12–14, 17, 204, 240–242, 252, 257, 273–290, 389, 398, 400, 423, 425 Impedance ............................................................. 314, 317 Incident angle ......................4, 5, 10, 14, 16, 19, 31, 50, 168 In-fiber .................................................................. 403–421 In-fiber grooves ..................................................... 403–421 Injecting system ......................................................... 30, 34 Injection valve.........................................405, 409, 410, 417 Integrated circuit (IC) .................... 314, 317, 318, 325, 361 Integrating Waveguide Biosensor (IWB) .........................................389–401, 423, 434 Integration time...............................................15, 227–229, 231, 232, 234, 247, 255, 444, 445 Interference filter ......................................... 8, 15, 102, 133, 294, 311, 318, 361–363, 365–366, 368, 372, 373, 383, 401, 414 Interferometer ................................................180–182, 184 Interferometry ................................................. 22, 179–186 Iris......................................... 4, 7, 10, 38, 41, 42, 255, 296, 298–301, 303, 369 Isoelectric point (PI)...............................104, 108, 119, 133
K Kinetic.. ..................................21, 28, 33, 49, 60–62, 71, 89, 90, 93, 102, 104, 119, 120, 124, 129, 139, 158, 161, 175, 176, 178, 180, 185, 186, 199, 362 Kinetics analysis .............................................................. 33 Kretchmann ....................................................................... 4
M Magnesium fluoride (MgF2) crystal ............................. 381 Magnet ........................................ 6, 39, 41, 66–68, 72, 141, 190, 222, 262, 314, 347, 351, 381–382, 409 Magnetic beads.................................................................. 6 Mask................................................. 39, 315, 363, 369–370 Mask aligner .................................................................. 363
BIOSENSORS AND BIODETECTION 450 Index Media..... ..........................................................55, 67, 76, 78, 79, 81, 205, 212, 232, 319 Medium ..............................51, 53, 56, 59, 66, 82, 101, 168, 181, 190, 205, 327, 394, 427 Membrane ..............................................109, 128, 223, 362 Metal film...............................................................4, 66, 69 Microchannel ..........361–362, 367–369, 371, 382, 403–421 Microfluidic .........76, 90, 139, 143, 171, 361–373, 403–421 Monochromatic light.....................................21, 28, 70, 72, 85, 90–91, 102, 172 Monochromators ...................................186, 233, 293–294, 307, 311, 376 Monte Carlo simulations ....................................... 393, 427 Motion controller ............................... 42, 43, 141, 145, 147 MUA solution ........................................................... 22, 23 Multiplexer .....................................................170, 340, 343 Mumps Virus ............................................................ 37–46 Mycotoxin aflatoxin ............................................... 281, 285 Mylar.... .......................................... 406, 412–414, 420, 421
N Nanobeads ............................................................... 3, 6, 13 Nanoparticles................................................................... 22 Neutral density filter................8, 14, 19, 191, 195, 441, 444 NHS/EDC ................................................................ 12, 17 N-hydroxylsuccinimide (NHS) .....................12–13, 19, 23, 52, 54, 76, 77, 79, 96, 97, 99, 100, 102, 105–108, 110, 111, 114, 116–118, 122, 130, 133, 142, 157, 171, 173, 193–195, 197, 252, 277, 282, 445 Noise sources .......................... 246, 247, 249, 250, 257, 309
O Objective ............................8, 15, 24, 28, 54, 182–183, 213, 301–303, 407, 414–416 Ochratoxin .....................................................274, 281, 285 Oligo probe immobilization .................................. 158–159 One base extension .........................................341, 354, 356 Op Amp .................................................313, 314, 318–320 Optical fiber .........................41, 43, 44, 171, 226, 234, 376, 377, 382–383, 386, 403–407, 410–417, 419, 420 Optical path length (OPL)............. 140, 168, 179–181, 235 Optical ring resonator............................................ 139–164 Optical table ............................................. 7, 10, 38, 42, 297 Oriented immobilization ................................................. 77 Oxide layer .............................................................. 54, 329
P Pathogen.........................................................260, 325–334 PDMS. ........................... 274, 276, 278–281, 284, 286–287 Peltier................................................................93, 250, 262 Peptide purification ....................................................... 100 Peristaltic pump ............................. 7, 10, 12, 141, 150, 163, 170, 171, 196, 276, 280 Pesticide........................................................................... 84
Phase detection.......................................................... 21–35 Phase difference....................................................25, 26, 91 Phase distribution ......................... 21, 22, 26–28, 31–32, 34 1,4-Phenylene diisothiocyanate (PDITC) ............................................193, 195–198 Phospholipid ....................................................98, 115, 116 Photobleaching.........................................19, 249, 417, 420 Photodetector .................................. 70, 142, 147, 149, 150, 163, 221, 242, 307, 311, 314, 317, 318, 320–322, 386, 408, 416 Photodiode .........................................4, 7, 8, 10, 14, 15, 55, 56, 71, 172, 240, 243, 259, 260, 307–323, 325–334, 337, 342–344, 361–364, 369, 373 Photo-diode array.................................................. 337–359 Photodiode detector ............................. 7, 8, 10, 14, 15, 314 Photolithography............................328–329, 364–366, 369 Photomultiplier ....................... 17, 222, 240, 259–260, 269, 308, 321, 375–385, 389, 391 Photonic crystal ......................................................... 49–63 photoreceiver ............................................21, 142, 317, 318 Photoresist ..............................................315, 363–366, 395 Phycobiliprotein ............................. 221, 222, 231, 233, 234 PID........................................................................ 251, 254 Piezoelectric .................................................................. 295 Pinholes lens............................................................ 24, 102 Piranha solution ...................................... 11, 18, 22, 23, 54, 86, 172, 194, 196, 198, 364 Planar waveguide ............................ 238, 241, 251, 254–256 Plano-convex lens .......................................92, 95, 102, 103 Plasma.................................9, 11–13, 16–19, 86, 242, 283, 364–366, 369 Plasmid .......................................................................... 206 Platinum electrode......................................................... 371 Polarized light ............................... 4, 10, 25–26, 28, 40–42, 91, 167, 300, 301 Polarizer.......................... 25, 31, 40, 41, 72, 90–92, 95, 103, 294, 295, 302, 304, 307 Polarizing filter .......................................................8, 14, 15 Polarizing prism ................................. 25, 26, 28, 29, 38, 40 Poly(diallyldimethylammonium chloride) (PDDA)..............................................37, 39, 44, 45 Polyethylene glycols (PEG) ....................171, 173, 278, 284 Poly L-lysine ............................................................. 39, 44 Polymerase chain reaction (PCR) ................................. 160, 260, 338, 345–346, 349–351, 362, 423, 425 Polymethyl methacrylate (PMMA).......................... 16, 30, 171, 404–406, 410–414, 416, 418, 420 Prism..................................... 4, 7, 8, 10, 11, 15, 16, 19, 23, 25, 26, 28, 29, 31, 34, 38, 40, 42, 46, 52, 69, 71, 72, 85, 90–92, 95, 103, 162, 295, 300 Prostate-specific antigen ................................................ 105 Protease ................................................................. 139, 141 Protein A ........................................ 102, 126, 129–131, 133 Protein array ....................................... 37, 38, 40, 41, 43–46 Protein coupling ............................................................ 189
BIOSENSORS AND BIODETECTION 451 Index PVC.... ......................................................... 7, 10, 16, 196 Pyrosequencing....................... 337–349, 352–354, 356–358
Q Quartz tungsten......................................................... 38, 46
R Raman spectroscopy ...................................................... 180 Reactive surface ................................................76, 177, 193 Reflectometric Interference Spectroscopy ............. 167–178 Reflectometry ........................................................ 167–178 Refractive index ......................................... 3, 4, 6, 8, 11, 16, 18, 19, 21–23, 25–28, 32, 35, 37, 43, 50, 54, 62, 65–70, 72, 73, 85, 89–93, 95, 103, 118, 139–141, 152–155, 163, 169–171, 177, 180, 181, 185, 189–192, 197–199, 310, 372 Resonance............................................ 3–19, 21–35, 37, 40, 42–45, 54–56, 58, 62, 65–86, 91, 92, 103, 125, 133, 134, 150, 157, 180, 189, 222, 242 Resonance wavelength ....................................37, 40, 42–45 Resonant Mirror (RM)...............................49, 89–135, 189 RF generator ..................................................297–299, 302 RNA.........................................................................425 Rotation stage................................... 4, 8, 10, 14, 15, 38, 42
S Salmonella ........................................275, 289–291, 398–400 Sample loop .................................... 172, 405, 409, 410, 417 Sandwich ..................................................... 3, 5, 12, 14, 17, 83, 84, 94, 127, 128, 174, 177, 241, 254, 256, 273, 281, 285–287, 295, 316, 325, 326, 370, 389, 390, 398, 400, 443 Sandwich assay .............................. 12, 14, 93–94, 127, 128, 273, 281, 285, 390, 400 Screening ................................ 172, 189, 190, 273, 274, 362 Secondary antibody .......3, 5–7, 9, 12–14, 16, 17, 19, 84, 95 Self-assembled monolayers (SAMS) .............................. 5, 14, 18, 23, 74, 76–79 Semiconductor laser ....................... 241–244, 251, 363, 367 Serodiagnosis ..............................................37–40, 189–200 Serum...............................................9, 13, 23, 37, 44, 45, 76, 78, 79, 82, 98–102, 121, 122, 142, 177, 184, 189, 190, 192–195, 197–200, 205, 277, 281, 283, 358 Signal amplification ......................................3, 5, 6, 13, 312 Signal digitalizing .......................................................... 391 Silanization .................................... 104, 133, 172, 173, 195, 196, 326, 332, 399 Silanized chips ................................ 104, 172, 173, 196, 332 Silicon 7, 10, 11, 54, 119, 249, 254, 328, 329, 361, 362 Silicon oxide ............................................................ 54, 329 single-stranded DNA (ssDNA)............................. 338, 339 SNPs 337, 344, 354–357 Spatial modulation .................................................... 21–35 Spectral filtering ............................................................ 297
Spectral SPR ............................................................. 37–46 Spectrofluorometer ........................................................ 423 Spectrometer ..................................... 37, 38, 40–43, 45, 70, 73, 85, 170–172, 182, 183, 186, 223–228, 233, 277, 423, 425, 426, 429–430, 434, 444, 445 Spectroscopy ........................................ 3–19, 167–178, 180, 189, 190, 293 Spin coater..................................................................... 363 SPR array ................................................ 22, 27, 37, 38, 40, 41, 44–46, 71, 73, 76 Sputter. .......................................................39, 52, 363–365 Stability .18, 67, 73, 74, 78, 79, 85, 185, 194, 222, 314, 318, 375, 384, 443 Stabilization of baseline ................................................... 12 Staining ..... 46, 203, 204, 206, 210, 212, 214, 217, 218, 255 Staphylococcus aureus enterotoxin B (SEB) ..........................................274, 277, 288, 390 Streptavidin ........................................... 3, 6, 13, 49, 51, 52, 57–63, 76, 78–80, 97, 98, 100, 106, 111–112, 123–125, 134, 193, 347, 351 Striper 328 Surface functionalization ....................... 76, 78, 81, 96, 106, 108–111, 113, 156, 158–160, 173, 179, 193 Surface passivation .................................189, 190, 193, 199 Surface plasmon field-enhanced fluorescence.............. 3–19 Surface Plasmon Resonance (SPR) .................3–19, 21–35, 37, 49, 51, 65–86, 167, 180, 189, 242, 312 Syringe pump ............75, 141, 150, 151, 405, 409, 414, 417
T Temperature control ......................... 62, 149, 150, 186, 316 Thermal noise......................... 242, 247, 260, 262, 269, 309 Thermister ............................................................. 142, 150 Thermoelectric cooler.....................................239, 250, 253 Thrombin .............................................................. 165, 187 Titering ................................................................... 46, 344 Total internal reflection (TIR).........................4, 14, 40, 50, 51, 53, 140, 241, 257, 273, 280 Toxin........................................ 84, 259–261, 267, 274, 277, 281, 283–286, 288, 290, 387, 390 TracePro simulation.......................... 392, 395, 396, 427429 Translation stages ................................................ 39, 41–44 Triton X-100 ........................................................39, 44, 46 Trypsin .......................................................................... 156 Tumor.. ................................................................ 3–19, 128 Tunable filter ......................................................... 293–304 Tunable laser................................... 141, 147, 149, 150, 162 Tween... ....................................9, 12, 13, 17, 18, 44, 76, 81, 96–101, 261, 391, 399 Tween-20 .................................... 13, 39, 44, 75, 76, 95–99, 101, 102, 125, 135, 194, 197, 277, 278
U UV glass ........................................................................ 381
BIOSENSORS AND BIODETECTION 452 Index V Valve.... ...............................30, 34, 170, 171, 276, 279, 280, 287, 340, 341, 405, 409, 410, 417 Vector....................................................................... 67, 181
W Wafer... ........................................... 363–366, 369, 370, 372 Waveguide ...........................................................49–51, 54, 89–91, 102, 140, 162, 190, 193, 195, 198, 221, 239, 241, 242, 248, 250, 251, 254–258, 260, 273
Waveguide sensor .................................. 241, 242, 254, 256, 389–391, 423–434 Wavelength interrogation .................................... 37–38, 50 Wavelength scan ................................................91, 92, 102, 103, 109, 133 Wollaston prism .......................................25–26, 28, 29, 34
Z Zoom lens .............................................................203, 204, 208, 212, 214
