Microarray Methods for Drug Discovery (Methods in Molecular Biology)

Methods

in

Molecular Biology™

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

Microarray Methods for Drug Discovery Edited by

Sridar V. Chittur Department of Biomedical Sciences, School of Public Health, Center for Functional Genomics, University at Albany-SUNY, Rensselaer, NY, USA

Editor Sridar V. Chittur, Ph.D. Department of Biomedical Sciences Schools of Public Health Centre for Functional Genomics University at Albany-SUNY Rensselaer, NY USA [email protected]

ISSN 1064-3745 e-ISSN 1940-6029 ISBN 978-1-60761-662-7 e-ISBN 978-1-60761-663-4 DOI 10.1007/978-1-60761-663-4 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2010921137 © Humana Press, a part of Springer Science+Business Media, LLC 2010 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. While the advice and information in this book are believed to be true and accurate at the date of going to press, ­neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Printed on acid-free paper Humana Press is a part of Springer Science+Business Media (www.springer.com)

Preface The postgenomic era presents a multitude of challenges for scientists in all areas of science. The information overload from new discoveries in genomics and proteomics highlight how little we really know about the functioning of a cell. The advent of Next-Generation Sequencing technologies promises to make our genetic blueprint available to the common man. The availability of the plethora of biological information has lead to the development of new areas of science and the coining of new “omics” terms including transcriptomics, methylomics, toxicogenomics, pharmacogenomics, metabolomics, lipidomics, and so on. Remarkable research is being conducted to understand the various aspects of human health and how processes like histone modifications, promoter usage, alternative splicing, posttranscriptional, and posttranslational modifications contribute to disease. The advent of systems biology has unified chemists and biochemists alike in the struggle to eradicate or treat human disease. Microarrays have blossomed into a fast developing and cutting-edge technology that promises to become a major component of personalized medicine. The 1990s witnessed a boom in many areas including genome sequencing, combinatorial chemistry, and computers, all of which have contributed to the development of microarray technology from its infancy into a mature tool. The growing potential of this tool is evident from the number of publications since 1991 when Fodor et  al. of Affymax (now Affymetrix) first described the microarray prototype. The number of publications using microarrays in 1990–1999 was approximately 300, while over 8,200 journal articles have been published in the first half of this year alone. The usage of microarrays in experiments designed to identify differential gene expression is well accepted now. Since the seminal work of Pat Brown’s group at Stanford, microarrays became a technology that could be developed by any individual researcher using simple spotting robots. Currently, few laboratories make their own arrays due to the availability of commercial cost-effective solutions that are less prone to variation. Microarrays have evolved from traditional oligonucleotide arrays for gene expression into tools that have even more fascinating applications. Today, one can find arrays containing not only DNA oligonucleotides but antibodies, carbohydrates, small molecules, and enzymes. The diversity of these applications makes this field exciting and limited only by imagination. This, however, makes it challenging for an inexperienced scientist wishing to enter this arena. I am still surprised by the lack of general information amongst individuals regarding how to design and conduct a microarray experiment. As we get exposed to the concept of “Personalized Medicine”, we find ourselves confounded by the myriad of platforms and applications attributed to microarrays. This book aims at enlightening individuals with all levels of experience about some of the

v

vi

Preface

most common applications of microarrays in drug discovery and development. I hope that this book will serve as a reference for students and scientists alike who would like to enter this exciting field but are a bit intimidated. I am especially grateful to the many friends, colleagues, and family who encouraged me in this effort. Rensselaer, NY

Sridar V. Chittur

Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

v ix

  1 Multicenter Clinical Sample Collection for Microarray Analysis . . . . . . . . . . . . . . Tony S. Mondala, Daniel R. Salomon, and Steven R. Head   2 Isolation of Total RNA from Transgenic Mouse Melanoma Subsets Using Fluorescence-Activated Cell Sorting . . . . . . . . . . . . . . . . . . . . . . . Scott Tighe and Matthew A. Held   3 Microarray Analysis of Embryonic Stem Cells and Differentiated Embryoid Bodies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alexander C. Zambon and Christopher S. Barker   4 Determination of Alternate Splicing Events Using the Affymetrix Exon 1.0 ST Arrays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sita Subbaram, Marcy Kuentzel, David Frank, C. Michael DiPersio, and Sridar V. Chittur   5 Profiling microRNA Expression with the Illumina BeadChip Platform . . . . . . . . . Julissa Tsao, Patrick Yau, and Neil Winegarden   6 TaqMan® Array Cards in Pharmaceutical Research . . . . . . . . . . . . . . . . . . . . . . . . David N. Keys, Janice K. Au-Young, and Richard A. Fekete   7 DMET ™ Microarray Technology for Pharmacogenomics-Based Personalized Medicine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . James K. Burmester, Marina Sedova, Michael H. Shapero, and Elaine Mansfield   8 The Use of Microarray Technology for Cytogenetics . . . . . . . . . . . . . . . . . . . . . . Bassem A. Bejjani, Lisa G. Shaffer, and Blake C. Ballif   9 PCR/LDR/Universal Array Platforms for the Diagnosis of Infectious Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Maneesh Pingle, Mark Rundell, Sanchita Das, Linnie M. Golightly, and Francis Barany 10 RIP-CHIP in Drug Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ritu Jain, Francis Doyle, Ajish D. George, Marcy Kuentzel, David Frank, Sridar V. Chittur, and Scott A. Tenenbaum 11 ChIPing Away at Global Transcriptional Regulation . . . . . . . . . . . . . . . . . . . . . . . Kelly Jackson, James Paris, and Mark Takahashi 12 HELP (HpaII Tiny Fragment Enrichment by Ligation-Mediated PCR) Assay for DNA Methylation Profiling of Primary Normal and Malignant B Lymphocytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rita Shaknovich, Maria E. Figueroa, and Ari Melnick

1

vii

27

45

63

73 87

99

125

141

159

173

191

viii

Contents

13 High-Throughput Screening of Metalloproteases Using Small Molecule Microarrays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mahesh Uttamchandani 14 Metabolic Enzyme Microarray Coupled with Miniaturized Cell-Culture Array Technology for High-Throughput Toxicity Screening . . . . . . . . . . . . . . . . . Moo-Yeal Lee, Jonathan S. Dordick, and Douglas S. Clark 15 Use of Tissue Microarray to Facilitate Oncology Research . . . . . . . . . . . . . . . . . . Panagiotis Gouveris, Paul M. Weinberger, and Amanda Psyrri 16 Small Molecule Selectivity and Specificity Profiling Using Functional Protein Microarrays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Peter R. Kraus, Lihao Meng, and Lisa Freeman-Cook 17 Production and Application of Glycan Microarrays . . . . . . . . . . . . . . . . . . . . . . . . Julia Busch, Ryan McBride, and Steven R. Head

203

221 239

251 269

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283

Contributors Janice K. Au-Young  •  Molecular Biology Division, Life Technologies Corporation, Foster City, CA, USA Blake C. Ballif  •  Signature Genomic Laboratories, Spokane, WA, USA Francis Barany  •  Department of Microbiology, Weill Medical College, Cornell University, New York, NY, USA Christopher S. Barker  •  Gladstone Institute of Cardiovascular Disease, University of California, San Francisco, CA, USA Bassem A. Bejjani  •  Signature Genomic Laboratories, Spokane, WA, USA James K. Burmester  •  Center for Human Genetics, Marshfield Clinic Research Foundation, Marshfield, WI, USA Julia Busch  •  DNA Array Core Facility, The Scripps Research Institute, La Jolla, CA, USA Sridar V. Chittur  •  Department of Biomedical Sciences, Center for Functional Genomics, School of Public Health, University at Albany-SUNY, Rensselaer, NY, USA Douglas S. Clark  •  Department of Chemical Engineering, University of California, Berkeley, CA, USA Sanchita Das  •  Department of Microbiology, Weill Medical College, Cornell University, New York, NY, USA C. Michael DiPersio  •  Center for Cell Biology and Cancer Research, Albany Medical College, Albany, NY, USA Jonathan S. Dordick  •  Department of Chemical and Biological Engineering, Rensselaer Polytechnic Institute, Troy, NY, USA Francis Doyle  •  NanoBio Constellation, College of Nanoscale Science and Engineering, University at Albany-SUNY, Albany, NY, USA Richard A. Fekete  •  Molecular Biology Division, Life Technologies Corporation, Foster City, CA, USA Maria E. Figueroa  •  Division of Hematology/Oncology, Department of Medicine, Weill Medical College, Cornell University, New York, NY, USA David Frank  •  Department of Biomedical Sciences, Center for Functional Genomics, School of Public Health, University at Albany-SUNY, Rensselaer, NY, USA Lisa Freeman-Cook  •  Life Technologies, Carlsbad, CA, USA Ajish D. George  •  Department of Biomedical Sciences, Gen*NY*Sis Center for Excellence in Cancer Genomics, School of Public Health, University at Albany-SUNY, Rensselaer, NY, USA Linnie M. Golightly  •  Department of Microbiology, Weill Medical College, Cornell University, New York, NY, USA Panagiotis Gouveris  •  Division of Hematology Oncology, Department of Internal Medicine, Yale University, New Haven, CT, USA

ix

x

Contributors

Steven R. Head  •  DNA Array Core Facility, The Scripps Research Institute, La Jolla, CA, USA Matthew A. Held  •  Departments of Dermatology and Pathology, Yale University School of Medicine, New Haven, CT, USA Kelly Jackson  •  UHN Microarray Center, Toronto, ON, Canada Ritu Jain  •  NanoBio Constellation, College of Nanoscale Science and Engineering, University at Albany-SUNY, Albany, NY, USA David N. Keys  •  Molecular Biology Division, Life Technologies Corporation, Foster City, CA, USA Peter R. Kraus  •  Life Technologies, Carlsbad, CA, USA Marcy Kuentzel  •  Department of Biomedical Sciences, Center for Functional Genomics, School of Public Health, University at Albany-SUNY, Rensselaer, NY, USA Moo-Yeal Lee  •  Solidus Biosciences, Inc., Troy, NY, USA Elaine Mansfield  •  Application Sciences Department, Affymetrix, Inc., Santa Clara, CA, USA Ryan McBride  •  DNA Array Core Facility, The Scripps Research Institute, La Jolla, CA, USA Ari Melnick  •  Division of Hematology/Oncology, Department of Medicine, Weill Medical College, Cornell University, New York, NY, USA Lihao Meng  •  Life Technologies, Carlsbad, CA, USA Tony S. Mondala  •  The Scripps Research Institute, La Jolla, CA, USA James Paris  •  UHN Microarray Center, Toronto, ON, Canada Maneesh Pingle  •  Department of Microbiology, Weill Medical College, Cornell University, New York, NY, USA Amanda Psyrri  •  Division of Hematology Oncology, Department of Internal Medicine, Yale University, New Haven, CT, USA Mark Rundell  •  Department of Microbiology, Weill Medical College, Cornell University, New York, NY, USA Daniel R. Salomon  •  The Scripps Research Institute, La Jolla, CA, USA Marina Sedova  •  Assay and Application Product Development, Affymetrix, Inc, Santa Clara, CA, USA Lisa G. Shaffer  •  Signature Genomic Laboratories, Spokane, WA, USA Rita Shaknovich  •  Division of Immunopathology, Department of Pathology, Weill Medical College, Cornell University, New York, NY, USA; Division of Hematology/Oncology, Department of Medicine, Weill Medical College, Cornell University, New York, NY, USA Michael H. Shapero  •  Assay and Application Product Development, Affymetrix, Inc, Santa Clara, CA, USA Sita Subbaram  •  Center for Cell Biology and Cancer Research, Albany Medical College, Albany, NY, USA Mark Takahashi  •  UHN Microarray Center, Toronto, ON, Canada Scott A. Tenenbaum  •  NanoBio Constellation, College of Nanoscale Science and Engineering, University at Albany-SUNY, Albany, NY, USA Scott Tighe  •  Microarray Core Facility, University of Vermont, College of Medicine, Burlington, VT, USA

Contributors

Julissa Tsao  •  UHN Microarray Center, Toronto, ON, Canada Mahesh Uttamchandani  •  Defense Medical and Environmental Research Institute (DMERI), DSO National Laboratories, Singapore; Department of Chemistry, National University of Singapore, Singapore Paul M. Weinberger  •  Department of Otolaryngology, Medical College of Georgia, Augusta, GA, USA Neil Winegarden  •  UHN Microarray Center, Toronto, ON, Canada Patrick Yau  •  UHN Microarray Center, Toronto, ON, Canada Alexander C. Zambon  •  Department of Pharmacology, University of California, San Diego, La Jolla, CA, USA

xi

Chapter 1 Multicenter Clinical Sample Collection for Microarray Analysis Tony S. Mondala, Daniel R. Salomon, and Steven R. Head Abstract In this chapter, we describe numerous methods to extract RNA, DNA, and protein from tissue, represented by kidney transplant biopsies, and from peripheral blood cells collected at various clinical sites. Gene expression profiling and SNP-based genome-wide association studies are done using various microarray platforms. In addition, protocols that enable simultaneous protein purification from these clinical samples, enable additional strategies for understanding of the molecular processes involved in organ transplantation, immunosuppressive drug regimens, and the elements determining allograft success and failure. Successfully establishing a multicenter clinical study was essential to meet our objectives for subject enrollment and transplant outcomes. This chapter focuses on our experience setting up and coordinating clinical sample collection from multiple transplant centers for the purpose of microarray analysis. Key words: Microarrays, Genomics, Transplantation, Multicenter clinical study, Nucleic acid extraction, Protein extraction

1. Introduction The analysis of clinical samples utilizing microarray technology has advanced the field of clinical research including transplan­ tation medicine (1–6). Our research group, the Transplant Genomics Collaborative Group (TGCG; http://www.genetics. ucla.edu/transplant-genomics/index2.php) is involved in a large study of kidney transplantation outcomes with an emphasis on defining genomic biomarkers that could be used to monitor and individualize the adequacy and efficacy of immunosuppressive drug therapy. Organizing the Transplant Genomics project has provided a better understanding of the challenges facing anyone

Sridar V. Chittur (ed.), Microarray Methods for Drug Discovery, Methods in Molecular Biology, vol. 632, DOI 10.1007/978-1-60761-663-4_1, © Humana Press, a part of Springer Science+Business Media, LLC 2010

1

2

Mondala, Salomon, and Head

planning a large, multicenter clinical project that involves collection of multiple sample types, at multiple predefined time points, with multiple sample handling protocols and where obtaining precise clinical data and outcomes are necessary. The key to success of any study involving multiple clinical centers is the efficient collection, preservation, and transport of clinical material from collection sites to a central processing facility where samples can be prepared for analysis using microarrays and other analytical techniques. Organizing a multicenter research study requires effective training and support of physicians, nurse coordinators, and laboratory personnel in order to guarantee adherence to enrollment (inclusion/exclusion) criteria, proper sample collection, preshipment specimen processing, and documentation of patient/subject data. A key insight is that physicians, nurses, and laboratory personnel all have different jobs, training, and work environments so that strategies to effectively communicate study objectives and monitor sample collection and data integrity must be developed for each. As with any research study involving human subjects, Institutional Review Board approval of a Human Subjects Protocol is required as well as informed consent for study participants. Setting up a central processing center is necessary to create, test, and then provide kits for sample collection and transport. The central processing center is also tasked with the tracking of all collected specimens from the various clinical centers as well as coordinating sample preparation and archiving. Finally, it is critical to have a highly secure clinical database that is readily accessible to all the participating centers and a parallel, but integrated, specimen tracking database in the central processing center. The wealth of scientific information that can be obtained through the establishment of a well-organized system to collect, document, and process clinical research samples provides a foundation for advancing clinical research and translational medicine. The purpose of this chapter is to discuss our experience in a large multicenter clinical study, specifically the Transplant Genomics project. 1.1. Setting up the Clinical Centers

Before recruitment of candidate clinical centers can begin, it is imperative to have reagents and protocols thoroughly established and validated for sample collection, processing at the collection sites, storage and shipment to the central processing center. In order to collect kidney biopsies and blood samples from transplant recipients and donors, we prepared kits containing suitable containers to hold the specimens along with detailed instructions written in plain language as well as illustrations for collecting, processing (when needed), and temporarily storing specimens at the clinical sites. A second set of kits were prepared containing all the necessary components to facilitate shipping samples from the clinical sites to the central processing center. Finally, a specimen labeling system, which includes bar coding, was implemented

Multicenter Clinical Sample Collection for Microarray Analysis

3

to facilitate linking specific specimens with the clinical data for each patient. In our experience, every clinical center is going to have specific needs and support requirements. We have found it helpful to have a single person in the central processing team responsible for any given center to evaluate the capabilities of the center, provide on-site training, phone, and email support. In addition, this person is tasked with monitoring problems in sample collection, enrollment, processing, and shipping as well as ensures patient data associated with each sample are properly entered into a clinical database and timely follow-up and outcome data are also obtained per the study protocols. The role of supporting clinical centers is a key component and should not be under resourced. For our Transplant Genomics project, we ask each clinical center to collect whole blood and core needle kidney biopsies from kidney transplant recipients at specific time points or at the time of specific events such as acute rejection. We also require the clinical centers to process a portion of the whole blood into purified lymphocytes and plasma. This requires centrifugation of whole blood within 2 h of collection followed by separation of the lymphocyte and plasma fractions. Therefore, to be considered for inclusion in our study, we required each candidate clinical center to be capable of performing this procedure. Clinical centers are frequently staffed by nurses with varying levels of laboratory experience and competing demands on their time. In some cases, all on-site sample processing is done by clinical laboratory staff and in other cases, it is done by nursing staff. This adds complexity to the process of establishing clinical center-specific procedures required to efficiently collect, process, and store samples. However, an enthusiastic and highly motivated staff at the clinical site can often find creative solutions to complete the tasks within the required parameters. One key component of collecting samples from clinical centers is the parallel collection of accurate and complete patient data and records. We designed an online database accessible via a secure web portal using a 256-bit Secure Socket Layer (SSL) encryption protocol. On-site training in clinical data entry for the project was a key and final step in launching the study at each clinical center and is repeated whenever local staff changes. The clinical database, maintained by a database administrator at our center, is accessible only by authorized personnel at each clinical center with individual usernames and passwords. To further protect information confidentiality, clinical center staff can only access patient-identified data from their own clinical center. It is critical to continuously monitor the patient data entry to ensure that every patient and sample has completed records for all sample collections as well as any follow-up clinical information relevant to the study. Again, this should not be under resourced as it is often impractical to retroactively fill in required patient data and records.

4

Mondala, Salomon, and Head

1.2. Institutional Review Boards, HIPAA, and Compensation for Participation

Institutional Review Boards (IRBs) are critical to the integrity of clinical research. A comprehensive discussion of the importance and functions of the IRB is beyond the scope of this chapter. For an excellent starting source of information, the reader is referred to the FDA website (http://www.fda.gov/oc/ohrt/irbs/). In the case of a multicenter clinical trial, the IRB of the primary project center takes the lead in the first review and approval of the research protocols. But this approval is based on all the procedures agreed upon by the local IRB of the primary center. Next, the research protocols must be reviewed and approved by each participating center’s IRB and again, this involves all the local procedures for these different IRBs. Thus, the format of the protocol submissions can be very different in every center, the requirements for the informed consents can be different, and a local IRB can raise issues with any element of the research independent of any other IRB’s approval. Our experience is that several candidate clinical centers were never able to complete their IRB reviews. A common challenge was our study’s requirements for DNA collections that raised concerns over unwanted dissemination of genetic information, an issue that many IRBs had never confronted. Another critical element of planning are the provisions of the Health Insurance Portability and Accountability Act (HIPAA). Again, it is beyond the scope here to detail HIPAA’s provisions and the website of U.S. Department of Health and Human Services is an excellent first source (http://www.hhs.gov/ocr/ privacy/index.html). However, as a starting point, it is necessary to create a strategy to protect patient-identified data so that no one can access this information beyond the authorization given by the IRBs to the principal investigators and key personnel of the project as well as relevant regulatory governmental agencies (e.g., FDA). Such strategies include assigning coded patient identification numbers to decouple sample IDs from patient names. Thus all downstream processing of samples and analysis of data by the technicians and scientists is done with no patient-identifiable information, just anonymous alpha-numeric codes. We purposefully created two project databases to facilitate our compliance with HIPAA. The first was the highly secure clinical database that is accessible by password protection at each site via a web portal. This database contains all the sensitive patient identified information. The second database is designed only for sample tracking, disposition, and archiving. It has no patient-identified data and all samples are listed as anonymous alpha-numeric codes assigned to each sample collected and bar-coded. The two databases are linked through the clinical database where the key to the alpha-numeric codes are kept. Finally, we encountered a number of issues involving compensation of research subjects. For one example, different transplant

Multicenter Clinical Sample Collection for Microarray Analysis

5

programs had different standards of practice. In programs like ours that does serial monitoring transplant biopsies in all patients, we could not compensate our research subjects for the biopsy due to issues of fairness. In contrast, in clinical centers where such serial biopsies are not standard of practice, the local IRBs did allow subject compensation for research participation as long as it was reasonable, not coercive, and covered acceptable things such as loss of work time, travel expenses, childcare etc. In general, the subject of compensating subjects for research remains a very sensitive issue and every IRB has different views of what is reasonable. 1.3. Central Processing Laboratory

Setting up a central processing laboratory is essential for coordinating the shipment, tracking, and processing of clinical samples. In addition, the preparation and validation of sample collection and sample shipping kits is most efficiently accomplished by the same laboratory facility that will be receiving and processing the samples. A key point is that all protocols, reagents, and kits must be fully developed, tested, and ready for implementation by every clinical center before on-site training and study initiation. It is a major error to launch a project and then find out that one or more elements of the sample collection protocol require significant changes. Another important role of the central processing facility is to ensure that the clinical centers are always adequately supplied with kits for sample collection and shipping. In kidney transplantation, newly enrolled patients are frequently in need of urgent care and if the clinical center is missing specimen collection kits we miss a valuable opportunity. Frequent communication with the clinical center staff ensures adequate supplies are always available. Since most biological specimens are perishable, specimen shipments by clinical centers should be made early in the week to prevent weekend deliveries. All shipments must be immediately unpacked and safely stored until they can be logged into a specimen tracking database and processed. We created a separate database (linkable to the clinical database by several key variables) for tracking the arrival of samples and following their progress through extraction of RNA, DNA, and protein from blood, cell pellets, and kidney biopsy material. In addition, the sample tracking database records how much and where aliquots of RNA, DNA, and protein are shipped for further analysis. Finally, the archived sample amounts and locations (freezer, drawer, box, and position) are recorded in the specimen tracking database. The use of a bar coding system facilitates tracking the processing and archiving of clinical specimens. We use a single 5 digit barcode number to label all collection tubes within a specific kit. Each kit is used for all specimens collected from a single patient at a single collection time point. The kit barcode is recorded into the clinical database by clinical center staff at the time of sample

6

Mondala, Salomon, and Head

procurement in order to provide a unique identifier that allows the clinical database data to be linked to the data in the specimen tracking database. When specimens are processed into RNA, DNA, and protein at the central processing laboratory, a one letter prefix is added to the barcode (new labels are printed) to identify the sample type and source (i.e., RNA derived from whole blood vs. RNA derived from biopsy core). An additional aliquot number as a suffix is also added (5 digits) to the barcode, which is now one letter and 10 digits long, to identify each specific tube and facilitate tracking the shipping or archiving of each aliquot. We use several protocols for extraction of RNA, DNA, protein, and plasma from whole blood, cell pellets, and biopsies, which are described below in Subheadings 2 and 3.

2. Materials 2.1. Specimen Collection Kits

1. PAXgene Blood RNA tube, 2.5 ml (Qiagen). 2. Vacutainer Cell Preparation tube (CPT) with sodium citrate, 8 ml (Becton-Dickenson). 3. Vacutainer Plasma Preparation tube (PPT) with EDTA, 5 ml (Becton-Dickenson). 4. RNAlater (Ambion). 5. Phosphate buffered saline, pH 7.2 (Invitrogen). 6. 2.0 and 4.0 ml cryovials (USA Scientific). 7. Color cap inserts (USA Scientific). 8. 15 and 50 ml conical tubes (USA Scientific). 9. 3 ml disposable transfer pipet (VWR). 10. Cardboard boxes (Office Depot). 11. Barcode label printer (TLS PC Link, Brady Worldwide). 12. Barcode labels (PTL-76-461, Brady Worldwide) (see Note 1).

2.2. RNA Extraction from PAXgene Blood Samples

1. PAXgene Blood RNA Kit (Qiagen). 2. Ethanol (100%) (Sigma). 3. GLOBINclear Kit – Human (Ambion). 4. Isopropanol (100%) (Sigma). 5. Magnetic Stand (Ambion).

2.3. RNA, DNA, and Protein Extraction from Mononuclear Cells

1. AllPrep DNA/RNA/Protein Mini Kit (Qiagen). 2. 14.3 M b-mercaptoethanol (Sigma). 3. Ethanol (100%) (Sigma). 4. 5 ml syringes and 18G needles (Becton-Dickenson).

Multicenter Clinical Sample Collection for Microarray Analysis

2.4. Plasma Separation from Whole Blood Samples (CPT)

1. 3 ml disposable transfer pipet (VWR).

2.5. RNA, DNA, and Protein Extraction from Biopsies

1. Trizol Reagent (Invitrogen).

7

2. 2  ml Wheaton Plastic Coated Tissue Grinder with Teflon Pestles (VWR). 3. Chloroform (Sigma). 4. Isopropanol (100%) (Sigma). 5. Ethanol (100, 80, and 75%) (Sigma). 6. DEPC-treated water (Ambion). 7. RNeasy Kit (Qiagen). 8. 14.3 M b-mercaptoethanol (Sigma). 9. 0.1 M NaCitrate in 10% ETOH. 10. 8 mM NaOH. 11. Phase Lock Gel tube (Eppendorf). 12. Phenol Chloroform (Ambion). 13. 3 M Sodium Acetate. 14. Glycogen (5 mg/ml) (Ambion). 15. 0.3 M Guanidine HCl.

2.6. DNA Extractions from Whole Blood

1. QIAamp DNA Blood Midi Kit – 100 (Qiagen). 2. Ethanol (100%). 3. 15 ml Centrifuge tubes (USA Scientific).

2.7. DNA Extractions from Mononuclear Cells

1. QIAamp DNA Mini Kit – 50 (Qiagen). 2. Ethanol (100%). 3. Phosphate buffered saline, pH 7.2 (Invitrogen).

3. Methods The proper procurement and handling of human blood and tissues for translational studies is critical to ensure that the quality of the specimen remains intact and suitable for downstream analysis. We discuss below in detail how the blood and biopsies are collected and handled in transport and the extraction of RNA, DNA, and protein. 3.1. Blood and Tissue Procurement

Each clinical center is supplied with multiple sample collection kits for procurement of specimens. Each kit (7 × 4 × 3  in. cardboard box) contains two 2.5 ml PAXgene tubes, one 8.5 ml CPT

8

Mondala, Salomon, and Head

(Cell Preparation Tube) tube, two 4.5 ml PPT (Plasma Preparation Tube) tubes, one 2 ml cryovial containing 1 ml of RNAlater for holding the biopsy core, one 50 ml conical tube containing 35 ml of PBS for washing the cell pellet obtained from the CPT tube, one empty 15 ml conical tube for washing the cell pellet from the CPT tube, one empty 4 ml cryovial tube to hold the plasma from the CPT tube, one 2 ml cryovial tube containing 1.5 ml RNAlater to resuspend the washed cell pellet, one empty 2 ml cryovial to hold the cells resuspended in RNAlater, and one transfer pipet used to transfer plasma from the CPT tube to the 4 ml cryovial. In addition, a detailed set of instructions for specimen collection, processing, and temporary storage at the clinical center are included as well as a laboratory requisition form used to document the date, time, and person collecting and processing the specimens. Finally a set of barcode labels are provided to attach to each collection tube and vial. This barcode information specifies a 5 digit number assigned to the specific collection kit and allows us to uniquely label specimens from a specific patient and collection time point. This barcode is subsequently entered into both the clinical and specimen tracking databases. Standard phlebotomy technique is utilized in collecting the whole blood samples. 2.5 ml of blood is collected into each of two PAXgene tubes and immediately inverted several times to efficiently mix the blood with the RNA stabilizing reagent. This mixing step is critical. Nursing and phlebotomy staff must be made aware of the importance of mixing the tubes immediately after draw. 8.5 ml of blood is then collected into the CPT tube and 4.5 ml of blood into each of two PPT tube. As with the PAXgene tubes, the CPT and PPT tubes must be inverted immediately several times to ensure the blood and anti-coagulant is thoroughly mixed (see Note 2). Core needle kidney biopsies are collected by trained transplant physicians or radiologists. Biopsies are immediately submerged in 1 ml of RNAlater in a 2 ml cryovial and stored at 4°C overnight and then frozen at –20°C the following day (see Note 3). The blood collection tubes are transported to the onsite processing laboratory for further processing and storage until shipment to the Central Processing Laboratory. 3.2. Transport

Each clinical center is provided with sample shipping supplies. This includes detailed shipping instructions, biohazard plastic zip lock bags, absorbent tube sleeves, FedEx forms, dry ice labels, diagnostic specimen labels (UN3373), styrofoam shipping boxes contained within a secondary cardboard box. All frozen samples are shipped to the central processing center with a sufficient amount of dry ice for next day delivery. Packages are shipped only on Mondays, Tuesdays, and Wednesdays to lessen the chance that the shipment would be delayed over a weekend.

Multicenter Clinical Sample Collection for Microarray Analysis

9

When packaging the frozen specimens for shipment, it is very important to make sure that the frozen tubes are properly protected from damage by rough handling during transport. The frozen tubes are very brittle and prone to crack thus they should be packaged carefully cushioned in a smaller box that is then placed in the larger styrofoam box that also contains the dry ice. This also prevents the larger pieces of dry ice from crushing the frozen tubes. The clinical center documents in the clinical database when specimens are shipped to the central processing center. The database then automatically sends an email notification to the central processing center that a shipment is on its way and to expect arrival the following day. 3.3. RNA, DNA, and Protein Extraction

Once the samples are received at the central processing center they are immediately unpacked and placed in –20°C storage for up to 60 days until further processing. When samples are ready for processing the specimens are removed from the freezer and thawed on ice. The following protocols are used to extract RNA, DNA, and protein from whole blood, cell pellet, and biopsy tissue. Once the extraction procedure is completed each sample type (RNA, DNA and protein) is assigned a new barcode consisting of the original 5-digit barcode plus an additional letter prefixed which identifies the sample type as well as assigning a unique 5 digit aliquot identification number. At this point, the samples are archived at –80°C and ready for downstream analysis by microarray and other analytical techniques. At the time this project started, we utilized Trizol reagent to extract RNA, DNA, and protein from both cells and biopsies. Subsequently,thecreationoftheQiagenAllPrepDNA/RNA/Protein Mini kit allowed us to extract all three fractions with improved efficiency. This is a simplified protocol, takes less time, avoids the use of toxic phenol and chloroform, does not require ethanol precipitation, and it allows greater ease when processing multiple samples simultaneously. The DNA extracted using AllPrep also proved to be better suited for genotyping using the Affymetrix SNP microarrays than DNA extracted using Trizol reagent. This was observed in the increased SNP call rates for DNA samples extracted using the AllPrep method. In our experience biopsies processed using the AllPrep did not yield RNA, DNA, and protein in amounts comparable to the Trizol method. Thus for biopsies we continued to use the Trizol method and for mononuclear cells we have adapted the AllPrep method. Table 1 shows the different extraction protocols used for both biopsies and cells and the downstream applications for each fraction. In our Transplant Genomics project, we also collect extracted donor and recipient DNA specimens for analysis on genome-wide Affymetrix SNP microarrays. In cases where only peripheral blood

10

Mondala, Salomon, and Head

Table 1 Extraction methods utilized based on starting tissue type, desired purified fraction, and downstream analysis platform Tissue type Kidney biopsy

Mononuclear cells

Extraction method

Sample type Application

Invitrogen Trizol reagent

RNA

Obtained higher yields Gene expression: (Affymetrix than the Qiagen Human Genome U133 AllPrep Method Plus 2.0 Array) Whole-transcript gene expression and alternative splicing: (Affymetrix Human Exon 1.0 ST Array; Human Gene1.0 ST Array)

DNA

SNP based genome-wide association studies: (Affymetrix Genome-Wide Human SNP Array 6.0)

Affymetrix SNP call rates were not as high compared to DNA derived using Qiagen AllPrep Method

Protein

Proteomic analysis: (MudPIT Tandem Mass Spectrometry, quantification, expression)

Obtained higher yields than the Qiagen AllPrep Method

RNA

Gene expression: (Affymetrix Human Genome U133 Plus 2.0 Array) Whole-transcript gene expression and alternative splicing: (Affymetrix Human Exon 1.0 ST Array; Human Gene1.0 ST Array)

Obtained yields comparable to the Trizol method. All QC metrics similar to that derived using the Trizol method

DNA

SNP based genome-wide association studies: (Affymetrix Genome-Wide Human SNP Array 6.0)

Obtained yields comparable to the Trizol method. Affymetrix SNP call rates were higher than the Trizol method

Protein

Proteomic analysis: (MudPIT Tandem Mass Spectrometry, quantification, expression)

Obtained yields similar to the Trizol method

SNP based genome-wide association studies: (Affymetrix Genome-Wide Human SNP Array 6.0)

Extraction method used when only DNA is required for SNP genotyping

Qiagen Allprep DNA/ RNA/ Protein Kit

DNA Qiagen, Whole blood QIAamp, or DNA Blood mononuclear Midi and cells DNA Mini Kit

Notes

Multicenter Clinical Sample Collection for Microarray Analysis

11

lymphocytes or frozen anti-coagulated whole blood is available such as those found in most tissue typing laboratory archives we extract DNA using the QIAamp DNA kits from Qiagen. From a Vacutainer tube, 4 ml draw with EDTA we use the QIAamp Blood Midi Kit to extract DNA. When we have isolated lymphocytes to extract DNA from we use the QIAamp Mini Kit. The specific protocols for extraction of RNA, DNA, and protein are described in a following section. 3.3.1. RNA Extraction from PAXgene Blood Tubes

The manufacturer (Qiagen) recommended protocol was followed. PAXgene Blood RNA tubes are intended for the collection of whole blood and the stabilization of cellular RNA for up to 3 days at 18–25°C or up to 5 days at 2–8°C. Tubes should be stored at –20 to –80°C for longer periods of time. Draw 2.5 ml of blood directly into PAXgene tube and invert the tube 10 times immediately, do not shake (see Note 4). 1. Buffer BR4 is supplied as a concentrate. Before using for the first time, add 4 volumes of ethanol (100%) as indicated on the bottle to obtain a working solution. Buffer BR2 may form a precipitate upon storage, warm to 37°C to dissolve if necessary. 2. Prepare DNase I stock solution when using the RNase-Free DNase set for the first time. Dissolve the solid DNase I in 550  ml of RNase-free water provided. Take care that no DNase I is lost when opening the vial. Mix gently by inverting the tube. Do not vortex (see Note 5). 3. All centrifugation steps for this protocol are done at room temperature. 4. Centrifuge the PAXgene tube containing 2.5 ml of blood for 10 min at 3,500 × g, brake on using a swing-out rotor with adapters for round-bottom tubes. 5. Remove the supernatant by decanting and discard the supernatant. Dry the rim of the tube with a Kim wipe. Add 4 ml of RNase free water to the pellet and cover the tube using a new, fresh secondary Hemogard closure provided with the kit. 6. Vortex until the pellet is visibly dissolved. Centrifuge for 10 min at 3,500 × g, brake on using a swing-out rotor. Remove the supernatant by decanting and discard the supernatant. Dry the rim of the tube with a Kim wipe. 7. Add 350 ml Buffer BR1 and vortex until the pellet is visibly dissolved. 8. Pipet the sample into a 1.5  ml microcentrifuge tube. Add 300 ml Buffer BR2 and 40 ml proteinase K. Mix by vortexing for 5  s and incubate for 10  min at 55°C using a shakerincubator at 1,400 rpm.

12

Mondala, Salomon, and Head

9. Pipet the lysate directly into a PAXgene Shredder column placed in a 2 ml processing tube and centrifuge for 3 min at maximum speed (not to exceed 20,000 × g). Carefully transfer the entire supernatant of the flow through fraction to a fresh 1.5 ml microcentrifuge tube without disturbing the pellet in the processing tube. 10. Add 350  ml ethanol (100%). Mix by vortexing and quick spin for only 1–2 s to collect the droplets from inside of the tube lid. 11. Add 700 ml of the sample to a PAXgene RNA spin column placed in a 2 ml processing tube and centrifuge for 1 min at maximum speed. Place the spin column in a new 2 ml processing tube and discard the old processing tube containing the flow-through. 12. Add remaining sample to the spin column and centrifuge for 1  min at maximum speed. Place the spin column in a new 2  ml processing tube and discard the old processing tube containing the flow-through. 13. Add 350 ml Buffer BR3 to the spin column. Centrifuge for 1 min at maximum speed. Transfer the spin column to a new processing tube and discard the old processing tube containing the flow-through. 14. Add 73.5  ml Buffer RDD to each thawed 10.5  ml DNase I stock solution aliquot. A single DNase I aliquot per PAXgene tube being processed. Mix by gently flicking the tube, do not vortex. Centrifuge briefly to collect residual liquid from the sides of the tube. 15. Add 80 ml of the DNase I incubation mix directly onto the PAXgene RNA spin column membrane and incubate at room temperature for 15 min. 16. Add 350 ml Buffer BR3 to the spin column. Centrifuge for 1  min at maximum speed. Place the spin column in a new 2  ml processing tube and discard the old processing tube containing the flow-through. 17. Add 500 ml Buffer BR4 to the spin column and centrifuge for 1  min at maximum speed. Place the spin column in a new 2  ml processing tube and discard the old processing tube containing the flow-through. 18. Add another 500 ml BR4 to the spin column and centrifuge for 3 min at maximum speed. Place the spin column in a new 2  ml processing tube and discard the old processing tube containing the flow-through. Centrifuge for 1 min at maximum speed to dry the spin column membrane. 19. Place the spin column in a 1.5 ml microcentrifuge and discard the old processing tube. Add 40 ml Buffer BR5 directly onto

Multicenter Clinical Sample Collection for Microarray Analysis

13

the spin column membrane. Centrifuge for 1 min at maximum speed to elute the RNA. Do not discard the eluate. 20. Add another 40 ml of Buffer BR5 directly onto the spin column membrane. Centrifuge for 1 min at maximum speed. Discard the spin column. 21. Incubate the eluate for 5 min at 65°C in a constant temperature incubator. Do not exceed incubation time or temperature. After incubation immediately chill on ice. 22. Quantify RNA using NanoDrop or other spectrophotometer, blank instrument with Buffer BR5. A 260/280 absorbance ratio between 1.8 and 2.0 is typical of pure RNA. Assess quality of the RNA using the Agilent Bioanalyzer. If the RNA will not be used immediately, store at –80°C. Typical RNA yield from a single PAXgene tube is 4–10 mg. 3.3.2. Globin Reduction of RNA Derived from PAXgene Blood Tubes

It is known that the presence of globin mRNA in total RNA samples derived from whole blood can reduce detection sensitivity when using gene expression arrays as seen as a decrease in present calls and an increase in signal variation (7–10). We followed the manufacturer (Ambion) recommended protocol for reducing the presence of globin mRNA in our total RNA samples derived from PAXgene RNA blood tubes. This protocol utilizes magnetic beads and biotin/streptavidin binding to remove 95% or more of alpha and beta globin mRNA from whole blood derived total RNA samples. 1. Set constant temperature incubators to 50 and 58°C. 2. Prior to starting the procedure, prepare the following reagents. Add 2 ml isopropanol (100%) to the bottle labeled RNA Binding Buffer. Concentrate, mix well, and mark the label to indicate that the isopropanol was added. Add 4 ml ethanol (100%) to the RNA Wash Solution Concentrate bottle, mix well, and indicate on the label that the ethanol was added. 3. Prepare Bead Resuspension Mix prior to starting procedure by combining in a 1.5 ml microcentrifuge tube, 10 ml of RNA Binding Beads (mix thoroughly by vortexing before dispensing) and 4 ml of RNA Bead Buffer for a single reaction, mix briefly, then add 6 ml of isopropanol (100%), mix by vortexing. Scale volumes for multiple reactions include 5% overage for pipetting error. 4. Prepare Streptavidin Magnetic Beads prior to starting procedure by warming the 2× Hybridization Buffer and the Streptavidin Bead Buffer to 50°C for at least 15 min and vortex well before use. Vortex the tube of Streptavidin Magnetic Beads and aliquot in to a 1.5 ml microcentrifuge tube 30 ml for each sample to be processed. Briefly centrifuge for less than 2 s at low speed to collect the mixture at the bottom of the tube.

14

Mondala, Salomon, and Head

Place the tube on a magnetic stand to capture the Streptavidin Magnetic Beads. Leave the tube until the mixture becomes transparent (~5 min). Carefully aspirate the supernatant using a pipet without disturbing the Streptavidin Magnetic Beads. Discard the supernatant and remove the tube from the magnetic stand. Add Streptavidin Bead Buffer to the Streptavidin Magnetic Beads, use a volume equal to the original volume of Streptavidin Magnetic Beads. Vortex vigorously until beads are resuspended and keep at 50°C for at least 15 min before being used later in procedure. 5. Warm Elution Buffer to 58°C prior to using later in procedure. 6. Combine 1–10 mg human whole blood total RNA (in a maximum volume of 14 ml) with 1 ml of Capture Oligo Mix in a 1.5 ml microcentrifuge tube. Add nuclease-free water to the sample mixture as necessary to a final volume of 15 ml. 7. Add 15 ml of 50°C 2× Hybridization Buffer, vortex briefly to mix, and centrifuge briefly for less than 2 s at low speed to collect contents in the bottom of the tube. Incubate at 50°C for 15 min. 8. Remove the prepared Streptavidin Magnetic Beads from the 50°C incubator and resuspend them by gentle vortexing. Briefly centrifuge for less than 2 s at low speed. Add 30 ml of prepared Streptavidin Magnetic Beads to each RNA sample, vortex to mix well, and centrifuge briefly for less than 2 s at low speed. Flick the tube very gently to resuspend the beads, being careful to keep the contents at the bottom of the tube. Incubate at 50°C for 30 min. 9. Remove sample and vortex briefly to mix, centrifuge for less than 2 s at low speed. Capture the Streptavidin Magnetic Beads on a magnetic stand. Leave the tube until the mixture becomes transparent (~5 min). Carefully draw up the supernatant, which contains the globin mRNA depleted RNA, and transfer the RNA to a new 1.5 ml microcentrifuge tube. Place RNA on ice, discard the tube with the Streptavidin Magnetic Beads. 10. Add 100 ml RNA Binding Buffer to each sample. Vortex the Bead Resuspension Mix to resuspend the beads thoroughly and immediately dispense 20 ml to each sample. Vigorously vortex the sample for 10 s, briefly centrifuge for less than 2 s at low speed. 11. Capture the RNA Binding Beads by placing the tube on a magnetic stand. Leave the tube until the mixture becomes transparent (~5 min). Carefully aspirate the supernatant without disturbing the RNA Binding Beads and discard the supernatant (it is important to remove as much of the supernatant as possible). Remove the tube from the magnetic stand.

Multicenter Clinical Sample Collection for Microarray Analysis

15

12. Add 200 ml RNA Wash Solution to each sample and vortex for 10  s. Briefly centrifuge for less than 2  s at low speed. Capture the RNA Binding Beads on a magnetic stands as in the previous magnetic bead capture steps. Carefully aspirate and discard the supernatant and remove the tube from the magnetic stand. Briefly centrifuge the tube as in previous steps and place it back on the magnetic stand. Remove any liquid in the tube with a small-bore pipet tip, remove the tube from the magnetic stand, and allow the beads to air-dry for 5 min with the caps left open (see Note 6). 13. Add 30  ml warm Elution Buffer to each sample and vortex vigorously for 10 s to thoroughly resuspend the RNA Binding Beads. Incubate at 58°C for 5 min. Vortex the sample vigorously for 10  s to thoroughly resuspend the RNA Binding Beads and centrifuge for less than 2 s at low speed. 14. Capture the RNA Binding Beads on a magnetic stand as in the previous magnetic bead capture steps. Be especially careful at this step to avoid disturbing the RNA Binding Beads when collecting the supernatant. The purified RNA will be in the supernatant, and transfer to a new 1.5  ml microcentrifuge tube (frequently some of the RNA Binding Beads are carried over to the eluate, tinting it brownish but this does not affect absorbance or downstream applications). 15. Quantitate RNA using the NanoDrop or any other spectrophotometer, blank the instrument with Elution Buffer. A 260/280 absorbance ratio between 1.8 and 2.0 is typical of pure RNA. Assess quality of the RNA using the Agilent Bioanalyzer. If the RNA will not be used immediately, store at –80°C. Processing whole blood total RNA with the GLOBINclear Human Kit can reduce RNA yield by as much as 30% though we have observed average decrease in yield of about 15% (see Note 7). 3.3.3. Isolation of Mononuclear Cells and Separation of Plasma from CPT Tubes and the Processing of the PPT Tubes

The CPT tube and the 2 PPT tubes are centrifuged (swing-bucket rotor) at room temperature, 20 min at 1,700 × g. It is important to centrifuge the CPT tube within 2 h of draw as cell yields considerably decline and red blood cell contamination of the cell fraction increases thereafter. (http://www.bd.com/vacutainer/ products/molecular/citrate/limitations.asp). After centrifugation the 2 PPT tubes are frozen upright in a –20°C freezer. As for the CPT tube, the stopper is removed and approximately 3 ml of clear plasma from the upper phase is transferred to a 4 ml cryovial and then frozen at –20°C. The CPT tube is then recapped with the stopper and inverted a couple of times to mix the remaining plasma with the mononuclear cells. This cell suspension is then decanted into a 15 ml conical tube and PBS is added to bring the volume to 15  ml and mixed. This tube is then centrifuged

16

Mondala, Salomon, and Head

(swing-bucket rotor) at room temperature, 15  min at 300 × g. The supernatant is then removed and the cell pellet is resuspended with a few microliters of PBS and the volume is brought up to 15  ml with more PBS (see Note 8).This cell washing step is repeated a second time after which the cell pellet is resuspended in 1  ml of RNA later, transferred to a 2  ml cryovial, and then frozen at –20°C. 3.3.4. Extraction of RNA, DNA, and Protein from Mononuclear Cells

The manufacturer (Qiagen) recommended protocol was followed. The AllPrep DNA/RNA/Protein Mini Kit allows the simultaneous purification of genomic DNA, total RNA, and total protein from a single sample. In this protocol the RNA is first extracted to completion then the protein is processed to the point that it is pelleted and finally the DNA is purified to completion (see Note 9). 1. Prior to starting the procedure prepare the following reagents. Add 10 ml b-mercaptoethanol per 1 ml Buffer RLT. This is stable at room temperature for 1 month. Buffer RLT may form a precipitate during storage, redissolve by warming, and then return to room temperature. 2. Buffer RPE, Buffer AW, and Buffer AW2 are supplied as a concentrate, add the appropriate volume of ethanol (100%) as indicated on the bottle to obtain a working solution. 3. Thaw frozen cells resuspended in RNA later on ice. Pellet cells by centrifuging at maximum speed for 2 min. Carefully remove all the supernatant by aspiration. 4. Flick the tube to loosen the pellet and then disrupt the cells by adding 600 ml Buffer RLT, vortex lightly or pipet to mix till no cell clumps remain. 5. Homogenize the lysate by passing it through an 18-gauge needle attached to a 5 ml syringe at least 5 times. 6. Transfer the lysate to an AllPrep DNA spin column placed in a 2 ml collection tube, centrifuge for 30 s at 8,000 × g. 7. Place the DNA spin column in a new 2 ml collection tube and store at 4°C for later DNA purification. Use the flow-through for RNA and protein purification. 8. To the flow-through add 400 ml ethanol (100%), mix well by pipetting. Do not centrifuge. Proceed immediately to next step. 9. Transfer 700 ml of the sample, including any precipitate that may have formed to an RNeasy spin column placed in a 2 ml collection tube. Centrifuge for 15 s at 8,000 × g. Transfer the flow-through to a 2 ml microcentrifuge tube for protein purification. Add the remaining sample to the same RNeasy spin column, centrifuge for 15 s at 8,000 × g. Combine the flowthrough in the 2 ml microcentrifuge tube. 10. Add 700 ml Buffer RW1 to the RNeasy spin column, centrifuge for 15 s at 8,000 × g, discard the flow-through.

Multicenter Clinical Sample Collection for Microarray Analysis

17

11. Add 500 ml Buffer RPE to the RNeasy spin column, centrifuge for 15 s at 8,000 × g, and discard the flow-through. 12. Add 500 ml Buffer RPE to the RNeasy spin column, centrifuge for 2 min at 8,000 × g, and discard the flow-through. 13. Carefully remove the RNeasy spin column and transfer to a new 2 ml collection tube, centrifuge at maximum speed for 1 min. 14. Place the RNeasy spin column in a new 1.5 ml microcentrifuge tube, carefully add 50 ml RNase-free water directly on the spin column membrane, centrifuge for 1 min at 8,000 × g to elute the RNA. 15. Quantitate RNA using the NanoDrop or any other spectrophotometer, blank the instrument with water. A 260/280 absorbance ratio between 1.8 and 2.0 is typical of pure RNA. Assess quality of the RNA using the Agilent Bioanalyzer. If the RNA will not be used immediately, store at –80°C. Typical RNA yield from a single 8 ml draw CPT tube is 3–7 mg. 16. To start the total protein precipitation, add 1 volume (~1,000 ml) of Buffer APP to the flow-through from step 7. Mix vigorously and incubate at room temperature for 10 min to precipitate protein. 17. Centrifuge at full speed for 10 min and carefully decant the supernatant. 18. Add 500 ml 70% ethanol to the pellet, centrifuge at full speed for 1 min, and then remove as much of the supernatant by carefully decanting followed by a pipet. 19. Dry the protein pellet for about 20 min at room temperature. Protein pellet can now be stored at –80°C. This protein fraction can now be quantified and is suitable for tandem mass spectrometry analysis. Typical protein yield from a single 8  ml draw CPT tube is 50–90 mg. 20. To complete the purification of genomic DNA, add 500 ml Buffer AW1 to the DNA spin column from step 5. Centrifuge for 15 s at 8,000 × g to wash the membrane. Discard the flowthrough. 21. Add 500 ml Buffer AW2 to the DNA spin column. Centrifuge for 2 min at full speed to wash the membrane. 22. Carefully remove the spin column avoiding contact with the flow-through. Place the spin column in a new 1.5 ml microcentrifuge tube. Add 100  ml Buffer EB directly to the spin column membrane, incubate at room temperature for 1 min, and then centrifuge for 1 min at 8,000 × g to elute the DNA. 23. Quantitate DNA using the NanoDrop or any other spectrophotometer, blank the instrument with Buffer EB. A 260/280 absorbance ratio between 1.7 and 1.9 is typical of pure DNA.

18

Mondala, Salomon, and Head

Check size distribution (15–30 kb) by running an aliquot on an agarose gel. If the DNA will not be used immediately store at –80°C. Typical DNA yield from a single 8 ml draw CPT tube is 4–12 mg. 3.3.5. Extraction of RNA, DNA, and Protein from Kidney Biopsies 3.3.5.1. RNA Extraction: Trizol

1. Bring Trizol reagent to room temperature. 2. Turn on constant temperature incubator to 60°C. 3. If precipitate has formed in Buffer RLT, redissolve by warming and place at room temperature. 4. Buffer RPE is supplied as a concentrate. Add 4 volumes of ethanol (100%) before using for the first time. 5. In a chemical fume hood/biosafety cabinet, thaw biopsy core submerged in RNAlater on ice. Add 1 ml of Trizol reagent into a properly decontaminated, RNase-free 2 ml Wheaton Plastic Coated Tissue Grinder. 6. Carefully and quickly transfer core tissue and all smaller pieces into the grinder using forceps. Manually homogenize tissue using the Teflon coated pestle until completely homogenized as determined by visual inspection. Wear eye protection during this step. Incubate at room temperature for 5 min. 7. Transfer sample to a 1.5  ml microcentrifuge tube and add 200 ml chloroform, cap securely and vortex lightly for 20 s. Incubate at room temperature for 3  min. Centrifuge at 12,000 × g, 4°C for 15 min. 8. Carefully remove upper aqueous layer down to the interphase using a P200 pipet tip and transfer into a new 1.5 ml microcentrifuge tube (~500 ml volume). Save the tube containing the Trizol/Chloroform mixture for subsequent DNA and protein extraction (freeze the sample if the isolation of DNA and protein will be done another day). 9. Add 500  ml of room temperature isopropanol (100%) and mix by inversion. 10. Incubate at room temperature for 10 min. 11. Spin at 12,000 × g, 4°C for 10 min. 12. Carefully decant the supernatant. 13. Add 500 ml ETOH (70%), do not resuspend pellet. 14. Spin at 7,000 × g, 4°C for 5 min. 15. Carefully decant the supernatant and turn tubes upside down on Kim wipes. Using a P10 pipet tip, aspirate any remaining ETOH and then immediately add 100 ml DEPC water. Do not allow the RNA pellet to dry completely, do not speedvac, do not resuspend the pellet. 16. Incubate the tube at 55°C in a constant temperature incubator for 10 min.

Multicenter Clinical Sample Collection for Microarray Analysis

19

17. Pellet usually dissolves by itself, tap the tube a couple times, and then quick spin. 18. In a fume hood, make up enough Buffer RLT with b-mercaptoethanol (10 ml bME: 1 ml Buffer RLT). 19. Add 350 ml Buffer RLT containing bME to each tube containing sample (total RNA in 100 ml DEPC-treated water). 20. Add 250 ml cold ethanol (100%), mix. 21. Add sample to an RNeasy spin column. 22. Spin at 10,000 × g, room temperature for 1 min. 23. Reapply flow-through to the column. 24. Spin at 10,000 × g, room temperature for 1 min. 25. Place column in a new 2 ml collection tube. 26. Add 500 ml Buffer RPE buffer. 27. Spin at 10,000 × g, room temperature for 1 min. 28. Discard flow-through. 29. Add 500 ml Buffer RPE buffer. 30. Spin at 10,000 × g, room temperature for 1.5 min. 31. Discard flow-through. 32. Spin at 10,000 × g, room temperature for 2 min. 33. Heat an aliquot of water to 70°C. 34. Place column in a new 1.5 ml tube. 35. Add 50 ml of 70°C water to the membrane, incubate at room temperature for 1 min. 36. Spin at 14,000 × g, room temperature for 2 min. 37. Quantitate RNA using the NanoDrop or any other spectrophotometer, blank the instrument with water. A 260/280 absorbance ratio between 1.8 and 2.0 is typical of pure RNA. Assess quality of the RNA using the Agilent Bioanalyzer. If the RNA will not be used immediately, store at –80°C. Typical RNA yield from a single 18G kidney biopsy core is 3–10 mg. Yield varies greatly depending on the cellular make up and overall size of the needle core. 3.3.5.2. DNA Extraction: Trizol

1. Total RNA Isolation must be performed prior to DNA Isolation. 2. Set constant temperature incubator to 37°C. 3. Prepare wash buffer that is, 0.1 M NaCitrate in 10% ethanol, 2 ml is used per sample. 4. Spin the tube containing Trizol reagent and sample from step 8 above, at 12,000 × g, room temperature for 2 min to separate phases.

20

Mondala, Salomon, and Head

5. Carefully remove any remaining upper aqueous phase using a pipet and discard. 6. Add 400 ml of ethanol (100%), mix by inversion. 7. Spin at 2,000 × g, 4°C for 5 min. 8. Aliquot supernatant equally into two separate tubes (~500 ml each) and set aside for the protein extraction later. 9. Add 1 ml 0.1 M NaCitrate in 10% ethanol (wash buffer) to pellet, vortex lightly. 10. Incubate at room temperature for 30 min, mix by inversion periodically. 11. Spin at 2,000 × g, 4°C for 5 min. 12. Remove supernatant and add 1 ml of wash buffer. 13. Incubate at room temperature for 30 min, mix by inversion periodically. 14. Spin at 2,000 × g, 4°C for 5 min. 15. Remove supernatant and add 1 ml ethanol (75%). 16. Incubate at room temperature for 20 min, mix by inversion periodically. 17. Spin at 2,000 × g, 4°C for 5 min. 18. Remove supernatant, dry down in speed-vac on medium heat for ~30 s, do not over dry. 19. Add 300 ml 8 mM NaOH, pass the pellet through a pipet tip a few times, and incubate overnight at 37°C. 20. Pre-spin Phase Lock Gel (PLG) tube at 14,000 rpm for 20 s. Add equal volume of phenol/chloroform as NaOH (~300 ml) to sample, vortex, transfer phenol/sample mix to PLG tube and spin at 14,000 rpm for 2 min. Transfer top, clear aqueous phase to new 1.5 ml microcentrifuge tube. 21. Add 0.1 volumes 3  M Sodium Acetate (~30  ml). Add 1  ml 5 mg/ml Glycogen. Add 2.5 volumes ice-cold 100% ethanol (~830 ml). Incubate 1 h at –80°C. 22. Spin at 14,000 rpm, 4°C for 20 min. 23. Remove and discard supernatant and add 1 ml ice-cold ethanol (80%). 24. Spin at 14,000 rpm, room temperature for 2 min. 25. Remove and discard supernatant and add 1  ml ice-cold ethanol (80%). 26. Spin at 14,000 rpm, room temperature for 2 min. 27. Remove and discard supernatant, let pellet air dry, and resuspend in 22 ml of water. 28. Quantitate on NanoDrop or other spectrophotometer, blank instrument with water. A 260/280 absorbance ratio between

Multicenter Clinical Sample Collection for Microarray Analysis

21

1.7 and 1.9 is typical of pure DNA. Check size distribution by running an aliquot on an agarose gel. Typical DNA yield from a single 18G kidney biopsy core is 1–15 mg. This can vary greatly depending on the cellular make up and overall size of the needle core. 3.3.5.3. Protein Extraction: Trizol

1. Total RNA Isolation and DNA Isolation must be performed prior to Protein Isolation. 2. Add 750 ml isopropanol (100%) into each of the 2 tubes from step 8 above, invert 20 times. 3. Incubate at room temperature for 10 min. 4. Spin at 12,000 × g, 4°C for 7 min. 5. Remove supernatant, wash with 750 ml 0.3 M guanidine HCl in 95% ethanol, vortex lightly. 6. Incubate at room temperature for 20 min. 7. Spin at 7,500 × g, 4°C for 5 min. 8. Wash a total of 3 times. 9. After the last wash, add 1  ml ethanol (100%) to the pellet, vortex lightly. 10. Incubate at room temperature for 20 min. 11. Spin at 7,500 × g, 4°C for 5 min. 12. Remove supernatant. 13. Store protein pellet at –80°C. This protein fraction can now be quantified and is suitable for tandem mass spectrometry analysis. Typical protein yield from a single 18G kidney biopsy core is 50–100 mg. This can vary greatly depending on the overall size of the needle core.

3.3.6. DNA Extraction from Whole Blood

The QIAamp DNA Blood Midi Kit (100 reaction kit) provides a simple, fast method for purifying DNA from blood. The separation of leukocytes is not necessary and no phenol/chloroform or alcohol precipitation is required. DNA purified using this method ranges in size up to 50 kb (see Note 10). 1. Equilibrate samples to room temperature, thoroughly mix by inversion. 2. Set constant temperature incubator to 70°C. 3. Before starting the DNA purification prepare the Protease Working Solution by pipetting 5.5 ml water into the vial of lyophilized Qiagen Protease. Freeze aliquots of the unused protease at –20°C for later use. 4. Add 125 ml of ethanol (100%) to Buffer AW1. 5. Add 150 ml of ethanol (100%) to Buffer AW2.

22

Mondala, Salomon, and Head

6. If a precipitate has formed in Buffer AL, dissolve by incubating at 56°C. 7. Pipet 400 ml of Protease into the bottom of 15 ml centrifuge tube. 8. Add 4 ml of blood and mix briefly (add appropriate amount of PBS if sample is less than 4 ml). 9. Add 4.8 ml Buffer AL to the samples, mix thoroughly by vortexing at least 3 times for 5 s each time. Do not add Protease directly to Buffer AL. 10. Incubate at 70°C for 10  min in a constant temperature incubator. 11. Add 4 ml of ethanol (100%) to the sample and mix again by vortexing. 12. Carefully transfer 3.3 ml of sample onto a QIAamp Midi column placed in a 15 ml centrifugation tube. Close the cap and centrifuge for 3 min at 1,900 × g, room temperature, brake on using a swing-out rotor with adapter for round-bottom tubes (do not over tighten caps, if the caps are tightened until they snap they may loosen during centrifugation and damage the centrifuge). 13. Remove the Midi column, discard the filtrate, and wipe off any spillage from the thread of the 15 ml centrifugation tube before re-inserting the column in to the 15 ml centrifugation tube. Load another 3.3 ml of sample onto the column, close the cap, and centrifuge for 3 min at 1,900 × g, room temperature, brake on. 14. Remove the Midi column, discard the filtrate and wipe off any spillage from the thread of the 15 ml centrifugation tube before re-inserting the column in to the 15 ml centrifugation tube. Load remaining sample onto the column, close the cap, and centrifuge for 3  min at 1,900 × g, room temperature, brake on. 15. Remove the Midi column, discard the filtrate, and wipe off any spillage from the thread of the 15 ml centrifugation tube before re-inserting the column in to the 15 ml centrifugation tube. Carefully without moistening the rim add 4 ml Buffer AW1 to the column, close the cap and centrifuge for 5 min at 3,500 × g, room temperature, brake on. 16. Remove the Midi column, discard the filtrate, and wipe off any spillage from the thread of the 15 ml centrifugation tube before re-inserting the column in to the 15 ml centrifugation tube. Carefully without moistening the rim add 4 ml Buffer AW2 to the column, close the cap, and centrifuge for 5 min at 3,500 × g, room temperature, brake on.

Multicenter Clinical Sample Collection for Microarray Analysis

23

17. Wipe off any spillage off the Midi column and place the column in a clean 15 ml centrifugation tube and discard the tube containing the filtrate. Add 600 ml of water, close the cap, and incubate at room temperature for 5 min. Centrifuge at 3,500 × g for 10 min, room temperature, brake on. 18. Reload the 600 ml eluate containing the DNA onto the membrane of the Midi column. Close the cap and incubate at room temperature for 5  min. Centrifuge at 3,500 × g for 10 min, room temperature, brake on. 19. Quantitate on NanoDrop, blank instrument with water. A 260/280 absorbance ratio between 1.7 and 1.9 is typical of pure DNA. Check size distribution (15–30 kb) by running an aliquot on an agarose gel. If the DNA will not be used immediately store at –80°C. Typical DNA yield from 4 ml blood is 40–60 mg. 3.3.7. DNA Extraction from Mononuclear Cells

The QIAamp DNA Mini Kit (50 reaction kit) provides a simple, fast method for purifying DNA from cells. DNA purified using this method ranges in size up to 50  kb. The manufacturer (Qiagen) recommended protocol was followed. In our Transplant Genomics Project we typically would process 2–5 million cells suspended media. 1. Equilibrate samples to room temperature. 2. Set constant temperature incubator to 56°C. 3. Before starting the DNA purification prepare the Protease Working Solution by pipetting 1.2 ml protease solvent in to the vial of lyophilized Qiagen Protease. Freeze aliquots of unused protease at –20°C for later use. 4. Add 25 ml of ethanol (100%) to Buffer AW1. 5. Add 30 ml of ethanol (100%) to Buffer AW2. 6. If a precipitate has formed in Buffer AL, dissolve by incubating at 56°C. 7. Centrifuge sample containing cells at maximum speed (20,000 × g) to pellet cells, remove supernatant, and resuspend pellet with 200 ml PBS. 8. Pipet 20  ml Qiagen Protease into a 1.5  ml microcentrifuge tube. 9. Add 200 ml sample to the microcentrifuge tube. 10. Add 200 ml Buffer AL to the samples, mix by pulse-vortexing for 15 s. Note: Do not add Protease directly to Buffer AL. 11. Incubate at –56°C for 10 min. 12. Briefly centrifuge the sample to collect all drops from the inside of the lid.

24

Mondala, Salomon, and Head

13. Add 200  ml of ethanol (100%) to the sample and mix by pulse-vortexing for 15 s. After mixing briefly centrifuge the tube to collect all drops from the inside of the lid. 14. Carefully apply sample mixture directly on to a Mini spin column placed in a 2 ml collection tube without wetting the rim. Close cap and centrifuge at maximum speed for 1 min. Place the spin column in a clean 2 ml collection tube and discard the tube containing the filtrate. 15. Carefully open the spin columns and add 500  ml Buffer AW1 without wetting the rim. Close the cap and centrifuge at 6,000 × g for 1 min. Place the spin column in a clean 2 ml collection tube and discard the tube containing the filtrate. 16. Carefully open the spin column and add 500 ml Buffer AW2 without wetting the rim. Close the cap and centrifuge at maximum speed for 3 min. 17. Place the spin column in a new 2 ml collection tube and discard the old tube with the filtrate. Centrifuge at full speed for 1 min. 18. Place the spin column in a clean 1.5 ml microcentrifuge tube and discard the tube containing the filtrate. Carefully open the spin column and add 200 ml water. Incubate at room temperature for 5  min and then centrifuge at 6,000 × g for 5 min. 19. Reload the eluate containing the DNA onto the membrane of the spin column and close the cap and incubate at room temperature for 5  min and then centrifuge at 6,000 × g for 5 min. 20. Quantitate on NanoDrop, blank instrument with water. A 260/280 absorbance ratio between 1.7 and 1.9 is typical of pure DNA. Check size distribution (15–30 kb) by running an aliquot on an agarose gel. If the DNA will not be used immediately store at –80°C. Typical DNA yield from 2 to 5 million mononuclear cells is 4–10 mg.

4. Notes 1. Various labels were tested, these labels passed requirement that they adhere, stay intact and that the printed barcode not smear (remain scanable using a handheld barcode reader) under extreme cold temperature such as dry ice and after numerous freeze-thaw cycles. 2. We instruct all clinical centers to collect blood tubes in a specific order to ensure all collection procedures are done as

Multicenter Clinical Sample Collection for Microarray Analysis

25

uniformly as possible. PAXgene tubes are collected first followed by the CPT tube and then the PPT tubes. 3. It is important that detailed instruction must be given to the physicians or radiologists performing the biopsies to immediately and completely submerge the kidney core into RNAlater to eliminate the impact of RNA degradation. 4. Before starting the PAXgene RNA purification procedure, incubate the tube at room temperature for at least 2 h in order to ensure complete blood cell lysis. If the tube was immediately frozen or stored at 2–8°C after blood collection, then after removal from storage, first thaw to room temperature for at least 2 h, invert 10 times and then incubate at room temperature for an additional 2 h. After incubating, invert the tube another 10 times. 5. Divide DNase I into single-use aliquots of 10.5 ml and store at 2–8°C for up to 6 weeks or at –20°C for up to 6 months. Thaw appropriate number of DNase I stock solution aliquots for on-column DNase digestion. Do not refreeze the aliquots after thawing. 6. Do not air-dry the RNA Binding Beads for more than 5 min, in our experience over-drying at this step resulted in lower yield. 7. Initially we had observed as much as a 30% reduction in total RNA after globin reduction. Subsequently, yield was reduced on average 15% only. This was attributed to an increase in familiarity with the magnetic bead technique decreasing the unintentional loss of binding beads when removing supernatant from the captured beads. 8. The cell pellet can be easily resuspended by flicking the tube a couple of times. 9. Qiagen now has a supplemental protocol for the purification of miRNAusingtheAllPrepDNA/RNA/ProteinMiniKitandRNeasy MinElute Cleanup Kit (http://www1.qiagen.com/products/ RnaStabilizationPurification/AllPrepDNARNAProteinMiniKit. aspx#Tabs=t2). 10. A modified version of the manufacturer (Qiagen) recommended protocol was followed. We increased the starting blood volume from 2 to 4 ml.

Acknowledgments This work is supported by the National Institute of Allergy and Infectious Diseases (NIAID) Program Project Grant U19 AI63603 Genomics for Kidney Transplantation.

26

Mondala, Salomon, and Head

References 1. Copland JA, Davies PJ, Shipley GL, Wood CG, Luxon BA, Urban RJ (2003) The use of DNA microarrays to assess clinical samples: the transition from bedside to bench to bedside. Recent Prog Horm Res 58:25–53 2. Al-Mulla F (2007) Utilization of microarray platforms in clinical practice. Methods Mol Biol 382:115–136 3. Flechner SM, Kurian SM, Head SR, Sharp SM, Whisenant TC, Zhang J, Horvath S, Mondala T, Gilmartin T, Cook DJ, Kay SA, Walker, JR, Salomon DR (2004). Chara­ cterizing acute kidney transplant rejection by gene profiling of biopsies and peripheral blood lymphocytes. Am J Transplant 4(9): 1475–1489 4. Flechner SM, Kurian SM, Solez K, Cook DJ, Burke JT, Rollin H, Hammond JA, Whisenant T, Lanigan CM, Head SR, Salomon DR (2004) De novo kidney transplantation without use of calcineurin inhibitors preserves renal structure and function at two years. Am J Transplant 4(11):1776–1785 5. Kurian SM, Flechner SM, Kaouk J, Modlin C, Goldfarb D, Cook DJ, Head S, Salomon DR (2005) Laparoscopic donor nephrectomy gene expression profiling reveals upregulation of

6.

7.

8.

9.

10.

stress and ischemia associated genes compared to control kidneys. Transplantation 80(8): 1067–1071 Kurian S, Grigoryev Y, Head S, Campbell D, Mondala T, Salomon DR (2007) Applying genomics to organ transplantation medicine in both discovery and validation of biomarkers. Int Immunopharmacol 7(14):1948–1960 Burczynski ME et  al (2005) Transcriptional profiles in peripheral blood mononuclear cells prognostic of clinical outcomes in patients with advanced renal cell carcinoma. Clin Cancer Res 11(3):1181–1189 Cobb JP et al (2005) Application of genomewide expression analysis to human health and disease. Proc Natl Acad Sci USA 102(13): 4801–4806 Tsuang MT, Nossova N, Yager T, Tsuang MM, Guo SC, Shyu KG, Glatt SJ, Liew CC (2005) Assessing the validity of blood-based gene expression profiles for the classification of schizophrenia and bipolar disorder: a preliminary report. Am J Med Genet B Neuropsychiatr Genet 133(1):1–5 Feezor RJ et al (2004) Whole blood and leukocyte RNA isolation for gene expression analyses. Physiol Genomics 19:247–254

Chapter 2 Isolation of Total RNA from Transgenic Mouse Melanoma Subsets Using Fluorescence-Activated Cell Sorting Scott Tighe and Matthew A. Held Abstract The majority of tumors, including melanoma, are phenotypically heterogeneous in that they contain various cell populations with differential expression of cell surface antigens such as CD133/Prominin-1. We have used fluorescence-activated cell sorting (FACS) technology to purify CD133+ and CD133− cellular subsets from mouse melanoma models for high-quality total RNA practical for downstream applications such as expression profiling. Implementation of this strategy can lead to higher resolution of transcripts that are potentially important for the survival and functionality of one cancer cell population relative to another. Suboptimal extraction of RNA after FACS is common and can ultimately result in misinterpretations that impede the effective design of novel therapies. Here, we describe a number of methods that have been amenable to the successful isolation of high-quality total RNA after FACS of CD133+ and CD133− mouse melanoma cell fractions. Key words: Melanoma, Mouse models of cancer, FACS, Cell surface markers, Cell subsets, RNA isolation, RNA FACS sorting

1. Introduction Methods for genome-wide expression analyses, such as DNA microarrays (1), can be used to delineate global RNA expression differences between cancer cell subsets that show variations in function such as their abilities to resist chemotherapy or propagate tumors. The cell surface antigen CD133 has been demonstrated to identify cancer cells from a variety of solid-tissue cancers such as melanoma that display higher tumorigenicity or treatment resistance (2–5) and can be characterized through, for example, gene expression profiling of the CD133+ and CD133− subset phenotypes. To accomplish such a task requires well-established flow cytometric sorting methods and RNA extraction protocols (6, 7). Sridar V. Chittur (ed.), Microarray Methods for Drug Discovery, Methods in Molecular Biology, vol. 632, DOI 10.1007/978-1-60761-663-4_2, © Humana Press, a part of Springer Science+Business Media, LLC 2010

27

28

Tighe and Held

It is well known that RNA is a sensitive nucleic acid that can easily degrade as a result of erroneous introduction of ribonucleases (RNases) either from instrumentation, immunostaining procedures, end user, or endogenously from the sample itself (8, 9). In addition, improper extraction and storage of RNA can decrease its overall half-life, compromising future utility. Here we establish several workflows for fluorescence-activated cell sorting (FACS) of CD133+ and CD133− melanoma cell subsets for highquality total RNA purification including instrument decontamination, cell surface marker labeling, cell sorting procedures, and RNA handling and extraction methods. In addition, we discuss quantification techniques and integrity analyses used for validating the RNA quality of these cellular subsets after FACS.

2. Materials 2.1. Cell Culture and Antibody Staining

1. Dulbecco’s Modified Eagle’s Medium-F12, 1:1 (Gibco). 2. Fetal-bovine serum (FBS), US-origin, irradiated, heatinactivated (Hyclone). 3. Modified Eagle’s (Cellgro).

Medium

Nonessential

amino

acids

4. Trypsin 0.25%/2.2 mM EDTA (Cellgro). 5. Penicillin streptomycin (Pen/Strep), 1 × 104 U/ml each (Cellgro). 6. Phosphate-buffered saline (PBS), RNase-free (Ambion). 7. Bovine serum albumin (BSA), RNase-free (Equitech-Bio). 8. RNase inhibitor, e.g., RiboLock (Fermentas Corp). 9. Dulbecco’s PBS containing 100  U/ml RiboLock (PBSRIBO). 10. RNase-free 1.5  ml microcentrifuge tubes (Axygen, #MCT175c). 11. RNaseZap (Ambion). 12. Rat anti-CD133 mouse monoclonal antibody (eBioscience). 13. AlexaFluor488 chicken antirat IgG secondary antibody (Invitrogen). 2.2. Flow Cytometry and Sorting

1. BD FACSAria flow cytometer or equivalent. 2. RNase-free water (VWR Scientific). 3. Sterile polystyrene round-bottom tubes for flow 5  ml (BD Falcon). 4. Bleach 10% (0.525% sodium hypochlorite).

Isolation of Total RNA from Transgenic Mouse Melanoma Subsets Using Fluorescence

29

5. Sterile sheath fluid (saline) RNase-free. 6. Propidium iodide (1  mg/ml) solution-ultra-high purity (Enzo, #enz-52403R). 7. Bovine serum albumin, RNase-free (Equitech-Bio). 8. RNaseAlert RNase detection system (Ambion). 2.3. RNA Isolation

1. Trizol or Trizol LS or equivalent. 2. RNeasy Micro Kit (Qiagen). 3. Beta-mercaptoethanol. 4. Chloroform (100% ACS Grade). 5. 100% Ethanol (Electron Microscopy Sciences). 6. MaxyClear RNase-free tubes 1.5, 15, and 50 ml (Axygen). 7. QIAvac-24 Plus Vacuum manifold (Qiagen). 8. Nanodrop ND1000 spectrophotometer. 9. Qubit Spectrofluorometer (Invitrogen). 10. Quant-IT RNA reagents (Invitrogen). 11. Agilent 2100 Bioanalyzer or equivalent.

3. Methods 3.1. Quality Control of the FluorescenceActivated Cell Sorter

Before proceeding with FACS of cell subsets, stringent quality control of the instrumentation is mandatory to ensure the success of good quality RNA isolation from sorted cell populations. This involves thorough decontamination followed by empirical validation of FACS machine sanitation. Decontamination time will depend on the instrument type, age, and degree of contamination. However, procedures for sanitizing any FACS instrument are similar, and so a review of the following steps is warranted. It is urged to perform all steps with RNaseZap-treated gloves in a low contamination environment (see Note 1). Once decontamination is complete, a test sort using noncritical cells with a known viability >80% should be performed to test the instrument (see Note 2). 1. Ensure the dip tube, septa, flow cell, tubing lines, and nozzles have been decontaminated with 10% bleach, 100% ethanol, RNaseZap, autoclaving, or other suitable qualifying technique prior to the sort. 2. Ensure sheath tank and fluid are RNase-free. Quality control sampling of each may be tested with RNase-detecting reagents such as RNaseAlert.

30

Tighe and Held

3. Replace all contaminated fluid lines and filters as outlined by the manufacturer. 4. Prior to sort, run several tubes of 10% bleach through the flow cytometer including the sorting components followed by flushing with sterile RNase-free sheath fluid or PBSRIBO. 5. Perform a sort using a bead solution containing 400 U/ml of an RNase inhibitor just prior to sorting critical samples. 3.2. Preparation of Melanoma Cells and Antibody Staining for FACS

Melanoma cell lines were derived from transgenic conditional mouse melanoma tumors as previously described (10). Tumors were finely minced using aseptic technique and enzymatically dissociated with 0.05% trypsin/0.55mM EDTA for 30 min at room temperature, with thorough mincing every 10 min. Dissociated tumors were then lightly triturated 15–20 times, and the resulting suspensions were transferred to tissue culture treated 10 cm adherent dishes. Melanoma cultures were grown in 1:1 DMEM:F12 media with 5% FCS and 1% Pen/Strep (media complete) in a cell culture incubator at 37°C with 5% CO2 and allowed to grow until approximately 75% confluent. The following protocol was then followed with proper RNA handling in a biosafety cabinet or PCR hood for indirect antibody labeling of cells for the surface marker CD133 followed by FACS of CD133+ and CD133− cellular subsets using a BD FACSAria flow cytometer (see Note 3). 1. Aspirate media from 10 cm adherent melanoma culture dishes and detach cells by briefly incubating (2–3 min) with 1 ml of 0.25% trypsin/2.2 mM EDTA, followed by neutralization of trypsin with 10 ml of media complete. 2. Centrifuge cell suspensions at 800 × g for 5  min, aspirate supernatant, and resuspend cell pellet in 1  ml of 1× PBSRIBO with 2% BSA (PBS-RIBO-BSA) (see Note 4). 3. Perform a viability count using a hemocytometer and Trypan Blue dead-cell discrimination dye (see Note 5). 4. For each sample and control, transfer 5 × 105 cells to a new tube. Controls should include samples with primary antibody only, secondary antibody only (or isotype-control only), unstained cells, and propidium iodide only. These are required for fluorescent compensation and proper gate positioning. 5. All samples and controls are centrifuged at 800 × g, aspirated, and resuspended in 100  ml of PBS-RIBO-BSA followed by staining with 1  mg/ml final concentration anti-CD133 primary antibody for 30 min at 4°C (in fridge, not on ice). 6. Samples are quenched with 900  ml PBS-RIBO-BSA, centrifuged, aspirated, and resuspended in 100 ml PBS-RIBO-BSA.

Isolation of Total RNA from Transgenic Mouse Melanoma Subsets Using Fluorescence

31

7. Each sample is stained with a species-matched, AlexaFluor-488 chicken antirat IgG secondary antibody at 1:1,000 for 20 min in a dark fridge at 4°C. 8. Centrifuge cells, aspirate supernatant, and resuspend cells in 500 ml PBS-RIBO-BSA with a final concentration of 1 µg/ml RNase-free propidium iodide for dead-cell discrimination. Transfer samples to FACS machine-compatible, sterile 5 ml round-bottom tubes and cap. 9. During the flow cytometric procedure, exclude all propidium iodide-positive signals (i.e., dead cells). Whenever possible, use forward scatter (FSC), side scatter (SSC) height, width, and area measurements to exclude any potential doublets or putative apoptotic/dead cells. Live, single cells are then analyzed and sorted by FACS on CD133 signal into precooled, RNase-free 1.5 ml microcentrifuge tubes for subsequent total RNA extraction of purified cell subsets (see Note 6). 3.3. Methods for Sorting Cell Subsets for Total RNA Extraction

There are a variety of procedures for recovering total RNA from sorted cells. The choice of any one protocol depends on two factors: (1) whether the type of FACS machine used for cell purification is mechanical or electrostatic and (2) whether high or low dispensed sort volumes are expected. Mechanical sorters, such as the BD FACSCalibur, use a mechanical sorting device called the “catcher tube” positioned near the flow cell and sort relatively slowly (e.g., 300 events/s) with a relatively high sort volume (e.g., 100 nl–10 µl per event). Therefore, direct sorting of cells using mechanical sorters is not ideal for sorting large numbers of cells directly into RNA extraction buffer as the high dispensed volumes will dilute the buffer substantially and impede RNA recovery. When using mechanical sorters, it is recommended to first centrifuge the sorted cells to form a cell pellet followed by addition of the chosen RNA extraction buffer as outlined below in Subheading 3.3.1. It is important to consider that any additional handling before adding the RNA extraction buffer, such as centrifuging, may lead to consequential gene expression changes (see Note 7). Electrostatic sorters or “stream-in-air” FACS machines can operate at much higher speeds (e.g., 25,000 events/s or more) and involve a vibrating nozzle by which cells exit within single droplets resulting in much smaller dispensed sort volumes (11). Electrostatic sorters are also capable of fitting various sized nozzles in order to accommodate for cell size and maximize cell viability during the procedure. For example, a 70 mm nozzle decreases flow stream width, thereby resulting in droplet volumes of 1 nl drops per event – an approach applicable to sorting small cell types (e.g., T lymphocytes). In contrast, a 100 mm nozzle will relax flow stream width slightly to accommodate larger sized cells

32

Tighe and Held

(e.g., tumor cells) which results in volumes of 5–10 nl drops per event. Although the exact droplet size may vary slightly based on the system settings of each individual flow cytometer, smaller dispensed volumes allow for sorting directly into RNA extraction buffer such as Trizol LS or RLT buffer from the RNeasy system (see Note 8). Although other alternative methods for RNA isolation from sorted cells are available, they will not be described here (see Note 9). 3.3.1. RNeasy System for RNA Isolation After Centrifugation

Isolation of cells from high sorted volumes, such as those from a mechanical sorter, will require a centrifugation step to collect the cell pellet followed by RNA isolation using a silica column approach, such as the RNeasy microcolumn, or a standard Trizol precipitation method as described by the manufacturer (see Note 10) (12). When sorting cells for RNA, it is important to consider adding an RNase inhibitor to the sort recovery tube prior to the sort and adjust to 5–20 U/ml following the sort whenever possible (see Note 11). Sorting directly into a cell preservation reagent for future RNA isolation should be avoided (see Note 12). 1. Immediately following the sort, aseptically centrifuge cells to a pellet at 1,000 × g for 10 min using a refrigerated centrifuge. 2. Using a sterile aspirator, remove all supernatant from the cell pellet. 3. Add 100 ml of RNase-free water and 350 ml of RLT buffer and vortex for 30 s (see manufacturer’s protocol) (13). 4. Add 250 ml of 100% EMS grade ethanol and vortex. 5. Using a micropipettor with aerosol resistant tip, transfer sample to the RNeasy microcolumn and centrifuge at >10,000 × g for 15  s. Replace waste capture tube containing the passthrough liquid. 6. A DNase treatment (steps 8–10) may be required when downstream methods involving random hexamer priming such as in the case of exon microarrays, RT-qPCR, or equivalent are used. If no DNase treatment is required, proceed to step 10 (see Note 13). 7. Apply 200 ml of RW1 buffer to the column and centrifuge at >10,000 × g for 15 s. 8. For each sample, prepare the DNase solution from the Qiagen RNase-free DNase kit by combining 70  ml of RDD buffer with 10 ml of DNase I (27.3 Kunitz units total) and applying 80 µl to the column’s silica membrane. Incubate at room temperature for 20 min. 9. Add 200 ml of RW1 buffer to the column and centrifuge at >10,000 × g for 15 s. Replace the waste capture tube containing the pass-through liquid.

Isolation of Total RNA from Transgenic Mouse Melanoma Subsets Using Fluorescence

33

10. Apply 0.5  ml RPE buffer to the column and centrifuge at >10,000 × g for 15 s. Replace the waste capture tube containing the pass-through liquid. 11. Repeat step 10. 12. Using a 20 ml pipette, remove the remaining liquid that may be caught up on the edge of the column’s inner O-ring. 13. Perform an extended centrifugation for 3–5 min to remove as much liquid from the membrane as possible. Do not centrifuge with column open as described in the manufacturer’s protocol. 14. Replace waste tube with a new standard RNase-free 1.5 ml microcentrifuge tube. 15. Apply 15 ml of 60°C RNase-free water directly to the center of the RNeasy microcolumn membrane and incubate at room temperature for 30 s. 16. Centrifuge at >10,000 × g for 15 s. 17. Carefully remove the 15 ml of sample from the tube and reapply it to the same RNeasy membrane again. Close column and centrifuge at >10,000 × g for 15 s. This reelution is performed with the same 15 ml aliquot to assure complete recovery of RNA from the entire surface of the column’s silica membrane. 18. Remove the RNeasy microcolumn from the microcentrifuge tube containing the 15 ml of sample, and add the equivalent of 20 U of RNase inhibitor and vortex. Store sample at −20°C. 19. Quantify the RNA using a high resolution spectrometer such as the Nanodrop ND-1000 and Qubit fluorometer (see Subheading 3.4.1). 20. Analyze the RNA quality using an Agilent 2100 Bioanalyzer or equivalent (see Subheading 3.4.2). 3.3.2. Direct Sort Method

When low sort volumes are expected, it is advantageous to sort directly into extraction reagent such as Trizol LS or RLT buffer in order to minimize downstream handling and inadvertent gene expression changes. Regardless of the method selected, it is imperative to maintain the exact ratio of aqueous sorted volume to extraction reagent consistent with the manufacturer’s recommendations and to extract RNA promptly. If immediate extraction is not possible, then short-term storage in dilute extraction reagent may be considered (see Note 14). Although direct sorting into extraction buffer is optimal for RNA recovery, secondary analyses such as microscopy or postsort cell purity validation will require additional steps (see Note 15).

34

Tighe and Held

3.3.2.1. Direct RNA Extraction Using a Combined Trizol LS-RNeasy Method

1. Start with 500 ml of Trizol LS in a sterile RNase-free FACS tube if choosing to sort directly into the Trizol LS. Otherwise, sort into 1× PBS-RIBO and spin cells down at 800 × g for 10  min, aspirate supernatant, and then add 500 ml of Trizol LS. 2. Multiple sort tubes may be used to collect cells if dispensed sort volumes exceed the volume capacity of the sort collection tubes. If so, use 500 ml starting Trizol LS volume for the extra sort tubes as well. 3. After the sort, use a pipette equipped with an aerosol resistant tip to measure the final volume in the tube. Subtract the amount of Trizol LS to determine the amount of dispensed liquid. 4. Adjust the amount of Trizol LS required to maintain the sample at a Trizol:dispensed volume ratio of at least 3:1. This may require the solution to be transferred to a larger RNase-free tube (see Note 16 for a mathematical example). 5. Add 200 ml of chloroform for every 750 ml of Trizol LS to the tube and mix. Let the samples sit on bench top for 3  min. Alternative organic phases may be used in place of chloroform but are not preferred by the authors (see Note 17). 6. Centrifuge at >10,000 × g for 10  min at 4°C to separate the top aqueous layer from the bottom layer and interface. If the volume of solution is too large to fit into a microcentrifuge tube, it can be transferred to a 15 or 50 ml centrifuge tube and spun down with a larger centrifuge (see Note 18). 7. Carefully remove samples from the centrifuge and transfer the top aqueous layer to an RNase-free tube. Determine the exact volume of the aqueous layer and add 1.5 times the volume of 100% RNase-free ethanol and mix (see Note 19). 8. Filter the entire volume through an RNeasy microcolumn. For larger volumes (e.g., >5 ml), a vacuum manifold is suggested for faster sample processing (step 9a). Smaller volumes can be processed as individual 700 ml applications to the same RNeasy column and centrifuged (step 9b). 9a. Vacuum manifold technique: Using the QiaVac manifold (see Fig.  1) or equivalent (14), turn on the vacuum pump, and open the selected receiver ports to allow suction. Saturate receiver ports with RNaseZap for 30  s followed by rinsing with 100% ethanol. Turn off pump and aseptically install RNeasy microcolumn to selected receiver port(s). Turn vacuum pump on and repeatedly load 700  ml aliquots of the same sample into the RNeasy column until all of the sample volume has been filtered through. Remove column from vacuum manifold and place in a standard 2 ml capture tube and continue to step 10.

Isolation of Total RNA from Transgenic Mouse Melanoma Subsets Using Fluorescence

35

9b. Centrifugation technique: Apply no more than 700 ml to the RNeasy microcolumn and spin at >10,000 × g for 15 s. If the total volume is greater than 700 ml, multiple loadings to the same column will be required. 10. Perform DNase I treatment if required as per Subheading 3.3.1, step 7. If a DNase treatment is not needed, proceed to step 11 below. 11. Apply 700 ml RPE buffer to each column and centrifuge at >10,000 × g for 15 s. Discard and replace the waste capture tube containing the pass-through liquid. 12. Repeat step 11 a total of four times. This is required to remove any remaining Trizol that may otherwise be bound to the silica membrane when a Trizol-based lysis protocol is performed. Any residual Trizol contamination will lead to inaccurate UV-based RNA quantitation at 260  nm (see Note 19). 13. Using a 20 ml pipette, remove the remaining liquid that may be caught on the inner edge of the column’s O-ring. 14. Perform an extended “dry” centrifugation at >10,000 × g for 2  min to remove as much residual liquid from the RNeasy microcolumn as possible. Do not centrifuge with column cap open. 15. Replace waste tube with a new standard RNase-free 1.5 ml microcentrifuge tube. 16. Apply 15 ml of 60°C RNase-free water directly to the center of the RNeasy microcolumn membrane, and incubate at room temperature for 30 s. 17. Centrifuge at >10,000 × g for 15 s. 18. Carefully remove the 15 ml of sample from the tube and reapply it to the same RNeasy membrane again. Close column

Fig. 1. Standard configuration for a vacuum manifold system fitted with RNeasy microcolumns. This approach allows the processing of large volumes of RNA extraction buffer (e.g., >5 ml) through the silica membrane without the use of a centrifuge.

36

Tighe and Held

and centrifuge at >10,000 × g for 15  s. This reelution is performed with the same 15  ml aliquot to assure complete recovery of RNA from the entire surface of the column. 19. Remove the RNeasy microcolumn from the microcentrifuge tube, and add the equivalent of 20 U of RNase inhibitor and vortex. At this point, samples may be stored at −20°C for short-term use or at −80°C for long-term storage. 20. Using both a UV spectrophotometer and fluorometer, such as the NanoDrop ND1000 and Qubit, determine the concentration of each sample. Make note of possible Trizol contamination as noted by a 270 nm absorbance peak on the UV spectrometer (see Subheading 3.4.1). In most cases, quantitative results for the fluorometer are lower than that of the UV spectrophotometer, but are considered more accurate. 3.3.2.2. Direct RNA Extraction Using RNeasy Microcolumn Method

When sorting directly into RLT buffer (guanidium isothiocyanate), a ratio of 100 ml of sorted sample to 350 ml of RLT should be maintained. In general, the Trizol LS method has a greater RNA recovery on cells with more resistant cell membranes, aggregated cells, or organisms with a cell wall, but is more costly and involves more reagents. RNA recovered by directly sorting into RLT buffer is typically much cleaner than that recovered with Trizol and does not require additional quantitation with a Qubit spectrofluorometer because there is no interfering 270 nm absorbance from trace amounts of Trizol carryover. 1. Start with 500 ml of RLT buffer with 5 ml BME in a sterile RNase-free FACS tube. 2. While sorting, periodically mix to get liquid off sides of the tube. Keep sample cold whenever possible. 3. After the sort, using a pipette with sterile tip, measure the final volume and calculate the exact volume of sample sorted into the RLT. 4. Adjust the amount of RLT required, so that a ratio of 350 ml of RLT buffer to every 100 ml of sorted sample is maintained and then vortex. Samples may need to be transferred to larger RNase-free tubes if final volumes are high. 5. Add 250 ml of 100% ethanol for every 350 ml RLT buffer, and then mix samples. 6. If volumes from step 5 are high (e.g., >5 ml), then use of a vacuum manifold is suggested. Smaller volumes can be processed as individual 700 ml applications to the same RNeasy microcolumn and centrifuged at >10,000 × g for 15 s. 7. Complete protocol by referring to steps 9a–19 in Subheading “Direct RNA Extraction Using a Combined Trizol LS-RNeasy Method”.

Isolation of Total RNA from Transgenic Mouse Melanoma Subsets Using Fluorescence

3.4. Analyzing RNA from Sorted Melanoma Cell Fractions 3.4.1. Quantitation of RNA

3.4.2. Assessing RNA Quality

37

After extracting RNA from CD133+ and CD133– subsets using methods described above, the concentration of RNA was determined using a Nanodrop spectrophotometer and Qubit spectrofluorometer (15). Both methods are necessary because UV absorbance from the Nanodrop or other similar instruments alone cannot effectively discriminate some contaminants from true RNA; therefore, additional quantitation using a fluorescent RNA intercalation dye along with the Qubit spectrofluorometer is required (16). If residual Trizol carryover is present, an absorbance at 270 nm (Fig. 2) may be observed and interface with the absorbance value at 260 nm used for RNA and other nucleic acids resulting in erroneous quantification data. In cases where this carryover is problematic, further purification steps may be necessary. This may include an adjustment to the Trizol procedure to include an additional chloroform wash or a subsequent RNA cleanup step using a standard RNeasy MinElute column provided there is sufficient RNA available (see Note 19). RNA integrity was analyzed using the Agilent 2100 Bioanalyzer by loading 1 ml of sample RNA into the appropriate analysis cassette according to the manufacturer’s protocol. For low RNA recovery samples (e.g., <10 ng/ul), the low-range Agilent Picochip cassette was required along with a 1  ng/ml RNA calibration ladder and three known RNA standards at a predetermined

Fig. 2. Nanodrop ND-1000 trace of RNA recovered from FACS by a combined Trizol LS-RNeasy microcolumn strategy. Shown for illustrative purposes is the absorbance peak at 270 nm indicating the presence of residual Trizol contamination. This sample was further quantified using the Qubit spectrofluorometer, which uses the ribo-green nucleic acid dye to specifically intercalate RNA providing a more accurate RNA quantitation.

38

Tighe and Held

concentration of 1, 2, and 3 ng/ml to help empirically determine sample concentration (17). This method not only provides RNA integrity data but also provides quantitative data for samples that are below the effective range of the Nanodrop spectrophotometer. This was especially important for the rarer CD133+ cell subset, where less than 10,000 cells were recovered. Figure  3a shows high-quality RNA as indicated by the two sharp and well-defined ribosomal RNA peaks corresponding to the 18s and 28s subunits of total RNA recovered from the sorted CD133+ melanoma cells. Figure 3b reveals RNA degradation as indicated by the reduction in the 28s subunit and appearance of smaller RNA fragments represented as smaller peaks to the left of the 18s signal. Figure 3c shows a separate sort from mouse liver CD45+ cells with genomic DNA contamination illustrated by a “shoulder” between the 18s and 28s peaks. Figure 3d shows substantial removal of the genomic DNA after DNase I treatment. 3.4.3. Expected RNA Yields

We had previously determined RNA yield from various concentrations of cultured melanoma cells without FACS for later comparison of RNA yield after sorting. Presort RNA yields resulted in approximately 2–5 pg/cell. RNA extraction after FACS purification of CD133 melanoma subsets using the aforementioned protocols resulted in similar RNA yield, indicating comparable RNA

Fig. 3. RNA integrity profiles for FACS-purified cell samples using the Agilent 2100 Bioanalyzer. (a) Example of good quality RNA from purified mouse melanoma CD133+ cells, (b) partially degraded RNA, (c) RNA contaminated with gDNA before and (d) after DNase treatment.

Isolation of Total RNA from Transgenic Mouse Melanoma Subsets Using Fluorescence

39

isolation efficiency when sorting cells with FACS. For comparison, we also sorted CD45+CD1d+ and CD45+CD1d− mouse liver cells on the same BD FACSAria machine which resulted in similar dispensed sort volumes and RNA yields. Interestingly, a FACSCalibur mechanical sorter was used to sort human neuroblastoma cells and although dispensed sort volumes were far greater relative to the stream-in-air sorter, differences in RNA yield were unremarkable (see Table 1). While this data is not typically documented by most flow cytometry core facilities, its use for predictive sort volume recovery and quality control is vital and should be implemented as a routine performance indicator during each experiment.

4. Notes 1. During these procedures, it is assumed that all handling of tubes is performed using RNaseZap treatment for gloves and all surfaces, tubes, centrifuges, and pipettes. Flame sterilization should be adopted for any utensils where applicable. It is also recommended to perform all steps in a low-travel area and biosafety cabinet or PCR hood. At no point should RNaseZap be used in place of an RNase inhibitor. These types of surface decontamination solutions are composed of alkali hydroxides or chlorine derivatives such as sodium dichlorocyanurates and if exposed to samples will kill cells and degrade RNA; therefore, they are intended for sanitizing solid surfaces and utensils only. 2. Before sorting critical cells for RNA isolation, perform a FACS instrument QC test run using one test sample with a known viability greater than 80%. This sample is divided into two tubes: (1) a no sort control and (2) a postsort control. Both samples should be extracted for RNA simultaneously and analyzed for RNA integrity on the Agilent 2100 bioanalyzer. If both test samples result in degraded RNA (see Subheading 3.4.2), a review of handling and reagents is necessary. If RNA quality is good for the no sort control only, then RNase contamination of the flow cytometer is likely. Reagents and sheath fluids can be tested using such reagents as RNaseAlert (or equivalent) available from several vendors. 3. Fixation of cells using formalin and other aldehydes should be avoided because it causes nucleic acid cross-linking and contributes to RNA degradation. Although ethanol fixation does not negatively affect RNA, it does cause cell membrane permeability and possible mRNA leakage. For these reasons, it is preferable to use cell surface markers in order to avoid cell

Subheading “Direct RNA extraction using a combined Trizol LS-RNeasy method”

Subheading 3.3.1

Stream-in-air sorter

Mechanical sorter

Mouse liver

Human neuroblastoma

RNA isolation method

Subheading “Direct RNA extraction using RNeasy micro-column method”

Sorter type

Mouse Stream-in-air melanoma sorter

Cell type

Catcher tube

100

45 ml 550 ml 60 ml

CD133− CD133+ CD133− CD133+

1.0 × 105 7.5 × 103 4

4

4.5 ml

90 ml 90 ml 90 ml

CD45+ CD1d+

n/a n/a n/a n/a

5

1.4 × 104 4

2.0 × 104 5.5 × 104

1.2 × 10

5.7 × 10

90 ml

13.6 ml

CD45+ CD1d−

6

1.7 × 10

2.8 ml

CD45+ CD1d+

3.0 × 105

7.8 ml

CD45+ CD1d−

1.0 × 106

1.0 × 10

5.0 × 10

550 ml

30 ml

CD133+

5.0 × 10

550 ml

CD133−

3

100

1.0 × 105

Sorted volume

Nozzle size (mm) #Cells sorted Targeted subset

160

450

740

690

   7.8

   8

   9.1

   7.7

   6

  11

   6

   5.5

   6

   5.5

  44

  64

  19

  44

322

900

172

408

  40

145

  28

270

  16

250

nl/sorted cell RNA (ng)

0.44

3.2

1.6

3.3

0.56

0.53

0.57

0.41

4

2.9

3.7

2.7

3.2

2.5

pg RNA/cell

Table 1 Summary of FACS data for RNA isolated from mouse melanoma CD133+ and CD133– cell fractions sorted using a BD FACSAria compared to normal mouse liver CD45+CD1d– and CD45+CD1d+ cells and human neuroblastoma cells sorted using either a BD FACSAria or BD FACSCalibur, respectively

40 Tighe and Held

Isolation of Total RNA from Transgenic Mouse Melanoma Subsets Using Fluorescence

41

fixation–permeabilization before FACS for optimal RNA yield and integrity. 4. The use of RNase-free BSA (acetylated BSA) in place of serum in sample buffers is preferred as some serum sources may be inherently contaminated with RNases. 5. Samples with viability less then 80% viability are likely to result in partially degraded RNA and may benefit from the addition of 100 U of RNase inhibitor/ml of sample before performing cell sorts. 6. When sorting into precooled sterile RNase-free FACS tubes, it is important to consider adding an RNase inhibitor prior to initiating the sort. During the sort, the sample should be periodically mixed. 7. The physical stresses of FACS and centrifugation may lead to artifactual gene expression pattern changes of the sorted cells and it is recommended to work quickly when centrifuging cells after the sort and lysing cells with the RNA isolation buffer immediately following the centrifugation while keeping samples cold. Additionally, sort all samples in an experiment using the same reagents, cytometer settings, centrifugation forces, and temperatures throughout the procedure. 8. The total dispensed volume can be predicted beforehand by taking the target number of cells desired for sorting and multiplying by the approximate volume of each cell droplet. For example, if using an electrostatic sorter with a 100 mm nozzle, and it is wished to sort 105 cells, then the total dispensed liquid during the FACS run will be between 5  nl × 105 and 10 nl × 105, or 500 µl to 1 ml. However, empirical testing to determine the exact dispensed FACS volumes is recommended. 9. Alternative silica columns, such as RNeasy Mini, Invitrogen’s PureLink RNA Micro kit, and Ambion’s RNAqueous microcolumns, have similar silica technology and are amendable to the procedures outlined above with minor changes. Magnetic bead RNA isolation procedures (e.g., Dynabeads® from Dynal Corp. or MACS® from Miltenyi Biotec) have also been adapted for isolating RNA from FACS samples and recent data suggest improved RNA recoveries over those observed with standard silica-based columns (7). However, large volume extractions have yet to be investigated. 10. Unless small RNA species are required from the sorted cells, a silica column-based approach is advantageous due to its capacity for recovering small amounts of RNA from a limited number of cells. Recovery of small RNA species, such as miRNAs, is best accomplished using a standard Trizol–chloroform

42

Tighe and Held

precipitation procedure which involves centrifuging the sorted cells to a pellet and extracting using the manufacturer’s recommended protocol (12). However, when small numbers of cells are expected, the precipitation must be performed using a nucleic acid coprecipitate such as PelletPaint or GlycoBlue in an Axygen MaxyClear MCT175C centrifuge tube for maximum RNA pellet formation. 11. Before using an RNase inhibitor, it is necessary to review downstream applications of the RNA because some methods cannot tolerate the presence of an RNase inhibitor. Understandably, when sort volumes are high (>25 ml), it is not economical to maintain a final RNase inhibitor concentration at 20 U/ml. Regardless of the final concentration selection, it is most important to maintain consistency for samples belonging to the same experiment. 12. During a sort, it is not recommended to sort into RNAlater or other ammonium sulfate solutions as the resulting viscosity will be too high to centrifuge the cells properly and result in poor cellular recovery and compromised RNA quality. This is not unexpected as this reagent is designed for tissue preservation and not for purified cells from FACS (18). 13. A DNase I treatment will be required when downstream methods involving random hexamer priming such as in the case of exon microarrays, RT-qPCR, or equivalent. If no DNase I treatment is required, it should be omitted as results from our laboratory indicate that an expected loss of 30–40% of RNA may be observed when performing an on-column digestion (unpublished data). 14. Freezing directly sorted extracts in Trizol or RLT buffer often results in degraded RNA and is not recommended. However, we have observed that samples that are maintained at 4°C overnight in a dilute (~20%) Trizol LS solution followed by proper RNA extraction the next day have resulted in good quality RNA. Any storage method should be evaluated on each sample type prior to beginning an experiment because some cell types do not tolerate any lengthy Trizol or RLT exposure. 15. Although direct sorting of cells into RNA extraction buffer will negate a postsort cell purity check, this can still be performed by separately sorting a fraction of the cells into another tube containing PBS with 2% BSA, so that purity analysis can be performed after FACS is complete. 16. The example below indicates the amount of each reagent required to process a sample from the method outlined in Subheading 3.3.2. In this example, the cell lysis, nucleic acid

Isolation of Total RNA from Transgenic Mouse Melanoma Subsets Using Fluorescence

43

separation, and ethanol steps must be done in either a 15 or 50 ml RNase-free centrifuge tube and a vacuum manifold will be needed for processing the RNeasy microcolumn: Original Trizol LS in FACS Tube (presort)

  0.5 ml

Sorted volume (Total volume postsort minus 0.5 ml Trizol LS above)

  2.2 ml

Trizol LS needed to maintain 3:1 ratio (Trizol:sample ratio, 6.6:2.2)

  6.1 ml

Amount of chloroform needed (0.2 ml/0.75 ml Trizol LS)

  1.8 ml

Total volume for centrifugation

10.6 ml

Recovered aqueous phase (AQP)

  4.5 ml

100% ethanol needed (1.5 × AQP v/v)

  6.8 ml

Total volume to be applied to column

11.3 ml

17. The use of alternative organic phases in Trizol precipitations, such as 1-bromo-3-chloropropane (BCP) and 4-bromoanisole (4BA), has proven to be less desirable in our facility as their vapor pressures are low and do not benefit by evaporating from the final sample such as in the case of chloroform. 18. When using larger centrifuge tubes to processing larger volumes of the Trizol sample mix, it is not possible to centrifuge at 12,000 × g, and we have found that spinning as low as 1,000 × g results in good quality RNA. 19. Unfortunately, the Nanodrop and Qubit instruments cannot effectively discriminate RNA from DNA and other 260 nm absorbing contaminants. Any resulting DNA contamination must either be characterized or digested before proceeding to downstream reactions. Trizol carryover (absorbance at 270 nm) can sometimes be minimized by adding an additional chloroform cleanup step. This is done by combining the recovered aqueous phase with an equal volume of fresh chloroform at step 7 in Subheading “Direct RNA Extraction Using a Combined Trizol LS-RNeasy Method”. The tube is mixed and incubated at room temperature before centrifuging at full speed. The resulting aqueous phase is then processed exactly as the original aqueous phase at step 7 of Subheading “Direct RNA Extraction Using a Combined Trizol LS-RNeasy Method” by combining with a 1.5× volume of ethanol.

44

Tighe and Held

References 1. Gershon D (2005) DNA microarrays: more than gene expression. Nature 437:1195–1198 2. Rappa G, Fodstad O, Lorico A (2008) The stem cell-associated antigen CD133 (Prominin-1) is a molecular therapeutic target for metastatic melanoma. Stem Cells 26:3008–3017 3. Zabierowski SE, Herlyn M (2008) Melanoma stem cells: the dark seed of melanoma. J Clin Oncol 26:2890–2894 4. Klein WM, Wu BP, Zhao S, Wu H, KleinSzanto AJ, Tahan SR (2007) Increased expression of stem cell markers in malignant melanoma. Mod Pathol 20:102–107 5. Mizrak D, Brittan M, Alison MR (2008) CD133: molecule of the moment. J Pathol 214:3–9 6. Barrett MT, Glogovac J, Prevo LJ, Reid BJ, Porter P, Rabinovitch PS (2002) High-quality RNA and DNA from flow cytometrically sorted human epithelial cells and tissues. Biotechniques 32:888–896 7. Mack E, Neubauer A, Brendel C (2007) Comparison of RNA yield from small cell populations sorted by flow cytometry applying different isolation procedures. Cytometry A 71:404–409 8. D’Alessio G, Riordan JF (1997) Ribonucleases: structures and functions. Academic, San Diego, CA

9. Beintema JJ (1998) Introduction: the ribonuclease A superfamily. Cell Mol Life Sci 54:763–765 10. Bosenberg M, Muthusamy V, Curley DP, Wang Z, Hobbs C, Nelson B, Nogueira C, Horner JW, Depinho R, Chin L (2006) Characterization of melanocyte-inducible Cre recombinase transgenic mice. Genesis 44:262–267 11. Introduction to flow cytometry: a learning guide (2002) Becton, Dickinson and Company. 11-11032-03 rev. A 12. TRIzol reagent and TRIzol LS reagent technical note. Invitrogen Corp. Carlsbad, California 13. RNeasy® Micro Kit handbook (2007) QIAGEN sciences. Germantown, MD 14. QIAvac® 24 Plus handbook (2005) QIAGEN sciences. Germantown, MD 15. Qubit™ fluorometer instruction manual (2007) Invitrogen Corp. Carlsbad, California 16. Quant-iT™ RiboGreen RNA assay kit. Invitrogen Corp. Carlsbad, California 17. Kuschel M, Ausserer W (2000) Characterization of RNA quality using the Agilent 2100 Bioanalyzer. Agilent Technologies Application Notes 18. RNAlater® handbook (2006) QIAGEN sciences. Germantown, MD

Chapter 3 Microarray Analysis of Embryonic Stem Cells and Differentiated Embryoid Bodies Alexander C. Zambon and Christopher S. Barker Abstract By altering the cellular microenvironment and culture media composition, embryonic stem cells (ESCs) can be induced to differentiate in vitro into somatic cell types from the three primitive germ layers. ESC differentiation is regulated by an intricate series of signaling events that result in their transcriptional reprogramming, asymmetric cell division, and differentiation. Using various pharmacological agents and/or genetic manipulations, one can drive and enrich ESC differentiation to specific cell lineages. Identifying the transcriptional fingerprint during ESC differentiation could yield novel targets for genetic or pharmacological regulation to increase the yield of desirable cell types. We discuss here how to culture undifferentiated mouse ESCs (E14 line from 129/Ola) and generate embryoid bodies (EBs). We also discuss in detail how to prepare Affymetrix samples, how to hybridize and scan arrays, and how to quality control data and generate signal values and permutation based P-values. Key words: Embryonic stem cells, Stem cell differentiation, Embryoid bodies, Expression profiling

1. Introduction Culturing mouse ESCs (1) in vitro was a major scientific breakthrough that led to a series of significant biomedical research advances in transgenic (2) and knockout mouse models (3, 4) and provided valuable insight for the subsequent culture of human ESCs (5). Cultured ESCs have been used as a developmental model system to study gene and signaling networks that drive stem cells to differentiate into specific somatic cell types (e.g., cardiac myocytes (6)). Microarrays provide an opportunity to make unbiased genome-wide surveys to identify the transcriptional fingerprints of the gene networks that drive ESC differentiation (7) into somatic cell types. The use of genetically engineered

Sridar V. Chittur (ed.), Microarray Methods for Drug Discovery, Methods in Molecular Biology, vol. 632, DOI 10.1007/978-1-60761-663-4_3, © Humana Press, a part of Springer Science+Business Media, LLC 2010

45

46

Zambon and Barker

ESCs that express selectable transgenes (e.g., neomycin) (8) or fluorescent marker genes with tissue-specific promoters enables one to enrich and purify desired cell types. Such lines are well suited for genome-wide expression profiling. Several considerations should be made when choosing an ESC line, including the strain of mouse from which the line was derived from (considering that the ESC line will be used for genetargeted mutations and the generation of chimeric mice), whether or not the line requires coculture with mitotically inactive embryonic fibroblast feeder layers (i.e., culture of “feeder free” ESCs is less labor-intensive), and the potential for in vitro differentiation into desired cell types. This last consideration is supported by evidence of variability in the cardiogenic potential of various human (9) and mouse (8) ESC lines. It is important to note that variations in the culture conditions and genetic background of ESC lines can have a dramatic effect on gene expression signatures and should be taken into consideration when planning and interpreting expression profiles of ESCs and ESC-derived cells (10). A variety of microarray platforms and sample preparations have been described (for review (11)). The most commonly used array platforms available today are Affymetrix, Agilent, and Illumina microarrays. When selecting an array it is important to keep in mind that while different array suppliers may detect the same RNA transcript, the exact probe sequences used on each array can be quite different and located on different exons within a transcript. As a result, it can be problematic to directly compare data from similar samples that were run on different kinds of arrays. We recommend that array users use the same array across multiple data sets to facilitate future meta-analyses. While we discuss the classical sample preparation for Affymetrix microarrays by in vitro transcription, in many cases, it is not possible to obtain the amounts of RNA required for the described protocol. Numerous commercial sample preparation kits available also work quite well, including Affymetrix GeneChip One-Cycle Target Labeling kit (³1 µg of total RNA needed per sample), Applied Biosystems MessageAmp II – Biotin Enhanced kit (³100  ng of total RNA/sample), NuGEN Technologies WT-Ovation Pico kit (³500  pg of total RNA/sample), Molecular Devices Arcturus RiboAmp HS Plus kit (³100 pg total RNA/sample), and NuGEN Technologies WT-Ovation FFPE kit (³50 pg total RNA/sample). When selecting sample amplification methods, it is best to scale the reactions for the study to the sample with the least amount of RNA, and then pick the sample preparation kit that best meets those needs while ensuring that the sample preparation is compatible with the microarray platform used. In analyzing microarray results, a variety of strategies and techniques can be employed that are beyond the scope of this

Microarray Analysis of Embryonic Stem Cells and Differentiated Embryoid Bodies

47

chapter (for reference (12, 13)). In planning microarray studies, typically, the largest source of variation in a study is due to biological variation. As a result, we highly recommend designing studies with multiple (three or more) experimental, not technical, replicates to allow the use of statistical analysis to reduce experimental noise and allow the researcher to validate these changes with alternate methods (e.g., real-time PCR). We also recommend isolating RNA from extra experimental replicates (when possible) in case of RNA contamination or problems during microarray processing. The protocol below highlights one of these cases, in which there was unusable data generated by an array despite multiple quality control checks of the sample. Once expression signal values are generated, multiple testing procedures or other statistical tests can be conducted to define which transcripts show different expression. For a two sample comparison (i.e., ESCs compared to EBs), we generated permutation based unadjusted and Westfall and Young multiple-testing adjusted P-values (14) and employ a greater than twofold and P < 0.05 cutoff to call a gene differentially expressed. Most of the basic analysis can be done directly in Excel or another spreadsheet with basic search and filtering functions. To generate permutation-based P-values and to quality control array images, we use several R-based (http://www.r-project.org/) statistical packages that are freely available from the Bioconductor (15) website (http://www.bioconductor.org). R is a free software environment for statistical computing and graphics and in conjunction with packages available from Bioconductor can be used to both visualize and analyze a variety of genomic data sets (e.g. SAGE, SNP arrays). For Affymetrix arrays, we use the following R packages. For quality control of scanned microarray images, we use the affyQCReport. To generate log2 expression signal values, we use the gcRMA package. It has greater accuracy and precision than other available algorithms (16). To generate permutation-based unadjusted and Westfall and Young adjusted p-values, we use the multtest package.

2. Materials 2.1. Mouse E14 Feeder-Independent ESC Culture

1. Dulbecco’s phosphate-buffered saline (PBS) without calcium and magnesium. 2. 1000× beta-mercaptoethanol (b-ME) stock solution: Add 70 µl of b-ME to 20  ml of distilled, deionized water. Filter sterilize with a 22 mm Steriflip (Millipore, Billerica, MA), and store at 4°C for up to 2 weeks.

48

Zambon and Barker

3. ESC culture medium: Glasgow MEM/BHK21 medium (Sigma, St. Louis, MO) supplemented with 10% ES cell-characterized FBS (Hyclone, Logan, UT), 1× MEM nonessential amino acids (Invitrogen, Carlsbad, CA), 2 mM l-glutamine (Invitrogen), 1  mM Na-Pyruvate (Invitrogen), 1× b-ME, 1 × 106 units/L of ESGRO (Millipore). Culture medium can be stored at 4°C for up to 4 weeks; after that, resupplement the remaining medium with l-glutamine. 4. ESC trypsin solution: Add 100 mg of EDTA tetrasodium salt to 500 ml of PBS. Filter-sterilize and add 10 ml of 2.5% porcine trypsin solution (Invitrogen) and 5 ml of chicken serum (Invitrogen). Store as 20 ml aliquots at −20°C (avoid multiple freeze-thawing cycles). 5. 0.1% gelatin solution: Add 25 ml of a 2% bovine gelatin solution (Sigma) to 500 ml of PBS. Store at 4°C. 2.2. Embryoid Body (EB) Formation by Hanging Drops

1. EB differentiation medium (EBDM): Glasgow MEM/ BHK21 medium supplemented with 20% ES cell characterized FBS, 1× MEM nonessential amino acids, 2  mM l-glutamine, 1 mM Na-Pyruvate, and 1× b-ME. 2. 96-well sterile conical bottom polypropylene plates (E&K Scientific). 3. Sterile 96-well plate lids (E&K Scientific). 4. Wide orifice tips (Rainin RT-L250WS).

2.3. Total RNA Extraction

1. TRIzol Reagent (Invitrogen). 2. Phase Lock Heavy Gel Tubes (2 ml) (Eppendorf). 3. RNAeasy Mini Kit (Qiagen).

2.4. cDNA Synthesis

1. T7-(dT)24 Primer, HPLC Purified (Operon Technologies). 2. Superscript double-stranded cDNA Synthesis kit (Invitrogen).

2.5. cRNA Synthesis and Labeling

1. BioArray High Yield DNA Transcript kit (Affymetrix). 2. RNAeasy Mini Kit (Qiagen). 3. 5× fragmentation buffer: 200  mM Tris–acetate, pH 8.2, 500 mM potassium acetate, 150 mM magnesium acetate.

2.6. Genechip Hybridization

1. U430 2.0 GeneChip (Affymetrix). 2. GeneChip Eukaryotic Hybridization Control Kit including 20× hybridization controls and control oligonucleotide B2 (Affymetrix). 3. 12× MES stock: Resuspend 70.4 g MES hydrate (Sigma) plus 193.3 g MES sodium (Sigma) in 800 ml of molecular biology grade H2O (Gibco), mix and adjust volume to 1 l. The pH

Microarray Analysis of Embryonic Stem Cells and Differentiated Embryoid Bodies

49

should be 6.5–6.7. Filter through a 0.2 µM filter. Store at 4°C in a foil-covered bottle (protect from light). 4. 2× hybridization buffer: 19.9  ml molecular biology grade H2O + 8.3  ml of 12× MES stock + 17.7  ml of 5 M NaCl (Ambion) + 4.0 ml of 0.5 M EDTA (Sigma) + 0.1 ml of 10% Tween 20 (Pierce). Store at 4°C in a foil-covered bottle (protect from light). 5. Herring sperm DNA 10 mg/ml (Promega). 6. DMSO (Sigma). 2.7. GeneChip Wash, Stain and Scan

1. Prepare antibody and stain solutions immediately before use. 2. Wash buffer A (nonstringent): 300 ml  of  20× SSPE (Fisher) + 1 ml of 10% Tween 20 + 699 ml of molecular biology grade H2O to final volume 1 l and filter through a 0.2-µm filter. 3. Wash buffer B (stringent): 83.3 ml 12× MES Buffer + 5.2 ml 5 M NaCl + 1 ml 10% Tween 20 + 910.5 ml Molecular Biology Grade H2O to final volume 1 l and filter through 0.2-µm filter and store at 4°C, protected from light. 4. 2× stain buffer: 41.7  ml 12× MES Buffer + 92.5  ml 5  M NaCl + 2.5  ml 10% Tween 20 + 113.3  ml molecular biology grade H2O to a final volume of 250 ml and store 4°C protected from light. 5. Goat IgG stock 10 mg/ml (Sigma): resuspend 50 mg in 5 ml of 150 mM NaCl and store at 4°C. 6. SAPE stain solution: 600 µl of 2× stain buffer + 48 µl of 50  mg/ml BSA (Invitrogen) + 12 µl of 1  mg/ml streptavidin/phycoerythrin (SAPE) (Invitrogen) + 540 µl of DI H2O. Mix well and divide into two aliquots of 600 µl each. 7. Antibody solution: 300 µl of 2× stain buffer + 24 µl of 50 mg/ml BSA + 6 µl of 10 mg/ml goat IgG stock + 3.6 µl of 0.5 mg/ml biotinylated antibody (Vector Laboratories) + 266.4 µl of DI H2O.

3. Methods We use a feeder-independent ESC line derived from the 129/ Ola strain of mice as shown in Fig. 1 (17). These cells are easy to maintain and significantly reduce the amount of tissue culture required. The parental cell line E14Tg2A (denoted as E14 herein) was established from delayed blastocysts on gelatinized tissue culture dishes in ES cell medium containing leukocyte inhibitory factor (LIF) (17). Sublines were isolated by plating cells at a single-cell density, picking and expanding single colonies,

50

Zambon and Barker

Fig. 1. Phase images of (a) mouse E14 ESCs and (b) 5-day-old embryoid bodies. Images captured at 20×

and testing several clones for germline competence. To differentiate ESCs into embryoid bodies that contain spontaneously beating cardiac myocytes, we have made slight modifications to the method of Boheler et al. (18). We isolated RNA from seven T25-cm flasks of undifferentiated E14 cells and seven 10-cm dishes each containing EBs recovered from one 96-well plate of EBs. Eight days after the initiation of hanging drops, beating cardiomyocytes could be visualized with a microscope. We then prepared five samples of either ESC or EB total RNA for quality control and array hybridization and analyzed the resultant data using the protocols described. Yields at various steps in the protocol are reported in Table 1 for reference. 3.1. Mouse E14 Feeder-Independent ESC Culture

1. Coat a 25-cm2 tissue-culture flask with 0.1% gelatin and aspirate off the excess immediately before use. 2. Thaw ES cells (approximately 2.5 × 106 cells, equivalent to ½ of a confluent 6-well or 1/4 of a confluent 25-cm2 flask) in a 37°C water bath and dilute into 10 ml of prewarmed ES cell medium.

Microarray Analysis of Embryonic Stem Cells and Differentiated Embryoid Bodies

51

3. Collect the cells by spinning for 3 min at 1,100 rpm (130 g) in a bench-top clinical centrifuge. 4. Aspirate off medium and gently resuspend cells in 10 ml of prewarmed medium. 5. Transfer cell suspension to a 25-cm2 flask and grow at 37°C in a humidified 7% CO2 incubator. 6. Change medium the following day to remove dead cells and residual DMSO (see Note 1). 7. ES cells are routinely passaged every 2 days, and the medium is changed on alternate days. Thus, ES cells require daily attention (see Fig. 1a for an example of subconfluent E14 ES cells) (see Note 2). To passage every other day, aspirate the culture media, rinse with 5  ml PBS, aspirate and add 1  ml of ESC trypsin and incubate in the tissue culture incubator for ~3 min. Neutralize with 9 ml of ESC media and passage at 1/10 split. 3.2. Embryoid Body Formation by Hanging Drops

1. Day 0: From a confluent 25-cm2 flask of cells, aspirate off the medium and wash with 5  ml of room temperature PBS, pipetting it away from the cells. Rock flask gently and aspirate medium. 2. Cover cells with 1 ml of 1× trypsin solution, and return to 37°C incubator for 2  min or until cells are uniformly dispersed into small clumps. 3. Add 9 ml of EBDM to inactivate the trypsin and pipette up and down gently to create a single cell suspension. 4. Count cells and dilute to 25,000  cells/ml (approximately 40-fold) in EBDM. Using a multichannel pipet, transfer 20 ml (500 cells) to the center of each well of a 96-well polypropylene plate with conical bottoms. Each plate will require 2 ml of cell suspension. Invert plates gently and incubate at 37°C for 2 days. 5. Day 2: Invert plate right side up and use multichannel pipet to add 200 µl/well fresh 20% EBDM. 6. Day 5: Use multichannel pipet to remove 100 µl of medium, being careful not to disturb developing EB at bottom of well, and replace with 100 µl of fresh EBDM. 7. Day 7: EBs are collected from the 96-well plate by rinsing/ scraping the V-bottom wells with a multichannel pipet set at 150 µl and wide orifice tips. Transfer to a sterile reservoir with ~5 µl of EBDM (see Fig. 1b for an example of a Day 7 EB). After all the EBs have been collected, transfer to a 50 ml conical tube, and allow the EBs to sediment by gravity flow (~5 min). After the EBs have settled to the bottom of the conical tube, aspirate all but ~5  ml of medium, where the settled EBs resided and transfer to a 10-cm tissue culture dish with 10 ml of fresh EBDM (1 dish/96 well plate = sample replicates).

52

Zambon and Barker

Change the medium every 2–3 days and look for beating foci with a microscope (you should see beating areas). 3.3. Total RNA Extraction

1. Add 1 ml of TRIzol Reagent per 10-cm dish of EBs or ES cells. 2. Scrape cells with a cell scraper and pipet up and down several times with a 1 ml pipet. 3. Incubate the homogenized sample for 5 min at R/T. 4. Add 0.2  ml of chloroform per ml of TRIzol reagent. Cap tubes securely. 5. Handshake for 15 s and incubate for 2–3 min at R/T. 6. Transfer aliquots of 500 µl (up to 750 µl) homogenates to prespin (12,000 rpm (15,300 g) for 30 s) heavy phase lock tubes. 7. Centrifuge for 10 min at 12,000 rpm (15,300 g) at 4°C. 8. Remove upper colorless aqueous phase remaining the RNA to a fresh tube. 9. Precipitate RNA with 0.5 ml of isopropanol per ml of TRIzol reagent. 10. Incubate for 10 min at R/T. 11. Centrifuge at 12,000 rpm (15,300 g) for 10 min at 4°C. 12. Remove the supernate carefully. 13. Wash pellet with 1 ml of 75% ethanol per ml TRIzol Reagent. 14. Vortex and centrifuge at 9,000 rpm (8,600 g) for 5 min at 4°C. 15. Remove the supernate and briefly dry the RNA pellet by air-dry or vacuum-dry for 5–10 min. Do not dry RNA by centrifugation under vacuum. It is very important not to let the RNA pellet dry completely. 16. Dissolve RNA in RNase-free water by passing the solution a few times through a pipet tip and incubating for 10 min at 55–60°C. 17. Take 1 µl or an aliquot for quality and quantity measure­ ments. 18. Store sample at −80°C, if necessary, or go on to next step. 19. Adjust the volume of the total RNA sample to 100 µl with RNase-free water. 20. Add 350 µl buffer RLT (with b-ME) (RNAeasy Mini Kit) to the sample and mix thoroughly (see Note 3 and Table 1 for yield). 21. Add 250 µl of 100% ethanol to the lysate and mix by pipetting. 22. Apply sample (700 µl) to an RNAeasy column sitting in a 2-ml collection tube. Spin for 15 s at max speed. Discard the flow-through.

Microarray Analysis of Embryonic Stem Cells and Differentiated Embryoid Bodies

53

Table 1 Yields for microarray sample preps Total RNA Sample # Sample yield (mg)

Post cleanup Adjusted A260/A280 [RNA] mg/ml A260/A280 IVT yield [IVT]

1

EB

137.2

1.47

0.86

1.90

47.90

43.75

2

ES

89.4

1.52

2.21

1.70

45.98

41.83

3

EB

142.4

1.38

0.79

1.90

53.09

48.94

4

ES

103.8

1.53

1.27

1.90

58.50

54.35

5

EB

143.7

1.30

0.76

1.90

51.55

47.40

6

ES

96.9

1.55

1.92

1.90

59.74

55.59

7

EB

52.8

1.83

0.84

1.90

39.49

35.34

8

ES

75.1

1.79

1.37

1.90

47.62

43.47

9

EB

53.5

1.83

0.97

1.90

47.01

42.86

10

ES

57.8

1.80

1.42

1.90

59.49

55.34

23. Add 500 µl RPE buffer (ethanol added) onto the column and spin for 15 s at max speed to wash. Discard the flow-through. 24. Add additional 500 µl RPE buffer and spin at max speed for 2 min to dry RNAeasy membrane. 25. Carefully transfer column to a new 1.5-ml tube and pipet 30–50 µl of RNase-free water directly onto the membrane. Wait for 3–4 min. Spin at max speed for 1 min to elute. 26. Repeat step 7 if more than 30 µg RNA yield is expected and elute into the same collection tube. 27. Take a 1-µl aliquot for quality assessment. We recommend using the Agilent 2100 BioAnalyzer (see Fig. 2a for a representative and acceptable tracing of total RNA). The output file will also generate an RNA integrity number (RIN). The RIN is generated on a scale of 1–10 (poor to excellent quality). We recommend using samples with RIN ³ 7. An approximation of the amount of RNA can also be derived from this tracing, but this measurement is frequently imprecise. 28. If there is no access to an Agilent 2100 BioAnalyzer, check the total RNA quality on 1% agarose (RNase-free) gel by loading 1 µl of the total RNA sample. Treat gel equipment with 3% peroxidase before use. Use RNase-free water when making TAE buffer needed for the agarose gel preparation and electrophoresis buffer. Run at 60 V for 30 min or until RNA bands are well separated. Look for 2 kbp (28S), 0.9 kbp (18S), and 200 bp (5S) ribosomal RNA bands.

Zambon and Barker

a 6

total RNA

Fluorescence

5 4 3 2

19

b

24

29

34

3

1

0

2

1

39 44 49 Time (seconds)

54

59

64

69

2.5

cRNA

Fluorescence

2.0 1.5 1.0 0.5 0.0 19

24

29

34

39

44

49

54

59

64

69

Time (seconds)

c 2.5

Fragmented cRNA

2.0 Fluorescence

54

1.5 1.0 0.5 0.0 19

24

29

34

39 44 49 Time (seconds)

54

59

64

69

Fig. 2. Agilent BioAnalyzer tracings of (a) total RNA (b) cRNA (c) fragmented cRNA

29. The amount of RNA should be determined spectroscopically by measuring the A260 value by standard methods. The A260/A280 OD ratio should be 1.8–2.0 for pure RNA when RNA sample diluted in 10 mM Tris–HCl, pH 7.5.

Microarray Analysis of Embryonic Stem Cells and Differentiated Embryoid Bodies

3.4. cDNA Synthesis

55

1. Add DEPC treated H2O to 16–24 µg of total RNA (no more than 24 µg) to achieve the final volume of 9 µl. 2. Add 1 µl of the T7-(dT)24 primer (100 pmol/µl)/sample. 3. Incubate for 10  min at 70°C in PCR machine with heated cover. 4. Quick spin and then put the samples on ice. 5. Add 4 µl of 5× first-strand cDNA buffer, 2 µl 0.1 M DTT, 1 µl of 10 mM dNTP mix. 6. Mix and incubate at 42°C for 2 min. 7. Add the 3 µl of the SSII RT (final volume = 20 µl). 8. Mix well and incubate at 42°C for 1 h. 9. Set the water bath to 16°C. Spin samples briefly to bring down condensation on side of tube. 10. On ice, add the following reagents, in the order shown, to the first strand reaction tube: 91 µl of DEPC-treated water, 30 µl of 5× second strand buffer, 3 µl of 10 mM dNTP mix, 1 µl of 10  U/µl DNA ligase, 4 µl of 10  U/µl DNA polymerase I, E. coli, 1 µl 2 U/µl RNase H, E. coli (final volume, 150 µl). 11. Tap the tube and mix. Spin briefly and incubate for 2 h at 16°C. 12. Add 2 µl (10 U) of T4 DNA polymerase and continue incubating for 5 min at 16°C. 13. Place the reaction on ice and add 10 µl of 0.5 M EDTA. 14. Store at −20°C or proceed with cleanup steps. 15. Pellet the material in a 1.5-ml green phase lock light tube (PLG) at max speed for 30 s. 16. Add 162 µl phenol:chloroform:isoamyl (25:24:1) alcohol to the final volume (162 µl) of the cDNA reaction (total volume 324 µl). Vortex thoroughly. 17. Transfer the entire volume to the PLG tubes. Do not vortex. 18. Spin at maximum speed for 2 min. 19. Transfer the aqueous upper phase to a new tube. 20. Add 0.5 volume of 7.5 M NH4Ac and 2.5 volume of 100% ethanol (stored at −20°C). 21. Vortex and spin at maximum speed for 20 min at R/T. 22. Remove the supernatant carefully. Wash with 1  ml of 80% ethanol (stored at −20°C). Spin for 5 min. Discard supernatant. Repeat once. 23. Air-dry pellet. Resuspend the pellet in 12 µl of DEPC-treated H2O. 24. Remove resuspended sample and place in fresh 200-µl PCR tube (proceed or store at −20°C).

56

Zambon and Barker

3.5. cRNA Synthesis and Labeling

1. Add the reagents in the following order to new RNase-free tubes (final volume: 40 µl) at R/T: 4 µl of cDNA, 18 µl of DEPC-treated H2O, 4 µl of 10× reaction buffer (BioArray High-Yield DNA Transcript kit), 4 µl of 10× biotin-labeled ribonucleotides, 4 µl of 10× DTT, 4 µl of RNase inhibitor mix, and 2 µl of T7 RNA polymerase enzyme. 2. Carefully mix and then spin briefly. 3. Immediately place the tube at 37°C. Incubate for 5  h (the longer, the higher yield). Gently mix the tube every 45 min during the incubation. 4. Store at −20°C if not purifying cRNA immediately. 5. Bring the volume of the IVT reaction to 100 µl with 60 µl of RNase-free water, then add 350 µl Buffer RLT (with b-ME) to the sample and mix thoroughly. 6. Add 250 µl of 100% ethanol to the lysate and mix well by pipetting. 7. Apply sample (700 µl) to an RNAeasy column with a new 2-ml collection tube (supplied in the kit). Spin for 15 s at max speed. Discard the flow-through. 8. Add 500 µl diluted RPE buffer and centrifuge for 15  s at maximum speed to wash. Discard the follow-through. 9. Add additional 500 µl diluted RPE buffer onto the RNAeasy column, and centrifuge for 2 min at maximum speed to dry RNAeasy membrane. 10. Carefully (without touching ethanol), transfer RNAeasy column into a new 1.5 ml collection tube (supplied) and pipet 30 µl of RNase-free water directly onto the RNeasy membrane. Wait for 3–4 min. Spin for 1 min at maximum speed to elute. 11. Usually repeat step 10 if more than 30 µg of cRNA yield is expected (50–100 µg expected). 12. Lightly vortex tubes before quantification. Save a 1 µl aliquot for quality and quantity measurement (see Note 4). 13. Use the same procedure as described for RNA above. Check for concentration (1  OD at 260  nm equals 40 µg/ml). A260/A280 ratio of 1.8–2.0 is acceptable purity. 14. The cRNA must be at a minimum concentration of 0.6 µg/µl. If it is not, it can be concentrated with ethanol precipitation or SpeedVac Concentrator. 15. If ethanol precipitation is required, add 0.5 volumes of 7.5 M NH4Ac and 2.5 volumes of 100% (−20°C) ethanol. Vortex. 16. Precipitate at −20°C for 1 h to overnight. 17. Spin for 30 min at maximum speed in a microfuge at 4°C.

Microarray Analysis of Embryonic Stem Cells and Differentiated Embryoid Bodies

57

18. Remove the supernatant carefully. Wash with 0.5 ml of 80% ethanol (−20°C). Spin for 5 min and then discard the supernatant. Repeat once. 19. Air-dry pellet. Resuspend the pellet in 10–20 µl DEPC-treated H2O. 20. Save 1 µl for quality control analysis with the Agilent 2100 BioAnalyzer (see Fig. 2b for a representative and acceptable tracing of cRNA). 21. Use 40 µg of adjusted cRNA for fragmentation: Adjusted cRNA yield = RNAm − (total RNAi)*(Y ); RNAm = amount of cRNA measured after IVT (µg), Total RNAi = starting amount of total RNA (µg), Y = fraction of cDNA reaction used in IVT. 22. Add 2 µl of 5× fragmentation buffer for every 8 µl of RNA plus H2O. The final concentration of RNA in the fragmentation mix can range from 0.5 µg/µl to 2 µg/µl. 23. Bring the volume of 40 µg cRNA to 64 µl with RNase-free H2O. 24. Add 16 µl 5× fragmentation buffer, final concentration is 0.5 µg/µl. 25. Incubate for 35  min at 94°C. Putting on ice after the incubation. 26. Save a 1 µl aliquot for analysis on the Agilent BioAnalyzer (see Fig. 2c for a representative and acceptable tracing of fragmented cRNA). Store at −20°C or at −80°C until ready to perform the hybridization. Fragmented cRNA is very stable at −80°C. 3.6. GeneChip Hybridization

1. Heat 20× Eukaryotic Hybridization Controls to 65°C for 5 min to completely resuspend before aliquoting. 2. Mix hybridization cocktail components at room temperature: 15  mg of fragmented cRNA + 5 µl control oligonucleotide B2 + 15 µl of 20× Eukaryotic Hybridization Controls + 3 µl herring sperm DNA + 3 µl BSA + 150 µl 2× hybridization buffer +  30 µl DMSO + molecular biology grade H2O to final volume 300 µl. 3. Warm GeneChip to room temperature immediately before use. Fill GeneChip with 1× hybridization buffer and incubate at 45°C for 10 min while mixing in hybridization oven. 4. Heat hybridization cocktail to 99°C for 5  min and cool to 45°C in heating block for 5 min. 5. Microfuge hybridization cocktail for 5 min at room temperature to remove any insoluble material. 6. Remove hybridization buffer from GeneChip and fill with hybri­ dization cocktail. Cover GeneChip septa with ToughTag spots.

58

Zambon and Barker

7. Place GeneChips in hybridization oven at 45°C and incubate rotating at 60 rpm for 16 h. 3.7. GeneChip Wash, Stain, and Scan

1. Load two 600 µl tubes of SAPE solution and one 600 µl tube of antibody solution into Affymetrix fluidics station. 2. Load GeneChip into fluidics station. 3. Select Fluidics Script EukGE-WS2v4_450 for Affymetrix Fluidics Model 450 Stations in Affymetrix Command Console software. 4. GeneChip wash and stain should take about 2 h. 5. Remove GeneChip from fluidics station and place in Affymetrix Scanner. 6. Start array scan using Command Console Software.

3.8. Quality Control of Hybridized Microarrays

The .cel files for each microarray were quality tested with the Bioconductor packages Affy and affyPLM. Please see the package help file for specifics regarding commands for running the analysis. Several plots in the reports of these analysis indicated that the data for sample 4 did not fit the RMA model well despite acceptable A260/280 ratios (Table 1) and Agilent 2100 BioAnalyzer tracings. RMA (19) is an algorithm that generates signal values from the .cel files from the scanned chip image. The chip pseudoimage (shown here in Fig. 3 in gray scale) function in the affyPLM package, which plots the weights and residuals from RMA signal

01.EB.Cel

02.ES.Cel

03.EB.Cel

04.ES.Cel

05.EB.Cel

06.ES.Cel

07.EB.Cel

08.ES.Cel

09.EB.Cel

10.ES.Cel

Fig. 3. ES and EB microarray data was quality controlled with the R program with Bioconductor plots

Microarray Analysis of Embryonic Stem Cells and Differentiated Embryoid Bodies

59

model fitting procedures. A white area indicates the probe fits the model well, and a dark area indicates the probe does not. Thus, sample 4 contained an even distribution of probes which did not fit the model well indicating a problem with either the microarray or the sample prep/hybridization. We did not detect any abnormalities of the sample at the RNA, IVT or fragmented IVT as indicated by quality control checks using the Agilent BioAnalyzer (see Fig. 2 for representative data). One can note the markings on samples 1 and 3 which are typical. These marks represent local chip or hybridization anomalies. Since the probes for a gene are distributed randomly over the chip, and since these probes are downweighted, local problems do not affect the final model and expression estimates in a significant manner, especially when there are lots of chips used in the model (10 is fairly large). We excluded sample four and proceeded to generate signal values with the gcRMA package from Bioconductor in R. Please see the package help file for specifics regarding commands for running the analysis. The resultant gcRMA signal values are log2 expression values and are converted to geometric folds with standard spreadsheet calculations. Permutation adjusted P-values were then generated with the multtest package in R. Figure 4 indicates the number of probe sets that are either up or down regulated with a permutation based P-value > 0.05 and an absolute fold change greater than 2. We would focus on these genes for downstream pathway analysis or validation and have shown previously a high degree of correlation in gcRMA generated fold changes and real-time PCR validated fold changes (20). 2000 1500

Number of probesets changed (Fold >2 and P<0.05)

3.9. Generation of Signal Values and Multiple Testing Procedures

1000 500 0

Downregulated Upregulated

500 1000 1500

Fig. 4. Number of probe sets differentially up or down regulated (|Fold| > 2, P < 0.05)

60

Zambon and Barker

4. Notes 1. ES cells are frozen in medium containing 10% DMSO. Since DMSO may induce the differentiation of ES cells, we advise thawing the cells late in the day and changing the medium the following morning to minimize the effects of residual DMSO. 2. In our experience, feeder-independent ES cells grow rapidly and quickly acidify the medium, turning it yellow. Allowing the cells to acidify the medium (by not changing the medium every day or by passaging the cells at too low a dilution) will cause the cells to undergo crisis, triggering excess differentiation and cell death, after which their pluripotency cannot be guaranteed. Plating cells at too low a density, insufficient dispersion of cells during passage, or uneven plating can cause similar problems, as the cells will form large clumps before reaching confluence, and the cells within these clumps will differentiate or die. Germline transmission is significantly reduced in cells that have been mistreated, even when they appear healthy at the time of injection. 3. Do not exceed 100 µg RNA/spin column. Add 10 µl b-ME per ml of RLT buffer. Make sure four volumes of 100% ethanol were added to the RPE buffer. All centrifugation steps should be performed at 20–25°C. We see a significant increase in RNA purity as noted by the improved A260/A280 ratios (Table 1) after RNA cleanup with RNeasy cleanup kit. 4. It is suggested to purify one half of the IVT product and check yields before purifying the second half.

Acknowledgments The authors would like to thank Drs. Whitmore Tingley and Roland Russnak for their contributions to the protocols for E14 ESC culture and embryoid body formations and Bruce Conklin and the late Karen Vranizan for their contributions to the design and interpretation of the data presented in this manuscript. We would also like to thank the Gladstone editorial department Gary Howard and Stephen Ordway for their contributions. Dedicated to the memory of Karen Vranizan. References 1. Evans MJ, Kaufman MH (1981) Establishment in culture of pluripotential cells from mouse embryos. Nature 292:154–156 2. Robertson E, Bradley A, Kuehn M, Evans M (1986) Germ-line transmission of genes

introduced into cultured pluripotential cells by retroviral vector. Nature 323:445–448 3. Capecchi MR (1989) Altering the genome by homologous recombination. Science 244: 1288–1292

Microarray Analysis of Embryonic Stem Cells and Differentiated Embryoid Bodies 4. Koller BH, Hagemann LJ, Doetschman T, Hagaman JR, Huang S, Williams PJ et  al (1989) Germ-line transmission of a planned alteration made in a hypoxanthine phosphoribosyltransferase gene by homologous recombination in embryonic stem cells. Proc Natl Acad Sci USA 86:8927–8931 5. Thomson JA, Itskovitz-Eldor J, Shapiro SS, Waknitz MA, Swiergiel JJ, Marshall VS et  al (1998) Embryonic stem cell lines derived from human blastocysts. Science 282: 1145–1147 6. Beqqali A, Kloots J, Ward-van OD, Mummery C, Passier R (2006) Genome-wide transcriptional profiling of human embryonic stem cells differentiating to cardiomyocytes. Stem Cells 24:1956–1967 7. Chang HY, Thomson JA, Chen X (2006) Microarray analysis of stem cells and differentiation. Methods Enzymol 420:225–254 8. Zandstra PW, Bauwens C, Yin T, Liu Q, Schiller H, Zweigerdt R et al (2003) Scalable production of embryonic stem cell-derived cardiomyocytes. Tissue Eng 9:767–778 9. Moore JC, Fu J, Chan YC, Lin D, Tran H, Tse HF et al (2008) Distinct cardiogenic preferences of two human embryonic stem cell (hESC) lines are imprinted in their proteomes in the pluripotent state. Biochem Biophys Res Commun 372:553–558 10. Allegrucci C, Young LE (2007) Differences between human embryonic stem cell lines. Hum Reprod Update 13:103–120 11. Hardiman G (2004) Microarray platforms – comparisons and contrasts. Pharmacogenomics 5:487–502

61

12. Durinck S (2008) Pre-processing of microarray data and analysis of differential expression. Methods Mol Biol 452:89–110 13. Page GP, Zakharkin SO, Kim K, Mehta T, Chen L, Zhang K (2007) Microarray analysis. Methods Mol Biol 404:409–430 14. Westfall PH, Young SS (1993) Resamplingbased multiple testing: Examples and methods for p-value adjustment. Wiley, NY 15. Gentleman RC, Carey VJ, Bates DM, Bolstad B, Dettling M, Dudoit S et  al (2004) Bioconductor: open software development for computational biology and bioinformatics. Genome Biol 5:R80 16. Wu Z, Irizarry RA (2004) Preprocessing of oligonucleotide array data. Nat Biotechnol 22:656–658, author reply 8 17. Nichols J, Evans EP, Smith AG (1990) Establishment of germ-line-competent embryonic stem (ES) cells using differentiation inhibiting activity. Development 110:1341–1348 18. Boheler KR, Czyz J, Tweedie D, Yang HT, Anisimov SV, Wobus AM (2002) Differentiation of pluripotent embryonic stem cells into cardiomyocytes. Circ Res 91:189–201 19. Irizarry RA, Hobbs B, Collin F, BeazerBarclay YD, Antonellis KJ, Scherf U et  al (2003) Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics 4:249–264 20. Zambon AC, Zhang L, Minovitsky S, Kanter JR, Prabhakar S, Salomonis N et  al (2005) Gene expression patterns define key transcriptional events in cell-cycle regulation by cAMP and protein kinase A. Proc Natl Acad Sci USA 102:8561–8566

Chapter 4 Determination of Alternate Splicing Events Using the Affymetrix Exon 1.0 ST Arrays Sita Subbaram, Marcy Kuentzel, David Frank, C. Michael DiPersio, and Sridar V. Chittur Abstract Alternative splicing plays an important role in regulation of normal cellular function. Alternative splicing of pre-mRNA leads to the diversity of downstream protein products in the cell. The Affymetrix Exon arrays allow for a high throughput evaluation of the differences in spliced mRNA expressed in a biological system. In this study, we describe a method using this technology to study the generation of alternative mRNA transcripts in breast cancer cells that differ in the levels of a particular integrin, a3b1. Key words: Alternative splicing, Gene regulation, Expression profiling, Microarray, Exon splicing, Integrins

1. Introduction a3b1 integrin belongs to a family of heterodimeric cell surface receptors that mediate cell adhesion to the extracellular matrix. Integrins can mediate both inside-out and outside-in signal transduction, and they have been demonstrated to be involved in many aspects of cellular biology such as adhesion, migration, and survival. Laminin-332 is the primary ligand for a3b1 that is expressed in a variety of epithelial cell types. a3b1 is overexpressed in a variety of human cancers and experiments conducted in breast cancer cells have indicated an important role for this integrin in invasion (1, 2). In addition, we have shown that a3b1 in epithelial cells can induce the expression of EMT and angiogenesis promoting genes such as MMP9 and Mrp3 (3–5). a3b1-dependent induction of MMP9 gene expression was established to occur via enhanced stability of the MMP9 mRNA transcript in mouse keratinocytes, Sridar V. Chittur (ed.), Microarray Methods for Drug Discovery, Methods in Molecular Biology, vol. 632, DOI 10.1007/978-1-60761-663-4_4, © Humana Press, a part of Springer Science+Business Media, LLC 2010

63

64

Subbaram et al.

resulting in increased protein expression (3). To investigate the role of a3b1 in regulating differential gene expression as well as gene splicing events in the breast cancer cell line MDA-MB-231, we performed microarray analysis using the Affymetrix Human Exon 1.0 ST array platform (6). We have identified various candidate genes that are differentially spliced in cells that stably express an shRNA that targets the a3 integrin subunit (a3-knockdown cells), compared to cells that express a control shRNA. One of these genes was identified as POLR2I, which encodes a subunit of RNA Polymerase II. POLR2I mRNA was found to be differentially spliced at the 3¢-end, where part of Exon 6 was excluded from mRNA isolated from control breast cancer cells, but was included in mRNA from the a3-knockdown cells. This difference in Exon 6 processing could be attributed to differential usage of the 3¢-untranslated region of the gene or variations in polyadenylation.

2. Materials 2.1. Equipment

1. Agilent Bioanalyzer 2100 system. 2. Nanodrop ND-1000 UV–Vis spectrophotometer. 3. Affymetrix Genechip® System.

2.2. Materials for Cell Culture

1. MDA-MB-231 breast cancer cell lines were stably infected with lentivirus expressing a control shRNA (control cells; MISSIONTM shRNA, Sigma). 2. MDA-MB-231 breast cancer cell lines stably infected with lentivirus expressing shRNA that targets the human a3 mRNA (a3-knockdown cells; MISSIONTM shRNA, Sigma). 3. Phosphate buffered saline (PBS) 137  mM NaCl, 2.7  mM KCl, 4.3 mM Na2HPO4 7H2O, 1.4 mM KH2PO4, pH 7.4.

2.3. Materials for RNA Isolation

1. All tips, tubes, and reagent bottles must be DNase and RNase free (see Note 1). 2. Tri-reagent (Molecular Research Inc, cat#TR118) or TRIzol (Invitrogen cat#15586-026). 3. 1-Bromo-3-chloropropane(MolecularResearchInc,cat#BP151) or chloroform. 4. Isopropanol. 5. We recommend the use of nuclease-free water (Ambion cat#AM9932) to prepare all buffers and solutions. 6. RNeasy mini RNA isolation kit (Qiagen cat#74104). 7. DNase I (Ambion cat#AM2222). 8. RNase Zap (Ambion cat#AM9780).

Determination of Alternate Splicing Events Using the Affymetrix Exon 1.0 ST Arrays

2.4. Materials for RNA QC and Microarray Experiment

65

1. RNA 6000 Nanokit (Agilent cat#5067-1511). 2. GeneChip WT Sense Target Labeling and Control reagents (Affymetrix cat#900652). This catalog number includes all kits required for this protocol including cDNA synthesis, amplification, labeling, cleanup and hybridization. 3. GeneChip® Human Exon 1.0 ST arrays (Affymetrix cat#900650). 4. RiboMinus™ Transcriptome Isolation Kit (Human/Mouse) (Invitrogen cat#K1550-02). 5. Magna-Sep™ Magnetic cat#K1585-01).

Particle

Separator

(invitrogen

6. Betaine 5 M (Sigma cat#B-0300).

3. Methods 3.1. Cell Culture and Harvesting of Cells for RNA Isolation

1. Indicated cell lines were cultured in DMEM (Gibco) supplemented with 10% fetal bovine serum (Hyclone) 100  U/ml penicillin and 100  mg/ml streptomycin and 2  mM l-glutamine. Cells were maintained in 10-cm2 dishes in a 37°C incubator under 5% CO2. 2. Wash the cells with PBS to remove any residual media prior to harvesting. 3. Add 1 ml Tri-reagent or TRIzol directly to the cells in each 10-cm2 dish. Do not trypsinize the cells prior to treatment with tri-reagent or TRIzol (see Note 2). Move the TRIzol around the flask and gently tap to slough off all attached cells. Pipette into a clean tube and store at −20°C till further use.

3.2. RNA Isolation

3.3. Qiagen RNEasy Mini-Cleanup

The specific RNA isolation method that you choose will depend on your downstream application. Generally either method is acceptable for microarray, RT-PCR, or Northern blotting. The Qiagen spin-column cleanup offers the advantage of performing an optional DNase I digestion while purifying the RNA so further processing is avoided. However, detection of RNA molecules of 200 bp or smaller will be limited if using the Qiagen cleanup procedure and hence not advised if you intend to use the RNA for miRNA analysis. While using arrays such as the Exon ST 1.0, Gene ST 1.0 or Tiling arrays, ensure that the RNA is DNase treated since DNA contaminants will be amplified and labeled in the array protocol. 1. Add nuclease free water to the 150–200  ml RNA from the RNA isolation to adjust the volume to 200 ml (see Note 3). 2. Perform the RNEasy micro-cleanup as per the manufacturer’s protocol.

66

Subbaram et al.

3.4. Assessment of RNA Quality

1. Using a NanoDrop® spectrophotometer, measure the optical absorbance characteristics of the sample (see Note 4). The A260/A280 as well as the A260/A230 ratio will ideally be close to 2.0, signifying the purification of nucleic acids away from protein and other organics, respectively. If either ratio is lower than 1.6, expect problems with downstream applications of the RNA (see Note 5). 2. Performance of a NanoChip assay using Agilent’s BioAnalyzer allows for measurement of the molecular weight profile of the isolated RNA. In this way, you may evaluate the 28S/18S ratio measurements. A total RNA ratio between 1.8 and 2.0 is desirable; however, ratios 1.6–1.8 may be acceptable. A RNA Integrity number (RIN) score should be between 7 and 10 if the samples are to be used in a microarray or QPCR experiment downstream (see Note 6).

3.5. Expression Analysis of mRNA from Cells

1. While we have used many different microarray platforms for standard gene expression analysis, we recommend the use of Affymetrix Exon 1.0 ST arrays for experiments where alternate splicing is of interest. 2. There are two methods recommended by Affymetrix to amplify and label the RNA for hybridization to Exon arrays starting with 100 ng or 1 mg of total RNA. We will demonstrate the use of 1 mg protocol in this example (7) (Note 7). We have had good results with both protocols and also with the Nugen protocol, which enables starting with small amounts of RNA as seen with LCM or flow sorted samples. Please remember that since data generated by each of these protocols are not directly cross-comparable, process all samples of a given study using the same protocol.

3.6. Synthesis of Labeled cDNA and Microarray Hybridization 3.6.1. Ribosomal Reduction of 1 mg Total RNA

1. Make serial dilutions of the GenChip PolyA controls (1:20; 1:50, and finally 1:50) using the polyA dilution buffer supplied with the kit. The final concentration of the polyA controls is 1:50,000 of the original stock. 2. Add 1 part 5 M betaine to 2 parts hybridization buffer supplied in the Invitrogen ribominus kit (162 ml/sample). 3. Aliquot 1 mg total RNA in RNase free tube. The total volume of the sample should not exceed 3.2 ml. Add 2 ml of the diluted poly A controls to the sample. 4. Prepare a master mix composing of 1  ml ribominus probe (100 pmol/ml) and 30 ml of the betaine buffer per reaction. Add this to the tube from step 3. Incubate at 70°C for 5 min and then place on ice. 5. Resuspend the bottle containing the magnetic beads by flicking it until no sediment is seen at the bottom. Aliquot 50 ml

Determination of Alternate Splicing Events Using the Affymetrix Exon 1.0 ST Arrays

67

of the resuspended bead solution per reaction to a fresh tube. Add 50 ml of RNase-free water, briefly spin and place the tube on the magnetic stand for 1 min. Gently aspirate and discard the supernatant. Repeat this wash again with 50 ml water and then resuspend the beads in the hybridization buffer with betaine prepared in step 2. Spin briefly and place on the magnetic stand. Aspirate and discard the supernatant. Resuspend the beads in 20 ml hybridization buffer with betaine and incubate at 37°C for 10 min mixing once during incubation. 6. Transfer the cooled sample mix from step 4 to the bead suspension from step 5. Mix gently and incubate at 37°C for 10 min, mixing once during incubation. Place on the magnetic stand and aspirate the supernatant into a clean labeled tube. Add 50 ml of hybridization buffer with betaine to the beads, mix, place on magnetic stand, and aspirate the supernatant and combine into the previously labeled tube. The total volume of this rRNA reduced sample is approximately 100 ml. 7. Add 350  ml of cRNA binding buffer (containing ethanol) from the GenChip IVT cRNA cleanup kit to each rRNA reduced sample. Vortex and then add 250 ml of 100% ethanol to each reaction. Mix well and apply the sample to the IVT cRNA cleanup column. Centrifuge 15 s at 8,000 × g, transfer column to a fresh tube, add 500 ml cRNA wash buffer and centrifuge again for 15  s at 8,000 × g. Discard the flow through, add 500 ml of 80% ethanol to the column and spin again for 15 s at 8,000 × g. Discard the flow through, open the column cap and centrifuge for 5 min at 20,000 × g with the cap open. Transfer the column to a fresh tube and add 11 ml of RNase-free water directly to the membrane. Spin at 20,000 × g for 1  min to elute the rRNA reduced total RNA/Poly A RNA control mix. 8. Check the sample from step 7 on a bioanalyzer to ensure that ribosomal peaks are reduced in the sample. We typically see greater than 80% reduction after this protocol (see Fig.  1). Samples with less than optimal reduction may be subjected to an additional ribosomal reduction step. 3.6.2. Synthesis of Labeled cDNA

1. Prepare a 1:5 dilution of the supplied T7-(N)6 primers and add 1 ml of the diluted solution to 4 ml of the rRNA reduced total RNA/Poly A RNA control mix from step 8 of Subheading 3.6.1. Flick the tube to mix, spin down, and incubate 5 min at 70°C followed by 2 min at 4°C. Place on ice. 2. Prepare the double-stranded cDNA using the GeneChip WT cDNA synthesis kit as per the manufacturer’s protocol (see Note 8).

68

Subbaram et al.

before

after

Fig. 1. Electropherogram traces of total RNA before and after ribosomal reduction

3. This is then converted to complimentary RNA (cRNA) by in vitro transcription using the GeneChip WT cDNA amplification kit as per the manufacturer’s protocol. This protocol should yield at least 15–30 mg cRNA. 4. This cRNA (10 mg) from the first cycle is then reverse transcribed back to cDNA using random primers and a 10 mM nucleotide mix containing dNTP and dUTP. Typical yields of the sense DNA is in the range of 6–7.5 mg. 5. The uridylated single stranded cDNA (5.5 mg) is then fragmented using Uracil DNA glycolase (UDG) and human apurinic/apyrimidinic endonuclease (APE 1). This procedure fragments the cDNA reproducibly at locations where dUTP is incorporated in the DNA during the second-cycle firststrand reverse transcription step. 6. The fragmented cDNA is end labeled using terminal deoxynucleotidyl transferase (TdT) and the kit supplied DNA labeling reagent that is covalently linked to biotin.

Determination of Alternate Splicing Events Using the Affymetrix Exon 1.0 ST Arrays 3.6.3. Hybridization and Scanning

69

1. The labeled cDNA (5.5  mg) is mixed with 20× eukaryotic hybridization controls, denatured and hybridized to Human Exon 1.0 ST arrays as recommended in the kit (Note 9). 2. After 18 h hybridization, the arrays are subjected to a fluidics protocol that washes and stains the array with streptavidin phycoerithrin. 3. The stained arrays are then scanned in a GeneChip 3000G scanner and the data is exported as CEL files.

3.7. Analysis of Human Exon 1.0 ST Array Data 3.7.1. Gene Level Analysis

1. Traditional microarray analysis methods present a steep learning curve for the average user. The problem resides primarily in the normalization techniques used to distribute the signal intensities on the array. To obtain a robustly confident list of genes associated with a given condition, we use the iterPLIER algorithm as the probe intensity summarization method (6, 8). We have successfully used Agilent GeneSpring GX v10, Biotique X-ray as well as Partek Genomics software to analyze Exon array data. 2. We strongly recommend the use of replicates in the experiments using microarray technology for gene expression profiling. While we realize that these experiments can be cost prohibitive, confidence in that data from microarray experiments requires the use of at least 2–3 biological replicates. While generating preliminary data, one could resort to pooling of multiple samples to neutralize the biological variance; however, this could lead to loss of meaningful important data. 3. After summarization, we routinely conduct a Principle Component Analysis to identify any outliers in the samples. We also evaluate the control spikes and hybridization metrics as described by Affymetrix (9). 4. Next, we filter the data to exclude probesets that fall in the bottom 20th percentile for signal intensity and do not show good signal in all replicates of any given condition. This reduces the noise in the data and makes it manageable. 5. A statistical test (Students t-test or ANOVA) with a p-value <0.05 and a false discovery rate correction (Benjamini Hochberg or Bonforoni) routine is most appropriate at this step. The stringency of the statistics will determine how many differentially expressed targets are identified. 6. We further reduce the data by applying a filter on fold change of expression values between the two conditions. While a twofold cutoff seems to be used in many microarray experiments, we prefer to use a 1.5-fold cut off. This enables us to have enough probe sets in our lists while performing secondary analysis such as Gene Ontology or Pathway Analysis.

70

Subbaram et al.

3.7.2. Exon Splicing Analysis (Core-Level)

Exon array analysis can be done on multiple levels; core probe set (17,800 transcript clusters of RefSeq and full-length GenBank mRNAs); extended probe set (core + EST and partial mRNAbased annotation) and full probe set (262,000 transcript clusters, including extended + ab initio gene predictions). The analysis described here is based on core level probes. 1. For Exon level analysis, the summarized probeset values are filtered using the DABG (detection above background) algorithm and p-value (probeset) £0.05. For a transcript to be called as Present, a substantial number of core probe sets should be “Present” (as designated by the DABG generated p-value). The default value specifies 50% of core probe sets to be “Present.” The percentage of samples (within a condition) in which a gene must be present for it to be retained is set at 50% and can be increased for more stringency. 2. This is followed by a Splicing ANOVA with a p-value <0.05. This uses a gene-normalized intensity value i.e., ratio of probe set intensity to expression level of the gene. 3. A Splicing Index value is then calculated. This is similar to a fold-change filter where the gene normalized intensity values are compared between the two experimental conditions. For a given transcript, this difference is computed for each probeset; if any of the probesets has an absolute value difference greater than the specified threshold (0.5 by default) then the transcript will pass this filter. 4. The list of probes that pass the above steps can then be visualized. See Fig. 2 for an example of alternative splicing in the 3¢-end of POLR2I.

a3 knockdown

Wild type

Fig. 2. Signal intensities of exons for POLR2I from control and a3b1 knockdown samples show that 3′ UTR in Exon 6 is alternatively spliced and expressed in the knockdown cells

Determination of Alternate Splicing Events Using the Affymetrix Exon 1.0 ST Arrays

71

The results of any microarray experiment should be verified using an independent technique such as quantitative PCR or sequencing. Additional functional analysis is also recommended.

4. Notes 1. All instruments, glassware and plastic-ware that touch cells or cell lysates should be certified DNase-free and RNase-free or should be prewashed with RNase Zap (Ambion, cat. #9780; 9782) or RNase Away (Molecular BioProducts cat. #7001) followed by DEPC water and allowed to air dry. 2. The number of cells required for each microarray experiment can vary from cell type to cell type. Typically, we utilize a 25 or 75 cm2 flask of confluent cells per condition. This corresponds to about 2–10 × 106 cells and provides enough material for both the microarray experiment as well as other validation and QC experiments. 3. If using this RNA for any miRNA analysis, AVOID the Qiagen cleanup step since it results in loss of small RNAs. 4. If limited in the amount of available sample, one can analyze the RNA via NanoDrop® and then recover material to use for BioAnalyzer runs. 5. Ambion and Affymetrix protocols and technical literature (and our experience) suggest that samples failing to meet either (or both) of these criteria may (or will) perform poorly in molecular techniques, which are based on reverse transcription followed by amplification. This is likely due to the interference of protein, carbohydrate, or phenolic contaminants on the reverse transcription process. 6. The Agilent BioAnalyzer is a preferable substitute to MOPSformaldehyde agarose gel analysis due to the reduced sample required, increased sensitivity, and reduced exposure to toxic reagents. 7. The Agilent 2100 Expert software provides a RIN or RNA integrity number (8) for the RNA nano and pico assays (series II). It is recommended that this RIN number be between 7 and 10 if the RNA sample is to be used in a microarray experiment. We generally use the RIN number as a secondary QC criteria along with 260/280, 260/230, and 28S/18S ratios. 8. All the reagents for this protocol are supplied in the Affymetrix WT Sense Target Labeling and Control reagents kit (7). It is recommended that polyA RNA controls be spiked into the starting RNA samples since this will allow QC for any degradation occurring during the protocol. The signals from these spikes can also be used for normalization. We have also

72

Subbaram et al.

successfully used the NuGEN WT-Ovation Pico and the WT-Ovation Exon module to generate data from exon arrays. 9. It is recommended that hybridization controls be prepared from a master mix. The signal from the controls (bioB, bioC, bioD, and Cre) can be used to qualitatively compare chips being hybridized over time. References 1. Carpenter PM, Dao AV, Arain ZS, Chang MK, Nguyen HP, Arain S, Wang-Rodriguez J, Kwon SY, Wilczynski SP (2009) Motility induction in breast carcinoma by mammary epithelial laminin 332 (laminin 5). Mol Cancer Res 7:462–475 2. Carpenter PM, Wang-Rodriguez J, Chan OT, Wilczynski SP (2008) Laminin 5 expression in metaplastic breast carcinomas. Am J Surg Pathol 32:345–353 3. Iyer V, Pumiglia K, DiPersio CM (2005) Alpha3beta1 integrin regulates MMP-9 mRNA stability in immortalized keratinocytes: a novel mechanism of integrin-mediated MMP gene expression. J Cell Sci 118:1185–1195 4. Lamar JM, Pumiglia KM, DiPersio CM (2008) An immortalization-dependent switch in integrin function up-regulates MMP-9 to enhance tumor cell invasion. Cancer Res 68:7371–7379 5. Mitchell K, Szekeres C, Milano V, Svenson KB, Nilsen-Hamilton M, Kreidberg JA, DiPersio CM

(2009) Alpha3beta1 integrin in epidermis promotes wound angiogenesis and keratinocyte-toendothelial-cell crosstalk through the induction of MRP3. J Cell Sci 122:1778–1787 6. Clark TA, Schweitzer AC, Chen TX, Staples MK, Lu G, Wang H, Williams A, Blume JE (2007) Discovery of tissue-specific exons using comprehensive human exon microarrays. Genome Biol 8(4):R64 7. Affymetrix. GenChip Whole Transcript Sense target Labeling Assay Manual, 701880. http:// www.affymetrix.com/support/downloads/manuals/wt_sensetarget_label_manual 8. Xing Y, Kapur K, Wong WH (2006) Probe selection and expression index computation of Affymetrix Exon Arrays. PLoS One 1:e88 9. Affymetrix. Alternative transcript analysis for Exon Arrays (2005) white paper v 1.1 http://www. affymetrix.com/support/technical/whitepapers/ exon_alt_transcript_analysis_whitepaper.pdf

Chapter 5 Profiling microRNA Expression with the Illumina BeadChip Platform Julissa Tsao, Patrick Yau, and Neil Winegarden Abstract The complex mechanisms involved in the regulation of both gene and protein expressions are still being understood. When microarray technology was first introduced during the early to mid 1990s, they heralded a tremendous opportunity to study transcription on a global scale. Despite this promise, however, one thing that has become clear is that the expression of protein coding genes is not the only aspect of the transcriptome that researchers need pay attention to. Small noncoding RNAs, such as microRNAs, are now known to play a pivotal role in the control of both gene and protein expressions. Each microRNA may act upon a plurality of different targets, which makes the measurement of their expression levels a highly important part of understanding the entire cellular response. It has only been recently, however, that advancements and modifications to microarray technology have allowed us to study these important molecules in a high throughput and parallel manner. Key words: miRNA profiling, miRNA microarrays, DASL assay, BeadChip

1. Introduction When microarray technology was first introduced in 1995, it was received with a great deal of optimism (1). Rather than looking at one gene at a time, it was now possible to potentially look at the expression of every gene in parallel and without a priori identification of genes of interest. While microarrays for the profiling of gene expression have been widely used with over 30,000 publications to date indexed in Pubmed, there have been several reports that have highlighted some of the challenges associated with mRNA-based gene expression studies. One of the main concerns is that mRNA expression in isolation does not provide the full picture as to how a cell is behaving at any point in time. As early as 1999 (2), papers were emerging, which showed that gene Sridar V. Chittur (ed.), Microarray Methods for Drug Discovery, Methods in Molecular Biology, vol. 632, DOI 10.1007/978-1-60761-663-4_5, © Humana Press, a part of Springer Science+Business Media, LLC 2010

73

74

Tsao, Yau, and Winegarden

(mRNA) and protein expression could be discordant. At the time, it was uncertain as to whether this was a result of an inability of microarray technology to accurately capture what was actually happening in the cell or if there was some biological under-­ pinning to this lack of correlation. As microarray technology has evolved, it has become clear that the technology is indeed robust and overall quite accurate, indicating that in fact there must be a biologically relevant reasoning between the discordance of gene and protein expression. mRNA gene expression arrays generally measure steady state RNA levels. As such, it is not possible to distinguish RNA degradation and de novo synthesis, both of which can lead to changes in perceived gene expression. A major piece of the puzzle seems to have been resolved with the reemergence of microRNA (miRNA) as a topic of interest. miRNA are single stranded, small (~21–23  nt) endogenous noncoding RNAs that have been shown to regulate expression level of hundreds of mRNAs (3). Approximately 20–30% of all transcripts are regulated by miRNA in mammalian genome (3). Interestingly, miRNA seem to operate in one of three possible modalities, the choice of which seems to be somewhat species specific (4). microRNAs can either bind with 100% match to an mRNA target or can bind with several mismatches. In the case of perfect matching, miRNA seems to function by targeting the bound mRNA for specific degradation. In this way, specific mRNAs can be removed from the total RNA population in the cell. In cases where the miRNA binds in a nonperfect manner, it appears that the miRNA exerts control on translation of the mRNA, preventing processing by the ribosomes, and thus preventing proteins from being formed for the mRNA (5). A final mode of action is less clear but seems to involve transcriptional control of the target mRNA gene. miRNAs are small entities and can exert their effect on a number of different genes. Each miRNA may control dozens or even hundreds of target mRNAs. The relative number of microRNAs in a cell is one or two orders of magnitude less than the number of mRNAs, and as such, these two facts combined make miRNAs interesting potential biomarkers as they can have multiple effects in the cell, and there are fewer potential markers to monitor. miRNA has also been associated with cell proliferation and cell death (6), apoptosis and fat metabolism (7), and cell differentiation (8). Disregulation of miRNA has been implicated in many diseases such as heart diseases (9) and cancer (10, 11). With more understanding on how miRNAs contribute to cell physiology and function, novel miRNAs are constantly added to the Sanger Institute miRBase database (http://microrna.sanger.ac.uk/). Several array-based methodologies exist for the screening of miRNAs. We have tried several and have found the Illumina

Profiling microRNA Expression with the Illumina BeadChip Platform

75

platform to be particularly appealing. Illumina has developed a highly sensitive, specific, and reproducible method for miRNA profiling using BeadArray technology covering miRNA described in the Sanger Institute miRNA database plus additional content identified with Illumina sequencing. This method is powered by the DASL assay (The cDNA-mediated Annealing, Selection, Extension, and Ligation) using 100–200 ng of intact or FFPEderived total RNA sample. This method can detect 1.2–1.3 fold differences between samples, and it has a 3.5–4 log dynamic range (12). In our hands, we have found the method to be robust and reliable and we present here the protocol as we carry it out for the analysis of up to 24 samples at a time.

2. Materials 2.1. Preparation of Poly-A Polymerase (PAP) Plate

1. Total RNA samples (see Note 1). 2. Polyadenylation Single Reagent (PAS; Illumina). Store at −15°C to −25°C. 3. 96-well 0.2 ml skirted microplates (MJ Research) (see Note 2). 4. High Speed microplate shaker (Illumina). 5. Hybex microsample incubator (SciGene).

2.2. Preparation of cDNA Synthesis (CSP) Plate

1. cDNA Synthesis Single Reagent (CCS; Illumina). Store at −15°C to −25°C.

2.3. Preparation of Allele-Specific Extension (ASE) Plate

1. Oligo Hybridization & DNA binding Buffer 1 reagent (OB1; Illumina). Store at −15°C to −20°C (see Note 3).

2. 96- well 0.2 ml skirted microplate (MJ Research) (see Note 2).

2. Human or Mouse MicroRNA Assay Pool (MAP; Illumina). Store at −20°C. 3. 96-well 0.2 ml skirted microplate (MJ Research).

2.4. Addition of Master Mix for Extension and Ligation (MEL)

1. “Add MEL 1” reagent (AM1; Illumina). Store at 4°C. (see Note 3). 2. Universal Buffer 1 Reagent (UB1; Illumina). Store at 4°C. 3. Master Mix for Extension Ligation reagent (MEL; Illumina). Store at −15°C to −25°C. 4. Dynal MPC-96 S (raised-bar magnetic plate; Invitrogen).

2.5. Preparation of PCR Plate

1. Titanium Taq DNA polymerase (Clontech). Store at −20°C. 2. Uracil DNA Glycosylase (UDG; Illumina). Store at −15°C to −25°C.

76

Tsao, Yau, and Winegarden

3. Single Color Master Mix Reagent (SCM; Illumina). Store at −15°C to −25°C. Prepare aliquots of 800 ml each. 4. 96-well 0.2 ml skirted microplate (MJ Research). 2.6. Inoculation of PCR Plate

1. Universal Buffer 1 Reagent (UB1; Illumina). Store at 4°C.

2.7. Binding of PCR Products

1. Magnetic Particle B Reagent (MPB; Illumina). Store at 4°C.

2.8. Preparation of INT Plate for BeadChip

1. 0.1 N Sodium hydroxide. Make aliquots of 3 ml and store at 4°C.

2. Inoc PCR Reagent (IP1; Illumina). Store at −15°C to −25°C.

2. 96-well filter plate with lid (Millipore).

2. Universal Buffer 2 reagent (UB2; Illumina). Store at room temperature. 3. “Make Hyb1” reagent (MH1; Illumina). Store at room temperature. 4. 96-well V-bottom plate (VWR). 5. Multiscreen centrifuge alignment frame (filter plate adapter; Millipore). 2.9. Hybridization to BeadChip

1. Chamber humidification Buffer (CHB; Illumina). Store at room temperature. 2. XStain BeadChip solution (XC4; Illumina). Store at −20°C (see Note 4). 3. 100% Ethanol. 4. Hyb chamber (Illumina). 5. BeadChip (12×1). Store at 4°C until ready to use.

2.10. Washing BeadChip

1. Universal Buffer 2 reagent (UB2; Illumina). Store at room temperature. 2. XStain BeadChip Solution (XC4; Illumina). Store at −20°C.

3. Methods 3.1. Preparation of Poly-A Polymerase (PAP) Plate (see Note 5)

1. Normalize total RNA samples to 40–200 ng/ml using nuclease-free water (see Note 6). 2. Label a new 96-well 0.2 ml PCR plate “PAP”. 3. Thaw the PAS reagent to room temperature. Vortex the contents of the tube followed by a quick spin in a centrifuge to pull down all the liquid. Add 5 µl PAS to each well of columns 1, 2, and 3 of the PAP plate.

Profiling microRNA Expression with the Illumina BeadChip Platform

77

4. Quickly add 5 ml of normalized RNA sample (100–200 ng) to each well of columns 1, 2, and 3 (each sample in a separate well). Final volume will be 10 µl (5 µl PAS and 5 µl RNA sample). 5. Heat seal PAP plate with microplate heat seal. (see Note 7). 6. Vortex the sealed plate at 2,300  rpm for 20  s on Illumina shaker. 7. Quick spin the PAP plate for 1 min at 250 × g. 8. Incubate PAP plate at 37°C for 60  min in preheated heat block/microsample incubator with lid closed. 9. Transfer the PAP plate to a preheated 70°C heat block for 10 min in order to inactivate the PAP enzyme. 3.2. Preparation of cDNA Synthesis (CSP) Plate

1. Label new 96-well microplate “CSP”. 2. Thaw the CSS tube at room temperature. Add 8 ml CSS to each well of columns 1, 2, and 3 of the CSP plate. 3. Carefully remove the heat seal from the PAP plate. 4. Transfer 8 µl of the polyadenylated RNA sample prepared in subheading 3.1 from each well of the PAP plate to the corresponding well of the CSP plate. Do this as quickly as possible to minimize the difference in reaction time between the first column and the third column. 5. Seal the CSP plate with the heat sealer. 6. Vortex sealed CSP plate at 2,300 rmp on the Illumina plate shaker. 7. Quick spin the CSP plate for 1 min at 250 × g. 8. Incubate the CSP plate at 42°C for 60 min in preheated heat block with the lid closed. 9. Do one of the following (see Note 8): (a) If you wish to proceed, transfer the CSP plate to a heat block at 70°C for 10  min and proceed to next Subheading  3.3 Preparation of Allele-Specific Extension plate (ASE). (b) If you do not plan to proceed immediately, then do the following: ●●

●●

●●

Transfer the sealed CSP plate to preheated heat block at 70°C and incubate for 10 min to inactivate the RT enzyme. Quick spin the sealed CSP plate at 250 × g for about a minute to remove any condensation from the walls of each well. Store the sealed CSP plate at −20°C overnight.

78

Tsao, Yau, and Winegarden

3.3. Preparation of Allele-Specific Extension (ASE) Plate

1. Label a new 96-well plate “ASE”. If CSP plate was frozen overnight, thaw the CSP plate at room temperature. 2. Do a quick spin at 250 × g for 1 min. 3. Thaw the MAP reagent to room temperature. Vortex the tube and do a quick spin in a centrifuge to pull down all liquid. Add 5 µl MAP to each well of columns 1, 2, and 3 of the ASE plate. 4. Thaw the OB1 tube at room temperature. Vortex the tube and ensure that all beads are resuspended. Do not centrifuge the OB1 tube as this will repellet the beads. Add 30 µl OB1 to each well of columns 1, 2, and 3 of the ASE plate. 5. Carefully remove heat seal from CSP plate and avoid splashing from the wells. 6. Transfer all biotinylated cDNA generated in Subheading 3.2 from the CSP plate to the corresponding well of the ASE plate (~15 ml). 7. Heat-seal the ASE plate with the microplate heat sealer. Ensure that all wells are completely sealed. 8. Quick spin the ASE plate at 250 × g for 1 min. 9. Vortex the ASE plate at 1,600 rpm for 1 min on the Illumina plate shaker. Ensure that all beads are completely resuspended. Repeat vortex step if necessary. 10. Put the ASE plate into a 70°C preheated heat block with the lid closed. 11. Change the temperature on heat block to 40°C. Leave the ASE plate on the heat block for 2–4 h while it is cooling to 40°C. If possible, leave for the full 4 h.

3.4. Addition of Master Mix for Extension and Ligation (MEL)

1. Remove ASE plate from heat block and set heat block to 45°C. 2. Put the ASE plate on the raised bar-magnetic plate for 1 min or until the beads are completely sequestered to the side of the wells. 3. Carefully remove the heat seal from ASE plate. 4. Keep the ASE plate on the raised magnetic plate. Using a multichannel pipette, remove all liquid (~50 ml) from all occupied wells and discard. Visually inspect the pipette tips after removing the liquid to ensure that not all beads have been removed. If there are beads visible in the pipette tips, return solution to the wells and repeat step 2 to recollect beads. Use new pipette tips. (see Note 9). 5. Take the AM1 bottle from the refrigerator and leave it at room temperature for 10 min.

Profiling microRNA Expression with the Illumina BeadChip Platform

79

6. Use a multichannel pipette with new tips to add 50 µl AM1 reagent to each occupied well. 7. Seal the ASE plate with clear adhesive film. 8. Vortex the ASE plate at 1,600 rpm for 20 s on the Illumina plate shaker. Ensure that all beads are resuspended. Repeat vortexing step if necessary. 9. Put the ASE plate on the raised-magnetic plate for 1 min or until the beads are completely sequestered to the side of the wells. 10. Carefully remove the seal from the ASE plate, avoiding splashing from the wells. 11. Remove all AM1 reagent from each occupied well. Ensure that beads are not disturbed. 12. Repeat steps 6 to 11 once. 13. Remove ASE plate from the raised bar-magnetic plate. 14. Using new pipette tips, add 50 ml UB1 to each occupied well on the ASE plate using multichannel pipette. 15. Put the ASE plate on the raised-bar magnetic plate for 1 min or until the beads are completely captured. 16. With ASE plate still on the raised bar-magnetic plate, remove all UB1 reagent. Ensure that beads are not disturbed. 17. Repeat steps 13–16 once. 18. Thaw the MEL reagent to room temperature. Add 37 ml MEL to each occupied well on the ASE plate using multichannel pipette. 19. Seal the ASE plate with clear adhesive film. 20. Vortex the ASE plate at 1,600 rmp for 1 min to resuspend the beads. Ensure that beads are completely resuspended. If necessary, repeat vortexing step. 21. Incubate the ASE plate on preheated heat block at 45°C for 15  min. DO NOT allow the ASE plate to incubate longer than 15 min. 3.5. Preparation of PCR Plate

1. Label new 0.2 ml microplate “miRNA PCR”. 2. Thaw one aliquot of SCM reagent. 3. Add 16 µl Clontech DNA polymerase to 800 µl SCM reagent. 4. Add 12.5 µl Uracil glycosylase to the SCM tube + DNA polymerase (see Note 10). The addition of UDG helps to prevent PCR contamination. 5. Invert tube several times to mix, add 30 µl of SCM mixture into each well of columns 1, 2, and 3 of the miRNA PCR plate.

80

Tsao, Yau, and Winegarden

6. Cover the PCR plate with a lid and keep it on ice. 7. As soon as the 15  min incubation is complete for the ASE plate, immediately proceed to the next step. 3.6. Inoculation of PCR Plate

1. Put the ASE plate on the raised bar-magnetic plate for 1 min or until the beads are completely sequestered. 2. Carefully remove the clear adhesive seal from plate to avoid splashing. 3. Remove and discard the supernatant (~50 µl) from the ASE plate using multichannel pipette. Ensure that the beads are undisturbed. If you can see beads on the pipette tips, repeat step 1 and remove supernatant using new tips. 4. Leaving the plate on the raised bar-magnetic plate, add 50 ml of UB1 to each well occupied on the ASE plate. 5. Seal the ASE plate with clear adhesive film. 6. Vortex at 1,600 rpm for 1 min on the Illumina plate shaker or until all the beads are resuspended. 7. Put the plate back on the raised bar-magnetic plate for 1 min or until all the beads are sequestered. 8. Remove and discard UB1 (~50 ml) from all occupied wells. Ensure that beads are undisturbed. 9. Thaw the IP1 tube and keep on ice. Add 35 µl IP1 reagent to each occupied well on the ASE plate using a multichannel pipette. 10. Seal the plate with clear adhesive film. 11. Vortex the ASE plate at 1,800 rpm for 1 min or until all the beads are resuspended. 12. Put the ASE plate on preheated heat block at 95°C for 1 min. 13. Transfer the ASE plate back into the raised bar-magnetic plate for 1 min or until all the beads are captured. 14. Remove the seal from the ASE plate. Avoid splashing as this will cross contaminate your samples. 15. Add 30 µl of the supernatant from the ASE plate into the corresponding well of the microRNA PCR plate using a multichannel pipette. Pipette the contents of the PCR plate up and down several times to mix. 16. Seal the PCR plate with the PCR plate sealing film appropriate for your thermal cycler. 17. Quick spin the PCR plate at 250 × g for 1 min.

Profiling microRNA Expression with the Illumina BeadChip Platform

81

18. Transfer the PCR plate to the thermal cycler with the following program:

×34

Temperature

Time

37°C

10 min

95°C

3 min

95°C

35 s

56°C

35 s

72°C

2 min

72°C

10 min

4°C

5 min

19. Once the thermal cycler program is complete (~2 h 50 min), do one of the following: (see Note 11) (a) Proceed immediately to next step. (b) Store the PCR at −20°C overnight. 3.7. Binding of PCR Products

1. Quick spin the PCR plate at 250 × g for 1 min. 2. Vortex the MPB tube several times. Ensure that all beads are resuspended. Pipette 20 µl of resuspended MPB into each well occupied on the PCR plate. You do not have to change tips until all the MPB has been transferred to all three columns. To avoid contamination, place the tips against the top edge of the wells. 3. Set your multichannel pipette to 85 µl. Transfer the content of the PCR plate to the corresponding location on the filter plate using new pipette tips between columns dispenses and discard the PCR plate. 4. Cover the filter plate with the filter plate lid provided. 5. Store the filter plate in the dark for 60 min at room temperature. 6. Proceed to next step.

3.8. Preparation of INT Plate for BeadChip

1. Label a new 96-well V-bottom plate “INT.” 2. Place the filter plate adapter onto an empty, unlabeled 96-well V-bottom plate (waste plate). 3. Place the filter plate containing the bound PCR product onto the filter plate adapter and waste plate. 4. Centrifuge at 1,000 × g for 5 min at room temperature. 5. Remove the filter lid. Add 50 ml UB2 to each well of columns 1, 2, and 3 of the filter plate using a multichannel pipette. Dispense slowly to avoid disturbing the beads. To avoid disturbing the beads, place the pipette tips against the top edge of the well.

82

Tsao, Yau, and Winegarden

6. Replace the filter plate lid. 7. Centrifuge at 1,000 × g for 5 min at room temperature. 8. Add 30 ml MH1 to each well of columns 1, 2, and 3 on the INT plate using a multichannel pipette. 9. Transfer the filter and filter plate adapter to the INT plate. Ensure that the filter plate and INT plate have the same orientation (well A1 of filter plate lines up with well A1 of the INT plate). Discard the waste plate. 10. Using a multichannel pipette, add 30 ml of 0.1 N NaOH to each well occupied on the filter plate. Cover the filter plate with the lid. 11. Centrifuge immediately at 1,000 × g for 5 min. Avoid prolong incubation with 0.1 N NaOH. The DNA is denatured almost instantly. No beads should be present in the wells of the INT plate. 12. Discard the filter plate (see Note 12). 13. Gently mix the content of the INT plate by moving it from side to side without splashing. 14. Seal the INT plate with a 96-well cap mat. Store in the dark at room temperature until ready to dispense sample onto BeadChip. If you are not planning to hybridize the BeadChip immediately, store the INT plate at −20°C. (see Note 13). 3.9. Hybridize BeadChip

This section utilizes the Illumina BeadChip Hyb chambers and hybridization-related equipment. 1. Preheat the Illumina hybridization oven to 60°C. Allow 30 min for it to equilibrate. 2. Take the BeadChips to be hybridized from 4°C storage and let them equilibrate to room temperature for 10 min. Do not open BeadChip from its package until you are ready to begin hybridization. 3. If the INT plate has been frozen, thaw the INT plate at room temperature in the dark and then centrifuge at 250 × g for 1 min. 4. Place the Hyb chamber gasket into the Hyb chamber. Ensure that gasket is properly seated. Each Hyb chamber holds 4 BeadChips. 5. Add 200 ml CHB into the humidifying buffer reservoir in the Hyb chamber on the slots you are planning to place your BeadChip for hybridization. 6. Close and lock the BeadChip Hyb chamber. Seat the lid securely on the bottom plate of the chamber. Snap two diagonally opposed clamps shut. Snap the other two clamps shut. Leave the Hyb chamber on the bench at room temperature until the BeadChips are loaded with DNA sample.

Profiling microRNA Expression with the Illumina BeadChip Platform

83

7. Place each BeadChip in a Hyb chamber insert, orienting the barcode end of the slide so that it matches the barcode symbol on the Hyb chamber insert. Avoid contacting the beadstripe area and sample inlet port. 8. Add 15 ml of sample onto the center of each inlet port. Load samples by placing tips directly onto the array surface, holding the pipette straight up to avoid wicking. Some residual sample will remain in the inlet port. This is normal. 9. Open the Hyb chamber. 10. Load the Hyb chamber inserts containing sample-laden BeadChips into each hyb chamber. 11. Close and lock the BeadChip Hyb chamber as per step 6. 12. Place the Hyb chamber into the 60°C Illumina hybridization oven so that the clamps face the left and right sides of the oven. The Illumina logo on top of the Hyb chamber should face you. 13. Incubate for exactly 30 min at 60°C then reset the temperature in the oven to 45°C. 14. Incubate for at least 14 h but no more than 20 h at 45°C. 3.10. Wash BeadChip

The XC4 reagent for washing BeadChip can either be prepared the day before and left overnight at room temperature on your bench top or 30 min prior to washing BeadChips. 1. Add 330 ml 100% Ethanol to the XC4 bottle. The final volume will be 350 ml. 2. Shake the bottle vigorously for 15 s and leave the bottle on the benchtop overnight at room temperature. 3. Prior to using XC4, ensure that it is completely resuspended. Shake the XC4 bottle again. The solution should be clear and homogeneous, with no gelatinous or stringy remains. If any coating is visible, vortex at 1,625 rpm until it is clear and suspension is complete. 4. If you are resuspending the XC4 solution prior to washing the BeadChips, add 300 ml 100% Ethanol to the bottle. Place the bottle on the rocker for at least 30 min, leave it on the rocker until ready to use. The XC4 solution should be clear and homogenous, with no gelatinous or stringy remains. 5. For each BeadChip to be washed: (a) Add 45 ml UB2 to two 50 ml centrifuge tubes and label them UB2. (b) Add 40 ml XC4 to one 50 ml centrifuge tube and label the tube XC4. 6. Remove the BeadChip from the Hyb Chamber insert. 7. Remove the IntellyHyb Seal from each BeadChip, using powder-free gloves, and hold the BeadChip in one hand with

84

Tsao, Yau, and Winegarden

your thumb and forefinger on opposing edges of the BeadChip away from your face. The barcode should be closest to you. Remove the entire seal with your other hand in a rapid motion by pulling it off in a diagonal direction. Do not touch the exposed active areas. 8. Submerge the BeadChip in the first UB2 tube. 9. Cap the tube and invert five times. Ensure that all the black seal from the IntellyHyb seal around the BeadChip has been removed (see Note 14). 10. Transfer the BeadChip to the second UB2 tube using tweezers, invert several times. Incubate for 5 min. 11. Transfer the Beadchip to the XC4 centrifuge tube, tighten cap, and invert ten times. 12. Incubate the BeadChip in XC4 for 5 min. 13. Using self-locking tweezers remove the BeadChip from the XC4 tube and place the BeadChip on the tube rack completely horizontal with the barcode towards you. Note XC4 is slippery and makes the BeadChip difficult to hold. 14. Place the tube rack with the BeadChips in a vacuum desiccator. 15. Close the lid of the vacuum desiccator. Start the vacuum. To ensure that the desiccator is properly sealed, gently lift the lid of the desiccator, and it should not lift off the desiccator base. 16. Dry under vacuum for 20–55 min (see Note 15). 17. Release the vacuum by turning the handle very slowly to allow air to enter the desiccator slowly to avoid damaging the BeadChip. 18. Use a 70% ethanol wipe or wet a lint-free tissue paper with 70% EtOH to wipe the underside of the BeadChip to remove excess XC4. Hold the BeadChip at a downward angle to prevent the EtOH from dripping onto the stripes and wipe the underside five or six times, or until the surface is clean and smooth. 19. Proceed to Image BeadChip on the iScan System.

4. Notes 1. As this protocol is for the profiling of small 21–24  nt long RNAs, it is critical that an RNA isolation/purification strategy that includes small RNAs in the final product be used.

Profiling microRNA Expression with the Illumina BeadChip Platform

85

Total RNA isolated using several column-based purification strategies have poor representation of small RNAs. We have found that Trizol is a good option for the purification of total RNA that is inclusive of miRNA. 2. If it is preferred, it is also possible to use 8-well strip PCR tubes instead of a 96-well microplate. 3. Reagents OB1 and AM1 contain formamide. As such, caution should be used when handling these reagents, and waste solutions should be disposed of appropriately. 4. Once ethanol has been added to the XC4 reagent, it can be stored at 4°C for up to 14 days. It is not recommended to freeze thaw this reagent as it affects stability. 5. This protocol has been written with the intent of processing 24 samples. The Illumina miRNA BeadChips have 12 arrays each and as such, this protocol is for the processing of 2 BeadChips. The protocol scales nicely allowing up to 96 samples to be handled at a time by a single technician. If greater or fewer than 24 samples are to be processed, then the necessary modifications to the protocol must be made. 6. While the labeling system works well with 40–200 ng/ml of total RNA, it is recommended that all samples in a single project be treated equally where possible, and thus normalization should be to the same concentration for all samples if at all possible. 7. Seal plate for about 2 s on the heat sealer, then turn the plate around and seal for another 2–3 s. 8. This is a convenient place to stop if the protocol is too long to finish in a single day or if it becomes necessary to pause. 9. To avoid tip contamination and loss of sample, slant pipette tips so that they are drawing liquid from the side of the well opposite of where the beads have collected. It is easier to draw liquid from all the odd numbered columns first; then turn the plate to do the even numbered column as two columns of beads collect to one column of magnets. 10. The official Illumina protocol lists this step as “optional.” We however recommend that this step always be taken to increase reproducibility and to control contamination. 11. We have comparable results by leaving the PCR plate on the thermal cycler overnight at 4°C instead of freezing the plate at −20°C overnight. 12. Although we have said to discard the filter plate in this step, we do not do so until we have used the entire plate. In order to make these analyses more cost effective, we use the unused filter columns for subsequent labeling. To store the partially

86

Tsao, Yau, and Winegarden

used filter plate, the columns that were used are clearly marked and the plate is wrapped in plastic wrap and stored at room temperature. Ensure that you add the NaOH and MH1 reagent into the next empty columns for subsequent labelings. 13. We seal the INT plate with foil instead of the cap mat for −20°C storage. 14. If you see some black glue from the coverseal on the slide, you can dip the BeadChip in 70% EtOH for about 30 s to 1 min to help remove the glue from the slide. Then continue on to step 10. 15. Illumina recommends drying for 50–55 min. In our hands, with our vacuum source, BeadChips are dry after 15–20 min drying under vacuum. References 1. Schena M, Shalon D, Davis RW, Brown PO (1995) Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 270:467–470 2. Gygi SP, Rochon Y, Franza BR, Aebersold R (1999) Correlation between protein and mRNA abundance in yeast. Mol Cell Biol 19:1720–1730 3. Lewis BP, Burge CB, Bartel DP (2005) Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets. Cell 120:15–20 4. Zhao Y, Srivastava D (2007) A developmental view of microRNA function. Trends Biochem Sci 32:189–197 5. Baek D, Villen J, Shin C, Camargo FD, Gygi SP, Bartel DP (2008) The impact of microRNAs on protein output. Nature 455:64–71 6. Brennecke J, Hipfner DR, Stark A, Russell RB, Cohen SM (2003) Bantam encodes a developmentally regulated microRNA that controls cell proliferation and regulates the proapoptotic gene hid in Drosophila. Cell 113:25–36

7. Xu P, Vernooy SY, Guo M, Hay BA (2003) The Drosophila microRNA Mir-14 suppresses cell death and is required for normal fat metabolism. Curr Biol 13:790–795 8. Dostie J, Mourelatos Z, Yang M, Sharma A, Dreyfuss G (2003) Numerous microRNPs in neuronal cells containing novel microRNAs. RNA 9:180–186 9. Yang B, Lin H, Xiao J, Lu Y, Luo X, Li B, Zhang Y, Xu C, Bai Y, Wang H, Chen G, Wang Z (2007) The muscle-specific microRNA miR-1 regulates cardiac arrhythmogenic potential by targeting GJA1 and KCNJ2. Nat Med 13:486–491 10. Calin GA, Croce CM (2006) MicroRNA signatures in human cancers. Nat Rev Cancer 6:857–866 11. Esquela-Kerscher A, Slack FJ (2006) Oncomirs – microRNAs with a role in cancer. Nat Rev Cancer 6:259–269 12. Chen J, Lozach J, Garcia EW, Barnes B, Luo S, Mikoulitch I, Zhou L, Schroth G, Fan JB (2008) Highly sensitive and specific microRNA expression profiling using BeadArray technology. Nucleic Acids Res 36:e87

Chapter 6 TaqMan® Array Cards in Pharmaceutical Research David N. Keys, Janice K. Au-Young, and Richard A. Fekete Abstract TaqMan® Array Cards are high-throughput, accurate, sensitive, and simple-to-use tools for quantitative analysis of mRNA or miRNA transcripts using a real-time PCR protocol. They utilize a microfluidic card with 384 reaction chambers and eight sample loading ports. For studies of coding transcripts, the reaction chambers are preloaded with user selected or predefined panels of Applied Biosystems TaqMan Gene Expression Assays. These assays enable real-time monitoring of a PCR reaction via hydrolysis of an oligonucleotide probe which has been dual labeled with fluorescent dye and quencher. Applications of TaqMan Array Cards include verification and follow on testing of microarray results, as well as hypothesis driven testing of panels of genes selected for their biological functions and relationships. This chapter describes a protocol for assaying transcription in cultured cells using methods optimized to minimize hands-on time and pipetting steps by skipping RNA isolation and generating cDNA directly in Ambion® Cells-toCTTM lysis solution. Key words: Gene expression, mRNA, TaqMan Array Card, TLDA, Microfluidic, Real-time PCR, Quantitative PCR, 5¢-exonuclease assay, Cells-to-CT

1. Introduction Recent advances in high-throughput genomic technology, in particular hybridization based microarrays, have greatly impacted pharmacological research. This technology can be used to screen the full complement of an organism’s transcriptome, identifying genes whose expression levels correlate with disease status, drug response, toxicological mechanisms, or experimental treatments. Relative to microarrays, real-time qPCR technologies assay smaller sets of target genes. However, they exhibit better overall accuracy and reproducibility, especially for low expressing targets and targets with small fold changes (i.e. <2-fold) (1, 2). Experiments on small gene sets using real-time qPCR technologies complement microarray research in two important ways. First, real-time qPCR Sridar V. Chittur (ed.), Microarray Methods for Drug Discovery, Methods in Molecular Biology, vol. 632, DOI 10.1007/978-1-60761-663-4_6, © Humana Press, a part of Springer Science+Business Media, LLC 2010

87

88

Keys, Au-Young, and Fekete

is widely used for independent verification of the results of microarray experiments (3–6). Second, it is ideal for hypothesis testing experiments which further characterize panels of genes identified in microarray screens, as well as panels of genes for whose biological relation to the experimental conditional are already well established (7). 1.1. TaqMan® Gene Expression Assays

TaqMan Gene Expression Assays (Fig. 1) are commercially available preformulated real-time qPCR assays, which have been designed for most coding genes in human, mouse, rat, and other model systems. Because these assays are predesigned to universal reaction conditions, it is possible to move directly from the identification of target transcripts to identifying TaqMan Assays capable

Fig. 1. TaqMan real-time quantitative PCR. TaqMan real-time qPCR requires site-specific binding of two PCR oligos and a probe oligo dual labeled with a fluorescent reporter molecule (R) and a quencher molecule (Q). The 3¢ terminus of the probe oligo is modified to prevent elongation. The exonuclease activity of Taq DNA polymerase is responsible for hydrolysis of the probe oligo during the polymerization of a new DNA strand. This separates the reporter and quencher molecules, leading to fluorescence. The level of fluorescence in the reaction is directly proportional to the number of amplicon molecules which have been produced

TaqMan® Array Cards in Pharmaceutical Research

89

of quantifying those transcripts. There are several bioinformatics tools available to assist in the selection of assays. For microarray validation studies, the UMapItTM Tool can be used to map batches of probe IDs from Affymetrix, Agilent, or Illumina microarrays directly to TaqMan Gene Expression Assay IDs (http://info. appliedbiosystems.com/umapit). The NCBI Probe Database can also be used to map microarray and other data types directly to TaqMan Gene Expression Assay IDs (http://www.ncbi.nlm.nih. gov/sites/entrez?db=probe). Fixed panels of preselected TaqMan Gene Expression Assays for drug targets such as GPCRs, nuclearreceptors, and kinases are available. For pathway-based studies, the Ambion GeneAssistTM Tool can be used to identify TaqMan Gene Expression Assays for genes, which are involved in biological processes. This tool can also be used to help design siRNA experiments (http://www.ambion.com/tools/pathway). For hypothesis testing experiments where the full panel of genes to be assayed is already known, assays can be individually selected inside the TaqMan Array Configuration Tool by querying with gene names, symbols, or RefSeq IDs. TaqMan Array Cards (Fig.  2) are microfluidic devices containing 384 reaction chambers for individual qPCR reactions. They offer multiple practical and experimental advantages for use in pharmacological research. Most importantly, the performance of TaqMan real-time qPCR in this format has been shown to be equivalent to TaqMan performed in more traditional 384-well

GR2156a

Fig. 2. TaqMan Array Cards. Loading ports, microfluidic channels, and reaction chambers are laid out on a flat surface. Structural stability of the card is provided by a plastic carrier mounted above the reaction chambers. The size, reaction chamber spacing, and dimensions of the TaqMan Array Card match with those of a standard 384-well plate

90

Keys, Au-Young, and Fekete

plate formats (8). Simplified workflows, especially when combined with lysate reaction protocols (rather than RNA purification) vastly reduce the overall time to results, the amount of hands-on time, and the number of pipetting steps. In addition to the time savings, this minimizes opportunity for experimental errors. The content and number of TaqMan Array Cards purchased can be customized to the experimental design. This can offer large cost savings relative to purchasing assays individually. Additionally, lower reaction volumes (100 µL per 48 reactions) result in significant reductions in enzyme costs. For a 20 card study using 96 assays, a 70% savings in reagents can be realized relative to the same experiment using 384-well plates (8). For large-scale experiments, a major benefit of using TaqMan Array Cards is achieving reproducible results. Assays are loaded using a quality-controlled manufacturing process and the microfluidics platform ensures that samples are loaded properly. This provides standardization when screening gene panels across multiple samples and enables direct comparison of results between multiple laboratories. 1.2. Experimental Design and Analysis

One feature of the protocol presented here is that precise quantification of input cell numbers or total RNA is not required. This is made possible by the accuracy of TaqMan based-chemistry over a broad dynamic range of template input and the use of the relative quantitation method to calculate fold change relative to a calibrator sample rather than attempting to directly measure template copy number. The geometric nature of PCR makes relative quantitation possible. The result of an individual TaqMan qPCR reaction is reported as a CT (Threshold Cycle) value, which is the fractional cycle number, at which the fluorescence from hydrolyzed probe exceeds a fixed threshold. Because PCR amplification is based on doubling the number of amplicons each cycle, a reaction which starts with twofold more target template than a calibrator reaction will reach the fluorescence threshold 1 cycle earlier than that calibrator reaction. Thus, each CT represents a twofold difference in starting template. Many genes will have transcripts whose levels do not change in response to experimental treatments. These endogenous control genes can be used to normalize results between samples because differences in their CT values reflect general sample input rather than biological differences between samples. The difference between a target assay’s CT and an endogenous control assay’s CT is referred to as the DCT and represents the target assay normalized to the endogenous control. For a target assay that is normalized to the same endogenous control assay in two samples, usually a calibrator sample and an experimental sample, the difference between the DCT values from the two samples is referred to as the DDCT and represents the difference in the target gene’s CT values between the two samples, normalized for

TaqMan® Array Cards in Pharmaceutical Research

91

differences in how much starting material was present in each sample. The fold-change of a target transcript in the experimental sample relative to the calibrator sample can be directly calculated using the formula 2−DDCT (9). Experimental designs should take into account the need for both an appropriate calibrator sample, typically an untreated sample, and an endogenous control gene (or genes) whose transcripts do not vary with the experimental treatment. Endogenous control genes can be identified from historical data or selected through separate experiments using the TaqMan Endogenous Control Array Cards and software tools such as RealTime StatMiner® (Integromics Software) or geNorm (PrimerDesign Ltd.) (10). A common experimental design for pharmacogenomics research involves a dose response experiment, in which one cell culture is split into eight samples, seven of which are treated with increasing concentrations of the test compound (experimental samples) and one of which is left untreated (the calibrator sample). Using a TaqMan Array Card design where 16 assays are plated in triplicate in the 48 reaction chambers connected to each of the eight loading ports allows all samples to be run simultaneously on a single card. The untreated sample serves as the calibrator sample. Two to four of the assays should be selected specifically to be used as endogenous controls. The use of triplicates enables the calculation of standard deviations. The following protocol assumes this experimental design, however, many different combinations of sample, assay, and replicate numbers are possible.

2. Materials 2.1. Test Materials

1. Eight samples of cultured cells.

2.2. Reagents

1. TaqMan Gene Expression Cells-to-CT Kit (see Note 1) (Ambion). 2. TaqMan Array Cards preloaded with TaqMan Assays (Applied Biosystems). 3. Nuclease-free water.

2.3. Equipment

1. Applied Biosystems 7900HT Fast Real-Time PCR System (Applied Biosystems). 2. The 7900HT TaqMan Array Upgrade Kit (see Note 2) (Applied Biosystems). 3. Centrifuge compatible with TaqMan Array Cards (see Note 3) (Applied Biosystems).

92

Keys, Au-Young, and Fekete

3. Methods 3.1. Cell Lysis

1. Prepare sterile 1× PBS and chill to 4°C. 2. Thaw Cells-to-CT stop solution, mix by inversion, and place on ice. 3. Prepare a sufficient amount of Lysis Solution for the number of samples to be processed plus 10% overage for pipetting loss. For eight samples with a final volume of 50 µL per sample, combine 434.6 µL Cells-to-CT Lysis Solution and 4.4 µL DNase. 4. For adherent cells, start with 10–105 cells in a 96-well culture plate. For nonadherent cells, start with a known number of pelleted cells (see Note 4). 5. Remove and discard culture media. 6. Add chilled PBS in a volume equivalent to the smaller of 0.5 mL or the volume of discarded culture media. For adherent cells, remove PBS immediately, leaving 5 µL or less of PBS (see Note 5). To wash nonadherent cells, gently resuspend, repellet by centrifugation, and resuspend in a volume of PBS sufficient to give 10–105 cells in 5 µL. 7. Add 50 µL of the previously prepared Lysis Solution with DNase. Mix by pipetting up and down five times. 8. Incubate for 5 min at room temperature. 9. Add 5 µL Stop Solution. Mix by pipetting up and down five times. 10. Incubate for 5  min at room temperature then place on ice (see Note 6).

3.2. Reverse Transcription

1. Prepare a sufficient amount of Reverse Transcription Master Mix for the number of samples to be processed plus 10% overage for pipetting loss: For eight samples with a final reverse transcription reaction volume of 50 µL, combine 220 µL 2× Reverse Transcription Buffer and 22 µL 20× Reverse Transcription Enzyme Mix. Mix and store on ice until use (see Note 7). 2. Aliquot 27.5 µL Reverse Transcription Master Mix into thermocycler tubes or plates and place on ice. 3. Add 22.5 µL of lysate to each aliquot of Reverse Transcription Master Mix for a final reaction volume of 50 µL. Mix thoroughly and pulse spin to collect the liquid at the bottom of the tube.

TaqMan® Array Cards in Pharmaceutical Research

93

4. Use a thermocycler with a heated lid to incubate the reverse transcription reactions: 37°C for 60  min, 95°C for 5  min, followed by a 4°C hold (see Note 8). 3.3. Real-Time PCR Set Up (see Note 9)

1. For each sample, use a fresh tube to combine 45 µL reverse transcription product, 50 µL 2× TaqMan Gene Expression Master Mix, and 5 µL nuclease free water. Mix thoroughly and pulse spin to collect thse liquid at the bottom of the tube (see Note 10). 2. Aliquot the entire 100 µL qPCR reaction mixture into the TaqMan Array Card loading port using a hand pipette. 3. Once all eight loading ports have been filled, place the TaqMan Array Card into the swing bucket centrifuge racks (see Note 11). 4. Centrifuge at 331 × g (1,200  rpm with the Sorvall Legend rotor) for 1 min. Repeat the centrifugation for a total of two consecutive centrifugations. 5. Prepare the TaqMan Array Card Sealer by placing it on a firm level surface with the start position is close to the user and the carriage in the start position. 6. Place the TaqMan Array Card into the sealer with the reaction chambers facing down and the loading ports positioned to the side furthest from the carriage start position. 7. Push the sealer carriage across the TaqMan Array Card using a slow steady motion (see Note 12). 8. Using scissors, trim the loading ports off the TaqMan Array Card.

3.4. Real-Time PCR Run

1. Open the SDS software (version 2.3 or later) on the computer attached to the 7900HT instrument. Open a new file and use the pull-down menus in the New Document Dialog Box to set the plate format to “384 Well TaqMan Low Density Array” and the analysis method to either “Standard Curve (AQ)” or “ddCT (RQ)”. Leave the template field as “Blank template”. Also enter the barcode number located on the side of the TaqMan Array Card (see Note 13). 2. Optional: Using the Import Dialog Box, import the TaqMan Array setup file from the CD supplied with the TaqMan Array Cards (see Note 14). 3. Save the file (file type *.sds). 4. Select the (Real-Time) tab and press the (Open/Close) button to rotate the plate tray out of the instrument. 5. Load the TaqMan Array Card onto the plate tray with reaction chamber A1 on the top left and the bar code toward the front of the instrument. Press (Start Run) (see Note 15).

94

Keys, Au-Young, and Fekete

4. Notes 1. Other than nuclease-free water, the TaqMan Gene Expression Cells-to-CT kit contains all of the enzymes and buffers required to compete this protocol. Specifically, it includes Lysis Solution, DNase, Stop Solution, 2× Reverse Transcription Buffer, 20× Reverse Transcription Enzyme, and 2× TaqMan Gene Expression Master Mix. 2. Due to their unique shape, TaqMan Array Cards are not compatible with standard 384-well thermocycling blocks. At this time, all compatible instruments are based on the Applied Biosystems 7900 hardware platform. These include the ABI PRISM® 7900HT Sequence Detection System, and its successor, the Applied Biosystems 7900HT Fast Real-Time PCR System. Either machine must be equipped with the 7900HT TaqMan Array Upgrade Kit. In addition to the thermal block and heated cover, this kit includes the required centrifuge swing buckets and rotor adapters, as well as the TaqMan Array Card Sealer. 3. Reaction mix is distributed from loading ports to individual wells via centrifugation. Unlike 384-well plates, which are commonly spun flat, TaqMan Array Cards are run standing on end. Swing buckets and adapters are supplied as part of the 7900HT TaqMan Array Upgrade Kit. They are compatible with the following families of centrifuge models: Sorvall Legend T+, Legend XT, and ST40; Fisher accuSpin Model 3; Thermo SL40; Heraeus Multifuge 3S+, Multifuge X3, and Megafuge 40. 4. Many protocols in pharmacological and toxicology research involve growing cells a set amount of time in the presence of a test compound, then assaying transcript levels. Experimental designs should factor in the approximate number of cells, which will be present at the end of the treatment time to ensure that the total count does not exceed 105. For most cell types, using more than 105 cells can overwhelm the lysis reaction, leading to incomplete lysis and inhibition of downstream reactions. If the cell count is over 105, cultures should be split or cells should be washed in suspension and resuspended with an appropriate dilution. Using small numbers of cells will have no effect on the chemical performance of the reactions. However, low-end sensitivity is bound by the number of template molecules in the qPCR reaction. Therefore, there are practical limits that will come into play if the number of cells becomes too low. For example, with the experimental design described in this protocol, for every 175 cells present at the

TaqMan® Array Cards in Pharmaceutical Research

95

start of the lysis, 1 cell equivalent of cDNA will be present in each of the final 48 qPCR reaction chambers. 5. Adherent cells grown in 96-well culture plates can be washed and lysed directly in the plate they have been cultured in, although care must be taken to avoid dislodging the cells when adding the 1× PBS wash. 6. After the addition of the stop solution and the 2 min room temperature incubation, the lysate can be held at room temperature up to 20  min, on ice up to 2  h, or stored frozen (−20°C or −80°C) for up to 5  months. Lysate can be put through multiple freeze thaw cycles without loss of qPCR signal. 7. After thawing, the 2× Reverse Transcription Buffer should be vortexed, spun down and kept on ice. The reverse transcription reaction may contain up to 45% lysate, which is the volume used here. 8. Completed reverse transcription reactions can be frozen at −20°C for storage, and may be put through multiple freeze thaw cycles without loss of qPCR signal. 9. For extremely low expressing genes or limited starting material, preamplification methods can be run before loading the TaqMan Array Card (11). TaqMan based preamplifications are optional multiplex-PCR reactions, which are run on each reverse transcription product using a pooled set of the assays present in the TaqMan Array Card. This amplification generates sufficient target sequences to counteract the effects of splitting the sample among a large number of low volume reaction chambers. By running with a limited number of cycles (<14), low oligo concentrations, and a master mix optimized for high multiplex reactions (TaqMan PreAmp Master Mix, Applied Biosciences), artifacts from nonspecific amplification and target specific biases are minimized. 10. The qPCR reaction may contain up to 45% reverse trans­ cription product, which is the volume used here. Extremely high expressers such as 18s may be overloaded if the number of starting cells is high, however low expressing transcripts will be better represented. On TaqMan Array Cards designed to use multiple loading ports per sample, the reverse transcription product should be divided evenly among the loading ports. 11. TaqMan Array Cards should be loaded into the centrifuge buckets with the reaction chambers and loading ports facing outward. The centrifuge holds 12 cards, with 3 in each bucket. To avoid damage to the rotor or bucket, they should be balanced with blank cards whenever loading fewer than two full buckets.

96

Keys, Au-Young, and Fekete

12. The TaqMan Array Card Sealer mechanically deforms the backing of the TaqMan Array Card, collapsing and sealing the microfluidic channels which connect the reaction chambers to the loading ports. Excess reaction mix which is in the microfluidic channels during sealing is pushed back into the loading ports. 13. Setting the file type to Absolute Quantitation (AQ) allows CT values to be directly exported from the SDS software for analysis in external programs, including spreadsheet programs such as Microsoft Excel® and specialized software for gene expression analysis such as RealTime StatMiner (Integromics Software). For studies using up to 10 TaqMan Array Cards, the program RQ Manager (Applied Biosystems) can import and collectively analyze SDS files. However, the files must be set to type Relative Quantitation (RQ) at the time they are created. Setting the plate type to 384 Well TaqMan Low Density Array also loads a hot-start thermal cycling program specific to TaqMan Array Plates and chemistry: 50°C for 2 min; 94.5°C for 10 min; 40× (97°C for 30 s; 59.7°C for 1 min). 14. Importing the Setup File will associate Assay IDs, gene symbol information, and generic sample identification with each reaction chamber. The Setup file does not need to be imported into the SDS file before running the TaqMan Array Card. When running more than one array, it is often more convenient to import the assay information files after the runs have completed. The sample information can also be changed from generic entries (sample 1, sample 2, etc) to actual sample names at any time after the Setup file is imported. 15. For 7900HT systems equipped with robotic loaders, groups of TaqMan Array Cards can be queued for automated loading using the SDS Automation Controller Software. References 1. Canales RD, Luo Y, Willey JC et  al (2006) Evaluation of DNA microarray results with quantitative gene expression platforms. Nat Biotechnol 24:1115–1122 2. Wang Y, Barbacioru C, Hyland F et al (2006) Large scale real-time PCR validation on gene expression measurements from two commercial long-oligonucleotide microarrays. BMC Genomics 7:59 3. Chuaqui RF, Bonner RF, Best CJM et  al (2002) Post-analysis follow-up and validation of microarray experiments. Nat Genet 32 suppl.:509–514

4. Provenzano M, Mocellin S (2007) Comple­ mentary techniques: validation of gene expression data by quantitative real time PCR. Adv Exp Med Biol 593:66–73 5. Sinicropi D, Cronin M, Liu M (2007) Gene expression profiling utilizing microarray technology and RT-PCR. In: Ozkan M, Heller MJ (eds) BioMEMS and biomedical nanotechnology volume 2: micro/nano technologies for genomics and proteomics. Springer, New York, NY 6. Wang Y, Barbacioru C, Keys D et  al (2008) Real-time polymerase chain reaction gene expression assays in biomarker discovery and

TaqMan® Array Cards in Pharmaceutical Research validation. In: Wang F (ed) Methods in pharmacology and toxicology: biomaker methods in drug discovery and development. Humana, Totowa, NJ 7. Györffy B, Surowiak P, Kiesslich O et  al (2006) Gene expression profiling of 30 cancer cell lines predicts resistance towards 11 anticancer drugs at clinically achieved concentrations. Int J Cancer 118:1699–1712 8. Goulter AB, Harmer DW, Clark KL (2006) Evaluation of low density array technology for quantitative parallel measurement of multiple genes in human tissue. BMC Genomics 7:34

97

9. Livak KJ, Schmittgen TD (2001) Analysis of relative gene expression data using real-time quantitative PCR and the 2−DDCT method. Methods 25:402–408 10. Vandesompele J, De Preter K, Pattyn F et al (2002) Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol 3:34.1–34.11 11. Mengual L, Burset M, Marín-Aguilera M et al (2008) Multiplex preamplification of specific cDNA targets prior to gene expression analysis by TaqMan Arrays. BMC Res Notes 1:21

Chapter 7 DMET ™ Microarray Technology for Pharmacogenomics-Based Personalized Medicine James K. Burmester, Marina Sedova, Michael H. Shapero, and Elaine Mansfield Abstract Human genome sequence variation in the form of single nucleotide polymorphisms (SNPs) as well as more complex structural variation such as insertions, duplications, and deletions underlies each individual’s response to drugs and thus the likelihood of experiencing an adverse drug reaction. The ongoing challenge of the field of pharmacogenetics is to further understand the relationship between genetic variation and differential drug responses, with the overarching goal being that this will lead to improvements in both the safety and efficacy of drugs. The Affymetrix® DMET™ Plus Premier Pack (DMET stands for Drug Metabolizing Enzymes and Transporters) enables highly multiplexed genotyping of known polymorphisms in Absorption, Distribution, Metabolism, and Elimination (ADME)-related genes on a single array. The DMET Plus Panel interrogates markers in 225 genes that have documented functional significance in phase I and phase II drug metabolism enzymes as well as drug transporters. The power of the DMET Assay has previously been demonstrated with regard to several different drugs including warfarin and clopidogrel. In a research study using an earlier four-color version of the assay, it was demonstrated that warfarin dosing can be influenced by a cytochrome P450 (CYP) 4F2 variant. Additionally, the assay has been used to demonstrate that CYP2C19 variants with decreased enzyme activity led to lower levels of the active clopidogrel metabolite, resulting in a decreased inhibition of platelets and a higher rate of cardiovascular events when compared to noncarriers of the DNA variant. Thus, highly multiplexed SNP genotyping focused on ADME-related polymorphisms should enable research into development of safer drugs with greater efficacy. Key words: ADME, Clopidogrel, DMET (drug metabolizing enzymes and transporters), Genetic testing, SNP genotyping, Warfarin

1. Introduction There is increasing interest both in the medical research community and in the lay press in the potential use of pharmacogenetics to increase the safety and efficacy of commonly prescribed medications (1). Sridar V. Chittur (ed.), Microarray Methods for Drug Discovery, Methods in Molecular Biology, vol. 632, DOI 10.1007/978-1-60761-663-4_7, © Humana Press, a part of Springer Science+Business Media, LLC 2010

99

100

Burmester et al.

Toxicity and lack of efficacy are considered to be the major reasons for drug failures and genetically-determined differences in pharmacokinetics are a major contributor (2). Increasingly, preclinical pharmacokinetic evaluation should be comprehensive enough to ensure that compounds do not fail in the clinic. Preclinical screening for variation in Absorption, Distribution, Metabolism, and Elimination (ADME), or pharmacokinetic profiling, enables early elimination of weak drug candidates; thus, ADME studies can direct the focus of drug development programs toward fewer lead candidates. Identifying which specific sets of enzymes metabolize candidate compounds also helps in predicting probable drug–drug interactions. Compounds with multiple metabolic pathways are less likely to exhibit clinically significant drug interactions (3). The therapeutic efficacy of most drugs is influenced by a number of different factors that in part include age, weight, and concurrent drug use (4, 5). These factors may vary between patients. In addition, fixed parameters such as gender and human genome sequence variation can contribute. This genetic variation underlies every individual’s response to drugs. The vast majority of the enzymes involved in drug metabolism are highly polymorphic (6) and allele frequencies of low-activity variants often differ by population (7, 8). Consequently, their activity may differ depending upon an individual’s genotype(s). For example, drugs may be metabolized more slowly in individuals who are carriers of a genetic polymorphism that results in a decreased or null activity of a given enzyme. These individuals are at particular risk for adverse drug reactions (ADR) or therapeutic failure (9–11). Conversely, drug therapy could be ineffective if the drug is metabolized too rapidly. Genetically determined variation particularly impacts drugs with narrow therapeutic indices, increasing the risk for the development of ADRs. Comprehensive genotyping could be helpful for choosing the right drug and the optimal dosage for individual patients. The use of genetic profiles to individualize drug therapy is the vision of personalized medicine (12, 13). According to a recent study conducted by the Food and Drug Administration (14), approximately one-quarter of the prescriptions written in the United States in 2006 contained pharmacogenetic labeling recommendations. The DMET™ Plus Premier Pack offers a comprehensive profiling of genetic diversity across known ADME markers on a single array (Figs. 1 and 2). The assay genotypes 1,936 markers in 225 high-value genes focused on drug metabolism (pharmacokinetic pathways), transcription regulators, selected drug targets (e.g., vitamin K 2,3 epoxide reductase complex 1 [VKORC1] and 3-hydroxy-3-methylglutaryl-coenzyme A reductase [HMGCR]), and several classes of drug transporters (Fig. 3). The polymorphisms represented on DMET Plus Arrays were chosen by virtue

DMET™ Microarray Technology for Pharmacogenomics-Based Personalized Medicine

101

Fig. 1. DMET™ Plus Panel genes and mutations. A partial list of well-known genes in each category is shown with the total number of genes in each category (top number). The total number of SNP markers in each category is also shown (bottom number)

Fig. 2. DMET™ Plus DNA variants in disease treatment. Genes in the DMET Plus Panel involved in drug metabolism of selected common therapeutics are shown. The last column indicates the number of markers that are located within the genes listed for each drug

of their functional significance, as documented in the scientific literature (15, 16). The core polymorphisms have been publicly reviewed and prioritized by a panel of experts made up from both the pharmaceutical industry and academia (ADME Consortium; https://pharmaadme.org/). The DMET Plus Assay Panel has been evaluated across a minimum of 1,200 individuals from multiple populations including 715 DNA samples from Caucasian,

102

Burmester et al.

Fig. 3. Gene ontology and pathway redundancy in the DMET™ Plus Panel. A partial list of genes in the DMET Plus Panel and their biochemical/physiological role are shown

African, and Asian populations from the International HapMap Consortium. The functional composition of the multiplex ADME assay and differences in allelic frequency in important genes regulating response to warfarin (VKORC1, cytochrome P450 [CYP] 2C9 and CYP4F2 (17)) and clopidogrel (CYP3A4, CYP2C9, CYP2C19 (18)) will be highlighted in this chapter.

2. Materials 2.1. Laboratory Requirements

The assay is performed in three separate laboratory areas in order to segregate pre-polymerase chain reaction (PCR) reagents from post-PCR amplification products, thereby eliminating any potential for sample contamination from PCR products. The three areas include (1) a multiplex PCR (mPCR) staging area (a separate laboratory or fume hood within the preamplification lab), (2) a preamplification lab where the genomic DNA is stored and the prePCR reactions are set up, and (3) a postamplification lab. For more information on the laboratory requirements, refer to the Affymetrix DMET™ Plus Site Preparation Guide.

DMET™ Microarray Technology for Pharmacogenomics-Based Personalized Medicine

2.2. Affymetrix Reagents and Arrays Required 2.2.1. Affymetrix DMET Plus Premier Pack (P/N 901268)

103

One DMET Plus Premier Pack is sufficient to process 48 reactions (42 DNA samples, three genomic DNA control samples, and three plasmid control DNA samples) and includes the following components (Boxes 1– 4):

Box 1 Affymetrix DMET™ Plus Pre-Amp Kit (Store at −20°C) Includes Enzyme A. Buffer A. Preamplification water. GapFill mix 1. GapFill mix 2. dNTP mix. Exo mix. Universal amplification mix. Cleavage enzyme.

Box 2 Affymetrix DMET™ Plus Labeling Kit (Store at −20°C) Includes PCR clean-up mix. Postamplification water. Fragmentation buffer. Fragmentation reagent. DNA labeling reagent. 5× TdT buffer. TdT enzyme.

Box 3 Affymetrix DMET™ Plus Hyb-Stain Kit (Store at −20°C) Includes Hybridization solution. Oligonucleotide control reagent. Stain buffer. Hold buffer.

104

Burmester et al.

Box 4 Affymetrix DMET™ Plus Panel Kit (Store at −20°C) Includes DMET mPCR primer mix. 1× TE buffer. PCR dilution buffer. DMET molecular inversion probe (MIP) panel. DMET gDNA control 1. DMET gDNA control 2. DMET gDNA control 3. DMET plasmid control A. DMET plasmid control B. DMET plasmid control C.

Table 1 Additional reagents to purchase Item

Vendor

AccuGENE water

Lonza group LTD 51200

QIAGEN Multiplex PCR kit QIAGEN

206143

Streptavidin, R-phycoerythrin conjugate (SAPE)

Invitrogen

S866 (1 mL)

TITANIUM™ Taq polymerase

Clontech

639208 (100 rxns) 639209 (500 rxns)

TE buffer, pH 8.0

TekNova

T0223

2× loading buffer

Sigma

G2526

3% agarose gel

Bio-Rad

161-3040

Low molecular weight DNA latter

New England Biolabs

N32335

Quant-iT™ PicoGreen® dsDNA assay kit

Invitrogen

P11496

®

2.2.2. Wash Solutions (Shipped Separately; Store at Room Temperature) Includes

Wash solution A (three bottles; P/N511715). Wash solution B (two bottles; P/N511716).

2.2.3. 48 DMET™ Plus Arrays (P/N 901317; Shipped Separately; Store at +4°C)

2.3. Required Reagents from Other Suppliers

Part number

See Table 1.

DMET™ Microarray Technology for Pharmacogenomics-Based Personalized Medicine

105

Table 2 Equipment and software Item

Part number

GeneChip® Fluidics Station 450 – two or more units required

00-0079

GeneChip® Hybridization Oven 640 or 645

800139 (640) 00-0331 (645)

GeneChip® 3000 Scanner with 7G upgrade

901153

GeneChip® Command Console Software DMET™ Console Software

2.4. Required Affymetrix Equipment and Software

See Table 2. For the full list of equipment, refer to the DMET Plus Premier Pack User Guide.

2.5. DNA Input Requirements

The DMET Plus Assay requires 1 mg of genomic DNA normalized to a single concentration of 60 ng/mL using 1× TE buffer. The use of Quant-iT™ PicoGreen® dsDNA assay kit from Invitrogen is recommended to determine DNA sample concentrations, as DNA sample concentrations determined by UV absorbance at 260  nm may not be as accurate and may lead to inconsistent assay results.

3. Methods 3.1. Introduction

The DMET™ Plus Assay uses MIP technology (19, 20) to amplify in a highly multiplexed manner the sequence-specific information at each SNP locus. In this assay, PCR product generated with MIP technology then undergoes enzymatic fragmentation and end-labeling followed by hybridization to an array containing allele-specific oligonucleotides used for SNP discrimination and genotyping. Array probe signal intensities are read out using a single color detection scheme with the fluorophore R-phycoerythrin. The following protocol describes the workflow for processing 48 samples over 2.5 days and consists of multiple stages. Multiplex PCR, MIP probe anneal reaction, MIP assay, and PCR amplification are all performed in the prelab. The remaining steps are all carried out in the post lab and include the following: PCR clean-up, DNA target fragmentation, labeling, hybridization, array washing, array staining, and array scanning.

106

Burmester et al.

3.2. Preparing Genomic DNA Plate 1 (GP1)

1. Aliquot 17 mL of each genomic DNA sample (at 60 ng/mL) to be processed into wells B01–E12. It is not recommended to use plate rows A or H to process samples in the DMET Plus Assay due to the potential for evaporation of reaction material from the corner wells of the plate. 2. If the genomic controls included in the DMET Plus Panel Kit are being processed, transfer genomic controls 1, 2, and 3 to three wells located in three different rows, for example C11, D11, and E11. This layout ensures the presence of high quality DNA at the correct concentration in three of four rows and helps troubleshooting efforts if required. If the plasmid DNA controls are being processed, leave three wells (C12, D12, and E12) empty at this point. Plasmid controls are added later at the anneal step. For more information on assay controls, see Note 1. 3. Seal the plate, spin down at 2,000 rpm (685 g) for 30–60 s, and place in an aluminum block on ice (or at +4°C) until ready to use. For general rules on handling the plates, see Note 2.

3.3. mPCR

A number of loci interrogated in the DMET Plus Assay are known to have pseudogenes with similar sequence as well as close homologs (for example, the CYP gene family). Because of this, probes designed specifically for these targets may also anneal to multiple undesired locations of the genome, thereby reducing the subsequent ability to accurately genotype this marker. This issue is resolved by performing an initial genomic amplification using locus specific primers in a mPCR reaction using QIAGEN Multiplex PCR kit. The 36-plex mPCR reaction requires 50 ng of genomic DNA and represents the first step of the DMET Plus Assay. Location–pre-amplification lab and mPCR staging area The preparation and running of the mPCR plate on the thermal cycler is conducted in the preamplification lab. The dilution of mPCR products and subsequent addition to the anneal reactions is conducted in the mPCR staging area. Input required: genomic DNA plate 1 1. Thaw the following reagents on the bench top. From the DMET Plus Panel Kit, thaw the mPCR primer mix, 1× TE buffer and PCR dilution buffer. From the QIAGEN Multiplex PCR Kit, thaw two tubes 2× QIAGEN Multiplex PCR master mix, 5× Q-solution and RNase-free water. Once thawed, place all reagents on ice until ready to use except for the PCR dilution buffer that should remain at room temperature. 2. Prepare genomic DNA plate  2 (GP2): After vortexing and spinning the 1× TE buffer, aliquot 10 mL into each well of a new plate labeled “GP2.” Transfer 2 mL of each sample (and genomic controls) from the GP1 plate to the corresponding

DMET™ Microarray Technology for Pharmacogenomics-Based Personalized Medicine

107

Table 3 Reaction components Reagent

1 Reaction

48 Reactions (>20% extra)

QIAGEN Multiplex PCR master mix

25 mL

1,500 mL

mPCR primer mix (3 mM)

  5 mL

   300 mL

5× Q-solution

  5 mL

   300 mL

RNase-free water

10 mL

   600 mL

Total

45 mL

2,700 mL

well of the GP2 plate. Pipet the contents up and down three times to rinse the tips. Tightly seal the plate, vortex, spin down, and return to the aluminum block on ice. DNA concentration in the GP2 plate is 10 ng/mL. Seal the GP1 plate (60 ng/mL) and store on ice or at +4°C until ready to perform the annealing step. If running the plasmid controls, keep the designated wells empty. 3. To prepare the mPCR master mix, combine the components from the QIAGEN Multiplex PCR Kit with mPCR primer mix from the DMET Plus Panel Kit according to Table 3. 4. Mix by pipetting up and down five times using a single-channel P1000 set to 900 mL. 5. Aliquot 45 mL of mPCR master mix to each well of four rows on 96-well PCR plate placed on an aluminum block on ice. 6. Add 5 mL of each sample at 10  ng/mL from GP2 to the corresponding well of the mPCR plate. 7. Tightly seal the plate, vortex, spin down, and run DMET Plus mPCR program on a thermal cycler (Fig. 4a). 8. While the program is running, allow the PCR dilution buffer to thaw at room temperature. Steps 9 and 10 yielding a 1,000-fold dilution of mPCR product should be performed in the mPCR stage area. Once the DMET Plus mPCR program has finished, spin down the plate, transfer it to the mPCR staging area, and place on an aluminum block on ice. 9. Perform the 1,000-fold dilution of mPCR product in two steps by creating dilution plates  1 and 2 (DP1 and DP2): Mix the PCR dilution buffer by inverting the bottle ten times and pour into a reagent reservoir. Aliquot 153 mL of the buffer into each well of rows B, C, D, and E (skip the wells corresponding to plasmid samples) of plates DP1 and DP2. To avoid introducing bubbles, do not blow out the pipet tips, dispense to the first stop only. Transfer 5 mL of each mPCR

108

Burmester et al.

Fig. 4. Temperature conditions for pre-amplification lab steps (a) mPCR; (b) annealing program and (c) assay program

product from the mPCR plate to the corresponding well of the DP1 plate. Slowly mix with a multichannel pipet set to 80 mL. 10. Transfer 5 mL of each mPCR product (diluted once) from the DP1 plate to the corresponding well of the DP2 plate. Slowly mix with multichannel pipet set to 80 mL. Tightly seal the GP2 plate and keep on ice until used in the anneal step. Discard the DP1 plate. Keep mPCR product plate at −20°C until the arrays are processed and data analysis is completed. 3.4. Annealing and Addition of DMET MIP Panel

At this stage, genomic DNA (plate GP1), diluted mPCR product from stage 3.1 (plate DP2), and DMET MIP Panel are combined for an overnight annealing reaction at 58°C. During this time, MIP terminal sequences hybridize to their complementary genomic templates, thereby creating circularized structures that

DMET™ Microarray Technology for Pharmacogenomics-Based Personalized Medicine

109

incorporate a single or multiple nucleotide gap located between the probe termini and directly across from the SNP intended for interrogation. 3.4.1. Location: Preamplification Lab

Input required: plate GP1 and plate DP2 1. Thaw preamplification water, buffer A, DMET MIP Panel, enzyme A (and plasmid controls A, B, and C, if running) on the bench top. Keep enzyme A on the bench top only as long as required to thaw (~5 min). Once thawed, spin down and keep on ice. See Note 3 for instructions on working with enzymes. 2. Prepare anneal mix by combining the reagents listed in Table  4 into 1.5  mL Eppendorf tube. Mix and transfer to reservoir. 3. Create an anneal plate by aliquoting 21.7 mL of anneal master mix to each well of rows B, C, D, and E on 96-well PCR plate. Add 13.4 mL of each DNA sample, as well as genomic controls from plate GP1 to the corresponding well on the anneal plate. If running the plasmids, add 18.4 mL of each plasmid control. 4. Transfer the plate to the mPCR staging area. Add 5 mL of diluted mPCR product from DP2 plate to the corresponding wells of the anneal plate. Do NOT add diluted mPCR product to the wells with plasmid controls. 5. Tightly seal the plate, transfer it to the prelab, vortex, spin down, load onto a thermal cycler, and start DMET Plus anneal program (Fig. 4b). 6. At the end of the first 95°C hold, pause the program, remove the plate, and place it in an aluminum block on ice for 2 min. 7. Aliquot the DMET MIP Panel to one strip of 12 tubes, 25 mL in each tube. Add 5 mL of MIP Panel to each reaction on the anneal plate.

Table 4 Anneal mix preparation Reagents

1 Reaction

48 Reactions (>25% extra)

Pre-amplification water

16.6 mL

996

Buffer A

5 mL

300 mL

Enzyme A

0.0625 mL

3.8 mL

Total volume

21.7 mL

1,299.8 mL

110

Burmester et al.

8. Tightly seal the plate, vortex, spin down, load on a thermal cycler, and resume DMET Plus Anneal Program. Optimal incubation time is 16–18 h. 3.5. MIP Assay and Amplification

During this stage, the missing nucleotide complimentary to the SNP of interest is filled in by the DNA polymerase, and the DNA ligase enzyme covalently closes the circular padlock structure. Any unreacted linear probe that remains, in addition to single-stranded genomic DNA, are digested by the Exonuclease mix. After cleavage of the circular covalently-closed circles, the linearized, inverted probes are amplified using universal primers.

3.5.1. Location: Preamplification Lab

Input required: anneal plate 1. Place dNTP mix and the universal amplification mix on the bench top at room temperature and thaw. 2. Prepare GapFill mix by combining 190 mL GapFill mix 2 and 10 mL of GapFill mix 1 in an Eppendorf tube. Aliquot the solutions slowly since both solutions contain 50% glycerol. Mix well by pipetting up and down 15 times using P200 set up to 150 mL. Aliquot 14 µL of GapFill mix to each tube of one strip of 12 tubes. 3. Remove the anneal plate from the thermal cycler. Spin it down, and place it in an aluminum block on ice. 4. Add 2.5 mL of GapFill mix to each well, seal, vortex, and spin down. 5. Create a new assay plate by transferring 12 mL of each reaction from anneal + GapFill mix plate. Seal, spin down, place the plate onto a thermal cycler, and start the following DMET Plus Assay Program (Fig. 4c). 6. Vortex dNTP mix, spin down, aliquot 25 mL to each tube of one strip of 12 tubes. After 11 min at 58°C, pause the program and remove the plate from the thermal cycler. Place in an aluminum block on ice for 2 min. Spin down. Add 5 mL of dNTP per well. Seal, vortex, spin down, place back on the thermal cycler, and resume the program. 7. Remove Exo mix from −20°C and spin down. Aliquot 25 mL to each tube of one strip of 12 tubes. When the thermal cycler temperature reaches 37°C, pause the program for the Exo mix addition. Remove the plate from the thermal cycler. Place in an aluminum block on ice for 2 min. Spin down. Add 5 mL of Exo mix per well. Seal, vortex, spin down, place back on the thermal cycler, and resume the program. 8. During the 5  min at 95°C period, prepare the universal amplification mix by adding 25 mL of cleavage enzyme and 70 mL of TITANIUM Taq polymerase to the universal

DMET™ Microarray Technology for Pharmacogenomics-Based Personalized Medicine

111

amplification mix tube. Set a P1000 pipet to 900 mL and mix by pipetting up and down ten times. Pour the universal amplification mix into a reagent reservoir on ice. 9. When the temperature reaches 60°C, press pause and remove the assay plate. Put on ice for 2 min, and then spin down. Add 30 mL universal amplification mix to each well. Tightly seal the plate, vortex, spin down, load on a thermal cycler, and resume DMET Plus Assay program (Fig. 4c) to allow 23 cycle amplification. 10. After the program has finished, transfer the sealed assay plate to the post-amplification lab. All the following steps–clean-up, fragmentation, labeling, hybridization, and array processing are performed in the postamplification lab 3.6. Clean-up

During this stage, PCR products are enzymatically cleaned up, and upon completion, the first quality control gel can be run to ensure the presence of the correct size PCR product (see Note 4 [Fig. 5a]).

Fig. 5. Agarose gel examples of the first (PCR product) and second (fragmentation) quality control (QC) gels. (a). First QC gel with 120–130 bp PCR products. At times, a ~75 bp second band of lower intensity is visible. Well 23 illustrates poor PCR results: lower intensity of higher molecular weight product band indicates a reduced amount of product of interset. Arrows correspond to 150, 100, and 75 bp molecular weight marker bands (top to bottom). (b) Fragmentation QC gel for the samples shown in panel A. The majority of generated fragments are in the range of 25–75 bp. Arrows indicate 75, 50, and 25 bp bands (top to bottom)

112

Burmester et al.

Fig. 6. Temperature conditions for post-amplification lab steps (a) clean-up program, (b) fragmentation program, (c) label program, and (d) denature program

Input required: assay plate 1. Spin down PCR clean-up mix tube. Aliquot 15 mL to each tube of one strip of 12 tubes. Add 2.5 mL PCR clean-up mix to each reaction on the assay plate. 2. Tightly seal the plate, vortex, spin down, and place on a thermal cycler, and run the following DMET Plus Clean-up Program (Fig. 6a). 3.7. Fragmentation

During this stage, PCR products are fragmented to improve sample hybridization onto DMET Plus Arrays. DNA fragment size can be checked on the second quality control gel (see Note 5, (Fig. 5b)). Input required: assay plate processed with PCR clean-up mix 1. Thaw postamplification water and fragmentation buffer. Once thawed, place on ice. 2. Create a new fragmentation/labeling plate by transferring 25 mL of each reaction from the assay plate to the corresponding wells on the new plate.

DMET™ Microarray Technology for Pharmacogenomics-Based Personalized Medicine

113

3. Combine 536 mL of postamplification water and 60 mL of fragmentation buffer into the 1.5 mL Eppendorf tube. Mix and cool the tube on ice. Add 4.1 mL of fragmentation enzyme, vortex, spin down, and keep on ice. Aliquot 45 mL of fragmentation mix to each tube of one strip of 12 tubes. 4. Add 10 mL of fragmentation master mix to each reaction. Tightly seal the plate, vortex, and then spin down and place on a thermal cycler, and run the DMET Plus Fragmentation Program (Fig. 6b). 3.8. Labeling

During this stage, fragmented material is end-labeled with biotin for post-hybridization detection with streptavidin, R-phycoerythrin conjugate (SAPE). Labeling is performed in the same plate (fragmentation/labeling) as fragmentation. According to the protocol, 10 mL of fragmented material should be removed from each well for the second quality control gel before proceeding with the labeling reaction. DO NOT FORGET to take out 10 mL from each well to bring the volume down to 25 mL, even if no quality control gel is run. Input required: fragmentation/labeling plate containing 25 mL fragmented material (10 mL removed from each well for quality control fragmentation gel) 1. Thaw DNA labeling reagent and 5× TdT buffer. Once thawed, keep all the reagents on ice until ready to use. 2. Prepare labeling master mix in 1.5  mL Eppendorf tube according to Table 5. 3. Vortex, spin, aliquot 45 mL of labeling master mix to each strip tube well to a strip of 12 tubes. 4. Add 10 mL labeling master mix to each reaction. Tightly seal the plate, vortex, and then spin down, place on a thermal cycler, and run the following DMET Plus Label Program (Fig. 6c).

Table 5 End-labeling master mix Reagent

1 Reaction

48 Reactions (>20% extra)

Post-amplification water

  0.4 mL

  24 mL

5× TdT buffer

  7 mL

420 mL

DNA labeling reagent

  0.9 mL

  54 mL

TdT enzyme

  1.7 mL

102 mL

Total

10 mL

600 mL

114

Burmester et al.

3.9. Hybridization

During this stage, each reaction is denatured then loaded onto a DMET™ Plus Array – one sample per array. The arrays are then placed into a hybridization oven that has been preheated to 49°C. Samples are left to hybridize for 16–18 h. Input required: fragmentation/labeling plate 1. Turn each oven on and set the temperature to 49°C. Set the rotation speed to 35 rpm. Turn the rotation on and allow the oven to preheat. 2. Unwrap the arrays and place on the bench top with barcode side up. Allow the arrays to warm to room temperature (10–15  min). Be sure to equilibrate the arrays to room temperature; otherwise, the rubber septa may crack and the array may leak. Mark the arrays with a meaningful designation (e.g., a number or plate well name) to ensure that the identity of each sample loaded onto the array is known. 3. Scan the array barcodes into a sample batch registration file (for more information, see Note 6). Turn the arrays over onto a clean surface and insert a 200 mL pipet tip into the upper right septum (Fig. 7a). 4. To prepare the hybridization master mix, add 50 µL of oligonucleotide control reagent directly to the hybridization solution bottle, and mix well by inverting the tube ten times. Pour the hybridization master mix into a reagent reservoir and place on ice. 5. Wet the pipet tips by aspirating and dispensing the hybridization master mix three times. Aliquot 92 µL to the appropriate wells of the new hybridization plate.

Fig. 7. (a) Preparing the DMET™ Plus Array for loading hybridization material, (b) covering the array septa with large Tough-Spots that overlap the window after loading hybridization material prior to placing the arrays into the hybridization trays, (c) covering array septa with smaller Tough-Spots prior to scanning. Do not overlap the window at this step

DMET™ Microarray Technology for Pharmacogenomics-Based Personalized Medicine

115

6. Add 8 mL of each reaction from the fragmentation/label plate to the hybridization plate for the total volume of 100 mL. Tightly seal the hybridization plate, vortex, then spin down and load onto a thermal cycler, and run the DMET Plus Denature Program (Fig. 6d). 7. Once the samples have been denatured at 95°C for 10 min, remove the hybridization plate from the thermal cycler, put it in an aluminum block on ice for 2 min. Spin down. 8. Keeping the plate in an aluminum block, load 95 mL from one well into an array. Remove the pipet tip from the upper right septum. 9. Repeat step 8 for all 48 arrays. 10. Cover both septa with Tough-Spots as shown in Fig. 7b. Using large Tough-Spot that overlap the window makes it easier to remove them the next day before loading the arrays on the fluidic stations. Place the arrays into the trays, and load the trays into the hybridization oven in a properly balanced manner. 11. Seal and store fragmentation/labeling plate at −20°C. Place the remaining hybridization master mix into an Eppendorf tube and store at −20°C until all arrays have been successfully processed and analyzed. This mix can be used to set up additional hybridizations if needed. To prepare for washing and staining, move the stain buffer and hold buffer from –20°C to 4°C. 3.10. Washing, Staining, and Scanning

During this stage, arrays are washed and stained using Gene Chip ® Fluidics Station 450 Instruments and then scanned on the GeneChip® Scanner 3000 7G. The fluidics station and the scanner are controlled by Affymetrix GeneChip® Command Console® Software (AGCC). In addition, Command Console Software provides tools for registering the samples (see Note 6), and organizing data files. For more information on this application, refer to the Command Console Software User’s Guide. 1. Allow the arrays to incubate at 49°C and 35 rpm for 16–18 h before proceeding with the first round of array washing. 2. Prime the fluidic stations. Place wash solutions A and B into the designated positions. Fill the third container with dH2O. Run PRIME_450 protocol (see Note 7). 3. Prepare stain solution by adding 90 mL of SAPE to the stain buffer bottle. Invert five times to mix. Do not place on ice. 4. Remove 4N arrays (where N is a number of fluidic stations used) from the hybridization oven, remove the Tough-Spots from the arrays, and load the arrays on the fluidic stations. Leave remai­ ning arrays in the hybridization oven until ready to wash. Be sure to rebalance the trays every time any arrays are removed.

116

Burmester et al.

5. Place 1.5 mL amber tube with 300 mL SAPE stain solution into position 1 and clear tube with 300 mL of hold buffer into position 2. Keep an empty tube in position 3. 6. Run the DMET_Plus_169_v2 protocol. Follow the prompts displayed in the fluidics station window. When “load cartridge” is displayed, engage the wash block. It will lock the array into place. When “load vials 1 and 2” is displayed, lock the tubes in place. 7. It takes about 30 min to wash and stain the array. When “eject and inspect cartridge” is displayed, remove and inspect each array for bubbles. The display on the fluidic station will read “reload cartridge to de-bubble or engage wash block.” If no bubbles are visible, engage the wash block and the protocol will advance to the next step. If there are bubbles visible, place the array back on the fluidics station and engage the wash block. The array is drained and filled with hold buffer from vial 2. Repeat this process up to five times to remove all bubbles. 8. When “remove vials 1 and 2” is displayed, remove and discard the vials from positions 1 and 2. Leave the empty vial in position 3. 9. When “load clean vials” is displayed, load empty vials in positions 1 and 2 and resume the script. The cleaning part of the protocol is executed. 10. When “remove vials 1 and 2” is displayed, remove the vials from positions 1 and 2. “Protocol done” is displayed. If previously not completed, prepare a batch registration file containing sample information and array barcodes, and upload this information to Command Console Software prior to scanning the arrays. If the arrays were accidentally washed and scanned without first registering them, the sample file (. ARR files) will not include two of the attributes required by DMET™ Console: sample type and consented markers list. Before the .CEL files can be genotyped, these files must be manually edited to include the required information. See Note 6 for details. 11. To prepare arrays for scanning, place the arrays face down and carefully cover the septa with small Tough-Spots (Fig.  7c) without overlapping the window. 12. Inspect the windows for dust and lint, and clean if needed with Kimwipes. 13. Load the arrays onto scanner carousel, starting at position 1. 14. To scan the arrays, open Command Console Software. Click the Start icon. If the arrays in positions 1–4, at room temperature select the appropriate check box. If any arrays in the carousel are to be rescanned, select the check box “allow rescans.”

DMET™ Microarray Technology for Pharmacogenomics-Based Personalized Medicine

117

15. To process the remaining arrays, repeat the steps listed under “wash, stain and scan the arrays.” IMPORTANT: Load new tubes of stain and HOLD cocktail for each array. Leave the empty tube in position 3. After all the arrays are washed and stained, shut down the fluidics station. To do that, replace wash solution A and B with dH2O and run SHUTDOWN_450 protocol. 3.11. Data Analysis

1. DMET Plus Array images are processed using the DMET Console Software tool. Users interact with the data analysis tool in a stepwise manner through the main navigation pane. There is a file tree that walks through the analysis. Steps in this process include: 2. Defining a workspace and setting analysis options: DMET Console keeps track of the locations of files and algorithm parameters used in the analysis of the array data. These are stored in array library files. Work can be saved in interim stages, so that data may be reanalyzed when more arrays are processed. The two primary analysis files (Affymetrix supplied) include selection of the array annotation file and the gene translation file. The user also defines the directory, where these files are stored. 3. Importing scanned array data: A dialogue box is used for users to import array data. Before samples can be analyzed, two files (created in Command Console Software upon registering the samples and scanning the arrays) are required: a sample file (.ARR) and chip intensity file (.CEL). Arrays that have previously been analyzed may also be imported in this window. Analyzed data are stored in a .CHP file. A common suffix is used to group these three files together. 4. Analyzing array intensity data: Analysis of fluorescence intensity across the array, signal summation, and the mapping of alleles and markers in the DMET Plus Panel occur at this stage. Review of the array intensity information includes a tabular display of user defined parameters for each array. These may include sample name, source plate and well comments, sample type, consented marker list, condition, protocol version, individual reagent lot, operator, facility, and other laboratory parameters that may be of interest for quality control such as tracking the oven, fluidics station, and scanner. These data are read from the .ARR file. 5. Perform genotyping: Once users have verified that the DMET Plus Array data is complete, they analyze each sample’s genotype for the SNP alleles using the perform genotyping command at the intensity menu. All selected arrays are analyzed and segregated into three results groups: all arrays; in bounds (³ 98% call rate across the full DMET Plus Array);

118

Burmester et al.

out of bounds (<98% genotyping calls). A CHP summary report is generated and additional review of data can be made including display and export of marker summaries and copy number reports. The marker call rate across the full assay panel is also provided as a convenient quality control metric. Arrays with <98% marker call rate are flagged and as an option can be removed from further analysis by the user. 6. Concordance check is one of the analytical steps that can be performed on genotyped data. Genotypes of the plasmid control pools (representing all three genotypes in ~273 core markers in the DMET Plus Panel) and genomic DNAs provided with the DMET Plus Assay kits are stored in DMET Console. These data may be used to verify genotyping accuracy on customer runs. 7. Genotyping results may be exported for downstream analysis such as association studies or statistical correlation with sample values. The software permits different file formats to be exported. Extended genotyping results include fluorescent signals for the allele probes along with DMET Console calls. Users may create custom marker subsets (restricting analysis and export to genes of interest) using the same dialogue box. Copy variation at five of the genes is also computed (CYP2A6, CYP2D6, GSTM1, GSTT1, and UGT2B17) and results may be saved to tab delimited reports. 8. Allele translation provides interpretation of genotypes across a core group of ADME genes in the DMET Plus Panel. This includes identification of haplotype-based allele calls such as the star-allele nomenclature across the CYP450 genes (21). In the current database (DMET Plus v1), a total of 746 sequence-based markers and five copy number markers are thus translated with refined biological annotations including:  (a) Reference link: PubMed or sequence reference for the alleles (b) Probe set ID: Affymetrix probe ID  (c) dbSNP RS ID: Reference ID for the marker in NCBI dbSNP database (d) Defining: Notation of the star-allele or other change for which the marker is a defining SNP (e) cDNA nucleotide position: Location of the marker/SNP in a reference cDNA   (f) Genome position: Genomic location in HG18 database build (in format Ch7:87067437) (g) Change: Notation of codon change, or location in gene (splice site, 5¢UTR, etc)

DMET™ Microarray Technology for Pharmacogenomics-Based Personalized Medicine

119

(h) Common name: Marker name also containing the star alleles, where it was first characterized (i) Haplotype: A flag whether the marker is used in haplotypebased star allele calling (j) Reference and variant: The alternative alleles in the reference genome defining the alleles 9. Specialized translated reporting: DMET Plus allele translation produces three analysis reports and a log tracking all analysis parameters used in the data conversion. The analytical reports include: comprehensive report – all marker data output including the biological annotations listed in step 8 of Subheading 3.11 and column interpreting whether the sample has reference or variant alleles; summary report – a rollup restricted to functional variants at each of the analyzed genes. Genetic evidence that alterations from the reference allele (generally named *1 for most CYP450 genes) is provided. Neutral SNPs, when in the homozygous state are reported. “All markers responsible for known functional changes are Ref/Ref.” The third translation report is an uncalled data report. This file tracks the markers that may have uncalled data or variant alleles not observed across ~1,200 independent DNA samples. They are annotated “possible rare allele.” The uncalled report is essentially a work list for review of the specific genotypes missing for individual samples (generally <1% in Affymetrix development studies, unpublished results). It is recommended that signal plots be reviewed for the markers although many users prefer to obtain independent genotyping results for selected important markers in the panel. 3.12. Use Cases

One example of the use of the DMET™ product for applied medical research and gene discovery was a recent study which used the first generation (four-color) version of the DMET Assay (Targeted Human DMET 1.0) to analyze the effects of SNPs on warfarin doses. This led to the observation that CYP4F2*3 (rs2108622) is a SNP that alters warfarin dosing (17). Warfarin is a drug commonly used to prevent blood clots and acts by blocking VKORC1, the enzyme that catalyzes the reduction of vitamin K epoxide to the enzymatically active form. Due to the large variability in the extent how each individual patient responds to warfarin, additional personal factors such as genetics, age, and body size are accounted for when determining appropriate dosages. To adjust for multiple testing, a Bonferroni correction was used to determine a valid statistical significance threshold (P < 0.05/1228 = 0.00004). A multiple regression model was developed with data from the Marshfield pharmacogenetic cohort. Fig. 8a shows that the SNP rs2108622, which resides in CYP4F2, is statistically significant. This polymorphism alters warfarin dose

120

Burmester et al.

Fig.  8. Association of CYP4F2*3 variant (rs2108622) to warfarin dosing requirements. (a) SNPs tested on the DMET Product and their P-value for impact on warfarin dose. The dashed line indicates the P-value that is significant after Bonferroni correction; (b) effect of rs2108622 genotype on therapeutic dose in three cohorts

by about 0.5  mg/allele/day. This study concluded that stable warfarin dose requirements are affected by and associated with polymorphisms of CYP4F2. Fig.  8b shows that the effect of rs2108622 on warfarin dose is consistent in two additional cohorts. The observed effect of rs2108622 on warfarin dose has since been validated in three additional cohorts (22–26). A second example of the use of Targeted Human DMET 1.0 was the observation that carriers of a CYP2C19 reduced-function allele had a reduced response to clopidogrel and a greater rate of ischemic events than noncarriers (27). DNA samples from 162 healthy subjects treated with clopidogrel were analyzed using the DMET product, and then compared to genetic variants among 1,477 subjects with acute coronary syndromes that had also been treated with clopidogrel. Plasma concentrations of the active metabolite of clopidogrel, as well as maximal platelet aggregation (DMPA), were assessed. The likelihood-ratio, Gehan-Wilcoxon, and log-rank tests were used, and the rates of outcomes were expressed as Kaplan–Meier cumulative estimates. Statistically significant results indicated that carriers of at least one CYP2C19 reduced-function allele had a 32.4% reduction of plasma exposure to the active metabolite in clopidogrel (P < 0.001), as well as a 25% reduction in DMPA in response to clopidogrel (P < 0.001), as compared with noncarriers. The primary efficacy outcome rate was 12.1% for carriers and 8.0% for noncarriers who had been treated with clopidogrel (hazard ratio for carriers, 1.53; 95%

DMET™ Microarray Technology for Pharmacogenomics-Based Personalized Medicine

121

confidence interval, 1.07–2.19; P = 0.01). None of the other genes encoding CYP enzymes indicated a significant association between genetic variants and cardiovascular outcomes. This study shows that the DMET product can be used to identify genetic variations in genes that have an effect on both pharmacologic and clinical responses to clopidogrel.

4. Notes 1. Use of DMET Plus Control Samples: The DMET Plus Panel Kit includes three genomic DNA controls (genomic controls 1, 2, and 3) and three plasmid DNA controls (plasmid controls A, B, and C). The genomic DNA controls provide a source of high quality genomic DNA that can be included in each DMET Plus Assay run to assess assay performance, as well as provide positive assay controls when processing DNA of marginal quality in terms of the level of purity, accuracy of concentration, or extent of degradation. The genomic control samples are processed in the same manner as all genomic DNA samples and should be added to the GP1 plate as described in subheading 3.2. The plasmid DNA controls are composed of target sequences for a set of 242 specific rare markers and serve to demonstrate the capability of a DMET Plus Assay run to detect signals from rare alleles that may not be observed among the specific set of genomic samples processed. Detecting different genotypes for a plasmid marker in a run reduces the uncertainty that there may be some undetected rare genotypes among the genomic samples. The DMET Plus Plasmid control samples do not require the addition of mPCR material. To include these samples in a DMET Plus Assay run, add 18.4 mL of plasmid control solutions A, B, and C to 21.7 mL of anneal master mix in wells C12, D12, and E12 of the anneal plate, respectively. During the addition of mPCR material to anneal reactions containing genomic DNA samples, please ensure that no additional volume is added to the plasmid DNA wells. After all additions have been made, the total volume of all anneal reactions should be 40.1 mL. 2. Handling the plates and plate additions: Keep the plates in an aluminum block on ice when not running the thermal cycler program. Use multichannel pipettes to perform reagent additions in the plate format for consistent and uniform results. After each addition, pipet up and down three times to rinse tips. For mixing, tightly seal the plate, vortex at high speed for 3 s, then spin down at 2,000 rpm (685 g) for 30–60 s.

122

Burmester et al.

3. Working with enzymes: Keep enzymes at −20°C until ready to use. Add them last and add them right before the master mix is used. Always spin the tubes down before use, so that the contents are uniform. Enzymes are in 50% glycerol. Pipet slowly to ensure the correct volume and to avoid bubbles. 4. Running quality control gel 1: The first quality control gel is used to assess MIP PCR products. After completion of the DMET Plus Clean-up Program, remove the assay plate from the thermal cycler, spin down and place in an aluminum block on ice. For each sample, combine 2 mL of cleaned up PCR material with 8 mL of 1× TE or water and 2 mL 2× loading buffer (Sigma, P/N G2526). Mix well and load 10 mL onto a 3% agarose gel (Bio-Rad Precast ReadyAgarose™ Wide-Mini Gel, P/N 161-3040). In order to assess MIP PCR product size, load at least one lane of low molecular weight ladder (New England Biolabs, P/N N3233S). Run the gel at 120V for 20 min. Examine the gel to ensure that PCR product is present for all samples and that the size of each is between 120 and 130 base pairs. See Fig. 5a for an example of a typical MIP PCR product gel. 5. Running quality control gel 2: The second quality control gel is used to assess fragmentation of MIP PCR product. After completion of the DMET fragmentation program, remove the fragmentation/label plate from the thermal cycler, spin down and place in an aluminum block on ice. For each sample, combine 10 mL of fragmented material with 2 mL 2× loading buffer (Sigma, P/N G2526). Mix well and load 10 mL onto a 3% agarose gel (Bio-Rad Precast ReadyAgarose™ Wide-Mini Gel, P/N 161-3040). In order to assess MIP PCR fragment size, load at least one lane of low molecular weight ladder (New England Biolabs, P/N N3233S). Run the gel at 120V for 24  min. Examine the gel to ensure that fragmented PCR product is present for all samples and that the fragment size of each is <120 base pairs with the majority of fragmented material centered around 50  bp. See Fig.  5b for an example of a typical MIP PCR fragmentation gel. 6. Sample registration for scanning and data analysis: Sample registration process results in creating sample files (.ARR files) using AGCC portal component of Command Console Software. The process consists of generating batch registration file and uploading this file to the software. To generate a sample batch registration file, open AGCC portal from Launcher menu of Command Console Software. Hold the cursor over samples tab and select batch register from the drop-down menu. Select a DMET template included with DMET Console, the file type (TSV or Excel), the number of samples

DMET™ Microarray Technology for Pharmacogenomics-Based Personalized Medicine

123

to be bar coded, a project name, and the array type from drop-down menus. Click download. Two attributes, required by DMET Console for genotyping, are sample type and consented marker list. Click in empty cell to open drop-down menu and select appropriate sample type (sample, control, or plasmid) and consented marker list (DMET_Plus All, DMET_Plus_Verified, DMET_Plus Plasmids). DMET_Plus_ All list is recommended. Select a more restricted marker list only if you are certain that you will never want to access results from the markers that are excluded. In addition, enter sample name, sample file name (name that Command Console Software will assign to the .ARR file), array name (name that Command Console Software will assign to image [DAT] file and intensity [CEL] data file). Usually, it is the same as sample file name, and any other tracking information (reagent lot, operator, date, sample location on the plate). Scan the array barcodes and save the file. To upload the batch registration file to Command Console Software, open batch register window from AGCC Portal, click browse and navigate to and open the batch registration file. Click upload, and then save. The message “batch array registration is complete” is displayed. IMPORTANT: If save is not clicked, then the information is not uploaded. 7. Control of the fluid station and scanner using the AGCC: Start AGCC software. In the Launcher window, double-click AGCC fluidics control. In AGCC fluidics control window, open the protocol drop-down menu and select the protocol to run (PRIME_450, DMET_Plus_169_v2, or SHUTDOWN_450). Select the check/uncheck all stations and modules check box. Deselect any fluidics stations or modules that will not be used. Click copy to selected modules button. Click the run all button on the top of the window. Follow the prompts displayed in the fluidics station window or in the current stage column (lower half of the control window). For example, when running PRIME_450 protocol, you will be prompted to load three empty vials into positions 1, 2, and 3. Once the vials are loaded and locked in place, priming will begin. Progress is also displayed in the current stage and time/cycle columns in the lower half of the control window.

Acknowledgments The authors thank Marshfield Clinic Research Foundation for its support through the assistance of Amy VanProosdy and Alice Stargardt in the preparation of this chapter.

124

Burmester et al.

References 1. Thacker SM, Grice GR, Milligan PE, Gage BF (2008) Dosing anticoagulant therapy with coumarin drugs: is genotyping clinically useful? Yes. J Thromb Haemost 6:1445–1449 2. Huang RS, Ratain MJ (2009) Pharmaco­ genetics and pharmacogenomics of anticancer agents. CA Cancer J Clin 59:42–55 3. Singh SS (2006) Preclinical pharmacokinetics: an approach towards safer and efficacious drugs. Curr Drug Metab 7:165–182 4. Yoshizawa M, Hayashi H, Tashiro Y et  al (2009) Effect of VKORC1–1639 G>A polymorphism, body weight, age, and serum albumin alterations on warfarin response in Japanese patients. Thromb Res 124:161–166 5. Kang JS, Lee MH (2009) Overview of therapeutic drug monitoring. Korean J Intern Med 24:1–10 6. Hamdy SI, Hiratsuka M, Narahara K et  al (2002) Allele and genotype frequencies of polymorphic cytochromes P450 (CYP2C9, CYP2C19, CYP2E1) and dihydropyrimidine dehydrogenase (DPYD) in the Egyptian population. Br J Clin Pharmacol 53:596–603 7. Mizutani T (2003) PM frequencies of major CYPs in Asians and Caucasians. Drug Metab Rev 35:99–106 8. Nakai K, Tsuboi J, Okabayashi H et al (2007) Ethnic differences in the VKORC1 gene polymorphism and an association with warfarin dosage requirements in cardiovascular surgery patients. Pharmacogenomics 8:713–719 9. Wilke RA, Lin DW, Roden DM et al (2007) Identifying genetic risk factors for serious adverse drug reactions: current progress and challenges. Nat Rev Drug Discov 6:904–916 10. Flockhart DA, Tanus-Santos JE (2002) Implications of cytochrome P450 interactions when prescribing medication for hypertension. Arch Intern Med 162:405–412 11. Bosch TM, Meijerman I, Beijnen JH, Schellens JH (2006) Genetic polymorphisms of drugmetabolising enzymes and drug transporters in the chemotherapeutic treatment of cancer. Clin Pharmacokinet 45:253–285 12. Lacaná E, Amur S, Mummanneni P, Zhao H, Frueh FW (2007) The emerging role of pharmacogenomics in biologics. Clin Pharmacol Ther 82:466–471 13. Frueh FW, Gurwitz D (2004) From pharmacogenetics to personalized medicine: a vital need for educating health professionals and the community. Pharmacogenomics 5: 571–579 14. Frueh FW, Amur S, Mummaneni P et al (2008) Pharmacogenomic biomarker information in drug labels approved by the United States

food and drug administration: prevalence of related drug use. Pharmacotherapy 28:992–998 15. Daly TM, Dumaual CM, Miao X et al (2007) Multiplex assay for comprehensive genotyping of genes involved in drug metabolism, excretion, and transport. Clin Chem 53:1222–1230 16. Dumaual C, Miao X, Daly TM et  al (2007) Comprehensive assessment of metabolic enzyme and transporter genes using the Affymetrix Targeted Genotyping System. Pharmacogenomics 8:293–305 17. Caldwell MD, Awad T, Johnson JA et  al (2008) CYP4F2 genetic variant alters required warfarin dose. Blood 111:4106–4112 18. Brandt JT, Close SL, Iturria SJ et  al (2007) Common polymorphisms of CYP2C19 and CYP2C9 affect the pharmacokinetic and pharmacodynamic response to clopidogrel but not prasugrel. J Thromb Haemost 5:2429–2436 19. Hardenbol P, Banér J, Jain M et  al (2003) Multiplexed genotyping with sequence-tagged molecular inversion probes. Nat Biotechnol 21:673–678 20. Hardenbol P, Yu F, Belmont J et al (2005) Highly multiplexed molecular inversion probe genotyping: over 10, 000 targeted SNPs geno­typed in a single tube assay. Genome Res 15:269–275 21. Robarge JD, Li L, Desta Z, Nguyen A, Flockhart DA (2007) The star-allele nomenclature: retooling for translational genomics. Clin Pharmacol Ther 82:244–248 22. Borgiani P, Ciccacci C, Forte V et  al (2009) CYP4F2 genetic variant (rs2108622) significantly contributes to warfarin dosing variability in the Italian population. Pharmacogenomics 10:261–266 23. Cooper GM, Johnson JA, Langaee TY et  al (2008) A genome-wide scan for common genetic variants with a large influence on warfarin maintenance dose. Blood 112:1022–1027 24. Perez-Andreu V, Roldan V, Anton AI et  al (2009) Pharmacogenetic relevance of CYP4F2 V433M polymorphism on acenocoumarol therapy. Blood 113:4977–4979 25. McDonald MG, Rieder MJ, Nakano M, Hsia CH, Rettie AE (2009) CYP4F2 is a vitamin K1 oxidase: an explanation for altered warfarin dose in carriers of the V433M variant. Mol Pharmacol 75:1337–1346 26. Takeuchi F, McGinnis R, Bourgeois S et  al (2009) A genome-wide association study confirms VKORC1, CYP2C9, and CYP4F2 as principal genetic determinants of warfarin dose. PLoS Genet 5:e1000433 27. Mega JL, Close SL, Wiviott SD et  al (2009) Cytochrome p-450 polymorphisms and response to clopidogrel. N Engl J Med 360:354–362

Chapter 8 The Use of Microarray Technology for Cytogenetics Bassem A. Bejjani, Lisa G. Shaffer, and Blake C. Ballif Abstract The use of microarray technology is revolutionizing the field of clinical cytogenetics. This new technology has transformed the cytogenetics laboratory by adapting techniques that have heretofore been the province of molecular geneticists. Intimate knowledge and comfortable familiarity with these techniques are now a must for the modern cytogeneticist, rather than a stimulating but discretionary intellectual exercise or an elective luxury. The cytogenetic laboratory of the future will likely have more scanners than microscopes, more software packages than darkrooms, and more technologists, supervisors, and directors with molecular training than ever before. This technical convergence between molecular diagnostics and clinical cytogenetics is exciting and has already resulted in many stimulating discoveries. However, the traditional skills of the cytogeneticist are needed now more than ever before. As our ability to inspect the genome increases, so does the variety of abnormalities that we uncover. Understanding the mechanisms of these aberrations to guide additional testing of the parents and genetic counseling of the patients and their families requires the expertise of individuals who are well-versed in meiotic mechanisms and chromosomal structures that may lead to these abnormalities. Cytogeneticists are uniquely positioned to understand these mechanisms and assist genetic counselors and clinicians in their daily interactions with patients and families. Key words:  Microarray, Comparative genomic hybridization, Array CGH, Oligonucleotide, Molecular cytogenetics

1. Introduction The use of microarray technology is rapidly becoming an essential part of the diagnostic cytogenetic laboratory. DNA microarrays are solid surfaces on which nucleic acids are immobilized (spotted, lithographed, or synthesized in situ) and used as targets for hybridization. These nucleic acid molecules can be large insert clones (such as bacterial artificial chromosomes (BACs) or P1-derived artificial chromosomes (PACs)) or short (25–60 bp) oligonucleotides. In typical microarrays, they are attached to the Sridar V. Chittur (ed.), Microarray Methods for Drug Discovery, Methods in Molecular Biology, vol. 632, DOI 10.1007/978-1-60761-663-4_8, © Humana Press, a part of Springer Science+Business Media, LLC 2010

125

126

Bejjani, Shaffer, and Ballif

solid surface by a covalent bond through a chemical matrix (e.g., amino-silane). Most solid supports to which DNA is attached are either glass or silica, although beads are also used. The hybridization to these targets can be performed through the process of comparative genomic hybridization (CGH), or through non-CGH methods. In essence, they allow multiplex testing of many targets across the genome. The essential elements of these hybridization methods have been extensively reviewed (1–4). This chapter details the materials and methods used in oligonucleotide-based comparative genomic hybridization using microarrays (array CGH) in our laboratory. Some of the fundamental concepts of array CGH methodologies described here may apply to other platforms with minimal changes. Other methodo­ logies are specific to the type of array used, or to the manufacturer and the specific application for which the array is being utilized. Regardless of the platform used, it is clear that microarrays are now solidly rooted in the clinical activities of cytogenetic laboratories and will remain a part of this discipline for years to come.

2. Materials 2.1. General

1. 95% ethanol. 2. Isopropyl alcohol (isopropanol) 1 L. 3. Acetic acid, glacial, 2.5 L. 4. Tris base, 1 kg (Fisher). 5. EDTA (ethylenediamine tetra acetic acid) 1 kg.

2.2. DNA Purification

1. 5 PRIME ArchivePure DNA Blood Kit (120 ml) (Fisher). 2. Red blood cell (RBC) lysis solution, 900 ml (Fisher). 3. Cell lysis solution, 300 ml (Fisher). 4. RNase A solution, 1.5 ml (Fisher). 5. Protein precipitation solution, 100 ml (Fisher). 6. DNA hydration solution, 50 ml (Fisher). 7. Glycogen solution, 20 mg/ml, 500 ml.

2.3. Digestion of Genomic DNA (If Required)

1. DpnII Restriction Enzyme (5,000  units) (New England Biochemical, Ipswich, MA). 2. Gel loading solution, 6× concentrate, 5  ml (Sigma-Aldrich, St. Louis, MO). 3. 1 kb DNA Ladder, 1 mg (Invitrogen, Carlsbad, CA).

The Use of Microarray Technology for Cytogenetics

2.4. DNA Washing and Hybridization

127

1. BioPrime Total Labeling Kit (Invitrogen). 2. Human Cot-I DNA 1 mg (Invitrogen). 3. 10× blocking reagent. 4. 2× hybridization buffer. 5. Ultrahyb hybridization buffer (Ambion, Austin, TX).

2.5. Washing

1. PBS (Invitrogen). 2. Oligo aCGH/ChIp-on-Chip Wash Buffer Kit (Agilent).

3. Methods 3.1. Genomic DNA Purification 3.1.1. DNA Extraction from Whole Blood 3.1.2. Cell Lysis

Purpose : To purify genomic DNA from peripheral whole blood using the 5 PRIME ArchivePure DNA Blood Kit. The purified genomic DNA will be used in the array CGH experiment (see Note 1). Other DNA purification methods may be used but could impact the final microarray results. 1. Aliquot 900  ml RBC lysis solution to three sterile 1.5  ml microcentrifuge tubes 2. Using a syringe and needle, draw 900 ml of whole blood into the syringe and add 300 ml to each tube containing 900 ml RBC lysis solution. Invert tubes to mix (invert and tap ends of two tubes together for uniform mixing) 3. Incubate 10 min at room temperature, inverting again at least once during the incubation 4. Centrifuge 13,000–16,000 × g for 30 s to pellet cells 5. Remove supernatant, leaving behind the visible white pellet and about 10–20  ml of liquid. Vortex tubes vigorously to resuspend cells (see Note 2) 6. Add 300 ml cell lysis solution and vortex or mix by pipette to lyse the white blood cells 7. Incubate samples for 5  min at 37°C to ensure sample resuspension (if cells clump then incubate at 37°C until the solution is homogenous). Samples are stable in cell lysis solution at room temperature for at least 18 months

3.1.3. RNase Treatment

1. Add 1.5 ml RNase A solution to the cell lysate 2. Mix samples by inverting tubes 25 times 3. Incubate at 37°C for a minimum of 5 min (incubation should not be longer than 60 min because DNA may degrade)

128

Bejjani, Shaffer, and Ballif

3.1.4. Protein Precipitation

See Note 3 1. Cool the samples by placing on ice for 8 min (see Note 4) 2. Add 100  ml protein precipitation solution to the RNase A-treated cell lysate 3. Vortex two tubes vigorously at high speed for 25 s to mix the protein precipitation solution uniformly with the cell lysate (required to completely remove contaminants) 4. Centrifuge at 13,000–16,000 × g for 3 min (see Note 5)

3.1.5. DNA Precipitation

1. For each tube, transfer the supernatant (containing the DNA) into a sterile 1.5 ml microcentrifuge tube containing 300 ml of room temperature 100% isopropanol (see Note 6) 2. Mix sample by gently inverting 50 times (see Note 7) 3. Centrifuge at 13,000–16,000 × g for 1 min. DNA should be visible as a small white pellet 4. Carefully aspirate any liquid from the pellet 5. Add 300 ml of 70% ethanol at room temperature and invert the tubes several times to wash the DNA pellet 6. Centrifuge at 13,000–16,000 × g for 1 min 7. Carefully aspirate any liquid from the pellet 8. Re-cap samples and pulse centrifuge. Use a 10  ml pipet to remove any additional supernatant 9. Let the pellet air dry for 3–5 min

3.1.6. DNA Hydration

1. Add 50 ml DNA hydration solution to each tube 2. Rehydrate DNA by incubating at 37°C for 60 min 3. Vortex the tubes and centrifuge for 2–3 s to collect 4. Working with one test sample at a time, combine the suspension of all three tubes into one tube 5. Store DNA at 4°C. For long-term storage, store at –20°C or –80°C

3.1.7. DNA Quantification

Purpose : To accurately determine the concentration of genomic DNA using the NanoDrop spectrophotometer. Other DNA quantification techniques may be used as long as both the test sample DNA and control DNA are quantified in the same manner. 1. Ensure all test and control DNA samples have been vortexed and centrifuged 2. Open the NanoDrop software program 3. To measure the concentration of genomic DNA, click the “Nucleic Acid” button

The Use of Microarray Technology for Cytogenetics

129

4. The software will request that the instrument be initialized. With the sampling arm open, ensure that the sample pede­s tals are dry (using a Kimwipe), and pipet 1.5  ml of sterile Milli-Q water onto the lower measurement pedestal (see Note 8) 5. Close the sampling arm and click “OK” 6. Next, a blank reading is required. Repeat step 4, and click the “Blank” button. Now the instrument is ready to use for DNA quantifications 7. Control Sample: Before measuring test samples, a control sample should be measured to ensure that the instrument is functioning normally. Our laboratory recommends using a purified genomic DNA sample normalized to 100 ng/ml as a control 8. Type “Control” in the “Sample” text-box located on the right-side of the screen 9. With the sampling arm open, ensure that the sample pedestals are dry using a Kimwipe, and pipette 1.5  ml of the DNA sample onto the lower measurement pedestal 10. Close the sampling arm and initiate a spectral measurement by clicking the “Measure” button. The sample column is automatically drawn between the upper and lower measurement pedestals and the measurement made 11. When the measurement is complete, open the sampling arm and wipe the sample from both the upper and lower pedestals using a Kimwipe 12. Repeat steps 9–11 to take a duplicate control measurement 13. Type the identification of the first test sample to be measured in the “Sample” text-box 14. Repeat steps 9–11 to measure the first sample 15. Samples need only be measured in single, not replicates, using the NanoDrop system, unless replicate measurements are warranted in special cases 16. Repeat steps 13 and 14 for the remainder of the test samples to be quantified 17. Clean the measurement pedestals upon completion of a testing series. Apply 5  ml of sterile Milli-Q water onto the bottom pedestal, and lower the upper pedestal to form a liquid column. Let sit for 2–3 min 18. Wipe the water from both pedestals using a Kimwipe, and close the sampling arm Optional: If connected to a printer, click “Print report” to print the data

130

Bejjani, Shaffer, and Ballif

3.1.8. Sonication of Genomic DNA (Recommended Method)

Purpose: To fragment genomic DNA into 500  bp–4  kb fragments to allow for more efficient labeling. Other DNA fragmentation methods may be used; if restriction enzymes are used, the DNA samples need to be purified prior to labeling (see protocol for “Digestion of Genomic DNA,” Subheading 3.1.4). See Note 9. 1. Prepare a clear locking tube with 3 mg of DNA in 50 ml of sterile Milli-Q water, or 6  mg of DNA in 100  ml of sterile Milli-Q water 2. Place the microcentrifuge tubes containing the DNA samples in the microtube holder (supplied), so the tubes are resting in the holes (do not press the tubes all the way into the holes) 3. The microtube holder has a central ring that can hold eight tubes and an outer ring that can hold 12 tubes. Place tubes in the central ring first and move to the outer ring when full. Space samples evenly when a ring is not full 4. Place the microtube holder into the center of the sonicator microplate horn 5. Ensure that the microplate horn is filled with distilled water, so the water is just touching the rim of the microtube holder. Close the lid on the sonicator cabinet 6. Switch the sonicator on and select the standard horn 7. Sonication is performed with an output of 8; however, the duration is dependent on the number of tubes in the rack. If there are eight tubes or less, sonicate for 1 min 30 s, with a 30 s pulse every 30 s. If there are nine tubes or more, sonicate for 3 min, with a 30 s pulse every 30 s. Important step: While the sonicator is resting for 30 s, press “Pause” and open the cabinet lid. Rotate the microtube holder a ¼ turn and press “Pause” to restart the sonication. Rotating the microtube holder prevents variability in the sonication of samples due to holder location 8. After sonication is complete, remove the tubes from the holder and place tubes on ice 9. Check sonication quality by loading 8  ml of the sonicated DNA and 2 ml of loading dye on a 1% agarose gel. In addition to running the test samples, run 10 ml of a 1 kb DNA ladder for comparison. Any controls should also be run on the gel to verify that the size and intensity of the samples are comparable. The sonicated DNA should appear as a lower smear (500 bp–4 kb) on the gel 10. If the sample is not sufficiently sonicated (DNA is greater than 4 kb), resonicate for 30 s and run the sample on a gel.

The Use of Microarray Technology for Cytogenetics

131

If the sample is too small in size (DNA is less than 500 bp), prepare another tube of DNA and sonicate as described above 11. After storage at 4°C overnight, the sonicated DNA is ready for labeling 3.1.9. Digestion of Genomic DNA (Alternative to Sonication)

Purpose : To fragment genomic DNA into 500 bp–4 kb fragments to allow for more efficient labeling. 1. The volumes used in a restriction digest reaction depend of the starting concentration (and volume) of DNA. Determine the quantity of DNA to digest, and add the following into a sterile 1.5 ml microcentrifuge tube: Reaction size

3.1 mg

6.25 mg

12.5 mg

A. D  NA (calculate based on quantification)

X ml

X ml

X ml

B. Sterile Milli-Q water

Y ml

Y ml

Y ml

Total (A + B)

66.75 ml

133.5 ml

267 ml

C. 10× buffer

7.5 ml

15 ml

30 ml

D. DpnII enzyme

0.75 ml

1.5 ml

3 ml

Total volume

75 ml

150 ml

300 ml

2. After addition of enzyme and DNA, mix briefly by vortexing and re-collect samples by brief centrifugation 3. Incubate samples for a minimum of 30 min in a 37°C water bath 4. Check the digestion by loading 8  ml from the reaction mix (4 ml if performing a 3.1 mg digestion) and 2 ml of loading dye on a 1% agarose gel. In addition to running the test samples, run 10 ml of a 1 kb DNA ladder for comparison. Any controls should also be run on the gel to verify that the size and intensity of the samples are comparable. The digested DNA should appear as a lower smear (500 bp–4 kb) on the gel. If the sample has not digested completely, add 1 ml DpnII enzyme and allow more time for the sample to digest. Rerun the sample on the gel to determine the size 5. Precipitate the DNA once with an equal volume of phenol/ chloroform. Vortex to mix and let stand at room temperature for 7 min 6. Centrifuge for 7 min at 9,740 × g 7. Transfer the upper phase to a labeled sterile 1.5 ml microcentrifuge tube and discard the lower phase 8. Add an equal volume of chloroform. Vortex to mix and let stand at room temperature for 7 min

132

Bejjani, Shaffer, and Ballif

9. Centrifuge for 7 min at 9,740 × g. Transfer the upper phase to a labeled sterile 1.5 ml microcentrifuge tube and discard the lower phase 10. Precipitate the DNA by adding 12.5% (1/8) of the total volume of 3 M sodium acetate (pH 5.2) and 110% the total volume of isopropanol. Invert tubes 25 times to ensure that the samples are mixed 11. Place the tubes at −80°C for 15 min 12. Centrifuge for 20 min at 9,740 × g to pellet the DNA 13. Remove supernatant by pipette and discard 14. Wash once with 400 ml of 70% ethanol at room temperature . Centrifuge for 7 min at 9,740 × g 15. Remove supernatant by pipette and discard. Allow the pellet to air dry on the bench 16. Resuspend the pellet in (see Note 10) ●●

5 ml of sterile Milli-Q water for a 3.1 mg digestion

●●

10 ml of sterile Milli-Q water for a 6.25 mg digestion

●●

20 ml of sterile Milli-Q water for a 12.5 mg digestion

17. Seal tube with parafilm and resuspend for 30 min to overnight in a 37°C waterbath 3.2. Microarray Labeling and Hybridization 3.2.1. DNA Labeling (Using the BioPrime Total Labeling Kit)

Purpose: By labeling test sample DNA in one fluorochrome and normal control DNA (same gender) in another fluorochrome (red:green) through random priming, analysis of red:green signal intensity after hybridization provides information of gains or losses in the test sample DNA (see Note 11). 1. Have water boiling in a 500  ml beaker before starting the random priming procedure 2. Prepare two labeled tubes, one labeled AF5 (test), and the other labeled AF3 (control). Use safe lock tubes to prevent the lids from opening when boiling 3. Add 500  ng of purified, sonicated (or digested) DNA into each tube, and add TE (from the BioPrime Total labeling kit) to bring the total volume to 22 ml 4. Thaw the Alexa Fluor 2× Reaction Mixes at room temperature, protected from light. Briefly vortex each 2× Reaction Mix and centrifuge briefly to collect. Store the tubes on ice until required 5. Add 25 ml of the 2× Reaction Mixes to their respective tubes 6. Briefly vortex the tube and centrifuge to collect all the liquid at the bottom of the tube 7. Place the tubes into a foam or plastic float and set them in the beaker with boiling water for 5 min

The Use of Microarray Technology for Cytogenetics

133

8. Immediately place the tubes on ice for 5 min 9. On ice, add 3 ml of Exo-Klenow to each tube 10. Vortex briefly and centrifuge to collect 11. Incubate the tubes in a 37°C oven for 2  h (incubation of 2–6 h or overnight is acceptable, and will result in a higher percentage of labeled DNA) 12. Following incubation, proceed directly to the purification step. If storing the reaction prior to purification, add 5 ml of 0.5 M EDTA to each tube to stop the reaction 3.2.2. Purification of Labeled DNA

Purpose: To remove any salts or impurities from the BioPrime Total labeled DNA before hybridization. To prepare working solutions of Binding Buffer (B2) and Wash Buffer (W1), add the following: Binding buffer (B2)

9 ml (entire bottle)

100% Isopropanol

6 ml

Final volume

15 ml

Wash buffer (W1)

10 ml (entire bottle)

100% EtOH

40 ml

Final volume

50 ml

Mix thoroughly and be sure to mark the appropriate checkbox on the bottles to indicate addition of isopropanol or ethanol. 1. Use the Purelink Filter Units that are made up of two parts – the capped filter column and the clear 1.5  ml collection tube. Assemble one unit for every sample to be purified 2. Add 200 ml of prepared Binding Buffer (B2) to each of the sample tubes, vortex to mix and centrifuge briefly to collect 3. Transfer the sample onto the filter and centrifuge for 1 min at 5,369 × g 4. After spinning the tubes, remove the filter column and discard the flow-through in a beaker. Place the filter column in the original collection tube 5. Add 650 ml of prepared Wash Buffer (W1) to the filter column. Centrifuge for 1 min at 5,369 × g 6. Discard the flow-through and centrifuge for an additional 3 min at 5,369 × g 7. Transfer the filter column to a new collection tube. Add 55 ml Elution Buffer (E1) to the filter column and centrifuge for 2 min at 16,100×g 8. Discard the column; the flow-through contains the purified labeled DNA

134

Bejjani, Shaffer, and Ballif

3.2.3. Hybridization of the Microarray

3.2.4. Sample Preparation

Purpose : By combining the control DNA labeled in one fluorochrome with the oppositely labeled test sample DNA and cohybridizing to the microarray, a CGH experiment is performed to look for gains or losses in the test sample DNA. 1. Combine the 55 ml Alexa Fluor 3-labeled control DNA sample with the 55 ml Alexa Fluor 5-labeled test DNA sample 2. Add 50 mg of human Cot I to the sample tube 3. Calculate the total volume (X ml of Human Cot I + 55 ml of Alexa Fluor 3 + 55  ml Alexa Fluor 5). Add 10% of the total volume of 5 M NaCl and vortex the tube to mix. Then add 100% of the total volume of room temperature isopropanol to the tube 4. Mix well by gently inverting the tube 50 times 5. Centrifuge the sample at 9,740 × g for 10 min 6. Using a micropipettor, remove the supernatant, avoiding the pellet 7. Rinse the pellet once with 500  ml of 70% ethanol at room temperature and centrifuge for 5 min at 9,740 × g. During the centrifuge time, use a hot plate to start boiling water in a beaker 8. After the 5 min centrifugation, remove the supernatant and allow the pellet to air dry for 2–5 min at room temperature (see Note 12) 9. Add 104 ml 1× TE (pH 8.0) and let stand at room temperature for 5 min. Vortex the sample until the pellet is completely dissolved and centrifuge briefly to collect 10. Add 26 ml of Agilent 10× blocking agent to the sample 11. Add 130 ml of Agilent 2× hybridization buffer to the sample. This brings the total sample volume to 260 ml 12. Vortex the sample and centrifuge briefly to collect 13. Place sample tubes in a floating rack, and transfer to 95°C boiling water for 5 min to denature 14. Incubate the samples at 37°C in a floating rack in a waterbath or an oven for 30 min

3.2.5. Microarray Hybridization

See Note 13 1. Load a clean gasket slide into the Agilent SureHyb chamber base, with the gasket slide labeled “Agilent” facing up and aligned with the rectangular section of the chamber base. To help prevent leakage, ensure that the gasket slide is flush with the chamber base and is not ajar 2. Remove the sample tube from the 37°C waterbath or oven and centrifuge briefly to collect the sample

The Use of Microarray Technology for Cytogenetics

135

3. Place a sample identification sticker on the top (A) area on the inactive side of the Agilent microarray slide 4. Slowly pipette 245 ml of hybridization sample mixture onto the “A” gasket well using a “drag and dispense” technique (see Note 14) 5. If assaying more than one test sample: Repeat steps 3 and 4 to set up a second hybridization sample mixture on the bottom (B) area of the array slide 6. Place the stickered microarray slide active side down onto the gasket slide, so the numeric barcode is facing up and the Agilent-labeled barcode is facing down (Agilent-to-Agilent). Ensure that the sandwich-pair is properly aligned and that there is no leakage between the two gaskets (see Note 15) 7. Place the SureHyb chamber cover onto the sandwiched slides and slide the clamp assembly onto both pieces. Finger-tighten the clamp onto the chamber, with the chamber laying flat on the bench (see Note 16) 8. Vertically rotate the assembled chamber to wet the slides. Lightly tap the chamber assembly on a hard surface to move any stationary bubbles 9. Place the assembled slide chamber in the rotator rack, in a hybridization oven set to 65°C. Set the hybridization rotator to rotate at 20 rpm. If all of the available positions on the rotator rack are not loaded, balance the loaded hybridization chambers on the rack to prevent unnecessary strain on the oven motor 10. Hybridize at 65°C for 40 h 3.2.6. Washing the Microarray

Purpose : Washing the microarray removes any unbound labeled DNA, providing a clean microarray for analysis. Prewarm 500 ml of wash buffer 2 in a 37°C waterbath at least 30 min prior to use. 1. Fill two slide staining dishes with room temperature wash buffer 1. Place a slide rack into slide staining dish #2 2. Remove one hybridization chamber from the incubator 3. Prepare the hybridization disassembly: (a) Place the hybridization chamber assembly on a flat surface and loosen the thumbscrew (b) Slide off the clamp assembly and remove the chamber cover (c) Remove the sandwiched slides from the chamber base by handling the slides by their ends (keep the microarray slide numeric barcode facing up) (d) Without releasing the slides, submerge the slide sandwich into slide staining dish #1 containing wash buffer 1

136

Bejjani, Shaffer, and Ballif

4. With the slide sandwich completely submerged in wash buffer 1, use forceps to carefully pry the slides open from the barcode end only. Let the gasket slide drop to the bottom of the dish 5. Handling the slide only by the edges, remove the microarray slide and quickly transfer it into the slide rack in slide staining dish #2 containing room temperature wash buffer 1 6. Repeat steps 2–5 until all slides have been transferred to the slide rack. A maximum of five slides is advised at one time to facilitate uniform washing 7. Wash the slides for 5 min by agitation. This may be performed by placing slide staining dish #2 on a magnetic stir plate, submerging a magnetic stir bar in the staining dish and applying a moderate – not too vigorous – stir setting. During this time, fill a clean slide staining dish with the pre-warmed 37°C wash buffer 2 8. Transfer the slide rack to slide staining dish #3 containing 37°C wash buffer 2. Wash for 1 min as in step 7 9. Centrifuge slides in a plate spinner at 110 × g for 3  min, or until dry 10. Slides should be scanned as soon as possible to minimize the impact of environmental oxidants on signal intensities 3.3. Microarray Scanning and Analysis 3.3.1. Scanning the Microarray

Purpose : Scanning the microarray allows for the visualization of the hybridization and enables the technician to quantify the intensity of red and green signals at each spot on the microarray during the analysis steps. Auto PMT Scanning Procedure for the GenePix 4000B scanner: 1. Turn off the computer connected to the scanner, turn on the scanner and then power up the computer again. Give the scanner 15 min to warm up before acquiring any images 2. Open the GenePix Pro 6.0 software 3. Place slide with printed “Agilent” side down and the slide numbers toward the front of the scanner. Ensure that the slide is flat in the compartment by using the adjustment arm on the left, then close the slide holder lid and slide the scanner door closed 4. Using the GenePix Pro software, select “Open settings” from the file menu (see Note 17) 5. Select the image settings for the array that is to be scanned (microarray area A or B). This will load the required scan area for the array 6. Open the “Hardware Settings” box by clicking on the blue scanner icon at the bottom right side of the screen. Ensure that the pixel size is set to 5 mm scan

The Use of Microarray Technology for Cytogenetics

137

7. Click the “Auto-PMT” button to open the “Select PMT Gain Automatically” dialog box. The saturation tolerance should be set at 0.03%. The box next to “Automatically apply calculated PMT gain values and then do a data scan” should be checked 8. Click the “Start” button to begin scanning 9. The scanner will preview scan a small area within the array and automatically adjust the PMT settings to adjust the ratio to 1. It is usually preferable to have the red channel be as large as possible (maximum value of 1,000). If the PMT gain value for the red wavelength is low (<700), the image will be dark. In this case, stop the scan, move the subarray to another position and restart the Auto-PMT 10. Once the computer has confirmed the desired PMT settings, the final scan will start and the dialog box will close. After the slide is scanned, GenePix will automatically display the “Save” dialog box 11. Create a unique folder for the test sample, and save the image as a Multi-Image TIFF (both the 635 and 532 wavelengths are saved as a single file). Name the slides in the following format, in ALL CAPS: CONTROL/FLUOROCHROME/SAMPLE ID/ FLUORO­CHROME/SLIDE NO./_A or B e.g., CM AF3 08-000001 AF5 251791810000_A CM AF3 08-000002 AF5 251791810000_B 12. Remove the slide and place another slide in the scanner. To scan the remaining slides click “Auto-PMT” in the “Hardware Settings” box and click “Start” 13. Scanning can be stopped by clicking on the red “Stop” button at any time. If the array is not scanning correctly (e.g., part of the array is not in the scan area), stop the scan and adjust the settings as necessary. If the scan area is too small, click the “View scan area” button and resize the box to fit the array. When ready, Auto-PMT can be reinitialized and scanning can recommence

4. Notes 1. Signature Genomic Laboratories generally extracts 3 × 300 ml tubes of blood for a total volume of 900 ml 2. Vigorous cell resuspension is important for efficient cell lysis 3. Do not stop at this step. DNA is not stable for storage in the protein precipitation solution

138

Bejjani, Shaffer, and Ballif

4. Not cooling sufficiently may result in a loose protein pellet and impure DNA 5. The precipitated protein forms a tight, dark brown pellet. If this is not observed, repeat step 12, follow with incubation on ice for 5 min, then repeat step 13 6. Add 1  ml glycogen solution per 100  ml isopropanol if low yield is expected – for example, if the sample is compromised by clots or age 7. Vigorous mixing may shear the DNA 8. Be careful to avoid pipetting air bubbles into the dispensed sample. Air bubbles in the sample will cause a reading error 9. The following protocol has been optimized for use at our laboratory. Due to the variable nature of the sonication process, it is recommended that the sonication protocol is optimized for each individual laboratory 10. Technically, the concentration should be ~625  ng/ml, but with a 20% loss during the phenol/chloroform step, assume the samples are 500 ng/ml 11. The Alexa Fluor dyes are more photo-stable than Cyanine dyes; consequently labeling and subsequent steps can be carried out in the light. Although the Alexa Fluor dyes are more stable, they are still photo-sensitive and therefore should be covered or stored out of the light whenever possible 12. It is important that the pellet is dry before proceeding to the next step 1 3. Important: The microarray is printed on the active side of the glass side with the Agilent barcode. The other side of the slide, with the numeric barcode, is inactive. The hybridization sample mixture is applied directly to the gasket slide, not to the active side of the microarray slide. Then, the active side of the microarray slide is placed on top of the gasket slide to form a “sandwich” slide pair. To avoid damaging the microarray, handle the glass slides by their edges. Wear powder-free gloves, and be careful not to touch the surface of the slides 14. To prevent pipetting excess air bubbles onto the gasket slide, do not over-aspirate 15. Important: Be sure not to move the slide once it is in place. Lifting the slide will compromise the seal, and the experiment needs to be repeated 16. The clamp must be tightened sufficiently to prevent leakage of the hybridization solution. However, over-tightening the clamp will cause the slides to crack

The Use of Microarray Technology for Cytogenetics

139

17. If the image settings have not been previously established: (a) Run a preview scan by selecting the double arrow at the top right of the screen (b) Click the “View Scan Area” button, and select the top (A) area of the microarray scan (c) Open “Hardware settings” and set the pixel size to 5 mm scan (d) Click “File, Save Setting As” to save the current image settings for area A (e) Repeat steps the save image settings for area B

Acknowledgments We thank Emily Hall and Richard Lloyd (Signature Genomic Laboratories, Spokane, WA) for their contributions to the protocol. References 1. Bejjani BA, Theisen AP, Ballif BC, Shaffer LG (2005) Array-based comparative genomic hybridization in clinical diagnosis. Expert Rev Mol Diagn 5:421–429 2. Albertson DG, Pinkel D (2003) Genomic microarrays in human genetic disease and cancer. Hum Mol Genet 12(2):R145–R152 3. Pinkel D, Segraves R, Sudar D, Clark S, Poole I, Kowbel D et  al (1998) High

resolution analysis of DNA copy number variation using comparative genomic hybri­di­ zation to microarrays. Nat Genet 20: 207–211 4. Vermeesch JR, Melotte C, Froyen G, Van Vooren S, Dutta B, Maas N et  al (2005) Molecular karyotyping: array CGH quality criteria for constitutional genetic diagnosis. J Histochem Cytochem 53:413–422

Chapter 9 PCR/LDR/Universal Array Platforms for the Diagnosis of Infectious Disease Maneesh Pingle, Mark Rundell, Sanchita Das, Linnie M. Golightly, and Francis Barany Abstract Infectious diseases account for between 14 and 17 million deaths worldwide each year. Accurate and rapid diagnosis of bacterial, fungal, viral, and parasitic infections is therefore essential to reduce the morbidity and mortality associated with these diseases. Classical microbiological and serological methods have long served as the gold standard for diagnosis but are increasingly being replaced by molecular diagnostic methods that demonstrate increased sensitivity and specificity and provide an identification of the etiologic agent in a shorter period of time. PCR/LDR coupled with universal array detection provides a highly sensitive and specific platform for the detection and identification of bacterial and viral infections. Key words:  Ligase detection reaction, Microarray, Multiplexed detection, Bacterial identification, Viral identification

1. Introduction Molecular diagnostic methods for the identification of infectious agents are increasingly being used in clinical microbiology labo­ ratories, supplementing, and in some cases replacing, classical microbiological techniques (1–4). A majority of these methods are based on detecting specific nucleic acid sequences of the infectious agents. Initial methods relied on the use of nucleic acid probes to identify cultured organisms or for direct detection of the organisms from clinical samples. Nucleic acid amplification based methods offer greater sensitivity and are now being used for diagnosis, identification, and quantitation of infections agents, and in the case of bacteria and fungi, for evaluation of antimicrobial resistance profiles. The FDA has approved several commercial

Sridar V. Chittur (ed.), Microarray Methods for Drug Discovery, Methods in Molecular Biology, vol. 632, DOI 10.1007/978-1-60761-663-4_9, © Humana Press, a part of Springer Science+Business Media, LLC 2010

141

142

Pingle et al.

diagnostic kits that utilize a variety of amplification techniques (3). Additionally, some clinical laboratories offer homebrew assays for diagnosis of viral infections. The ligase detection reaction is a linear amplification technique, where two adjacent oligonucleotides hybridize to a single DNA strand and are ligated by a high fidelity thermostable ligase when there is a perfect match at the junction (5). LDR is an ideal technique for multiplexing as multiple primer pairs can hybridize and ligate along a given template simultaneously without interfering with each other. When coupled with an initial PCR amplification of the target sequence, LDR allows for high sensiti­ vity detection of single base variants with extremely high specificity (6–8). Furthermore, coupling of two orthogonal detection methods reduces the risk of allele dropout in multiplexed PCR amplification and false positives arising from spurious and target independent amplifications. Attaching a fluorescent label to one of the ligation primers enables the detection of LDR products. The products may be separated by gel or capillary electrophoresis. If the other primer bears a unique zip-code complement, the ligation products can be detected by hybridization to a universal microarray spotted with the unique zip-codes (Fig. 1). Thus, the universal microarray provides a standard platform for the detection of ligation products arising from any LDR reaction in any assay (9). The zip-code oligonuc­ leotides spotted on the array are designed with Tm values that fall within a narrow range to ensure that all ligation products bearing the different zip-code complements hybridize to the array with equal efficiency. The specificity of the array is derived from the LDR, thus false positives and false negatives associated with direct DNA hybridization arrays are obviated. Spotting or printing the zip-code oligonucleotides on a 3-dimensional matrix (such as that available on Codelink activated slides) rather than directly on the glass surface of a slide, increases signal intensities up to a 100-fold when compared to conventional arrays and reduces hybridization times to between 30 min and 2 h (10). PCR/LDR/universal array hybridization was first demon­ strated for the multiplexed detection of mutations and single base variations in genes implicated in cancer such as BRCA1, BRCA2, K-ras, and p53, both from tumor samples as well as from stool, demonstrating the ability of the technique to detect mutations in the presence of large quantities of normal background DNA (8, 11–15). For example, our laboratory has shown the ability to detect 1 in 100 for a p53 mutation in a wild type sequence, where standard direct hybridization chips are unlikely to be as sensitive (10, 14). Additionally, PCR/LDR/universal array has been used for detection of HLA polymorphisms (16), 21-hydroxylase alleles (7), and determination of CpG island methylation status (17).

PCR/LDR/Universal Array Platforms for the Diagnosis of Infectious Disease

143

A or C

PCR amplify gene(s) of interest using Taq DNA polymerase

C

Perform ligase detection reaction with thermostable ligase, upstream ligation primers with zipcode complements and downstream ligation primers with fluorescent labels.

cZip2 T cZip1

Fluor

G C

Hybridize ligation products to universal microarray containing zipcodes

Fluor Fluor Fluor

G

G

Zip1

G Zip1

Zip2

Zip3

Zip4

Zip5

Zip6

Zip7

Zip8

Zip9

Zip10

Zip11

Zip12

Zip2

Fig. 1. Schematic representation of a PCR/LDR/universal array assay. In the example shown, the assay is used to detect the identity of a single base variant (A/C) within a gene of interest. The target gene is amplified using gene specific PCR primers. The PCR amplicon is subjected to an LDR reaction using two upstream primers, one ending in T (complementary to the A allele) with the complement to zip-code 2 attached to its 5¢-end and the other ending in G (complementary to the C allele) with the complement to zip-code1 attached to its 5¢-end. If the allele present in the sample is C (as shown), then only the “T” primer will ligate to the downstream fluorescently labeled primer. When the ligation products are hybridized to the universal array, a positive signal is seen at zip-code address 1

144

Pingle et al.

The PCR/LDR/ universal array platform is ideally suited to develop assays for the detection and identification of infectious disease. For bacterial (or fungal) pathogens, initial PCR amplifi­ cation of a gene or genes that are conserved across all bacteria (or fungi respectively), can be followed by LDR for the detection of single base variations at multiple loci to identify the pathogen. We have previously demonstrated this technique for the multi­ plexed identification of a panel of 20 blood borne bacterial pathogens, where the ligation products are detected by capillary electrophoresis (18). The assay is directly transferrable to a universal array format when one set of the ligation primers are redesigned to bear zip-code complements. Others have used PCR/LDR/ universal array for identification of bacteria (19) as well as for determining diversity in target bacterial populations (20). RNA viruses such as hemorrhagic fever viruses present a different challenge. These viruses mutate rapidly and undergo significant sequence drift, making it difficult to design PCR primers that can amplify variant sequences. In these cases, it is prudent to PCR amplify more than one target using multiple PCR primers to account for sequence variations at the primer binding sites. LDR can then be performed to detect the presence of these PCR amplicons using multiple ligation primers at each ligation site to account for sequence variations. We have reported PCR/ LDR/universal array assays for the detection of two different RNA viruses, West Nile virus (21), and Dengue virus (Fig. 3) (22). The PCR/LDR/universal array platform is amenable to incor­ poration into microfluidic devices. These devices can perform the entire assay, including array hybridization and readout in less than 1  h, making the platform suitable for the development of point of care diagnostic devices (23, 24).

2. Materials 2.1. Microarray Printing and Coupling DNA Probes

1. Zip-code Oligonucleotides (see Note 1). 2. aQu Ultrasonic Microarray Pin Cleaning Solution (Genetix, Boston, MA). 3. 3 × 1 in Microscope slides. 4. 384 well plate. 5. Codelink Activated Slides (SurModics, Inc.). 6. 6× Print buffer: 300 mM sodium phosphate, pH 8.5. 7. Saturated NaCl humidification chamber.

2.2. Postcoupling and Hybridization

1. 10% Sodium dodecyl sulfate. 2. 20× standard sodium citrate (SSC).

PCR/LDR/Universal Array Platforms for the Diagnosis of Infectious Disease

145

3. Blocking solution: 50 mM ethanolamine, 0.1 M Tris (pH 9). 4. Postcoupling wash solution: 4× SSC, 0.1% SDS. 5. Posthybridization wash solution 1: 4× SSC. 6. Posthybridization wash solution 2: 2× SSC, 0.1% SDS. 7. Posthybridization wash solution 3: 0.2× SSC. 8. Posthybridization wash solution 4: 0.1× SSC. 9. 1× Wash Buffer: 0.3  M Bicine pH 8.0, , 0.1% SDS, filter sterilize. Store at room temperature. 10. Salmon sperm DNA. 11. ProPlate adhesive seal-strips (Grace Bio-Labs) (see Note 2). 12. Razor blade. 13. ProPlate multiarray chamber system (Grace Bio-Labs) (see Note 2). 14. Shaker. 15. Heat blocks. 16. Microcentrifuge. 17. Slide Spinner (Labnet International, Inc.). 18. Thermocycler. 19. Hybridization oven with rotary rocker. 20. QArrayMini robotic array printer (Genetix, Boston, MA). 21. ProScanArray microarray scanner (Perkin Elmer, Boston, MA). 2.3. Gene Specific PCR

1. PCR buffer (10×): 100 mM Tris–HCl buffer, 500 mM KCl, pH 8.0. Store at −20°C. 2. MgCl2, 25 mM. Store at −20°C. 3. Deoxynucleoside triphosphates (dNTPs), 0.8  mM. Store at −20°C. 4. Gene specific PCR primers. 5. AmpliTaq Gold DNA polymerase (Applied Biosystems, Foster City, CA). Store at −20°C. 6. Nuclease free water. Store at room temperature. 7. DNA templates extracted from blood or serum samples. For detection of viruses, extracted RNA is first converted to cDNA using any commercially available reverse transcription kit.

2.4. Ligase Detection Reaction (LDR) for Detection of Pathogen

1. 10× LDR buffer: 20 mM Tris–HCl buffer, pH 7.6, 100 mM KCl, 10 mM MgCl2. Store at −20°C (see Note 3). 2. Nicotinamide adenine dinucleotide (NAD+) 10 mM. Store at −20°C. 3. Dithiothrietol (DTT), 200 mM. Store at −20°C.

146

Pingle et al.

4. Ligation primers. The downstream primers are labeled with fluorescent dye at the 3¢ end. The upstream primer bears a specific zip-code complement at the 5¢ end. The 5¢ end of the upstream primer is also blocked with blocking group. Store stock primer solutions (100 µM) at −20°C. 5. Product obtained from gene specific PCR detailed in Subheading 2.1. 6. AK 16D ligase enzyme (expressed from Thermus spp. AK 16D), store at −20°C. The working concentration of the enzyme is 1 µM (see Note 4). 7. T4 ligase buffer: 50 mM Tris–HCl, 10 mM MgCl2, 10 mM dithiothreitol, 1 mM ATP, 25 µg/ml bovine serum albumin (New England BioLabs, Ipswich, MA). Store at −20°C (see Note 5). 8. T4 polynucleotide kinase (PNK) enzyme (10 U) (New England BioLabs, Ipswich, MA). Store at −20°C.

3. Methods The general procedure for carrying out a PCR/LDR/universal array experiment involves the following steps: (1) Array printing and quality control (2) PCR amplification (3) LDR and (4) Array hybridization and scanning. Steps 2 and 3 are outlined in Fig. 1. Arrays are generally printed in batches and a random selection of printed arrays is assayed for print quality. PCR reactions are set up using standard procedures. Sample preparation for PCR should be performed in an area separate from where the reaction mixture for amplification is being assembled. Assembling PCR reactions in a laminar flow cabinet equipped with a UV lamp is recommended to prevent contamination. If possible, a separate set of pipettes designated for PCR setup as well as using pipette tips with aerosol filters for both DNA sample and reaction mixture preparation should be used to further reduce the chance of contamination. PCR primers are designed based on alignments of the target gene sequence, either the same conserved gene from multiple species of organisms, or from multiple isolates of the same species. PCR primers are designed to have Tm values between 65 and 75°C and incorporate the use of degenerate nucleotides to cover for sequence variations (see Note 6). LDR primers are designed to detect single base variations or specific nucleotide positions within the PCR amplicons. As with the PCR primers, the LDR primers use degenerate nucleotides where required to cover for sequence variations (see Note 7).

PCR/LDR/Universal Array Platforms for the Diagnosis of Infectious Disease

3.1. Printing and Coupling of Zip-Code Addresses

147

1. Prepare the print plate containing a final concentration of 1× print buffer and 25 µM zip-code oligonucleotide in a 384 conical well plate. Add 1 µM fiducial oligonucleotide to the printing mix in each well with each zip-code address. This will be used as a quality control to determine the position and quality of each spot. 2. Remove Codelink activated slides from sealed package. Unused slides must be stored in a sealed desiccator. Place slides in the QArrayMini robotic array printer (To easily identify the location of the zip-codes spotted in each subarray, place the slides so that the text “Codelink” etched into the slide is at the bottom and each slide is placed flush up and right). Position 1 is for a blotting slide. Fill all empty spaces for each vacuum region. Turn on vacuum. Set the relative humidity <50% and the temperature to 10ºC. 3. Place the print plate in the QArrayMini robotic array printer. Print DNA onto activated slides to produce arrays in desired layout. 4. Leave the print plate in the QArrayMini robotic array printer overnight to dry out oligonucleotides in wells. Seal plate and store in desiccator between print runs (Resuspend in dH2O at least 1 h prior to print run for each subsequent use). 5. Place printed slides in slide storage box. Place uncovered slide storage box in saturated NaCl chamber. Seal the NaCl chamber and incubate at room temperature for 4–72 h.

3.2. Postcoupling Processing

1. Place printed slides in glass coplin jar with prewarmed blocking solution at 50°C for 30 min. 2. Rinse the slides 2× with ultrapure filtered water. 3. Wash the slides with 4× SSC, 0.1% SDS (prewarmed to 50°C) for 30 min on the shaker. 4. Discard wash solution and rinse briefly 2× with ultrapure filtered water. 5. Spin dry the slides using the Slide Spinner. 6. Store coupled slides in slide storage boxes at ambient temperature until use. For long-term storage, maintain the slides in a desiccated environment.

3.3. Array Hybridization Quality Control

1. Randomly select printed slides from each print batch (one from the beginning of the batch, one from the end of the batch, and one or more from the middle of the batch). 2. Prepare hybridization solution containing 10 fmol fluorescentlabeled zip-code complements, 500  fmol competition

148

Pingle et al.

a

b

Fig. 2. Quality control and validation of arrays post spotting. (Panel a) Hybridization of ROX-labeled fiducial complement to the array. Since the fiducial oligonucleotide is included at every location, all spots should provide a positive fluorescent signal of equal intensity. Note that we double spot each address on the array. (Panel b) Hybridization of FAM-labeled zip-code complements to the addresses in every fourth column (Columns 4, 8, 12, 16, 20, and 24) of the array (addresses 4, 8, 12, 16, 20, 24, 28, 32, 36, 40, 44, 48, 52, 56, 60, 64, 68, 72, 76, 80, 84, 88, 92, and 96), When no other columns or addresses provide a positive signal, the print run is considered successful. Three other hybridization experiments are carried out using FAM-labeled zip-code complements to the addresses in Columns 1, 5, 9, 13, 17, and 21 combined, Columns 2, 6, 10, 14, 18, and 22 combined, and Columns 3, 7, 11, 15, 19, and 23 combined (data not shown)

primers, 2.5 fmol fiducial complement, 0.1 mg/ml ssDNA, 5× SSC, 0.1% SDS in a total volume of 30 µl. A 6-carboxy-Xrhodamine-labeled fiducial complement included in the hybridization mixture serves as an internal positive control to determine the position and quality of each printed address (Fig. 2). 3. Prepare the ProPlate multiarray hybridization chamber by affixing the ProPlate multiarray chamber to the Codelink slides, so the printed side is facing the empty chambers. 4. Add 30 µl hybridization solution to each chamber. 5. Seal the ProPlate multiarray chamber system with ProPlate adhesive seal-strips. 6. Incubate the slides in the dark on a rocker platform at 60ºC for 2 h. 7. After incubation, dismantle the ProPlate multiarray chamber system and briefly rinse the slides with 4× SSC. 8. Wash the slides with prewarmed wash solution I (2× SSC, 0.1% SDS) for 10 min at 60ºC. 9. Wash the slides with wash solution II (0.2% SSC) for 1 min at room temperature. 10. Wash the slides with wash solution III (0.1% SSC) for 1 min at room temperature. 11. Spin dry the slides using the Slide Spinner. 12. Scan the slides and analyze the signal intensities using ProScanArray microarray scanner. The quality control data

PCR/LDR/Universal Array Platforms for the Diagnosis of Infectious Disease

149

should provide specific fluorescent signals in the absence of extraneous signals on adjacent addresses (Fig. 2). The require­ ments for signal intensities from spotted addresses that are statistically significant should be determined (see Note 8). 3.4. Setup of PCR Amplification

1. Nucleic acids extracted from human specimens (blood or serum) are used as a template for PCR amplification. For templates stored at −20ºC, gently vortex and centrifuge samples and thaw on ice. 2. Dilute the PCR primers in nuclease-free water to the working concentration (usually 2–5 µM) (see Note 9). 3. Prepare a master mix containing 2.5 µl of 10× PCR buffer, 2.5 µl of MgCl2, 2.0 µl of dNTPs, 2–5 µl of a 1 µM primer mix containing all primers, 1.25  U of AmpliTaq Gold DNA polymerase, and nuclease-free water. The total volume of the master mix for the PCR reaction is dependent on the volume of DNA template to be added. The recipe provided is for 25 µl reactions. It is best to prepare a sufficient quantity of master mix for multiple reactions and aliquot the mix into separate reaction tubes. Individual DNA templates can then be added to the respective tubes. To set up reactions with larger volumes, such as 50 µl, simply double the volume of reagents used. 4. Pipette 20–22 µl of the master mix in to thin walled PCR reaction tubes on ice. Add template DNA (amount to be added may be optimized, usually; 10 pg to 1 µg of DNA). 5. Carry out the PCR reaction in a thermocycler using the following protocol: 10 min at 94°C, followed by 30–40 cycles of 15 s at 94°C, 1 min at 60°C, 1 min at 72°C, followed by a final extension step of 7 min at 72°C, followed by 30 min at 99°C to destroy the Taq polymerase. The PCR products may be stored at 4°C until further processing (see Note 10).

3.5. S  etup of LDR

The downstream primers used in an LDR reaction need to be phosphorylated for ligation to occur. Since the upstream ligation primers are blocked at their 5¢-termini, a primer mix containing both the upstream and downstream primers is treated with T4 polynucleotide kinase and ATP prior to use (see Note 11). 1. Prepare a LDR primer mix such that each primer is at a final concentration of 250 or 500 fmol per reaction when used in the LDR. 2. Mix together, 3 µl of 10× T4 ligase buffer, 10 units of T4 PNK enzyme, and LDR primer mix. Make the volume up to 30 µl with nuclease free water. In a thermocycler, incubate the reaction at 37ºC for 1 h followed by 10  min at 65ºC, to inactivate the enzyme. Phosphorylated primers may be stored at 4ºC for up to a week.

150

Pingle et al.

3. Prepare a LDR master mix containing 2.0 µl of 10× LDR buffer, 0.1 µl of 200  mM DTT, 2 µl of 10  mM NAD+, sufficient phosphorylated primer mix for a final primer concentration of each primer of 250 or 500 fmol, 0.25 µl of 1 µM AK16D ligase, and nuclease-free water (for a total volume of 19 µl). The recipe provided is for a single 20 µl reaction. It is best to prepare a sufficient quantity of master mix for multiple reactions and aliquot the mix into separate reaction tubes. Individual DNA templates can then be added to the respective tubes. 4. Pipette 19 µl of the master mix into thin walled PCR reaction tubes on ice. Add 1 µl of PCR product from step 5 of Subheading 3.4 above. Carry out the LDR in a thermocycler using the following protocol: 94°C for 2 min followed by 20 cycles of 30 s at 94°C, 4 min at 64°C. The LDR product may be stored at 4°C until hybridization to the array. 3.6. Hybridization of LDR Products

1. Dilute the entire amount of the ligation products from step 4 of Subheading 3.5 in hybridization buffer: (5× SSC buffer, 0.1% SDS, 0.1  mg/ml salmon sperm DNA (Fisher Scientific), and 5  nM of the fiducial complement in a total volume of 30 µl). 2. Denature the mixture at 94°C for 3 min and chill on ice.

Dengue virus serotype 1

Dengue virus serotype 3

Dengue virus serotype 2

Dengue virus serotype 4

Fig. 3. Representative images of universal array detection of Dengue virus serotypes I–IV following hybridization of LDR products from a PCR/LDR assay. LDR primers for each serotype bear zip-code complements to distinct addresses. For correct identification only the addresses specific to a given serotype should provide a positive signal. In this assay, a minimum of two signals are required to identify a given serotype. (see ref. (22) for details)

PCR/LDR/Universal Array Platforms for the Diagnosis of Infectious Disease

151

3. Repeat steps 3–12 described in Subheading 3.3 above. 4. Determine the identity of the pathogen based on the zip-code addresses that present fluorescent signals. An example of the identification of four different serotypes of Dengue virus is shown in Fig. 3.

4. Notes 1. The zip-code oligonucleotides are synthesized with a 3¢-amino modifier and a spacer 18 modifier (phosphoramidite available from Glen Research, alternatively, most commercial oligonucle­ otide manufactures offer this modification) between the base at the 3¢-end and the amino modifier. The zip-code oligonucle­ otides must be either PAGE or HPLC purified for best results, however, reverse-phase cartridge purification may provide satis­ factory results. The sequences of the zip-code oligonucleotides are provided in Table 1. 2. The ProPlate multiarray chamber system with the ProPlate adhesive strips enables printing of 16 discrete arrays on a single slide. Thus, 16 individual LDR reactions may be

Table 1 Sequences of zip-code oligonucleotides Zip-code sequence

Zip-code complement

AGCGAGCGGGAACAGGCCAA

TTGGCCTGTTCCCGCTCGCT

GGAACACCACGCAGCGCAGG

CCTGCGCTGCGTGGTGTTCC

GCAGTGCTCACCGTCCGCGA

TCGCGGACGGTGAGCACTGC

CGGAGTGGCACCAGCGGGAA

TTCCCGCTGGTGCCACTCCG

GCAGCAGGCCAAAGCGAGCG

CGCTCGCTTTGGCCTGCTGC

GTCCGAGCCCTCACGCAGCG

CGCTGCGTGAGGGCTCGGAC

GCAGGACGACGCGGGTGGAA

TTCCACCCGCGTCGTCCTGC

TGGCGGTCTGCTGAGCGGTC

GACCGCTCAGCAGACCGCCA

GTGGGTCCCGGAAGCGTGCT

AGCACGCTTCCGGGACCCAC

GCCTCGAGCCAACACCGCCT

AGGCGGTGTTGGCTCGAGGC

TGGCCGGACAGGAGACACGC

GCGTGTCTCCTGTCCGGCCA

GCCTGCCTTCACGAGCCCAA

TTGGGCTCGTGAAGGCAGGC (continued)

152

Pingle et al.

Table 1 (continued) Zip-code sequence

Zip-code complement

GTGGGCGAAGCGGGAACCTC

GAGGTTCCCGCTTCGCCCAC

CGGAGCGATCACGTGGCACC

GGTGCCACGTGATCGCTCCG

ACGCGACGCACCTGCTCCAA

TTGGAGCAGGTGCGTCGCGT

CCTCCCTCACGCGCCTGCAG

CTGCAGGCGCGTGAGGGAGG

GTGGACTGAGCGCGGATGGC

GCCATCCGCGCTCAGTCCAC

CGAGGCAGACGCGTCCCACC

GGTGGGACGCGTCTGCCTCG

TGGCCGAGACTGCAGGAGCG

CGCTCCTGCAGTCTCGGCCA

AGCGGACGACTGCGGACGAG

CTCGTCCGCAGTCGTCCGCT

GCCTGCGAAGACCCAAGCGA

TCGCTTGGGTCTTCGCAGGC

GAGCAGCGACGCCGAGGCAG

CTGCCTCGGCGTCGCTGCTC

GCGAGTCCCGAGGGTCCCAA

TTGGGACCCTCGGGACTCGC

GGAAAGCGAGCGGCAGCCAA

TTGGCTGCCGCTCGCTTTCC

TGGCGGAACAGGACTGCGGA

TCCGCAGTCCTGTTCCGCCA

TGGCGGGTTGCTCCTCGTGG

CCACGAGGAGCAACCCGCCA

GACGGCCTTGCTAGCGCGGA

TCCGCGCTAGCAAGGCCGTC

GCCTGCAGTGCTGGTCCGGA

TCCGGACCAGCACTGCAGGC

CCTCCGGAAGACCCTCGCGA

TCGCGAGGGTCTTCCGGAGG

GCGAGCAGCAGGGTGGACCA

TGGTCCACCCTGCTGCTCGC

GCCTGAGCAGACGGTCGCGA

TCGCGACCGTCTGCTCAGGC

GGGTGCCTAGCGGTCCAGCG

CGCTGGACCGCTAGGCACCC

CAGGACGCACCAACGCCCAA

TTGGGCGTTGGTGCGTCCTG

AGCGCACCCGGAACTGGAGC

GCTCCAGTTCCGGGTGCGCT

ACGCGTGGACTGCCTCGAGC

GCTCGAGGCAGTCCACGCGT

ACGCCCTCCCAACCTCACGC

GCGTGAGGTTGGGAGGGCGT

CACCGCAGCCTCCCAACCAA

TTGGTTGGGAGGCTGCGGTG

GGGTTGGCGGAAGGTCGACG

CGTCGACCTTCCGCCAACCC

GCGAGCGAACCAGAGCGACG

CGTCGCTCTGGTTCGCTCGC

TGGCAGCGTCACGGGTCACC

GGTGACCCGTGACGCTGCCA

GGGTGACGAGCGCCAAGCCT

AGGCTTGGCGCTCGTCACCC

GCGATGGCAGCGGTGGAGAC

GTCTCCACCGCTGCCATCGC

GCAGGCGATGGCTCACGACG

CGTCGTGAGCCATCGCCTGC (continued)

PCR/LDR/Universal Array Platforms for the Diagnosis of Infectious Disease

153

Table 1 (continued) Zip-code sequence

Zip-code complement

GTCCTGCTGTGGGCGATGGC

GCCATCGCCCACAGCAGGAC

GGTCGTCCGGTCGCCTTGCT

AGCAAGGCGACCGGACGACC

ACCAAGCGGCCTCCTCGTCC

GGACGAGGAGGCCGCTTGGT

ACGCGGAAGGTCTGGCCAGG

CCTGGCCAGACCTTCCGCGT

CCAACGGAGCGACGAGCAGG

CCTGCTCGTCGCTCCGTTGG

CGGACAGGGACGGCGATCAC

GTGATCGCCGTCCCTGTCCG

GGTCGGGTCAGGCCTCGGAA

TTCCGAGGCCTGACCCGACC

CGGAAGCGCGAGACCACACC

GGTGTGGTCTCGCGCTTCCG

TCACCCTCTGGCGGAACGGA

TCCGTTCCGCCAGAGGGTGA

CCAAAGACAGCGGACGGCGA

TCGCCGTCCGCTGTCTTTGG

GGGTGGGTCGAGGCCTGGTC

GACCAGGCCTCGACCCACCC

CGGAGTCCTGGCAGCGTGGC

GCCACGCTGCCAGGACTCCG

GCGAAGCGACCAAGACCGGA

TCCGGTCTTGGTCGCTTCGC

CAGGCACCCACCGCGAAGAC

GTCTTCGCGGTGGGTGCCTG

GTCCGCAGCCAACCAAACGC

GCGTTTGGTTGGCTGCGGAC

GCAGAGCGTGGCCGAGGTCC

GGACCTCGGCCACGCTCTGC

GTGGCGGACGGACGAGTGGC

GCCACTCGTCCGTCCGCCAC

GCAGGTGGGACGGGTCGGGT

ACCCGACCCGTCCCACCTGC

AGACAGCGGCGAGAGCGGGT

ACCCGCTCTCGCCGCTGTCT

CGAGAGCGGTCCCGGAGGTC

GACCTCCGGGACCGCTCTCG

CCTCCGAGCACCGACGACGC

GCGTCGTCGGTGCTCGGAGG

ACGCCCAAACGCAGACCCAA

TTGGGTCTGCGTTTGGGCGT

GGTCCAGGTGGCGGTCGAGC

GCTCGACCGCCACCTGGACC

AGCGTCACGAGCCAGGCGGA

TCCGCCTGGCTCGTGACGCT

GAGCGTGGCGGAGGTCGGTC

GACCGACCTCCGCCACGCTC

GACGGCGAGGGTGCAGGCAG

CTGCCTGCACCCTCGCCGTC

CCTCGACGGTCCTGGCTGGC

GCCAGCCAGGACCGTCGAGG

GAGCTGCTTGGCGCGACACC

GGTGTCGCGCCAAGCAGCTC

GCGACAGGCGGAGAGCGGAA

TTCCGCTCTCCGCCTGTCGC

GCAGTGGCGTCCGGGTGAGC

GCTCACCCGGACGCCACTGC

CGAGGGAAGTGGGCAGCGGA

TCCGCTGCCCACTTCCCTCG (continued)

154

Pingle et al.

Table 1 (continued) Zip-code sequence

Zip-code complement

TGGCGAGCGCAGTGGCAGAC

GTCTGCCACTGCGCTCGCCA

CGAGGCCTGCAGGGAAAGCG

CGCTTTCCCTGCAGGCCTCG

CGAGGTCCGGGTGCGAGAGC

GCTCTCGCACCCGGACCTCG

GTGGGAGCGACGCAGGGCAG

CTGCCCTGCGTCGCTCCCAC

GGGTGCAGGCCTGTGGGTCC

GGACCCACAGGCCTGCACCC

GCCTACGCGAGCGACGGAGC

GCTCCGTCGCTCGCGTAGGC

CACCGAGCTGCTGCCTTGGC

GCCAAGGCAGCAGCTCGGTG

GGGTGGAAAGCGGAGCGTGG

CCACGCTCCGCTTTCCACCC

CAGGCCAAGCAGACGCGACG

CGTCGCGTCTGCTTGGCCTG

GCGAGGGTGCGAGGGTTGCT

AGCAACCCTCGCACCCTCGC

AGCGGTCCGACGGCCTTCAC

GTGAAGGCCGTCGGACCGCT

GTCCCGAGGCAGCGAGAGCG

CGCTCTCGCTGCCTCGGGAC

GCAGTCACGGTCAGCGGCCT

AGGCCGCTGACCGTGACTGC

GACGCCAACGGACGGAGGGT

ACCCTCCGTCCGTTGGCGTC

TCACGCGACACCCGGACACC

GGTGTCCGGGTGTCGCGTGA

CGAGCGGAGAGCGAGCCAGG

CCTGGCTCGCTCTCCGCTCG

AGCGTGCTGGTCGTGGGCCT

AGGCCCACGACCAGCACGCT

AGCGCCAAGGGTCCTCGGGT

ACCCGAGGACCCTTGGCGCT

CCAAAGCGAGACCGGAGCGA

TCGCTCCGGTCTCGCTTTGG

GACGCACCGAGCACGCACCA

TGGTGCGTGCTCGGTGCGTC

TCACCCAAGACGGCAGGCGA

TCGCCTGCCGTCTTGGGTGA

GGGTGGTCCGGAGCGAGCAG

CTGCTCGCTCCGGACCACCC

hybri­dized to the arrays on a slide without cross-contamination of the samples. 3. We recommend that the 10× LDR buffer be aliquoted in tubes and stored at −20°C. Once an aliquot is thawed for use, it may be stored at 4°C for up to 2 weeks and then discarded. 4. The AK16D ligase enzyme preparation is normally at a higher concentration than the 1 µM concentration required for the LDR. The enzyme may be diluted to a 1 µM concentration using 1× LDR buffer (tenfold dilution of the 10× LDR buffer)

PCR/LDR/Universal Array Platforms for the Diagnosis of Infectious Disease

155

just prior to setting up the LDR reaction. We do not recom­ mend storing the diluted enzyme. 5. The T4 ligase buffer composition is the same as that of the T4 PNK buffer supplied with the T4 PNK enzyme except that the ligase buffer already contains 1 mM ATP, the phosphate source. If using the PNK buffer, it is necessary to add ATP at a concentration of 1  mM to the buffer. We recommend storing aliquots of either buffer at −20°C for each use of the buffer. 6. We recommend using the software program Oligo 6 for designing PCR and LDR primers. Oligo 6 calculates Tm values for the primers using the nearest neighbor method. PCR primers should have GC content between 40 and 60% and Tm values of 65–70°C. Where possible, the last 4 bases at the 3¢-end of the primers should be SWWC, where S is a strong base such as G or C and W is a weak base such as A or T. The primer sequences should be checked for self-complementary regions or regions complementary to other primers in the reaction mixture, in order to avoid primer-dimer and hairpin formation. No primer should contain more than 3 degenerate positions. For multiplex PCR amplification, universal tail sequences may also be appended to the 5¢-ends of forward and reverse PCR primers to prevent the formation of primer dimers. The universal tail sequence commonly used in our laboratory for such applications is CGCTGCCAACTACCGCACATC. 7. LDR primers are designed to have Tm values of 65–70°C for the upstream primers and 70–75°C for the downstream primers. LDR primers should have no more than 3 degenerate positions in each primer, and there should be no degenerate base within at least 3 bases of the ligation site. In order to adhere to this requirement, it may be necessary to design multiple primers to cover all possible sequence variants. The last base on the upstream primer is the query base and must be perfectly complementary to the template for ligation to occur. As shown in Fig. 1, the upstream LDR primers have zip-code complements attached to their 5¢-ends. Additionally, the upstream primers are blocked at the 5¢-end with a blocking group such as a C3 spacer or a C6-amino linker. The downstream LDR primers are labeled with a fluorophore (commonly Cy3 or Cy5) at their 3¢-end. In a variation, the zip-code complements may be attached to the 3¢-ends of the downstream primers followed by a blocking group, and the fluorescent label attached to the 5¢-ends of the upstream primers. The sequences of the zip-code oligonucle­ otides are provided in Table 1.

156

Pingle et al.

8. In general, we consider a signal to be positive when its inten­ sity is at least tenfold higher than the background intensity. 9. For most PCR protocols, we use between 5 and 10 pmol of each PCR primer. For highly multiplexed PCR reactions, it may be necessary to reduce the amount of each primer to 2  pmol. The optimal amount of each primer is usually determined empirically. 10. The additional step of incubating the PCR reaction at 99°C is sufficient to inactivate the Taq polymerase. An alternate method involves incubating the PCR reaction with 1 µl of proteinase K (18 mg/ml) at 70°C for 10 min. The proteinase K is then inactivated by incubating at 95°C for 15 min. 11. While LDR primers may be synthesized with 5¢-posphate groups on the downstream primers, we have found that long-term storage of the primers results in degradation of the phosphate groups resulting in progressively weaker ligation signals over time. Therefore, we recommend that primers be freshly phosphorylated by treatment with kinase prior to each LDR reaction. There should be no more than 300 pmol of primers with free 5¢-OH ends per 30 µl phosphorylation reaction using 10 units of T4 PNK enzyme.

Acknowledgments Support for this work was provided by the National Cancer Institute (P01-CA65930) and the National Institute of Allergy and Infectious Diseases (UC1-AI062579). References 1. Boissinot M, Bergeron MG (2002) Toward rapid real-time molecular diagnostic to guide smart use of antimicrobials. Curr Opin Microbiol 5:478–482 2. Muldrew KL (2009) Molecular diagnos­tics of infectious diseases. Curr Opin Pediatr 21: 102–111 3. Pfaller MA (2001) Molecular approaches to diagnosing and managing infectious diseases: practicality and costs. Emerg Infect Dis 7: 312–318 4. Procop Gary W (2007) Molecular diagnos­ tics for the detection and characterization of microbial pathogens. Clin Infect Dis 45: S99–S111 5. Barany F (1991) The ligase chain reaction in a PCR world. PCR Methods Appl 1:5–16

6. Barany F (1991) Genetic disease detection and DNA amplification using cloned thermostable ligase. Proc Natl Acad Sci USA 88:189–193 7. Day DJ, Speiser PW, White PC, Barany F (1995) Detection of steroid 21-hydroxylase alleles using gene-specific PCR and a multiplexed ligation detection reaction. Genomics 29:152–162 8. Khanna M, Cao W, Zirvi M, Paty P, Barany F (1999) Ligase detection reaction for identi­ fication of low abundance mutations. Clin Biochem 32:287–290 9. Favis R, Gerry NP, Cheng YW, Barany F (2005) Applications of the universal DNA microarray in molecular medicine. Methods Mol Med 114:25–58 10. Gerry NP, Witowski NE, Day J, Hammer RP, Barany G, Barany F (1999) Universal DNA

PCR/LDR/Universal Array Platforms for the Diagnosis of Infectious Disease microarray method for multiplex detection of low abundance point mutations. J Mol Biol 292:251–262 11. Dong SM, Traverso G, Johnson C, Geng L, Favis R, Boynton K, Hibi K, Goodman SN, D’Allessio M, Paty P, Hamilton SR, Sidransky D, Barany F, Levin B, Shuber A, Kinzler KW, Vogelstein B, Jen J (2001) Detecting colorectal cancer in stool with the use of multiple genetic targets. J Natl Cancer Inst 93:858–865 12. Favis R, Barany F (2000) Mutation detection in K-ras, BRCA1, BRCA2, and p53 using PCR/LDR and a universal DNA microarray. Ann N Y Acad Sci 906:39–43 13. Favis R, Day JP, Gerry NP, Phelan C, Narod S, Barany F (2000) Universal DNA array detection of small insertions and deletions in BRCA1 and BRCA2. Nat Biotechnol 18:561–564 14. Favis R, Huang J, Gerry NP, Culliford A, Paty P, Soussi T, Barany F (2004) Harmonized microarray/mutation scanning analysis of TP53 mutations in undissected colorectal tumors. Hum Mutat 24:63–75 15. Fouquet C, Antoine M, Tisserand P, Favis R, Wislez M, Commo F, Rabbe N, Carette MF, Milleron B, Barany F, Cadranel J, Zalcman G, Soussi T (2004) Rapid and sensitive p53 alteration analysis in biopsies from lung cancer patients using a functional assay and a universal oligonucleotide array: a prospective study. Clin Cancer Res 10:3479–3489 16. Consolandi C, Busti E, Pera C, Delfino L, Ferrara GB, Bordoni R, Castiglioni B, Bernardi LR, Battaglia C, De Bellis G (2003) Detection of HLA polymorphisms by ligase detection reaction and a universal array format: a pilot study for low resolution genotyping. Hum Immunol 64:168–178 17. Cheng YW, Shawber C, Notterman D, Paty P, Barany F (2006) Multiplexed profiling of candidate genes for CpG island methylation status using a flexible PCR/LDR/Universal Array assay. Genome Res 16:282–289 18. Pingle MR, Granger K, Feinberg P, Shatsky R, Sterling B, Rundell M, Spitzer E, Larone D, Golightly L, Barany F (2007) Multiplexed identification of blood-borne

157

bacterial pathogens by use of a novel 16S rRNA gene PCR-ligase detection reactioncapillary electro­phoresis assay. J Clin Microbiol 45:1927–1935 19. Busti E, Bordoni R, Castiglioni B, Monciardini P, Sosio M, Donadio S, Consolandi C, Rossi Bernardi L, Battaglia C, De Bellis G (2002) Bacterial discrimination by means of a universal array approach mediated by LDR (ligase detection reaction). BMC Microbiol 2:27 20. Castiglioni B, Rizzi E, Frosini A, Sivonen K, Rajaniemi P, Rantala A, Mugnai MA, Ventura S, Wilmotte A, Boutte C, Grubisic S, Balthasart P, Consolandi C, Bordoni R, Mezzelani A, Battaglia C, De Bellis G (2004) Development of a universal microarray based on the ligation detection reaction and 16S rrna gene polymorphism to target diversity of cyanobacteria. Appl Environ Microbiol 70: 7161–7172 21. Rondini S, Pingle MR, Das S, Tesh R, Rundell MS, Hom J, Stramer S, Turner K, Rossmann SN, Lanciotti R, Spier EG, Munoz-Jordan J, Larone D, Spitzer E, Barany F, Golightly LM (2008) Development of multiplex PCR-ligase detection reaction assay for detection of West Nile virus. J Clin Microbiol 46:2269–2279 22. Das S, Pingle MR, Munoz-Jordan J, Rundell MS, Rondini S, Granger K, Chang GJ, Kelly E, Spier EG, Larone D, Spitzer E, Barany F, Golightly LM (2008) Detection and serotyp­ ing of dengue virus in serum samples by multiplex reverse transcriptase PCR-ligase detection reaction assay. J Clin Microbiol 46:3276–3284 23. Hashimoto M, Barany F, Soper SA (2006) Polymerase chain reaction/ligase detection reaction/hybridization assays using flowthrough microfluidic devices for the detection of low-abundant DNA point mutations. Biosens Bioelectron 21:1915–1923 24. Hashimoto M, Hupert ML, Murphy MC, Soper SA, Cheng YW, Barany F (2005) Ligase detection reaction/hybridization assays using three-dimensional microfluidic networks for the detection of low-abundant DNA point mutations. Anal Chem 77:3243–3255

Chapter 10 RIP-CHIP in Drug Development Ritu Jain, Francis Doyle, Ajish D. George, Marcy Kuentzel, David Frank, Sridar V. Chittur, and Scott A. Tenenbaum Abstract Microarrays are extensively used to evaluate the effects of compounds on gene expression in the cells. Most of the studies so far have analyzed the transcriptome of the cell. The basic assumption of this approach is that the changes in gene expression occur at the level of transcription of a gene. However, changes often occur at the posttranscriptional level and are not reflected in the analysis of whole transcriptome. We have pioneered the development of “ribonomic profiling” as a high-throughput method to study posttranscriptional regulation of gene expression in the cell. This method is also often referred to as RIP-CHIP. In this chapter, we describe how to use the RIP-CHIP technology to assess the posttranscriptional changes occurring in the cell in response to treatment with a drug. Key words: RIP-CHIP, Ribonomics, Posttranscriptional gene regulation, RNA-binding Protein (RBP), Immunoprecipitation (IP), Microarray expression profiling, Drug development

1. Introduction The process of drug development is a lengthy and laborious process in which a novel chemical entity, which has been identified to produce a beneficial effect in the human body, is subjected to rigorous testing to ensure safety and efficacy before it can be approved as a drug. In the very first phase of drug development (called preclinical testing), the investigational drug is extensively tested in the laboratory. These tests are done both in vitro (using cultured cells) and in  vivo (by administering the drug on small animals, e.g., rodents). Since changes in gene expression at the molecular level can be detected before any biological symptoms appear, in vitro studies designed to study the changes in molecular genetic events that occur in a cell in response to the drug can greatly enhance drug development. It also leads to a better Sridar V. Chittur (ed.), Microarray Methods for Drug Discovery, Methods in Molecular Biology, vol. 632, DOI 10.1007/978-1-60761-663-4_10, © Humana Press, a part of Springer Science+Business Media, LLC 2010

159

160

Jain et al.

understanding of the mode of action of the drug and any potential side effects it may generate. Most often, these genetic events are studied in terms of global changes in gene expression levels. High-throughput technologies such as microarrays can be used to quantitatively study the steady-state mRNA level of cells (also known as the transcriptome) and can greatly reduce the time and cost involved in the evaluation of a new drug (1–3). In fact, the US-FDA has set up a database for storing microarray gene-expression data specifically for pharmaco- or toxicogenomic studies (4). However, the accumulated level of RNA does not always directly correlate with the level of protein output of the cell. All events, beginning with the transcription of mRNA to processing, transport, localization, stability, degradation, and translation are regulated by various RNA-Binding Proteins (RBPs), and contribute significantly to the final amount of protein product produced (for a review see ref. (5–7)). In addition, RBPs can sequester functionally-related mRNAs together to allow coordinated expression of functionally related genes. These “posttranscriptional operons” may allow the cell to generate a rapid response to an incoming signal by regulating the stability and translation of specific mRNAs (8). The posttranscriptional operon model is able to explain the discordance that is often observed between the levels of mRNA and the protein of any given gene. Just because a gene is transcribed does not necessarily mean that it is translated; it may be kept on hold, or degraded by various RBPs. Experimental evidence from many different labs in recent years has demonstrated that mRNAs which either participate in a common function or subcellular localization tend to be part of the same RNP complex, thus providing further evidence for the existence of posttranscriptional operons. Often, these RBPs bind to mRNAs through elements that contain sequence and structural specificity and are present in their 5¢ or 3¢ untranslated regions. Each mRNA often contains binding-sites for more than one RBP, and each RBP has the potential to associate with more than one mRNA, giving rise to combinatorial, systems-level regulatory networks with the potential for tremendous complexity, yet elegant simplicity (9). This combinatorial posttranscriptional networking likely gives rise to much of the complexity of the human genome that has yet to be fully understood. Many RBPs e.g. HuR and IMP-1 are known to be involved in regulating the growth of cells by enhancing the expression of specific genes involved in cell proliferation (10–13). It is possible that some anticancer drugs may function through modulating the function of these RBPs to prevent cell proliferation. To study these events, we have helped pioneer the development of high-throughput technologies to analyze the entire subset of mRNAs associated with a particular RBP at a global level, using microarrays (14–16). This technique which employs

RIP-CHIP in Drug Development

161

immunoprecipitation of endogenously formed mRNP complexes using an antibody specific to the RBP in question, followed by purification of associated mRNAs and their global, quantitative analysis using a genomic read-out such as microarrays, sage analysis or deep-sequencing, has been termed RIP-CHIP (RNA-Binding Protein Immunoprecipitation-Microarray (Chip)), or “ribonomic profiling”. Ribonomic profiling studies in response to drug treatment of cells will further enhance our understanding of the effects of the drug on the cell. We hope that this will lead to better assessment of the safety and efficacy of a potential drug.

2. Materials We recommend that for the first part of this protocol (i.e. treatment of cells with a novel chemical compound), all standard precautions should be taken to minimize accidents and spills, and the compound should be handled in accordance with local EPA regulations. While we demonstrate this application using Dutasteride as our investigational drug, this protocol can be used for any drug of choice. For the second part of this protocol (RIPCHIP with lysate obtained from treated cells), precautions should also be taken to minimize RNAse contamination (see Notes 1 and 2). All reagents, glassware and plasticware should either be purchased RNAse–DNAse free or be treated with 0.1% DEPC, to ensure that they are RNAse free. All cell culture should be done with cell-culture grade media. 2.1. E quipment

1. Agilent Bioanalyzer 2100 system. 2. Nanodrop ND1000 Spectrophotometer. 3. Affymetrix Genechip System.

2.2. Materials and Reagents for Preparing PLB Lysates from Cells Treated with the Drug of Choice

1. LNCap cell line maintained in RPMI1640 containing 10% Fetal Bovine Serum (FBS), 100 mg/ml Penicillin and 100 U/ ml Streptomycin. 2. Dutasteride dissolved in DMSO to a final concentration of 10 mg/ml. 3. Phosphate-buffered Saline (PBS). 4. Teflon cell scrapers (Fisher cat # 08-773-2). 5. Nuclease-free Water (Ambion cat. No. 9932). 6. Polysome Lysis Buffer (PLB): 10  mM HEPES (pH 7.0), 100  mM KCl, 5  mM MgCl2, 0.5% NP-40, prepared in nuclease free water. Typically, we prepare a 10× stock of this

162

Jain et al.

buffer, from which a ready to use polysome lysis buffer can be prepared. At the time of use, 1 mM DTT, 100 Units/ml RNAse OUT (Invitrogen cat. No. 100000840) and CompleteTM proteinase inhibitor cocktail (Roche cat. No. 11836170001) should be added (see Note 3). 2.3. Materials and Reagents for RNP-IP (RIP)

1. Antibody to the RNA Binding Protein of interest. 2. Antibody binding matrix (e.g. Protein A Sepharose (Sigma cat. No. P3391) or Protein G Sepharose beads (Sigma cat. No. P7700), or the preswollen beads from GE Healthcare cat. No. 17-5280-01, or Magnetic beads (Dynabeads) from Invitrogen cat. Nos. 100.04D and 100.02D). Choice of beads depends on the antibody being used (see Notes 4 and 5). 3. 0.1 M DTT. 4. 0.5 M EDTA (pH 8.0), (Ambion Cat. No. 9260G). 5. 5 M Ammonium Acetate (Ambion cat. No. 9070G). 6. 7.5 M Lithium Chloride (Ambion cat. No. 9480). 7. Glycogen, 5 mg/ml (Ambion cat. No. 9570). 8. Proteinase K, 20 mg/ml (Ambion cat. No. 2546). 9. Acid-Phenol:Chloroform pH 4.5 (with IAA, 25:24:1), (Ambion cat. No. 9722). 10. Absolute Ethanol. 11. RNAse OUT, 40 U/ml (Invitrogen cat. No. 100000840) or RNAse Inhibitor (New England Biolabs cat. No. M0307) or Superase•IN™ (Ambion cat. #2696, 20 U/µL). 12. Nuclease-free Water (Ambion cat. No. 9932). We recommend using nuclease free water to prepare the following buffers: 13. NT-2 Buffer: 50  mM Tris–HCl (pH 7.4), 150  mM NaCl, 1 mM MgCl2, 0.05% NP-40. Typically, we prepare a 5× stock of NT-2 Buffer, from which a ready to use buffer can be prepared as and when desired. 14. NET-2 Buffer (Binding buffer for RNP to antibody): NT-2 buffer supplemented with 20  mM EDTA (pH 8.0), 1  mM DTT, 100 U/ml RNAse OUT. 15. Proteinase K digestion buffer: NT-2 buffer supplemented with 1% SDS, 1.2 mg/ml Proteinase K. 16. Antibody binding buffer: 5% BSA in NT-2 buffer.

2.4. Materials and Reagents for RNA QC and Microarray Profiling

1. RNA6000 Nanokit (Agilent cat No. 5067-1511). 2. GeneChip Human Gene 1.0 ST Array kit (cat No. 901147).

RIP-CHIP in Drug Development

163

3. Methods The method described here has been successfully used with human cells. A flow-chart explaining the various steps involved in RIP-CHIP is shown in Fig. 1. We have successfully immunoprecipitated RNA Binding Protein–mRNA complexes (RNP-IP) using both sepharose and magnetic beads. Both methods are detailed here. Typically, immunoprecipitations are performed in a 1.0 ml volume, and this allows for reproducible kinetics, reduced mRNA reassortment potential, and ease of calculations.

Fig. 1. Typical RIP-Chip workflow

164

Jain et al.

3.1. Treatment of Cells with the Experimental Drug and Preparation of PLB Lysate

1. Grow LNCaP cells to about 30–50% confluency, on 150 or 100 mm plates. 2. Aspirate the media and replace with media containing the drug of choice dissolved in DMSO or Ethanol (see Note 6). 3. For the untreated control plates, add media containing an equal volume of the solvent used to dissolve the drug (see Note 7). 4. Incubate at appropriate growth conditions for the desired amount of time. If treating for more than 24 h, it is important to replace the media with fresh drug containing growth media every 24 h. 5. At the end of incubation time, aspirate media, wash the cells twice with cold PBS. 6. Keep the plates on ice, add 10 ml of ice-cold PBS, and gently harvest the cells using a cell-scraper. 7. Transfer the cell suspension to a 50  ml conical tube, and centrifuge at 2,500×g to pellet the cells. 8. Wash the cells once with ice-cold PBS and transfer to an eppendorf tube. 9. Aspirate PBS from the cell pellet and add equal pellet volume of PLB (containing protease and RNAse inhibitors) to the cell pellet. Mix by pipetting up and down until the mixture looks homogeneous. Keep on ice for 5 min to allow the hypotonic PLB buffer to swell the cells. Dispense into 1 ml aliquots and store at −80°C. The cell lysates are ready to proceed to RBP immunoprecipitation (see Note 8).

3.2. Coating of the Beads with Antibody 3.2.1. Coating Sepharose Beads

1. Swell the desired amount of protein G or protein A agarose beads in NT2 buffer containing 5% BSA. Typically, we add 30 ml of buffer to 1 g of beads. This yields 20% slurry (1 ml of slurry yields about 200  ml of packed bead volume). Alternatively, preswollen beads equilibrated with NT2 buffer can be used. 2. Add the immunoprecipitating antibody or serum and incubate for at least 1 h at room temperature on a rotating device OR overnight (or longer) at 4°C (see Notes 9 and 10). Typically, for each immunoprecipitation reaction, we use 300  ml of slurry, dilute it with 1 ml of NT2 buffer containing 5% BSA, and add 5 mg of commercially purchased antibody.

3.2.2. Coating Magnetic Beads

For both Dynabeads protein A and protein G, the method is essentially as recommended by the manufacturer. 1. For each immunoprecipitation reaction, use 50 ml of Dynabeads A suspension, wash twice with 0.5 ml of NT-2 buffer using the magnet Dynal. Resuspend the beads in 100  ml of the same buffer and add 5 mg of the antibody.

RIP-CHIP in Drug Development

165

2. Incubate with rotation for 30 min at room temperature. 3. Wash three times with 0.5 ml of NT-2 buffer. Proceed to step 6. Unless cross-linking of antibody to the beads is desired. 4. For cross-linking the antibody to the beads (optional), resuspend the beads in 1.0 ml of 0.2 M triethanolamine (Sigma cat. No. T0449). Wash twice with 1.0 ml of 0.2 M triethanolamine. 5. Add 1.0  ml of freshly prepared 20  mM DMP (Dimethyl pimelimidate, Sigma cat. No. D8388) in 0.2  M triethanolamine and incubate with rotational mixing for 30 min at room temperature. 6. Stop the cross-linking reaction by replacing the DMP solution with 50  mM Tris–HCl (pH 8.0). Incubate with rotational mixing for 15 min at room temperature. 7. Wash three times with NT-2 buffer. The beads are now ready to proceed with RNP-binding. We have not attempted to store antibody coated magnetic beads for extended periods. 3.3. Immuno ­precipitation of RNA Binding Protein–mRNA Complex (RIP)

The method for immunoprecipitation on sepharose vs. Dynabeads is almost identical with minor variations. 1. Wash the sepharose beads from step 2 of Subheading 3.2.1, six times with 1.0 ml of cold NT-2 buffer. 2. Resuspend the antibody-coated beads (sepharose or Dynabeads) in 900  ml of NET-2 buffer. Typically for each immunoprecipitation, we mix:

850 ml NT-2, 35 ml 0.5 M EDTA, 10 ml 0.1 M DTT, 5 ml RNAse OUT to obtain 900 ml of NET-2 buffer.

3. Thaw the PLB lysate quickly by holding between your fingers and centrifuge at 14,000×g for 10  min at 4°C. Remove 100 ml of the supernatant and add to the beads in NET-2. The final volume of the immunoprecipitation reaction will now be 1.0 ml. 4. Mix and briefly centrifuge to bring the beads down. Remove 100 ml of the supernatant. This represents the starting material or “input”, which will be processed alongside the immunoprecipitation to monitor any RNAse contamination, and to compare with immunoprecipitated mRNAs at the end. 5. Tumble the reactions end-over-end for 3  h to overnight at 4°C (1 h at room temperature for Dynabeads) see Notes 11 and 12. 6. It is a good idea to also include a negative control for the immunoprecipitation by using an unrelated antibody, e.g. T7-tag antibody, generic IgG, or preimmune serum to monitor the nonspecific binding to the beads. 7. After incubation is complete, wash the beads six times with cold NT-2 buffer and resuspend in 150 ml of proteinase K buffer

166

Jain et al.

(see Notes 13–15). Typically for each immunoprecipitation we mix 124 ml NT-2 buffer, 15 ml of 10% SDS, 9 ml of 20 mg/ml proteinase K to obtain 150 ml of proteinase K digestion buffer. Also, add the same amount of SDS and proteinase K to the tubes labeled “input” and bring up the volume to 150 ml with NT-2. Incubate all tubes at 55°C for 30  min to digest the protein and release the RNA from RNP complex. 3.4. Purification of RNA

RNA is purified from the supernatants using standard phenol– chloroform extraction (see Note 16). 1. Add 150 ml phenol:chloroform:isoamyl alcohol to the tubes containing beads in proteinase K buffer. For Dynabeads, remove the supernatant using the magnet and place it in a separate tube. Add phenol–chloroform only to the supernatant. Vortex to mix and centrifuge at 14,000×g for 10 min to separate the phases. 2. Remove the aqueous phase carefully and place it in a new tube. Add 150 ml of chloroform and repeat the extraction. 3. To each tube add 50  ml of 5  M ammonium acetate, 15  ml 7.5 M Lithium Chloride (optional), 5 ml of 5 mg/ml glycogen, and 850 ml absolute ethanol. Mix and keep at −80°C from 1 h to overnight to precipitate the RNA. 4. To collect the RNA centrifuge at 14,000×g for 30  min at 4°C, pour off the supernatant, wash the pellet once with 80% ethanol, centrifuge again at 14,000×g for 30  min at 4°C. Pour off the supernatant, dry the pellet, and resuspend in 20 ml of RNAse-free water. 5. If desired, the RNA can be further purified using Qiagen RNEasy micro clean up, as per manufacturer’s instructions. However, we have not found it to be necessary.

3.5. Assessment of RNA Quality

1. Typically, optical absorbance of the RNA can be measured using a Nanodrop® spectrophotometer. Ideally, we expect both the A260/A280 and A260/A230 ratios to be close to 2.0, implying the purity of RNA and the absence of any contaminating proteins or chemicals. If this ratio is less than 1.8, there may be problems with further downstream applications. 2. The molecular weight profile of the subset of RNAs immunoprecipitated can be analyzed on a nanochip using Agilent’s BioAnalyzer. Nanochip is a convenient alternative to using formaldehyde–agarose gels. For the total RNA (or “input”), it is helpful to evaluate the 18s/28s rRNA ratio. A ratio between 1.6 and 2.0 indicates RNA of good integrity. However, for the immunoprecipitated material, such ratio cannot be obtained unless rRNAs are known to be targets of the RNA Binding Protein in question. And so, a bioanalyzer

RIP-CHIP in Drug Development

167

profile of immunoprecipitated RNA only serves to provide information about any extensive degradation of RNA. Usually, if the total RNA shows a good 18s/28s rRNA ratio, we can safely assume the immunoprecipitated RNA to be of good integrity. 3.6. Synthesis of Labeled cRNA and Microarray Hybridization

1. We recommend using 50–100  ng of immunoprecipitated material if available. The RNA is first converted to T7-random primed double stranded cDNA using the Affymetrix WT cDNA synthesis kit as per the manufacturer’s protocol (18) (see Note 17). 2. This is then converted to amplified RNA (aRNA) by in vitro transcription using GeneChip WT cDNA amplification kit as per the manufacturer’s protocol. 3. The aRNA is subsequently reverse transcribed back to cDNA in a manner similar to the first strand cDNA, but using random primers and a dNTP mix containing dUTP. 4. Then, the resulting cDNA is fragmented using Uracil DNA Glycosylase and Apurinic/apyrimidinic endonuclease 1, which causes strand breakage wherever UDP is incorporated. 5. The resulting fragments are end-labeled with terminal deoxynucleotidyl transferase and a biotinylated DNA labeling reagent. 6. Finally, the labeled fragments of cDNA are hybridized to the Affymetrix Gene ST 1.0 array.

3.7. Analysis of the Microarray Data

1. CEL files for the IP samples and the corresponding input samples were processed using the apt-probeset-summarize binary provided by Affymetrix. 2. One qualitative (DABG) and two quantitative (RMA, PLIER) measures of expression were derived for each gene-level metaprobeset on the chips using the Affymetrix Powertools apt-probeset-summarize command with the parameters given below (where Gene-ST-lib stands for the most up to date Affymetrix Gene_ST annotation library and genome_build stands for the current build of the relevant genome):

apt-probeset-summarize -a rma -a quant-norm,pm-gcbg, plier -a dabg.neglog10=true --precision 10 -p lib/(Gene-ST-lib).pgf



-c lib/(Gene-ST-lib).clf -b lib/(Gene-ST-lib).bgp



-m lib/(Gene-ST-lib).(genome_build).full.mps -o gene/ --cel-files cel_list

3. The output files for each of the three measures are imported into an R session for further analysis. 4. A DABG p-value cutoff of 0.001 was used for calling present probesets across the RBP IP chips.

168

Jain et al.

5. For the probesets present in all IP replicates, a t-test for overexpression in IP compared to input is performed. The subset of genes both present in all IP replicates and overexpressed in IP with a p-value of 0.05 or less across both RMA and PLIER methods are called IP targets.

4. Notes 1. For general precautions on working with RNA, all instruments, glassware, and plastic-ware that touch cells or cell lysates should be certified DNase-free and RNase-free or should be prewashed with RNase Zap (Ambion, cat. #9780; 9782) or RNase Away (Molecular BioProducts cat. #7001) followed by DEPC water and allowed to air dry. 2. Generally, it is more convenient to purchase solutions which are certified DNase-free and RNase-free from the manufacturer. In our experience, RNAse-free solutions from Ambion perform well. 3. One complete protease inhibitor tablet is sufficient for 10 ml of Polysome Lysis Buffer (PLB). Typically, we prepare 10  ml of ready to use solution at a time, which can be dispensed into 1 ml aliquots and stored at −80°C for extended periods of time without any deleterious effects (more than 6 months). 4. The type of beads used will depend on the binding affinity of the particular antibody to the beads. Most mouse monoclonal antibodies bind strongly to protein G beads, and rabbit polyclonal antibodies bind strongly to both protein A and G beads. Consult the binding chart for the beads from the manufacturer to determine the best choice of beads for the particular isotype of antibody being used. 5. We have observed lower background when using magnetic beads, however, different antibodies may perform differently on different beads. The user may want to determine the best choice of beads depending on their application. 6. Most pharmacological compounds are soluble in either DMSO or/and ethanol. However, DMSO can sometimes alter the growth of cells. Therefore, when possible, it is better to use ethanol as a carrier to administer the drug. In addition, the drug solution should be concentrated enough, so that final concentration of solvent in the culture medium does not exceed 0.01%.

RIP-CHIP in Drug Development

169

7. For example, if 10 ml of drug solution is to be added to 10 ml of culture media to achieve the desired concentration, the culture media for untreated cells should have an equal volume of solvent alone. 8. The volume of cell lysate obtained depends on the cytoplasmic volume of the cell; typically, we obtain 400–500  ml lysate from six 15 cm plates. The objective is to get an extremely concentrated lysate ranging in concentration from 20 to 50  mg/ml of total protein. On average, 1–5 × 106 cells will generate enough lysate for a typical immunoprecipitation reaction. 9. Typically, 2–20 µL sera or 5  mg of commercially available antibody per immunoprecipitation reaction is used, depending on the affinity of the antibody. We have found that using excess antibodies when possible greatly reduces nonspecific binding of proteins to sepharose beads, possibly by reducing the number of available binding sites on the beads. 10. Antibody-coated beads can be prepared in bulk and stored at 4°C with 0.02% sodium azide. 11. To minimize reassortment potential, the final volume of resuspended beads in NET2 buffer should correspond to approximately ten times the original volume of the RNP lysate being used (a 1:10-fold dilution of lysate). Since we typically perform our reactions in 1 ml, we use 100 µl lysate. Performing the immunoprecipitation reactions in larger volumes can help decrease background problems. 12. Provided there is no RNA degradation or RNAse problems, longer incubations will result in better recovery of RNA. 13. A concern when isolating mRNP complexes is the possibility of exchange of proteins and mRNAs. In principle, crosslinking agents, such as formaldehyde, could prevent this (16,17). However, we have found mRNA exchange to occur at a minimal level using these methods described here, and crosslinking therefore, to be unnecessary. In some cases, formaldehyde actually can interfere with subsequent mRNA detection methods and increase the background (15). 14. Several additional washes with NT2 buffer supplemented with 1–3 M urea can increase specificity and reduce nonspecific binding. However, it is important to first determine whether urea disrupts binding of the antibody to the target protein and/or the RBP-mRNA interaction. 15. It is a good idea to remove an aliquot during the last wash, to test the efficiency of immunoprecipitation by Western blotting. The proteins can be eluted off the beads by resuspending the beads in 1× SDS-PAGE loading buffer followed by heating at

170

Jain et al.

95°C. The beads can then be centrifuged down and the supernatant directly applied on SDS-PAGE. This is helpful when using a new antibody for the first time. Many commercially available antibodies are not of good quality and in our experience, sometimes we had to try as many as five different antibodies before we were able to successfully immunoprecipitate the target RBP. 16. We have tried to use the Qiagen RNEasy kit to avoid the phenol–chloroform extraction step. After the proteinase K digestion, the supernatant can be removed, mixed with buffer RLT/b-ME, and RNA can be purified from the spin column according to manufacturer’s directions. Although this process offers ease of use and saves time, we found the recovery of RNA to be at least ten-fold less when compared to phenol– chloroform extraction. 17. The ribominus step from the standard Affymetrix protocol (18) is omitted here since we are working with subset of mRNAs and rRNA is not expected to be present in the sample unless the RBP in question binds to rRNA. References 1. Braxton S, Bedilion T (1998) Integration of microarray information in the drug development process. Curr Opin Biotechnol 9:643–649 2. Gunther EC, Stone DJ, Gerwien RW, Bento P, Heyes MP (2003) Prediction of clinical drug efficacy by classification of drug-induced genomic expression profiles in vitro. Proc Natl Acad Sci USA 100(16):9608–9613 3. Lord PG, Nie A, McMillian M (2006) Applications of Genomics in preclinical drug safety evaluation. Basic Clin Pharmacol Toxicol 98(6):537–546 4. http://www.fda.gov/nctr/science/centers/ toxicoinformatics/ArrayTrack/index.htm 5. Moore MJ (2005) From birth to death: the complex lives of eukaryotic mRNAs. Science 309:1514–1518 6. Keene JD (2001) Ribonucleoprotein infrastructure regulating the flow of genetic information between the genome and the proteome. Proc Natl Acad Sci USA 98(13):7018–7024 7. Sanchez-Diaz P, Penalva LO (2006) Posttranscription meets post-genomic: the saga of RNA binding proteins in a new era. RNA Biol 3(3):101–109 8. Keene JD (2002) Tenenbaum SA: eukaryotic mRNPs may represent posttranscriptional operons. Mol Cell 9(6):1161–1167 9. Hieronymus H, Silver P (2004) A systems view of mRNP biology. Genes Dev 18:2845–2860

10. López de Silanes I, Lal A, Gorospe M (2005) HuR: post-transcriptional paths to malignancy. RNA Biol 2(1):11–13 11. Mazan-Mamczarz K, Hagner PR, Corl S, Srikantan S, Wood WH, Becker KG, Gorospe M, Keene JD, Levenson AS, Gartenhaus RB (2008) Post-transcriptional gene regulation by HuR promotes a more tumorigenic phenotype. Oncogene 27:6151–6163 12. Mazan-Mamczarz K, Patrick RH, Dai B, Wood WH, Zhang Y, Becker KG, Liu Z, Gartenhaus RB (2008) Identification of transformation-related pathways in a breast epithelial cell model using a ribonomics approach. Cancer Res 68:7730–7735 13. Kato T, Hayama S, Yamabuki T, Ishikawa N, Miyamoto M, Ito T, Tsuchiya E, Kondo S, Nakamura Y, Daigo Y (2007) Increased expression of insulin-like growth factor-II messenger RNA binding protein-1 is associated with tumor progression in patients with lung cancer. Clin Cancer Res 13: 434–442 14. Tenenbaum SA, Carson CC, Lager PJ, Keene JD (2000) Identifying mRNA subsets in messenger ribonucleoprotein complexes by using cDNA arrays. Proc Natl Acad Sci USA 97:14085–14090 15. Tenenbaum SA, Lager PJ, Carson CC, Keene JD (2002) Ribonomics: identifying mRNA subsets in mRNP complexes using antibodies

RIP-CHIP in Drug Development to RNA-binding proteins and genomic arrays. Methods 26(2):191–198 16. Penalva LO, Tenenbaum SA, Keene JD (2004) Gene expression analysis of messenger RNP complexes. Methods Mol Biol 257:125–134 17. Niranjanakumari S, Lasda E, Brazas R, GarciaBlanco MA (2002) Reversible cross-linking

171

combined with immunoprecipitation to study RNA-protein interactions in  vivo. Methods 26(2):182–190 18. Affymetrix. GeneChip WT sense strand target labeling assay manual, 701880. http://www. affymetrix,com/support/downloads/manuals/wt_sensetarget_label_manual

Chapter 11 ChIPing Away at Global Transcriptional Regulation Kelly Jackson, James Paris, and Mark Takahashi Abstract The regulation of gene expression impacts all aspects of cell biology and biochemistry. As we gain a greater understanding of the mechanisms involved in this process, we also begin to unveil its complexities. The delicate balancing act played out by the multitude of DNA interacting proteins can easily become unhinged. The implications of this may potentially lead to cell death or a diseased state. Recent microarray technologies are now allowing scientists to begin the journey into characterizing the relationship between gene expression and DNA modifying proteins. For example, genome-wide studies of protein–DNA interactions, such as Chromatin Immunoprecipitation on arrays (also referred to as ChIP-chip), allow for a global view of where and when DNA binding proteins interact. A number of microarray based genome wide methodologies have emerged based upon these same principles. Here, we outline a methodology that we have developed using the ChIP-chip technique. Application of this methodology is easily adaptable to different cell types, antibodies, and to a variety of array platforms. Key words:  Chromatin immunoprecipitation, Microarrays, Transcription factors, Gene expression, Whole genome profiling

1. Introduction Our understanding of gene expression has come a long way from early definitions of how a gene is transcribed and translated into a functional protein e.g. (1). For example, the current number of transcription factors known to this point has grown from the few that encompassed the basic transcriptional machinery to thousands. Many of these regulatory proteins play very specific roles and can have far reaching implications on the fate of the cell. Furthermore, the initial process of gene transcription can no longer be viewed as an isolated event. A multitude of regulatory events may occur that can influence not only whether or not a gene is even transcribed but also quantity and stability e.g., DNA methylation (2).

Sridar V. Chittur (ed.), Microarray Methods for Drug Discovery, Methods in Molecular Biology, vol. 632, DOI 10.1007/978-1-60761-663-4_11, © Humana Press, a part of Springer Science+Business Media, LLC 2010

173

174

Jackson, Paris, and Takahashi

Even the expression of one gene can have huge implications on the expression of others. How genes are transcribed can be as basic or complex as the number of genes in the genome. Until recently, much of this work had been carried out on a per gene basis. However, two advances that have occurred in the past decade changed the way we now view genomics and have pushed our knowledge of transcriptional regulation to new levels. The first was the completion of the human genome sequencing projects (3, 4). This has been pivotal in providing the necessary road map of all the genes including all of the intergenic regions. The second has been the advent of microarray technologies (5). By combining the information learned from sequencing the genome and applying it to microarrays, we are now capable of investigating gene expression at the genome wide scale. The question of what sequence to lay down on an array in order to profile transcriptional regulation of a gene has always been a difficult question to answer. However, the completion of both the mouse and human genomes have allowed for the generation of promoter arrays. These, in combination with CpG island arrays, have been used in a variety of genome-wide analyses of protein–DNA interactions including histones, transcription factors, and the machinery for both transcription and replication (6, 7). More recently, interest has been growing in the scientific community surrounding DNA modifications such as methylation acquired during one’s lifetime. Once again, by using antibodies raised against the methylated form of cytosine (5-methylcytosine), the same promoter arrays can be used to profile the extent and location of DNA methylation (8, 9). In the following protocol, we provide a robust method which we have used extensively with both spotted CpG island arrays developed in-house and Agilent CpG island arrays. However, the current protocol is versatile enough that it can be used in conjunction with a number of commercially available platforms. For example, with minor modifications, it can also be applied to profile the methylation pattern of the genome. Thus, we present a flexible protocol that is relatively inexpensive and a reliable means of conducting genome-wide analyses.

2. Materials 2.1. Cell Culture and Cell Pellet Collection and Lysis

1. Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% fetal bovine serum (FBS). 2. Sterile, purified water. 3. 25× Protease Inhibitor Cocktail made by dissolving one Complete tablet in 2 mL sterile, purified water.

ChIPing Away at Global Transcriptional Regulation

175

4. 37% formaldehyde used as supplied by manufacturer, monitored for freshness and lack of precipitate. 5. 2.5 M Glycine Quenching Solution; dissolving 187.675 g/L glycine in sterile, purified water. 6. RIPA Lysis Buffer: 0.1% SDS, 1% sodium deoxycholate, 150 mM NaCl, 10 mM phosphate buffer (pH 7.2), 2 mM EDTA, 0.2 mM NaVO3, 1% IGEPAL CA-630. 7. Cell Scrapers. 8. 2 mL microfuge tubes. 9. Liquid Nitrogen used to flash freeze cell pellets if storing cell pellets for subsequent lysis. 10. Cryostorage Unit for longer term storage of cell pellets. 2.2. Chromatin Shearing and Fragment Size Confirmation by Agarose Gel Electrophoresis

1. 1.5 mL microfuge tubes. Conical bottom tubes are essential to prevent the loss of material during sonication. 2. Packed ice cooler for tube support and cooling during sonication. 3. Sonicator (Branson, Danbury, CT) or a suitable equivalent with a probe designed to fit microfuge tubes. 4. Heating Block or 65°C Water Bath. 5. 5 M Sodium Chloride solution in sterile, purified water. 6. Apparatus for agarose gel electrophoresis, microwave for preparation of agarose gel. 7. UltraPure Agarose powder. 8. 50× TAE Tris/Acetate/EDTA Agarose Gel Running Buffer; dissolving 242 g/L Tris powder and 186 g/L EDTA in sterile purified water containing 57.1 mL/L glacial acetic acid. 9. UV light source and gel documentation system.

2.3. Cell Lysate Preclearing and Incubation with Antibody

1. Rocking platform at 4°C. 2. Bovine Serum Albumin. 3. Sonicated Salmon Sperm DNA. 4. RIPA Lysis Buffer: 0.1% SDS, 1% sodium deoxycholate, 150 mM NaCl, 10 mM phosphate buffer (pH 7.2), 2 mM EDTA, 0.2 mM NaVO3 , 1% IGEPAL CA-630. 5. Pansorbin Staph A cells either lyophilized (see Note 1) or supplied as a 10% suspension may be used. Cells are washed and resuspended in buffer as per manufacturer’s instructions. Staph A cells may be preblocked with 10 mg/mL BSA and 10 mg/mL sonicated salmon sperm DNA for a minimum of 3 h to overnight at 4°C with rocking prior to washing and resuspension in RIPA buffer containing protease inhibitors as a 10% suspension.

176

Jackson, Paris, and Takahashi

6. Alternative to Staph A cells: Protein A, G or Protein A/G agarose beads. 7. Antibody suitable for immunoprecipitation reactions. 2.4. Recovery and Washing of Antibody–Protein–DNA Complexes

1. Rocking platform at room temperature.

2.5. Elution of the Antibody–Protein–DNA Complex

1. Elution Buffer: 1% Sodium Dodecyl Sulphate, 50  mM NaHCO3 dissolved in sterile, purified water.

2.6. Decrosslinking the Protein–DNA Complex, RNAse A Digest

1. Heating Block or 65°C Water Bath.

2.7. Immunoprecipi­ tated DNA: Fragment Recovery and Proteinase K Digest

1. 70% ethanol.

2. 100 mg/mL yeast tRNA. 3. 1× Dialysis Buffer: 2 mM EDTA, 50 mM Tris–HCl pH 8.0. 4. IP Wash Buffer: 100 mM Tris–HCl pH 8.0, 500 mM LiCl, 1% IGEPAL CA-630, 1% sodium deoxycholate.

2. Horizontal shaker capable of holding microfuge tubes, or with a microfuge tube holder fastened to it.

2. 5 M Sodium Chloride solution in sterile, purified water. 3. 10 mg/mL RNAse A. 4. 100% Ethanol.

2. Proteinase K. 3. 5× Proteinase K Buffer: 50 mM Tris–HCl (pH 8.0), 25 mM EDTA, 1.25% SDS. 4. Heating Block or 42°C Water Bath. 5. QiaQuick PCR Purification Columns (Qiagen, Missisauga, ON). 6. Sterile, purified water. 7. Savant SpeedVac (GMI, Ramsey, MN).

2.8. Immunoprecipi­ tated DNA: Primer A Annealing and Extension

1. Thermocycler. 2. PCR tubes with cap strips. 3. Sequenase Version. 2.0 13 U/mL. 4. 5× Sequenase Buffer. 5. 0.1 M DTT. 6. 100 pmol/mL Primer A: GTT TCC CAG TCA CGA TCN NNN NNN NN. 7. Sterile, purified water. 8. 3 mM dNTP set. 9. 500 mg/mL BSA. 10. Round A Setup Reaction Mix (3mL for each immunoprecipitated sample): 2 mL 5× Sequenase Buffer, 0.6 mL Primer A,

ChIPing Away at Global Transcriptional Regulation

177

0.4 mL sterile, purified water. Prepare extra volume to account for losses due to pipetting. 11. Round A Reaction Mix: (5.05 mL for each immunoprecipitated sample): 1 mL 5× Sequenase Buffer, 1.5 mL 3  mM dNTP’s, 0.75 mL 0.1  M DTT, 1.5 mL 500 mg/mL BSA, 0.3 mL Sequenase. Prepare extra volume to account for losses due to pipetting. 2.9. Immunoprecipi­ tated DNA: Primer B Annealing, aa-dUTP Labeling and Amplification

1. 100 pmol/mL Primer B: GTT TCC CAG TCA CGA TC. 2. 25 mM MgCl2 (Sigma-Aldrich, Oakville, ON). 3. 10× PCR Buffer: 500 mM KCl, 100 mM Tris (pH 8.3). 4. 50× aa-dUTP/dNTP’s: 1  mg aa-dUTP resuspended in 19.11 mL sterile, purified water. Add 31.85 mL each of dATP, dGTP, dCTP (25 mM final concentration) and 12.74 mL dTTP (10 mM final concentration). Final aa-dUTP concentration is 15 mM. 5. 5 unit/mL Taq Polymerase. 6. Round B Reaction Mix: (85 mL for each immunoprecipitated sample): 8 mL 25 mM MgCl2, 10 mL 10× PCR Buffer, 2 mL 50× aadUTP/dNTP’s, 1 mL 100 pmol/mL Primer B, 1 mL 5 unit/mL Taq, 63 mL water. Prepare extra volume to account for losses due to pipetting.

2.10. Amplified Immunoprecipitated DNA: PCR Purification

1. CyScribe GFX Purification Kit (GE Healthcare, Piscataway, NJ). 2. 80% Ethanol. 3. 0.1 M Sodium Bicarbonate pH 9.0. 4. Savant SpeedVac (GMI, Ramsey, MN).

2.11. Amplified DNA: Alexa Fluor Labeling and Purification

1. CyScribe GFX Purification Kit (GE Healthcare, Piscataway, NJ). 2. 80% Ethanol. 3. 0.1 M Sodium Bicarbonate pH 9.0. 4. Savant SpeedVac (GMI, Ramsey, MN).

2.12. Microarray Hybridization and Wash Protocol

1. DIG Easy Hyb solution (Roche, Laval, QC). 2. Heating Block or 65°C Water Bath. 3. 10 mg/mL calf thymus DNA. 4. 10 mg/mL yeast tRNA. 5. 24 × 60 mm glass coverslip. 6. Hybridization chamber (microscope boxes using slides as rails to support the arrays). 7. 37°C incubator. 8. 1× SSC and 0.1% SDS (50 mL of 20× SSC stock, 10 mL of 10% SDS, bring to 1 L with sterile water).

178

Jackson, Paris, and Takahashi

9. 1× SSC. 10. 0.1× SSC. 11. Centrifuge with plate holder brackets.

3. Methods Using formaldehyde fixation, reversible protein–protein and protein–DNA linkages may be formed. Following cell lysis and chromatin shearing, DNA fragments associated with a specific transcription factor may be immunoprecipitated using a suitable antibody. Subsequent crosslink reversal, protein and RNA degradation, and DNA purification yields fragments which are randomly amplified to generate as representative a population as possible between two samples: i.e., + and − antibody pull-downs, or + antibody vs. total input chromatin in order to show the relative enrichment of bound species. Amplification with subsequent labeling and hybridization of the immunoprecipitated DNA fragments to an intergenic-region microarray allows for the determination of protein in vivo binding sites, leading to the possible identification of novel transcription factor–gene interactions, as well as confirming data obtained through sequence analysis. 3.1. Cell Culture and Cell Pellet Collection and Lysis

Depending on the endogenous expression levels of your protein of interest, and the relative efficiency with which the antibody binds and precipitates this protein, it may be possible to obtain a sufficient yield of immunoprecipitated DNA starting with adherent cells grown on a 100 mm dish. The following protocol utilizes the chromatin from 2× 150 mm dishes pooled for each immunoprecipitation reaction. Reagent volumes may be scaled up or down based on the relative surface area of the dishes that cells are grown on (see Note 2). 1. Prepare 1.5 mL per 150 mm dish of PBS for harvesting by adding protease inhibitor cocktail. Ice cold PBS used for wash steps may have EDTA added to a final concentration of 1 mM if desired. 2. To reduce any variability in fixation volume or pH, pool the medium from all like-treated dishes mix and realiquot, so that each 150 mm dish has a final volume of 13 mL for adequate coverage of the cells during fixation. 3. Cross link protein to DNA by adding 351 µL of 37% formaldehyde carefully to the 13  mL of growth medium and mixing to yield a final concentration of 1%. Incubate at 37°C for 10 min. Dishes may be placed on a shaker at low speed if desired.

ChIPing Away at Global Transcriptional Regulation

179

4. The cross-linking reaction is stopped by adding 650 µL of 2.5  M glycine to yield a 125  mM final concentration and mixing gently. Incubate at 37°C for 5 min. 5. The fixation medium is removed and the cells rinsed twice with 15 mL of ice-cold PBS. Multiple dishes may be kept on a bed of ice in the second PBS wash as the cells are collected. Care must be taken to work quickly to avoid loss of cells due to peeling away from the dish. 6. Aspirate away the wash PBS and add 1.5  mL PBS with protease inhibitor to harvest the cells. Using a cell scraper, thoroughly collect the entire volume into a 2 mL microfuge tube. 7. Centrifuge the cells at 250×g for 5 min at 4°C to pellet them and carefully remove the supernatant. At this point, cell pellets may be snap frozen in liquid nitrogen and stored at −80°C for subsequent sonication and immunoprecipitation reaction setup. In our experience, chromatin stability is limited when sonicated lysates are stored for extended periods, even at −80°C. 3.2. Chromatin Shearing and Fragment Size Confirmation by Agarose Gel Electrophoresis

Cell pellets are resuspended in a sufficient but not excessive volume of lysis buffer to allow adequate submersion of the probe to prevent foaming, and to allow some movement during sonication to increase the thoroughness of DNA shearing. Placing the microfuge tubes into packed ice and allowing “rests” between sonication intervals allows for greater heat dispersal and reduces the degradation of the chromatin fragments. Resuspension volumes, sonicator power settings, and interval durations must be optimized for cell type and plated volume. Ideally, electrophoresis should reveal a distribution of chromatin fragments with a peak quantity at 1,000–2,000  kb and a minimum of both degraded fragments and unsheared DNA in the loading wells. 1. Prepare 2 mL of RIPA lysis buffer per immunoprecipitation reaction (i.e., + or − antibody control) by adding the appropriate amount of protease inhibitor cocktail. 2. Resuspend the cell pellets in 200 µL RIPA lysis buffer and transfer to a 1.5  mL microfuge tube as the conical bottom improves sonication efficiency. Round or flat bottomed tubes are difficult to work with as often the cell lysate is pushed out of the tube explosively once sonication begins. Incubate the samples on ice for 10 min. 3. Keeping samples ice cold, sonicate the cell lysates in order to shear chromatin into lengths of approximately 1,000 base pairs. (6 intervals of 25 s, 2 min “rests” between intervals, power setting “3”, Branson 150 cell disruptor). Avoid foaming of proteins by keeping the tip sufficiently submerged during sonication.

180

Jackson, Paris, and Takahashi

Should foaming occur, allow samples to settle for at least 2 min before resonicating. 4. Remove cell debris by centrifuging 20  min at 10,000×g at 4°C. To minimize variability in sonication efficiency and quantity of chromatin between + and − antibody experiments, the cleared supernatants from the sonicated cell pellets may be pooled and equally divided between IP reactions in fresh 2 mL Sarstedt tubes. 5. Following sonication, a 3 µL aliquot of sheared lysate may be run on a 1% agarose gel to confirm shearing efficiency. For accurate determination of fragment size, samples should be decrosslinked by incubating at 65°C for 5  h with 200  mM final concentration NaCl and the DNA column purified before running. However, a rough estimate of fragment sizes can be determined without this step. 3.3. Cell Lysate Preclearing and Incubation with Antibody

Following the removal of sonication debris, the lysates are diluted to a volume that allows adequate movement and mixing of the antibody with the lysate. This in turn enhances the subsequent selective recovery of antibody bound protein–DNA complexes using protein A conjugated Staphylococcus aureus cells or protein A or G agarose beads. 1. The typical volume of lysate from two 150 mm dish cell pellets resuspended in 200 µL lysis buffer is approximately 600 µL following sonication. Additional lysis buffer with protease inhibitors is added to bring the IP reaction volume up to 2 mL. 2. Nonspecific or background protein binding of cells is reduced by preclearing the 2 mL IP reaction with 15 µL of preblocked Staph A cells. Allow samples to incubate for 15 min at 4°C with rocking. 3. Centrifuge samples at 10,000×g for 5  min and carefully transfer the supernatant to a new 2 mL Sarstedt tube avoiding any carry-over of cells. Add 20 µL (1% volume) of 10 mg/mL BSA to each IP reaction as a blocking agent. 4. Add 1 µg of antibody against the protein of interest to each “+ antibody” IP reaction and incubate overnight at 4°C with rocking. No antibody negative control samples are also incubated overnight at 4°C with rocking.

3.4. Recovery and Washing of Antibody–Protein–DNA Complexes (see Note 3)

1. 20 µL of 100  mg/mL yeast tRNA is added to each 2  mL immunoprecipitation reaction yielding a 1  mg/mL final concentration. This is added as an additional blocking agent before adding 10 µL of preblocked Staph A cells. 2. Incubate the reactions at room temperature for 15  min with rocking.

ChIPing Away at Global Transcriptional Regulation

181

3. The cells are pelleted by centrifuging at 10,000×g for 4 min. The resulting supernatant from the antibody treated samples is removed as this contains unbound, nonspecific DNA. 4. A 100 µL aliquot of the supernatant from the “– antibody” reaction may be taken as a total input chromatin positive control for the DNA amplification and labeling reaction. This sample is considered to be representative of the immunoprecipitation starting material and needs to have the protein–DNA crosslinks reversed by heating at 65°C for 5 h. 5. The cells are washed twice in 1× Dialysis Buffer and four times in IP Wash Buffer. 1.4  mL of buffer is used for each wash step: 200 µL is used to resuspend the cell pellet, 200 µL to rinse all cells out of the pipet tip and 1 mL is added before rocking for 3 min at room temperature. 6. At each step, centrifuge at 10,000×g for 4 min and remove as much wash buffer as possible without aspirating away Staph A cells. 3.5. Elution of the Antibody–Protein–DNA Complex

Following washing, the antibody-bound protein–DNA complex is eluted with a basic elution buffer. Fresh elution buffer and vigorous mixing are necessary for successful elution. 1. For each immunoprecipitation reaction, prepare 310 µL of fresh Elution Buffer. 2. Elute the antibody bound Protein–DNA complex by adding 150 µL of elution buffer to the cell pellet following the removal of the final wash. 3. Resuspend the cells in elution buffer by pipette mixing and elute by shaking as vigorously as possible for 15 min at room temperature. 4. Centrifuge at 10,000×g for 4 min and transfer the supernatant to a fresh microfuge tube avoiding any carryover of cells. 5. Resuspend the Staph A cell pellet a second time in 150 µL of Elution Buffer, centrifuge as done previously, and combine the supernatants. 6. Centrifuge the combined eluents at 10,000×g for 4 min and transfer to a fresh 1.5 mL microfuge tube in order to absolutely eliminate any carryover of Staph A cells. 7. If it is to be used, 200 µL of Elution Buffer may be added to the “Total Input Chromatin” sample collected earlier to bring its volume up to 300 µL.

3.6. Decrosslinking the Protein–DNA Complex, RNAse A Digest

Formaldehyde crosslinks are reversed by incubating the samples at 65°C under high-salt conditions. RNAse treatment eliminates any contaminating potentially competitive nucleic acid species from downstream amplification reactions.

182

Jackson, Paris, and Takahashi

1. Add 12 µL of 5  M NaCl to the combined eluents for a 200  mM final concentration. Add 1 µL of 10  mg/mL RNAse A . 2. Incubate the mixture at 65°C for 5 h to reverse the formaldehyde crosslinks. 3. Add 2.5 volumes of 100% ethanol (782 µL), mix by inversion and allow the decrosslinked DNA to precipitate at −20°C overnight. 3.7. Immunoprecipi­ tated DNA: Fragment Recovery and Proteinase K Digest

Decrosslinked proteins and antibody are Proteinase K digested. Immunoprecipitated DNA may be recovered from the mixture using phenol:chloroform isolation, but for convenience and consistency of recovery, centrifugation columns such as QiaQuick PCR Purification columns with binding buffers containing chaotropic salts may be used instead. 1. The samples are centrifuged at maximum rpm for 15–20 min. at 4°C. Carefully remove the supernatant and resuspend the pellet in 70% ethanol to wash. Spin again at maximum rpm 15–20 min. at 4°C. 2. Remove the supernatant and allow pellet to air dry completely. 3. Resuspend the pellet in 100 µL 1× TE (pH 7.5). Add 25 µL 5× Proteinase K Buffer and 1.5 µL of 22 mg/mL Proteinase K. Incubate at 42°C for 2 h. 4. For convenience, the DNA is recovered from the 126.5 µL Proteinase K digested sample using QiaQuick PCR Purification columns following the manufacturer’s protocol, but eluting with 100 µL Sigma water. 5. The eluted DNA volume may be brought down to 7 µL using a SpeedVac by heating the sample with medium heat for approximately 20 min.

3.8. Round A/B Random Amplification of Immunoprecipitated DNA: Primer A Annealing and Extension (see Note 4)

In this round, randomly annealed primers (Primer A) are extended using Sequenase in order to generate templates for subsequent PCR amplification. 1. Transfer the 7 µL volumes of + and − antibody immunoprecipitated DNA to PCR tubes. 2. Prepare Round A Setup and Round A Reaction master mixes as required, preparing extra volume to account for losses due to pipetting. To facilitate channel normalization during scanning of microarrays between + and − antibody conditions, the researcher may wish to introduce some control DNA into the Round A Setup mix (see Note 5). 3. Aliquot 3 µL of Round A Setup mix to each tube for a final volume of 10 µL and mix well.

ChIPing Away at Global Transcriptional Regulation

183

4. Incubate Round A Setup + immunoprecipitated DNA at 94°C for 2  min. Go to 10°C and hold for 5  min while adding Reaction Mix to each tube, mixing carefully. Ensure that full reaction volume is at the bottom of the tube, and that no droplets are on the wall of the tube. 5. After 10°C hold, ramp up to 37°C over 8 min. 6. Hold at 37°C for 8 min, go to 94°C for 2 min. 7. Go to 10°C and hold for 5 min. To each reaction, add 1.2 mL of fourfold diluted Sequenase in Sequenase Dilution Buffer and mix carefully. Ensure that full reaction volume is at the bottom of the tube, and that no droplets are on the wall of the tube. 8. After 10°C hold, ramp up to 37°C over 8 min. Hold at 37°C for 8 min. 9. Remove Round A and add 43.75 mL water to bring volume to 60 mL. 3.9. Round A/B Random Amplification of Immunoprecipitated DNA: Primer B Annealing, aa-dUTP Labeling and Amplification (see Note 6)

3.10. Amplified Immunoprecipitated DNA: PCR Purification

In this round, the specific primer Primer B is used to amplify the templates previously generated, and to incorporate amino allyl dUTP. 1. Transfer 15 mL of each Round A reaction to a fresh PCR tube. 2. Prepare sufficient Round B Setup master mix, allowing for losses due to pipetting, and add 85 mL to each PCR tube. 3. Run Round B amplification/nucleotide incorporation program: 92°C for 30  s → 40°C for 30  s → 50°C for 30 s → 72°C for 1 min (see Note 7). PCR Amplified DNA is purified using the CyScribe GFX Purification Kit. 80% ethanol is substituted for the Wash Buffer provided with the kit as it contains Tris, as does the provided Elution Buffer. Tris contains primary amines that will interfere with the dye-coupling reaction. Purified DNA must be eluted with 0.1 M sodium bicarbonate (pH 9.0) NOT water! 1. Place one GFX column in a collection tube for each purification to be performed. 2. Add 500 mL Capture Buffer to the GFX column. 3. Transfer the 100 mL PCR amplification reaction to the GFX column. 4. Mix thoroughly by pipetting sample 4–6 times. 5. Centrifuge at 10,000×g for 30 s. 6. Discard flow-through by emptying collection tube. Place GFX column back into the collection tube.

184

Jackson, Paris, and Takahashi

7. Add 600 mL of 80% ethanol to the column. Centrifuge at 10,000×g for 30 s. Repeat this wash step twice more. Empty the collection tube and spin column for an additional minute to remove all traces of 80% ethanol. 8. Discard the collection tube and transfer GFX column to a fresh 1.5 mL microfuge tube. 9. Apply 60 mL of 0.1 M sodium bicarbonate (pH 9.0) directly to the top of the glass fiber matrix in the GFX column, ensuring that it is completely covered. 10. Incubate sample at room temperature for 5 min. 11. Centrifuge sample at full speed for 1  min to recover the purified DNA. 12. Dry sample down to 8 mL final volume for the labeling reaction. SpeedVac for approximately 20 min at medium heat. 13. Should it be necessary the purified aa-dUTP labeled DNA may be frozen at −20°C before concentration, and the protocol continued the next day. 3.11. Amplified Immunoprecipitated DNA: Alexa Fluor Labeling and Purification

Just prior to use, resuspend one vial of Alexa 647 or Alexa 555 dye in 2 mL 100% DMSO. Vortex for 10 s to ensure that the dye is completely dissolved (see Note 8). 1. Add the 8 mL sample to the resuspended fluor and vortex briefly to ensure that the reaction is well mixed. As per manufacturer’s protocol, DO NOT spin sample down to bottom of tube, but tap it down gently if necessary. 2. Allow the reaction to incubate 1 h at room temperature in the dark. 3. Add 90 mL RNAse/DNAse-free water to each labeling reaction to bring volume up to 100 mL. 4. Place one GFX column in a collection tube for each purification to be performed. 5. Add 500 mL Capture Buffer to the GFX column. 6. Transfer the 100 mL labeled sample to the GFX column. 7. Mix thoroughly by pipetting sample 4–6 times. 8. Centrifuge at 10,000×g for 30 s. 9. Discard flow-through by emptying collection tube. Place GFX column back into the collection tube. 10. Add 600 mL of 80% ethanol to the column. Centrifuge at 10,000×g for 30 s. Repeat this wash step twice more. Empty the collection tube and spin column for an additional minute to remove all traces of 80% ethanol. 11. Discard the collection tube and transfer GFX column to a fresh 1.5 mL microfuge tube.

ChIPing Away at Global Transcriptional Regulation

185

12. Apply 60 mL of 0.1 M sodium bicarbonate (pH 9.0) directly to the top of the glass fiber matrix in the GFX column, ensuring that it is completely covered. 13. Incubate the GFX column at room temperature for 5  min. Centrifuge at 10,000×g for 1  min to collect the purified labeled DNA. At this point, paired reactions may be pooled together. 14. To reduce the 120 mL combined eluent volume to 5 mL for hybridization, SpeedVac for approximately 35  min at medium heat. 3.12. Microarray Hybridization and Wash Protocol

1. Prepare 100 mL of hybridization solution per slide. To each 100 mL of DIG Easy Hyb solution add 5 mL of 10 mg/mL calf thymus DNA and 5 mL of 10 mg/mL yeast tRNA. Mix thoroughly, avoiding formation of bubbles and incubate mixture at 65°C for 2 min. Cool to room temperature for at least 2 min. 2. Add 85 mL of hyb solution to the pooled Alexa 647 and Alexa 555 labeled DNA. Mix and incubate the solution at 65°C for 2 min. Cool to room temperature for at least 2 min. 3. Pipette the labeled probe onto an array. Place a 24 × 60 mm glass coverslip onto the droplet, angling it in such a way that any bubbles are pushed out from underneath as the coverslip settles. Place the loaded array into a hybridization chamber (microscope boxes using slides as rails to support the arrays). Hybridization chambers contain a small amount of DIG Easy Hyb solution in the bottom to maintain a humid environment. 4. Incubate on a level surface in a 37°C incubator for 8–18 h. 5. Preheat three slide staining boxes containing a wash solution of 1× SSC and 0.1% SDS by allowing to incubate 15 min in a 50°C water bath. Also, prepare one staining box containing a staining rack filled with 1× SSC, and two other boxes containing 0.1× SSC for the final rinses. These boxes are kept at room temperature. 6. Open the hybridization chamber and remove the coverslip by gently but quickly dipping the array into the slide box containing 1× SSC, allowing the coverslip to slide off on its own. Place slide into the staining rack (up to four slides) and agitate several times before transferring to box containing incubating wash solution. Agitate briskly and allow to incubate for 8 min. Agitate, incubate for an additional 8 min, agitate and transfer to the next wash box. Repeat with the next two wash boxes. 7. Transfer the staining rack into a staining box containing 0.1× SSC at room temperature. Rinse the slides by agitating

186

Jackson, Paris, and Takahashi

briskly about 10× in each of the rinse boxes and quickly remove the rack. 8. Remove the slide from rack, tap off excess rinse solution and quickly transfer to a microscope box lined on the bottom with blotting paper. Spin slides dry at 640 rpm for 15 min. Arrays can be stored in the dark until scanned for several hours to days.

4. Notes 1. Lyophilized Staph A Cell Preparation: (a) 1  g of lyophilized cells are resuspended in 10  mL 1× Dialysis Buffer: 2 mM EDTA, 50 mM Tris–HCl pH 8.0 (Sigma-Aldrich, Oakville, ON) and centrifuged 10,000 rpm at 4°C for 5 min. Repeat and remove buffer. (b) Resuspend cell pellet in 3  mL 1× PBS + 3% SDS + 10% beta mercaptoethanol (Sigma-Aldrich, Oakville, ON) and boil for 30 min. (c) Cells are pelleted at 10,000 rpm for 5 min and washed in 1× dialysis buffer twice. Cells are resuspended in 4 mL 1× dialysis buffer. Divide into 100 mL aliquots, snap freeze in liquid nitrogen, store in cryogenic storage until use. For each immunoprecipitation reaction, you will need 15 mL of preblocked cell suspension for preclearing and 10 mL for recovery (allow a little extra volume for losses due to pipeting). Alternatively, commercial preblocked ready to use bead slurrys work well, but have proven difficult to replicate using “home-made” preparations in the lab resulting in poorly selective “sticky” beads . Staph A cells are easier to preblock and have the added benefits of requiring shorter time intervals for preclearing and pull down versus beads, as well as being able to sustain much higher centrifugation speeds. This results in shorter spin times as well as tighter cell pellets that facilitate subsequent handling with less loss of immunoprecipitated material. 2. This protocol was optimized for the MEF2C transcription factor using C2C12 cells, and each immunoprecipitation reaction required a quantity of chromatin from adherent cells grown on two 150 mm dishes (5–15 × 106 cells depending on degree of confluence). It is necessary to determine what amount of antibody and starting chromatin are needed to give a well discernible signal from any nonspecific binding to Staph A cells or Protein A or G agarose beads following amplification. An approach we used was to perform

ChIPing Away at Global Transcriptional Regulation

187

immunoprecipitations with a gradient of both antibody and starting chromatin with no antibody negative control reactions. The resulting eluted DNA was then PCR amplified using primers to known transcription factor targets for 25–35 cycles, and the products run on a 1% agarose gel. The resulting bands will give some indication of the amount of signal (+ antibody) to noise (nonspecific binding in − antibody reaction). An excessive number of PCR cycles will result in a loss of discernible differences between the two conditions. 3. Washing, elution, decrosslinking and Proteinase K digest is based on the Farnham Lab Protocol: http://www.genomecenter. ucdavis.edu/farnham/protocol.html (previously found at: http://mcardle.oncology.wisc.edu/farnham/protocols/ chips.html) 4. Amplification protocol based on Round A/B/C Random Amplification of DNA Protocol, DeRisi Lab, UC San Francisco, June 2001: http://cat.ucsf.edu/pdfs/22_Round_A_B_C_ protocol.pdf 5. Slide Scanning Normalization using Arabidopsis Control DNA Spots During slide scanning, signals from the experimental (+ antibody) and reference (− antibody) samples may be normalized using Arabidopsis Control Spots printed on the array (Microarray Centre, University Health Network, Toronto, ON). To do this it is necessary to add a quantity of Arabidopsis Control DNA directly to the Round A Setup master mix for all samples to be amplified and labeled. A stock of Arabidopsis Control DNA is created by amplification of the Arabidopsis Chlorophyll Synthetase insert from a vector produced in-house (Microarray Centre, University Health Network, Toronto, ON). The product is size confirmed on an agarose gel, excised, and purified using the GFX PCR DNA and Gel Band Purification Kit (GE Healthcare, Piscataway, NJ) following manufacturer’s protocol. • Arabidopsis Control Primer sense: GAG CCA TAT CGT CCA ATT CC • Arabidopsis Control Primer antisense: GTT TCG GTG CCA AAA GCT AC The amount of Arabidopsis Control DNA to be added to the Round A Setup master mix is determined by the number of amplification cycles needed to give a good signal from the experimental samples while avoiding saturating the Control Spots. Typically, for amplifications of 22 cycles or so, 1 ng of Arabidopsis Control DNA is added to the Round A Setup mix resulting in 0.25 ng present in the Round B amplification reac­ tion is sufficient. To keep the Round A reaction volume at 10 mL, it may be necessary to dry down the immunoprecipitated

188

Jackson, Paris, and Takahashi

DNA to less than 7 mL, depending upon the concentration of the Control DNA. Alternatively, SpotReport Alien cDNA spots (Stratagene, La Jolla, CA) printed on the arrays may be used according to manufacturer’s guidelines. 6. By switching the 50× aa-dUTP/dNTPs within the Round B PCR step to 25  mM dNTPs, it is possible to switch to the Agilent Mammalian ChIP-on-chip (version 10) protocol following the Round B amplification. By starting at the point of the precipitation step which immediately precedes the Sample Labeling section, it is possible to carry through to hybridizing on the Agilent CpG Island and promoter arrays. 7. The optimal number of cycle numbers that gives a good dynamic range of spot intensitites while preserving differences between experimental conditions must be determined by the researcher. This can be done by tube-pulling experiments, where PCR reactions are removed every 2 cycles within a range of about 20–30 cycles. The protocol is then followed to completion, and the arrays scanned to determine what cycle number provides an optimal result avoiding over or under saturation of spots. In experiments where relative enrichment of immunoprecipitated targets is determined relative to total input chromatin, the IP’ed DNA and 20 ng of total chromatin are both amplified in a Round B reaction substituting 100× dNTP’s (25 mM each dNTP) for 50× aadUTP/dNTPs. Following 15–35 cycles depending on input IP’ed material, the experimental and control samples are normalized by DNA concentration and an equal amount (2–6 µg per array) then Round B amplified 10–25 cycles using 50× aa-dUTP/dNTPs. 8. Alternatively, cyanine dyes Cy3 and Cy5 may be used for indirect labeling. Should the researcher wish to directly incorporate Cy3 and Cy5 into the PCR amplicons, 15  mM Cy-dUTP may be substituted for aa-dUTP in the 50× aa-dUTP/dNTP mix, and the 72°C annealing step of the Round B amplification program may be increased to 2 min or greater duration. References 1. Ptashne M (1988) How eukaryotic transcriptional activators work. Nature 335: 683–689 2. Goll MG, Bestor TH (2005) Eukaryotic cytosine methyltransferases. Annu Rev Biochem 74:481–514 3. Venter JC, Adams MD, Myers EW, Li PW, Mural RJ, Sutton GG et  al (2001) The sequence of the human genome. Science 291: 1304–1351

4. Lander ES, Linton LM, Birren B, Nusbaum C, Zody MC, Baldwin J et  al (2001) Initial sequencing and analysis of the human genome. Nature 409:860–921 5. Brown PO, Botstein D (1999) Exploring the new world of the genome with DNA microarrays. Nat Genet 21:33–37 6. Kurdistani SK, Tavazoie S, Grunstein M (2004) Mapping global histone acetylation atterns to gene expression. Cell 117:721–733

ChIPing Away at Global Transcriptional Regulation 7. Paris J, Virtanen C, Lu Z, Takahashi M (2004) Identification of MEF2-regulated genes during muscle differentiation. Physiol Genomics 20: 143–151 8. Weber M, Davies JJ, Wittig D, Oakeley EJ, Haase M, Lam WL, Schubeler D (2005) Chro­mosome-wide and promoter-specific analyses identify sites of differential DNA methylation in

189

normal and transformed human cells. Nat Genet 37:853–862 9. Keshet I, Schlesinger Y, Farkash S, Rand E, Hecht M, Segal E, Pikarski E, Young RA, Niveleau A, Cedar H, Simon I (2006) Evidence for an instructive mechanism of de  novo methylation in cancer cells. Nat Genet 38: 149–153

Chapter 12 HELP (HpaII Tiny Fragment Enrichment by Ligation-Mediated PCR) Assay for DNA Methylation Profiling of Primary Normal and Malignant B Lymphocytes Rita Shaknovich, Maria E. Figueroa, and Ari Melnick Abstract The role of cytosine methylation in the regulation of gene expression during normal development and malignant transformation is currently under intense investigation. An ever increasing body of evidence demonstrates that carcinogenesis is associated with aberrant DNA methylation of the promoters of tumor suppressor genes (Chin Med J (Engl) 111:1028–1030, 1998; Leukemia 17:2533–2535, 2003), hypomethylation of oncogenes (Toxicol Appl Pharmacol 206:288–298, 2005; Toxicology 50:231–245, 1988), and concurrent loss of methylation in the intergenic areas and gene bodies, which may lead to genomic instability and chromosomal fragility (Cytogenet Cell Genet 89:121–128, 2000). Single locus methylation assays have focused largely on specific known tumor suppressor genes or oncogenes (Chin Med J (Engl) 111:1028–1030, 1998; Cancer Res 57:594–599, 1997; Hum Genet 94:491–496, 1994; Mol Cell Biol 14:4225–4232, 1994; Gastroenterology 116:394–400, 1999). Such approaches, while being useful, have clear limitations. With the advent of genome-wide microarray-based techniques, it has become possible to perform genome-wide exploratory studies to better understand genomic patterning of DNA methylation and also to discover new potential disease-specific epigenetic lesions (J Cell Biochem 88:138–143, 2003; Genome Res 16:1075–1083, 2006). In order to capture this type of information from primary human tissues, we have adopted and optimized the HELP assay (HpaII tiny fragment Enrichment by Ligationmediated PCR) to compare and contrast the abundance of cytosine methylation of genomic regions that are relatively enriched for CpG dinucleotides. While we have mainly used a custom NimbleGen-Roche high-density oligonucleotide microarray containing 25,626 HpaII amplifiable fragments, many other microarray platforms or high throughput sequencing strategies can be used with HELP. Key words: Epigenomics, DNA methylation, Genomic microarray, HELP assay, Gene regulation

1. Introduction DNA methylation is found on position five of the cytosine ring at CpG dinucleotides and plays a diverse set of roles, from embryogenesis and imprinting in development to gene silencing in Sridar V. Chittur (ed.), Microarray Methods for Drug Discovery, Methods in Molecular Biology, vol. 632, DOI 10.1007/978-1-60761-663-4_12, © Humana Press, a part of Springer Science+Business Media, LLC 2010

191

192

Shaknovich, Figueroa, and Melnick

disease (12, 13). Cytosine methylation is associated with repressed chromatin state and gene silencing (14–16). It cooperates with histone-tail modifications to achieve chromatin remodeling and to affect protein expression. Cells contain an evolutionaryconserved machinery of DNA methyltransferases and methylbinding proteins, which bind methylated cytosine and interact with the transcription machinery (17–19). DNA methylation undergoes dramatic changes during tumorigenesis, resulting in global losses in methylation from repetitive sequences as well as in promoter-specific gain of methylation (20–24). Khulan et al. demonstrated the utility of HELP assay by interrogating 6.2 Mb of mouse genome at 1,399 HpaII sites (25). By comparing mouse brain tissue and spermatogenic cells, they were able to identify 223 new tissue-specific differentially methylated regions. In addition, the assay allowed defining the profile of methylation: with most of the genome being methylated with clusters of hypomethylated areas at promoters and CpG islands. Figueroa et al. applied HELP to the study of human leukemias. In the subset of Acute Myelogenous Leukemias and Precursor B Lymphoblastic Leukemias, it was possible to correctly subclassify 2 subtypes of leukemias based on their HELP profiles, capturing the methylation signature of the two disease states (26). This study also demonstrated that integration of gene expression profiling with methylation profiling allowed to capture genes differentially expressed between two leukemia subtypes that were not identified by gene expression alone. This work demonstrates the utility of whole-genome profiling approach to study epigenomic signatures. Study of a subgroup of Myeloid Leukemias with the common gene expression profile, but with different CEBPA-locus status (mutated vs. hypermethylated wild type) also demonstrated the sensitivity and utility of the HELP assay in revealing the presence of underlying DNA methylation differences (27). The HELP assay is based on the principle of comparative isoschizomer profiling of cytosine methylation in the HpaII Tiny Fragment (HTF) fraction of the genome (25). This approach allows effective profiling of DNA methylation in CpG rich regions of the genome such as promoter associated CpG islands or CG clusters (28, 29). The assay is based on using 2 isoschizomer enzymes: HpaII, which is methylation-sensitive and only cuts unmethylated CCGG, and MspI, which cuts CCGG whether it is methylated or not. Purified, high molecular weight genomic DNA is digested to completion in 2 separate reactions with HpaII and MspI, after which linker oligonucleotides are ligated and Ligation Mediated – PCR (LM-PCR) is performed under conditions that allow preferential amplification of fragments from 200 to 2,000 bp. PCR products can then be labeled with 2 different fluorochromes and hybridized to genomic microarrays.

DNA Methylation Profiling of Primary Normal and Malignant B Lymphocytes

193

One of the array designs that we use most frequently is designed to cover two or more HpaII sites from ~14,000 promoters, which are each represented with ten oligonucleotide probes (total 385,000 features), along with 2,000 random sequence probe controls as well as mitochondrial DNA probes (mitochondrial DNA is never methylated and is present at high copy numbers so that both HpaII and MspI fluorescence intensities are high and equal). RocheNimbleGen arrays support two-color hybridizations and use 50-mer oligonucleotides that provide greater specificity and more uniform hybridization kinetics than shorter oligos. A more comprehensive human HELP array with 2.1 million features (HD2 platform) including the entire complement of gene-associated HpaII sites is also available (30). Users may create their unique design, which could include HpaII sites within promoters or/and gene bodies; focus on the whole genome or selected chromosomes, or even use a tiling design to assess every HpaII site in the continuous stretch of a chromosome. After hybridization and digital scanning of the array, the image file and raw data files are generated, which reflect the intensity of an individual spot and allow comparison of the signal in MspI and HpaII representations. The MspI representation serves as an internal control, allowing differentiation between the absent signal in HpaII due to the methylation from the absent signal in HpaII due to the loss of genomic material, the presence of SNPs, or the technical failure of the probe. This is of particular practical importance when studying cancer genomes with great genomic instability (26, 29).

2. Materials 1. 1% Agarose gels are prepared with distilled water. 2. 50× TAE buffer is diluted with distilled water to 1× con­ centration. 3. TE-saturated phenol. 4. Phenol–chloroform–isoamyl alcohol (PCI): 25 volumes of phenol: 24 volumes of chloroform: 1 volume of isoamyl alcohol. Make up fresh each time. 5. Chloroform. 6. Glycogen (Sigma, St Louis, MO) is dissolved in water and stored at −20°C freezer. 7. 100% Ethanol. 8. 3 M Sodium Acetate, pH 5.2. Dissolve 123.04 g of sodium acetate into distilled water and fill it up to 500 ml.

194

Shaknovich, Figueroa, and Melnick

9. 1 M EDTA: dissolve 186.12 g EDTA into distilled water, fill up to 1 L and autoclave. 10. 1 M Tris-HCl, pH 8.0 and pH 8.9: Dissolve 121.14  g Tris-base into about 800  ml of distilled water and adjust pH to 8.0 or 8.9 respectively with HCl. Fill the solution up to 1 L and autoclave. 11. TE pH 8.0: Add 10 ml of 1 M Tris-HCl pH 8.0 and 1 ml of 1  M EDTA and fill up to 1  L with distilled water and autoclave. 12. 20% SDS: Dissolve 20 g of SDS in autoclaved DIUF water. Bring the solution to a final volume of 100 ml. 13. 1 M Magnesium chloride. 14. Ammonium sulfate. 15. BSA. 16. b-Mercaptoethanol. 17. 4  mM dNTP mix (100  mM dNTP nucleotides from Invitrogen, Carlsbad, CA). Add 40 ml of each nucleotid (100 mM) and 840 ml of sterile water to make up 1 ml mix. 18. RNAse A (Sigma, St Louis, MO): resuspend in water to a final concentration of 10 mg/ml. Aliquot and store at −20°C. 19. HpaII (NEB, Ipswich, MA) stored at −20°C. 20. MspI (NEB, Ipswich, MA) stored at −20°C. 21. Native Taq Polymerase (Invitrogen, Carlsbad, CA) stored at −20°C. 22. T4 DNA Ligase (Invitrogen, Carlsbad, CA) stored at −20°C. 23. Primer/Linker JHpaII12 5¢-CGGCTGTTCATG-3¢ diluted with deionized water to 6 ODs/ml and stored at −20°C. 24. Primer/Linker JHpaII 24 5¢-CGACGTCGACTATCCAT GAACAGC-3¢ diluted with deionized water to 12 ODs/ml and stored at −20°C. 25. 5× RDA: mix 16.75 ml of 1 M Tris-HCl pH 8.9, 1 ml of 1 M magnesium chloride, 4 ml of 1M ammonium sulfate, 175 ml of b-mercaptoethanol and 25  mg of BSA and make up to 50 ml with distilled water and filter sterilize. Store at 4°C for up to 1 month (see Note 1). 26. Preannealing of linkers: Mix equal amounts of the 12mer and 24mer linkers (6 OD/ ml and 12  OD/ml, respectively) in a screw top eppendorf. Boil for 5  min and then allow cooling down at RT. The annealed linkers can then be stored at −20°C.

DNA Methylation Profiling of Primary Normal and Malignant B Lymphocytes

195

3. Methods 3.1. DNA Extraction from the Frozen Tissue Samples

1. 100 mg or 3 mm3 of tissue is pulverized using the mortar and pestle method: frozen tissue is cut into smaller pieces and submerged under the liquid nitrogen within the mortar dish. The tissue is carefully pulverized with the pestle, liquid nitrogen is allowed to evaporate, and the powder is quickly collected and transferred to the eppendorf tube on ice. The lysis solution from Puregene Gentra Tissue kit (Qiagen Valencia, CA) is added. We normally start with 300  ml of lysis solution. The amount is scaled up if more tissue is used (see Note 2).

3.2. DNA Extraction from the Cell Pellets

1. 1–10 million cells are pelleted at room temperature for 10 min at 300×g. The supernatant is removed and the cells are resuspended in 5 mls of cold 1× PBS and washed once by spinning 10 min at 300×g. Supernatant is discarded. 2. DNA purification is done according to the Qiagen Puregene Gentra cell kit (Qiagen, Valencia, CA) instructions.

3.3. HELP

1. The quality of genomic DNA is assessed on a 1% agarose gel. DNA should be of high molecular weight with no evidence of degradation. If a high molecular weight sharp DNA band is identified then proceed to step 2 (see Notes 3 and 4). An example is shown in Fig. 1a. 2. An overnight digestion of 1 mg genomic DNA with 2 ml of either HpaII or MspI is set up in separate 200 ml reactions at 37°C in a waterbath. Digestion is set up according to manufacturer’ instructions in NEB buffer #1 for HpaII and NEB buffer #2 for MspI. 3. On the following day 15  ml of the digest is run on a 1.5% agarose gel. There should be significant difference between the two digests: with HpaII most of the DNA remains high molecular weight, whereas with MspI, an almost even smear with no high molecular weight DNA remaining should be seen (Fig. 1a). 4. 200 ml of TE pH 8.0 buffer is added to the digested DNA and 400  ml of saturated phenol:chloroform:isoamyl alcohol mix (25:24:1) and vortexed briefly. The mixture is centrifuged at top speed in a microcentrifuge for 10 min. 5. The top aqueous phase from last step is removed (about 400 ml) and transferred into a new tube and 1 ml of glycogen and 40 ml of 3 M NaOAc pH 5.2 are added and mixed well. Then, 1,000 ml of 100% ethanol is added, vortexed, and spun at top speed for 45 min at 4°C.

b

c

Intensity

a

Hp aII Ms pI

Shaknovich, Figueroa, and Melnick

gD NA Hp aII Ms pI

196

2000 bp

Intensity

Fragment Size (bp)

200 bp

Intensity

Fragment Size (bp)

Fragment Size (bp)

Fig. 1. Major quality control steps are necessary to assure reproducible results. (a) Agarose gel reveals high molecular weight DNA of high quality with no degradation (1st lane and an arrow) and the products of 2 digestion reactions. HpaII results only in partial digestion with minimal smear and a small decrease in the molecular weight of the main band. MspI digestion results in complete disappearance of high molecular weight DNA and a smear. (b) PCR reaction is optimized to produce 200–2,000 bp products. MspI reaction normally gives higher yield due to the greater number of the templates in the digestion reaction. (c) Quality control steps for the data include examination of the dotplots and density plots for fragment length versus intensity of fluorescence. The top dotplot reflects MspI reaction with detectable intensity for all the studied fragments, while middle HpaII reaction did not produce significant signal for many fragments because of the methylation of HpaII site, resulting in multiple dots along the x axis. The lower dotplot illustrates the final read-out of logHpaII/logMspI intensities, revealing top unmethylated and bottom methylated fragments (black line divides 2 main methylation peaks). The density plots illustrate the same point after log transforming the values

6. The supernatant is removed, and the pellet is washed with 70% ethanol. Once all of the ethanol is carefully removed, the pellet is resuspended in 15.5 ml of 10 mM Tris pH 8.0. The ligation must be set up on the same day, since the digested DNA has sticky overhangs that may reanneal. 7. The ligation reaction is set up in a PCR tube as follows: 5× T4 DNA ligase buffer

  6 ml

DNA from last step

15.5 ml

Preannealed JHpaII linkers

  7.5 ml

T4 DNA ligase

  1 ml

Incubate overnight at 16°C in a PCR thermocycler with a heated lid.

DNA Methylation Profiling of Primary Normal and Malignant B Lymphocytes

197

8. On the following day, dilute with 970 ml of 10 mM Tris-HCL pH 8.0. The ligation mix can be stored at −20°C. 9. Set up the PCR reaction as follows in a 1.5 ml tube: Diluted ligation mix from last step 40 ml (for MspI)/80 ml (for HpaII) JHpaII 24 (12 OD/ml)

8 ml

5× RDA buffer

80 ml

4 mM dNTP mix

32 ml

Native Taq (Invitrogen)

3 ml

Water

237 ml (for MspI)/197 ml (for HpaII)

Divide the reaction mix into four PCR tubes and incubate in a thermocycler as follows: 1.

72°C for 10 min

2.

20 cycles of:

3.

10 min at 72°C

4.

Hold at 4°C

30 s at 95°C 3 min at 72°C

10. Run 10 ml of PCR product on a 1.5% agarose gel. A smear of PCR product from 200 to 2,000 bp is expected (Fig. 1b). 11. Clean the product using QIAquick PCR purification kit (Qiagen, Valencia, CA), eluting in 50 ml of the kit’s elution buffer. 12. We routinely use 1 mg of HpaII and MspI PCR products for labeling and hybridization. Roche-NimbleGen normally uses Cy-labeled random primers (9 mers) for the labeling of the HpaII and MspI reaction products. The Cy5-labeled HpaII and Cy3-labeled MspI fragments are then cohybridized to the microarray and scanned using a GenePix 4000B scanner (Axon Instruments). The choice of the best protocol for hybridization and scanning should be made according to the microarray manufacturer’s instructions for the best results. 3.4. Data Analysis

The specifics of data analysis will be unique to the array design and the experimental design, but some common principles of data analysis exist independent of the platform. 1. The output of the assay includes 4 files: 2 image files, and 2 data (intensity) files with quantitative data from 2 reactions (see Notes 6 and 7). 2. We routinely subject all our microarrays to a quality control pipeline in order to assess the technical quality of the assay, and to derive the logHpaII/logMspI ratio that we most

198

Shaknovich, Figueroa, and Melnick

commonly use as a relative methylation value (31). Currently, the downloadable software package compatible with R 2.8.0 is available through the Bioconductor 2.3 for the quality control analysis of data derived from the HELP promoter array (http://www.bioconductor.org/), and is called “HELP”. The same principle can be adapted for use with other microarray platforms. 3. Data summarization: Several oligonucleotides may be used to interrogate the same HTF, and thus the decision needs to be made whether to summarize the data and how to summarize it: to use mean or median intensity; and how to set up the failure threshold, i.e. how many oligos can fail for the data to be acceptably informative. 4. Designation of the failed fragments: Fragments may be designated as failed for biological or technical reasons. Biological reasons include the absence of genomic sequence due to deletion or mutations, which are common in cancer; or a presence of polymorphism at the restriction site, which renders sequence unrecognizable to the restriction enzyme and thus leads to the failure to generate the amplifiable fragment. Technical reasons may include failure to generate a hybridizable fragment at any step of the protocol: failure to digest DNA, or to amplify, label or hybridize the fragment to the array. Since HELP uses MspI as an internal control, probes failed for biological or technical reasons can be removed from HpaII channel. 5. Computation of log HpaII/log MspI: As a final step, it is necessary to compute the ratio of HpaII to MspI signal (Fig. 1c). MspI dotplot reveals positive signal in all amplifiable probes, with exception of few failed probes at the bottom of the graph below the horizontal black line. On the other hand, HpaII results in the similar dotplot with greater number of unamplified probes, likely due to the methylated status of HpaII site, which prevents generation of amplifiable fragments (all below the black horizontal line). Calculation of the log of HpaII to MspI ratio results in fragments with positive and negative values, reflecting lesser and greater degrees of methylation. 6. The illustration of methylation profiles of normal tissues can be seen in Fig. 2. In MspI channel, all fragments are amplified, except for technical failures, and result in similar “dot plots” in CBs and NBs. On the other hand, within HpaII channel, differentially methylated genes are clustering along the x-axis. 7. Once such data summarization is done, genome-wide comparison between different tissues and biological states can be carried out. Using different statistical methods, the differentially methylated genes can be identified. Fig. 3 illustrates the high reproducibility of HELP, as illustrated with high correlation coefficients of three technical repeats of DLBCL cell lines.

DNA Methylation Profiling of Primary Normal and Malignant B Lymphocytes Naïve B cells

Centroblasts MspI

HpaII

MspI

Intensity

Intensity

HpaII

199

Fragment Size (bp)

Fragment Size (bp)

Fig.  2. The dotplot of fragment size versus log of probe intensity reveals methylation profiles of Naïve B cells and Centroblast B cells. A subset of methylated probes in HpaII dotpplot is circled in black. The nature of differentially methylated probes can be identified after comparing he logHpaII/logMspI intensity for every fragment

DLBCL 1-2 0.99

DLBCL 1-2 DLBCL 2-1

0.87

0.64

0.66

0.63

0.98

DLBCL 2-2

0.65

0.63

0.64

0.66

DLBCL 3-1

0.59

0.57

0.61

0.64

0.97

DLBCL 3-2

Fig. 3. Technical repeats on three DLBCL cell lines (DLBCL 1-1 vs. DLBCL 1-2, DLBCL 2-1 vs. DLBCL 2-2, DLBCL 3-1 vs. DLBCL 3-2) reveal high reproducibility of HELP assay. The correlation coefficients are consistently above 0.97 for technical replicates, while comparison of various DLBCL cell lines yields lower correlations of 0.66 and less

4. Notes 1. Check for the presence of the “sulfur smell” before every use. Replace solution if smell is less than potent (“noxious”). Solution is good for approximately 1 month if stored at 4°C. 2. The success of the HELP assay depends on the quality of the purified DNA – unsheared DNA with no apoptotic

200

Shaknovich, Figueroa, and Melnick

degradation is necessary starting material. Poorly preserved tissue may not be of sufficient quality for this assay. 3. We use only desalting methods for DNA purification, avoiding other column methods. We also recommend avoiding excessive vortexing of the DNA. 4. After purification of DNA, do not over dry the pellet – redissolving it may be difficult. If you find unusually low DNA concentration or a varying concentration during use, it may signify the poor solubilization of the DNA. In this case, we dilute DNA with 2× or 3× volume and mix it ON at RT and rerun 1 ml on 1% agarose gel. 5. The consistency in processing of DNA ensures high reproducibility of the result 6. HELP is a double color hybridization assay: we normally request the brighter fluorophore to be used for the labeling of HpaII reaction, and a weaker one for the MspI reaction. 7. In addition to running a stringent QC algorithm (31), we also perform careful visual inspection of TIFF files to ensure universal hybridization. References 1. Chen W, Zhu J, Liu J, Tan S (1998) Methylation of p16 gene in hematological malignancies. Chin Med J (Engl) 111:1028–1030 2. Chim CS, Fung TK, Liang R (2003) Disruption of INK4/CDK/Rb cell cycle pathway by gene hypermethylation in multiple myeloma and MGUS. Leukemia 17:2533–2535 3. Benbrahim-Tallaa L, Waterland RA, Styblo M, Achanzar WE, Webber MM, Waalkes MP (2005) Molecular events associated with arsenic-induced malignant transformation of human prostatic epithelial cells: aberrant genomic DNA methylation and K-ras oncogene activation. Toxicol Appl Pharmacol 206:288–298 4. Bhave MR, Wilson MJ, Waalkes MP (1988) Methylation status and organization of the metallothionein-I gene in livers and testes of strains of mice resistant and susceptible to cadmium. Toxicology 50:231–245 5. Tuck-Muller CM, Narayan A, Tsien F, Smeets DF, Sawyer J, Fiala ES et  al (2000) DNA hypomethylation and unusual chromosome instability in cell lines from ICF syndrome patients. Cytogenet Cell Genet 89:121–128 6. Gonzalgo ML, Liang G, Spruck CH III, Zingg JM, Rideout WM III, Jones PA (1997) Identification and characterization of differentially methylated regions of genomic DNA by methylation-sensitive arbitrarily primed PCR. Cancer Res 57:594–599

7. Greger V, Debus N, Lohmann D, Hopping W, Passarge E, Horsthemke B (1994) Frequency and parental origin of hypermethylated RB1 alleles in retinoblastoma. Hum Genet 94: 491–496 8. Magewu AN, Jones PA (1994) Ubiquitous and tenacious methylation of the CpG site in codon 248 of the p53 gene may explain its frequent appearance as a mutational hot spot in human cancer. Mol Cell Biol 14:4225–4232 9. Matsuda Y, Ichida T, Matsuzawa J, Sugimura K, Asakura H (1999) p16(INK4) is inactivated by extensive CpG methylation in human hepatocellular carcinoma. Gastroenterology 116:394–400 10. Shi H, Maier S, Nimmrich I, Yan PS, Caldwell CW, Olek A et  al (2003) Oligonucleotidebased microarray for DNA methylation analysis: principles and applications. J Cell Biochem 88:138–143 11. Bibikova M, Chudin E, Wu B, Zhou L, Garcia EW, Liu Y, Shin S et  al (2006) Human embryonic stem cells have a unique epigenetic signature. Genome Res 16:1075–1083 12. Kawamata N, Inagaki N, Mizumura S, Sugimoto KJ, Sakajiri S, Ohyanagi-Hara M et  al (2005) Methylation status analysis of cell cycle regulatory genes (p16INK4A, p15INK4B, p21Waf1/ Cip1, p27Kip1 and p73) in natural killer cell disorders. Eur J Haematol 74:424–429

DNA Methylation Profiling of Primary Normal and Malignant B Lymphocytes 13. Heard E, Clerc P, Avner P (1997) X-chromo­ some inactivation in mammals. Annu Rev Genet 31:571–610 14. Herman JG, Baylin SB (2003) Gene silencing in cancer in association with promoter hypermethylation. N Engl J Med 349:2042–2054 15. Kay PH, Spagnolo DV, Taylor J, Ziman M (1997) DNA methylation and developmental genes in lymphomagenesis–more questions than answers? Leuk Lymphoma 24:211–220 16. Klangby U, Okan I, Magnusson KP, Wendland M, Lind P, Wiman KG (1998) p16/INK4a and p15/INK4b gene methylation and absence of p16/INK4a mRNA and protein expression in Burkitt’s lymphoma. Blood 91:1680–1687 17. Klose RJ, Bird AP (2006) Genomic DNA methylation: the mark and its mediators. Trends Biochem Sci 31:89–97 18. Klose RJ, Sarraf SA, Schmiedeberg L, McDermott SM, Stancheva I, Bird AP (2005) DNA binding selectivity of MeCP2 due to a requirement for A/T sequences adjacent to methyl-CpG. Mol Cell 19:667–678 19. Esteve PO, Chin HG, Smallwood A, Feehery GR, Gangisetty O, Karpf AR et  al (2006) Direct interaction between DNMT1 and G9a coordinates DNA and histone methylation during replication. Genes Dev 20:3089–3103 20. Fung MK, Au WY, Liang R, Srivastava G, Kwong YL (2003) Aberrant promoter methylation in gastric lymphoma. Haematologica 88:231–232 21. Gronbaek K, Hother C, Jones PA (2007) Epigenetic changes in cancer. Apmis 115: 1039–1059 22. Henrickson SE, Hartmann EM, Ott G, Rosenwald A (2007) Gene expression profiling

23. 24. 25.

26.

27.

28. 29.

30. 31.

201

in malignant lymphomas. Adv Exp Med Biol 593:134–146 Issa JP (2000) The epigenetics of colorectal cancer. Ann N Y Acad Sci 910:140–153, discussion 153–5 Issa JP (2004) CpG island methylator phenotype in cancer. Nat Rev Cancer 4:988–993 Khulan B, Thompson R, Ye K, Fazzari MJ, Suzuki M, Stasiek E, Figueroa ME et al (2006) Comparative isoschizomer profiling of cytosine methylation: the HELP assay. Genome Res 16:1046–1055 Figueroa ME, Reimers M, Thompson RF, Ye K, Li Y, Selzer RR et al (2008) An integrative genomic and epigenomic approach for the study of transcriptional regulation. PLoS One 3:e1882 Figueroa ME, Wouters BJ, Skrabanek L, Glass J, Li Y, Erpelinck-Verschueren CA, Langerak M et al (2009) Genome-wide epigenetic analysis delineates a biologically distinct immature acute leukemia with myeloid/T-lymphoid features. Blood 113:2795–2804 Fazzari MJ, Greally JM (2004) Epigenomics: beyond CpG islands. Nat Rev Genet 5:446–455 Glass JL, Thompson RF, Khulan B, Figueroa ME, Olivier EN, Oakley EJ et al (2007) CG dinucleotide clustering is a species-specific property of the genome. Nucleic Acids Res 35:6798–6807 Oda M, Greally JM (2009) The HELP assay. Methods Mol Biol 507:77–87 Thompson RF, Reimers M, Khulan B, Gissot M, Richmond TA, Chen Q et  al (2008) An analytical pipeline for genomic representations used for cytosine methylation studies. Bioinformatics 24:1161–1167

Chapter 13 High-Throughput Screening of Metalloproteases Using Small Molecule Microarrays Mahesh Uttamchandani Abstract The promise of rapid and cost-effective drug screening assays on solid support is one that may now be realized with the advent of small molecule microarrays. Many of the initial hurdles in library design and microarray fabrication have been overcome over the last decade, allowing this platform to become more accessible to researchers across both the academic and industrial spheres. Beyond pharmaceutical screening, microarrays reveal quantitative ligand-binding signatures that in the form of protein fingerprints provide a means to discriminate between closely related proteins. The value of protein fingerprinting in drug discovery is also highlighted through the identification of ligands that not only offer good potency, but also good selectivity. Herein, we describe the method for high-throughput screening and profiling of metalloproteases on small molecule microarrays. Metalloproteases are an important class of proteins, which are implicated in the pathogenicity of certain microbes and in the progression of cancer. We have introduced a novel two-colour labelling and application approach that directly elucidates functional ligands, reducing the burden of downstream revalidation of identified hits. Key words:  Small molecule microarray, High-throughput screening, Metalloproteases, Hydroxa­mate peptides, Two-colour labelling/application

1. Introduction One of the earliest and most important steps in any successful drug discovery programme is the identification of small molecule leads that are both potent and selective to the targets of interest (1). Small molecule microarrays represent a new paradigm for performing such high-throughput screening in a rapid and cost-efficient manner (2, 3). The miniaturization, automation and throughput offered on microarrays could significantly accelerate pharmaceutical and commercial screening efforts (4). Since the introduction of small molecule microarrays, a plethora of design Sridar V. Chittur (ed.), Microarray Methods for Drug Discovery, Methods in Molecular Biology, vol. 632, DOI 10.1007/978-1-60761-663-4_13, © Humana Press, a part of Springer Science+Business Media, LLC 2010

203

204

Uttamchandani

and fabrication issues have been addressed and overcome (5). A wide range of combinatorial synthesis protocols, immobilization chemistries, and application guidelines, together with strong infrastructure and commercial support is now available to drive research and future innovation. The platform has hence matured significantly and is now well poised for translation to more routine applications in industry, especially in the areas of pharmaceutical screening and drug discovery. In an effort to further extend the scope and utility of small molecule microarrays, we have developed a protocol for screening and identifying functional inhibitors immediately without the need for downstream revalidation (6–8). It has traditionally been very difficult to distinguish true activity-dependent binding from nonspecific/false positive/ functionally irrelevant binding on microarrays. Hits from microarray experiments frequently have to be retested to establish whether they actually bind to the intended target site. This can significantly limit throughput, especially when hundreds of molecules appear to be positive and each would require retesting. We have thus shown that by employing (a) a target-oriented library of individually synthesized compounds in which every member was characterized prior to spotting on the glass surface; (b) a two-color reciprocal protein labelling/screening strategy; and (c) simultaneous and quantitative measurements of multiple protein ligand-binding interactions, we were able to immediately and reliably elucidate the activity-dependent binding profiles of proteins using small molecule microarrays (6, 7).

2. Materials 2.1. Chemicals and Biochemicals

1. Rink amide-AM resin (GL Biochem).

2.1.1. Combinatorial Synthesis of Small Molecule Library

3. 20 proteinogenic Fmoc protected amino acids (GL Biochem).

2. Hydrochloric acid (HCl; Merck).

4. Fmoc-Lys(Biotin)-OH (GL Biochem). 5. O-Benzotriazole- N,N,N¢,N¢-tetramethyluronium hexafluorophosphate (HBTU; GL Biochem). 6. O-(7-Azabenzotriazole-1-yl)-N,N,N¢,N¢-tetramethyluro­ nium hexafluorophosphate (HATU; GL Biochem). 7. N-Hydroxybenzotriazole (HOBT; GL Biochem). 8. 2,4,6-Collidine (Sigma-Aldrich). 9. Piperidine (Acros Organics). 10. Trifluoroacetic acid (TFA; Sigma-Aldrich). 11. Triisopropylsilane (TIS; Sigma-Aldrich).

High-Throughput Screening of Metalloproteases Using Small Molecule Microarrays

205

12. Acetonitrile (Tedia). 13. N,N-Diisipropylethylamine (DIEA, Sigma). 14. Acetone (Tedia). 15. Dichloromethane (Merck). 16. Diethyl ether (Tedia). 17. Dimethylformamide (DMF; Fisher Scientific). 18. Methanol (Tedia). 2.1.2. Fabrication of Avidin Slides

1. Piranha solution (70% sulfuric acid: 30% hydrogen peroxide) (Tedia or Kanto Chemicals). 2. (3-Aminopropyl)triethoxysilane (Sigma-Aldrich). 3. Succinic anhydride (Sigma-Aldrich). 4. Acetic acid (Tedia). 5. Boric acid (Sigma-Aldrich). 6. Aspartic acid (Sigma-Aldrich). 7. Ethanol (Acros Organics). 8. Sodium bicarbonate (Sigma-Aldrich). 9. Dimethylformamide (DMF; Fisher Scientific). 10. O-Benzotriazole-N,N,N¢,N¢-tetramethyluronium hexafluorophosphate (HBTU; GL Biochem). 11. N-Hydroxysuccinimide (Sigma-Aldrich). 12. N,N-Diisipropylethylamine (DIEA, Sigma-Aldrich). 13. Nitrogen gas. 14. Distilled or MilliQ water. 15. Avidin (Pierce).

2.1.3. Enzyme Labeling and Screening

1. Phosphate-buffered saline (PBS), pH 7.4. 2. Cyanine-3 and Cyanine-5 N-hydroxysuccinimide esters (Cy3NHS, Cy5-NHS; Amersham). 3. Dimethylsulfoxide (DMSO; Fisher Scientific). 4. Hydroxylamine hydrochloride (Sigma-Aldrich). 5. MMP-7 fluorogenic substrate (MCA-Pro-Leu-Gly-Leu-DpaAla-Arg-NH2; Calbiochem). 6. Bovine serum albumin (BSA; Sigma-Aldrich). 7. MMP-7 (Calbiochem). 8. EnzChek Kit (Molecular Probes, Invitrogen). 9. Fluorescein isothiocyanate-conjugated biotin (FITC-biotin; Invitrogen). 10. Tween 20 (Sigma-Aldrich).

206

Uttamchandani

2.2. Equipment and Supplies

1. IRORI MicroKans.

2.2.1. Combinatorial Synthesis of Small Molecule Library

3. Common glassware for organic synthesis.

2. IRORI radiofrequency tag. 4. 60 ml fritted funnel with screw cap (Sigma-Aldrich). 5. Pyrex Buchner funnel with perforated plate (Sigma-Aldrich). 6. Vacuum evaporator (Genevac HT-4X). 7. Centrifuge (Hettich-Zentrifugen, model Rotina 35). 8. 1 L, 250 ml and 100 ml Duran glass bottles with screw caps (Schott). 9. 15 and 50 ml centrifuge tubes (Greiner). 10. HPLC column (analytical reverse phase): C18 (250 × 4.6 mm) (Phenomenex). 11. Waters 600E HPLC system equipped with a Waters 600 controller and a Waters 2487 UV detector. 12. 96-well polypropylene stock plates (Greiner Bio-One). 13. 384-well polypropylene microarray plate (Genetix). 14. Adhesive film (ABgene). 15. 8-channel robotic liquid handler (Precision XS, Biotek) (optional). 16. 25 × 75 mm microscope glass slides (Fisher Scientific). 17. 22 × 60 mm coverslips (Menzel-Glazer). 18. Polypropylene slide staining rack (Kartell). 19. Polypropylene slide staining jar (Kartell). 20. Glass slide staining dish (Electron Microscopy Sciences). 21. Metal slide rack (Electron Microscopy Sciences). 22. Dessicator/dry slide storage box. 23. Oven. 24. Magnetic stirrer. 25. Powder-free gloves. 26. Arrayed (Virtek Chipwriter, ESI SMA) fitted with Stealth Micro Spotting pins (Telechem International). 27. ArrayWoRx Microarray Scanner (Applied Precision). 28. Humid incubation chamber. 29. Microcentrifuge (Sorvall). 30. Circular shaker (Heidolph). 31. Microspin G-25 columns (Amersham, G.E. Healthcare). 32. Heat block (Lab-Line). 33. SpectraMax Gemini XS fluorescence plate reader (Molecular Devices). 34. Microplate shaker (GFL).

High-Throughput Screening of Metalloproteases Using Small Molecule Microarrays

2.3. S  oftwares

207

1. ArrayWoRx Microarray Software (or equivalent). 2. Softmax Pro Microplate Software (or equivalent). 3. Microsoft Excel (or equivalent). 4. TreeView (http://rana.lbl.gov/EisenSoftware.htm). 5. Accutag Synthesis Manager software.

3. Methods 3.1. Combinatorial Synthesis of Small Molecule Library

The design of the biotinylated small molecule hydroxamate library is shown in Fig. 1. The synthetic scheme is provided in Scheme 1. The synthesis of the panel of 14 different trityl protected hydroxamate warheads have been described previously (8–10). Here, the protocol is described for the synthesis of a 400 member library, with a single warhead, using Fmoc solid phase peptide synthesis (11). The procedure may be modified as desired to create various points of diversity. It is also manageable for one

Fig. 1. Design of the 1,400 member hydroxamate peptide library. The 20 natural amino acids were used in the P2¢ and P3¢ position for in the Leu library, while a representative set of 10 amino acids (Nonpolar: Ala, Leu, Phe, Trp, Charged polar: Glu, Lys, His and Uncharged polar: Gln, Ser, Tyr) was utilized for permuting all other sublibraries

208

Uttamchandani

Scheme 1. Synthesis scheme of small molecule library. (1) Fmoc-Lys(Biotin)-OH, HOBT, HBTU, DIEA, DMF; (2) 20% piperidine/ DMF; (b) (i) Fmoc-Gly-OH, HOBT, HBTU, DIEA; (ii) 20% piperidine/DMF; (c) (i) Fmoc-Gly-OH, HOBT, HBTU, DIEA; (ii) 20% piperidine/DMF; (d) (i) Fmoc-AA(P3¢)-COOH, HOBT, HBTU, DIEA; (ii) 20% piperidine/ DMF; (e) (i) Fmoc-AA(P2¢)-COOH, HOBT, HBTU, DIEA; (ii) 20% piperidine/DMF; (f) (i) 4, CPh3ONH-AA(P1¢)-COOH, HATU, 2,4,6-collidine; (ii) 20% piperidine/DMF; (g) 95% TFA/5% TIS, 2 h

person to handle this protocol with 400–500 MicroKans independently, which would consume a period of between 8 and 12 days. The procedure can be repeated to generate the large 1,400 member library set described. 1. Calculate the amount of rink amide resin required for library synthesis. The capacity for each MicroKan is 30 mg of resin. This would work out to a total of 12  g of resin for a 400 member library. The theoretical yield of each library member at this scale of synthesis would be 15  mmoles, when using rink amide resin with 0.5  mmol/g resin loading capacity. This quantity is more than sufficient for printing several thousand microarray slides. 2. Weigh 12 g of rink amide resin into a fritted funnel, with a screw cap. Add 35 ml of DMF, cap both ends of the funnel and shake for 2 h to allow the resin to swell. Drain the DMF using suction. 3. Repeat the wash step with 35  ml DMF for a further three times, for 15 min each time with shaking. 4. Deprotect the resin using 20% v/v piperidine in DMF for 1 h with shaking. Drain the piperidine solution using suction. 5. Wash the resin with DMF (35 ml) for three times, 15 min per wash with shaking

High-Throughput Screening of Metalloproteases Using Small Molecule Microarrays

209

6. Wash with DCM (35 ml) for a further three times for 15 min per wash with shaking. 7. Wash with DMF (35 ml) for a further three times for 15 min per wash with shaking. 8. Transfer the resin from the funnel into a 250 ml glass bottle. Rinse with 25  ml of DMF to ensure all resin is completely transferred. Repeat with another 10 ml of DMF. 9. Couple resin with Fmoc-Lys(biotin)-OH. Weigh out 4 equivalents (24 mmol) of Fmoc-Lys(biotin)-OH, HBTU and HOBt in a separate bottle. Dissolve in 80  ml of DMF and add 8 equivalents of DIEA (48 mmol). Mix well, and allow for preactivation by leaving the solution to stand for 15 min. 10. Add the preactivated solution to the resin. Seal the bottle with parafilm and shake overnight (~12 h). 11. Filter the resin using suction through a 350 ml fritted funnel. Ensure that all resin is transferred using small volume rinses with DMF. 12. Wash the resin using 3× DMF, 3× DCM and 3× DMF, as described in steps 5–7, using 100 ml of solvent per wash. 13. Isolate around 10 beads, and perform the ninhydrin test to ensure that the coupling is successful and complete (see Note 1). 14. Transfer the resin to a clean bottle and deprotect Fmoc using 150 ml of 20% piperidine in DMF. 15. Wash as detailed in step 12. 16. Repeat steps 8–15 using Fmoc-Gly-OH in place of FmocLys(biotin)-OH to couple a glycine residue. 17. Repeat step 16 to add on another glycine residue. 18. Dry the resin under vacuum, using an oil pump for 6 h. 19. Distribute ~30  mg of resin in each of 400 MicroKans reactors. Include an IRORI radiofrequency tag to each reactor and ensure that the cap is fitted on tightly. Load four extra reactors for ninhydrin tests to monitor coupling efficiency. 20. Program the Accutag Synthesis Manager software for a 20 × 20 aa = 400-member library. 21. Scan the 400 reactors to encode each tag and sort into twenty 100 ml bottles. Reactors in each bottle will be coupled with the same amino acid. 22. To couple the P3¢ residues, prepare preactivated solutions for each of the 20 Fmoc-protected proteinogenic amino acids, using the 4 equivalent molar excess of amino acid, HBTU,

210

Uttamchandani

HOBt and 8 equivalent molar excess of DIEA. Prepare these solutions in 50 ml of DMF. 23. Add the twenty different amino acid preactivated solutions to the 20 bottles containing the respective MicroKans. Ensure all bottles are appropriately labelled. 24. Shake bottles overnight (~12 h). 25. Drain the solutions in each bottle and wash twice (for 15 min) with 60 ml of DMF. 26. Pool all reactors into a 1 L bottle and rinse using 3× DMF, 3× DCM and 3× DMF, as described in steps 5–7, using 500 ml of solvent per wash. 27. Perform Fmoc deprotection using 400 ml of 20% piperidine in DMF. 28. Wash as described in step 26. 29. Repeat steps 21–28 to sort the MicroKans into 20 bottles couple the P2¢ position. 30. For the final coupling of the P1¢ position, consolidate all MicroKans into a clean 1 L bottle. Proceed with steps 22–28, except use 4 equivalents of a trityl protected hydroxamate warhead together with HATU and 2,4,6-collidine (in a 1:1:1.9 ratio) in a 100 ml DMF volume for the coupling step (see Note 2). 31. Perform a final wash with 500 ml of methanol for three times, with shaking each time for 15 min. 32. Dry the resin under vacuum, using an oil pump for 6 h. 33. Prepare the cleavage solution comprising TFA and TIS in the ratio of 19:1. Sort the 400 reactors into individually identified 15 ml tubes. Dispense 1.5 ml of the cleavage solution to each tube, and shake for 3 h. TFA is corrosive and generates fumes, so perform this step carefully with proper protection in a fume hood. 34. Organize 5 96-well deep well plates such that the identities of the samples will be preserved upon transferring all the solutions from each of the four hundred 15  ml tubes to the plates. Carefully transfer solutions from tubes to plate. 35. Concentrate and remove TFA and TIS using a vacuum evaporator, at a temperature of 35°C at a spin force of 300 g, until about 0.1 ml of the solution remains in each well. 36. Add 1.5 ml of cold ether to each well to precipitate the small molecule. Seal the plates with adhesive film and place in a −20°C freezer overnight. 37. Spin down the precipitated products at 1,000 g for 30 min, and decant the ether.

High-Throughput Screening of Metalloproteases Using Small Molecule Microarrays

211

38. Dissolve the products in 0.5  ml of DMF. This would give stock concentrations in the range of 1–10 mM. Plates may be sealed and stored for the long term at −80°C. 39. Perform analysis using LC-MS to determine the quality and purity of desired products. 3.2. Fabrication of Avidin Slides

Glass slides are derivatized as shown in Scheme 2 through processes of silanization and chemical activation in order to produce a stable and even avidin surface for microarray printing. Care has to be taken when handling glass slides (see Note 3). (This step takes about 2 days.) 1. Glass slides are cleansed and soaked in piranha solution (70% sulfuric acid: 30% hydrogen peroxide) to remove any organic contaminants and oxidize the silane surface. Use the polypropylene staining dish and racks. The slides may remain soaked in the solution until ready for use. 2. The slides are thereafter transferred to a stainless steel rack washed copiously with distilled water and dried. 3. Slides are then silanized using 400 ml of a solution containing 3% (aminopropyl)triethoxysilane, 2% water and 95% ethanol (sufficient for 30 slides). Premix the solutions in a glass staining jar for 10 min and add in the slides in the metal rack. Continuously stir the solution using a magnetic stirrer. 4. Incubate for 1–2 h. 5. Remove the slides from the solution and rinse copiously with ethanol. Air dry the slides to remove any residual ethanol. 6. Transfer the slide rack to a deep well dish and cure at 150°C for 4–8 h. Cool the slides till they reach room temperature and rinse then with ethanol. Thereafter, air dry the slides, or

Scheme 2. Fabrication of avidin coated slides. Glass slides are first silanized with amino-silane to generate a surface amino (-NH2) functionality. This is then reacted to produce a carboxylic acid functionality, which is then converted to an activated NHS ester. The slide is finally coated to produce an avidin surface

212

Uttamchandani

to speed up the drying process, use centrifugation or a stream of nitrogen/compressed air to dry the slides. 7. The resulting amine slides are then reacted with succinic anhydride to produce a surface carboxylic acid functionality. 8. Dissolve succinic anhydride to a final concentration of 180 mM in 650 ml of DMF. Once dissolved, add in 30 ml of 1  M solution of sodium borate at pH 9. Use a slide staining dish with a magnetic stirrer for continuous stirring. 9. Incubate the slides for 30  min, and thereafter transfer the slides to a boiling distilled water bath for 2 min to cleanse the slides. 10. Rinse the slides with ethanol and dry as described in step 6. 11. The resulting carboxylic acid slides are then esterified with NHS. Prepare a solution of NHS/HBTU/DIEA (100 mM/ 100 mM/200 mM) in DMF. 12 ml of the solution is required for coating 30 slides. 450–500 ml of the solution is applied to each slide (placed horizontally on a clean surface) under a 60 × 22 mm coverslip. 12. Cover the slides and incubate for 3 h. As only one surface of the slide is activated, care must be given to handling and washing the slides, so the orientation is preserved. 13. Slide the coverslip off each slide, and rinse the slides briefly with ethanol until all the reaction solution is removed. Slides are dried as described in step 6. 14. The resulting NHS slides are then reacted with avidin. Apply 50  ml of a 1  mg/ml solution of avidin in 10  mM sodium bicarbonate buffer, pH 9, to the NHS surface of the slides under coverslip (see Note 4). Take care to ensure no bubbles are introduced. 15. After 30 min incubation, gently slip off the coverslip and rinse slides with water to remove the unreacted avidin and dry. Regions coated with avidin will appear hydrophilic. 16. Quench the unreacted groups on the slides using a 2  mM solution of aspartic acid in 0.5 M sodium bicarbonate buffer, pH 9. 17. Rinse the slides with distilled water and dry. The avidin functionalized slides may be stored for extended periods at 4°C but are best prepared fresh before use. 3.3. Printing Chemical Libraries on Microarray Slides

(This step takes about 1–2 days.) 1. Prepare working stock solution of the small molecule library in Genetix 384-well. Dilute 8 ml of library stock to 8 ml of phosphate buffered saline (PBS), pH 7.4, to an approximate concentration of 2.5 mM. This is a very high saturating

High-Throughput Screening of Metalloproteases Using Small Molecule Microarrays

213

concentration, and lower concentrations may also be applied, (usually above 100 mM). This pipetting step should be carried out using multichannel pipettes and automatic liquid handlers. 2. Shake the 384 well plates to mix and spin down. These working plates may be prepared in advance and stored at −20°C or −80°C until ready for spotting. 3. Design grids to plan the arrangement of spots on the microarrays using the gridding software. Spot the library in duplicate, such that each sample is spotted twice, side by side. Incorporate sufficient redips in the programme, so that the sample does not run out of the pin during spotting. Blotting may also be used to enhance spot reproducibility. 4. It is useful to include biotin dye control on the plate, which will be spotted as alignment reference for the arrays. In this case, a 1–10 mM solution of FITC labelled biotin may be used for this purpose. 5. Ensure that the spotter is clean and dust free. 6. Load the required number of pins (see Note 5). Place the desired numbers of slides in the spotter, and ensure that the avidin surface is facing upwards. 7. Refill the washing buffers – distilled water and 70% ethanol. Two washing cycles for the pins, using 10 s wash with water and 10  s sonication with ethanol, is generally adequate to ensure there is no cross contamination. Include a drying step for the pins before returning to the plate to pick up the next sample. 8. Top up the washing buffers during the print run, if required. 9. Leave the slides in the spotter for at least 30 min upon completion of the print run. Printed slides may be stored in a dessicator at 4°C until ready for use. 10. Export the .gal file from the microarray software so that the spot arrangement and layout may be communicated to the microarray scanner (see Note 6). 3.4. Protein Labelling and Application

It has traditionally been difficult with small molecule microarrays to discern whether the spots that appear positive are true positives (indicative of activity-dependent binding) or false positives (a result of nonspecific binding events). Analysing these effects across different slides is challenging due to slide-slide variability and other differences. In order to immediately determine if the results on microarrays are due to actual protein activity, we have developed a two-colour labelling and application strategy that reduces the need for hit revalidation on microarrays (Fig.  2). Briefly, proteins to be screened are labelled in two channels. One channel is denatured. Both the active and inactive channels are mixed and applied on the same microarray. Comparison of the

214

Uttamchandani

Fig. 2. Reciprocal protein labelling and application strategy. Proteins are labelled in two channels (Cy3 – white, Cy5 – dark grey). Protein in one channel is inactivated by heat. Both the active and inactive proteins from different channels are mixed in equal amounts and applied to small molecule microarrays (top and bottom), to control for non-specific binding effects from the inactive channel. Results from the reciprocal experiments are averaged, to produce the active protein fingerprint

two-colour fingerprints allows the elucidation of activity-dependent binding events. (This step takes about 3 days.) 1. Prepare enzyme stock solutions in water or PBS buffer. Ensure enzymes are free of amine-based buffers such as Tris, which could interfere with the labelling procedure. Store these stock solutions at −20°C (avoid the use of glycerol, if possible). When in use, keep enzymes on ice to preserve activity. 2. Dissolve 1 vial of Cy3- and Cy5- conjugated N-hydroxy­ succinimide (NHS) esters in 100 ml of DMSO. 3. Minimally label the protein with Cy3 and Cy5 in two separate 45 ml reactions comprising 1.25 ml of each Cy3 and Cy5 dyes and 50 mg of protein in 0.1 M sodium carbonate buffer, pH 9.0 (see Note 7). 4. Mix and incubate for 30 min. 5. Quench the unreacted dyes with tenfold excess of hydroxylamine (pH 8.5) or using 5  mM Tris–HCl (pH 8.0), to a final volume of 50 ml. 6. Remove the unreacted dye using by size exclusion using Sephadex G-25 columns. Collect the enzyme enriched fraction in an eppendorf tube. The labelled protein may be stored at −20°C, but best results are obtained when it is prepared fresh before use. 7. Test protein activity to ensure the enzymes are still active before application onto the microarrays. Generic enzyme kits, like bodipy labelled casein (EnzChek Kit) or specific fluorogenic/calorimetric enzyme substrate assays may be adopted. Mild reduction in activity post-labelling is acceptable. Test enzyme activity using a fluorescence plate reader. 8. Wash the printed small molecule microarray slides with PBS for 20 min with shaking. Rinse with water and dry.

High-Throughput Screening of Metalloproteases Using Small Molecule Microarrays

215

9. Block the slide surface with 1% bovine serum albumin (BSA) in PBS for 30 min. 10. Heat denature enzyme in one channel. Heat the enzyme at 95°C for 10 min and immediately after plunge into ice. 11. Mix 10 mg of each active and inactive channels of protein in PBS containing 1% BSA, to a final volume of 100 ml. 12. Apply on the microarrays using the coverslip method (see Note 4). 13. Perform a reciprocated duplicate with active and inactive channels inverted. 14. Incubate slides for 1–2 h. 15. Remove the coverslip and rinse away unbound enzyme with distilled water. 16. Wash the slides in PBS buffer containing 0.1% Tween 20 (PBST) for 10–30 min, as long as required to obtain the best signal-to-noise ratio. 17. Wash slides with distilled water, dry and scan using a microarray scanner in both the Cy3 and Cy5 channels. 18. Process the data obtained. Normalize the signals by subtracting away the individual spot backgrounds. Then, subtract the background signals (from the inactivated enzyme channel). Average duplicate spots within the same slide, and remove data points that are inconsistent (greater than 20% difference in values). Perform the same procedure for the slide with the reciprocated dyes. 19. Plot the data against each other to showcase the reproducibility of data when crossing over the dyes, as shown in Fig. 3.

Fig. 3. A scatter plot of the data from the reciprocal experiment with thermolysin. A high degree of correlation is generally obtained (r = 0.92). Outliers shown in the open boxes are points that are significantly different in the two channels and can be excluded. The majority of points, (closed spheres) are generally consistent across both slides. (Reproduced from ref. (6) with permission from the American Chemical Society.)

216

Uttamchandani

Fig. 4. Activity based protein fingerprint for 5 enzymes; I – Thermolysin, II – Collagenase, III – Carboxypeptidase, IV – Anthrax Lethal Factor and V – Human Matrix Metalloproteases (MMP)-7. The 1,400 small molecules are sorted according to their P1¢ positions in the heatmap. As represented by the scale inset, the regions in black denote strong inhibitor potencies, whilst those in grey or white show weaker or no binding respectively. (Reproduced from ref. (6) with permission from the American Chemical Society.)

Analyse the resultant values from the two slides using Pearson correlation to establish the reproducibility of data. 20. Average the data from the reciprocated experiment. Present the data in the form of coloured heatmaps, using the Treeview software, as shown in Fig. 4. 21. Inhibitors may then be selected from the entire dataset that demonstrate high potency and selectivity against the target of interest. The inhibition potency may be quantitatively determined using IC50 or KD measurements (Fig. 5).

4. Notes 1. The presence of a blue colouration on the resin and/or solution implies the presence of free amines, and hence indicating incomplete coupling. On the other hand, a straw yellow colour indicates no free amines, and complete coupling. Repeat coupling, if necessary, until a straw yellow colour is obtained. Alternatively, if coupling remains incomplete after several tries, capping may be performed using acetic anhydride.

High-Throughput Screening of Metalloproteases Using Small Molecule Microarrays

217

Fig. 5. The structure of a potent and selective anthrax lethal factor inhibitor uncovered, with an IC50 of 2 mM and a KD of 0.81 mM

2. If additional warheads (or amino acids) are to be coupled, this may be programmed into the Accutag Synthesis Manager software and performed accordingly in separate reaction bottles. 3. Care must be taken when handling slides to be used in microarray experiments to ensure that at all stages of derivatisation, spotting and sample application no dust or dirt come into contact with the planar surfaces. Such particles may cause extraneous fluorescence or result in scratches that could affect the fluorescent readout when the slides are scanned. Ensure all surfaces and slide racks are clean and rinse these surfaces with ethanol before placing in direct contact with the slides. Gloves, if used, should be of the powder-free variety to ensure that the slides remain uncontaminated even after handling. 4. If there is sufficient reagent, it may be convenient to react both surfaces of the slides by placing them on slide racks and staining dishes. However for expensive reagents, where it is preferable to use a conservative volume of the chemical, coverslips may be used. For a 22 × 60 mm coverslip, a 50 ml preparation is sufficient for confluent coverage. Two methods

218

Uttamchandani

may be used to apply the reagent on the slide surface. Either the reaction mix is applied directly to the slide and the coverslip is applied, or the mix could be applied to the coverslip, and the slide may be inverted onto this. Both methods work equally well, but one ought to use the method that produces a uniform spread of the reagent across the slide surface without introducing any bubbles. The slides should then be placed in a humid, enclosed environment or slide hybridization chambers to prevent evaporation of the reaction mix. 5. Different pin sizes and volumes are available according to the type of spotting requirements. SMP3 stealth pins from Telechem are very commonly used for DNA microarray fabrication, which generally produce spot sizes around 100 mm. We found that when spotting viscous mixtures like proteins and small molecules (prepared in high concentrations of DMSO/DMF), it is preferable to use larger pins like SMP8B (this is the size we have used in this project) or even SMP15B that produce spots in the size range of 250 mm and 500 mm respectively. This nevertheless reduces the numbers of features that may be printed on a single microarray. For the SMP8B pin we used, we were generally able to print up to 3,000–3,500 spots on a single slide. These larger pins come with “bubble” features to cater for a larger uptake volume to reduce the need for excessive redips. Further details are available from the manufacturers of these pins. 6. It is useful to have pairs of microarray spotters and scanners that are able to communicate using the same file formats. This enables the scanner softwares to deconvolute the spot identities. Even if some spotter outputs may be incompatible, scanner softwares usually come with features allowing spot designs to be inputted, thereby assisting with spot identification and analysis. 7. It might be necessary to optimize the labelling ratios to determine the best concentration of dye to use for minimal labelling. We recommend trying dye (ml) to protein (mg) ratios in the range 1:60 to 1:30. However, the final ratio required may be dependent on different types of enzymes or proteins. The key is to ensure maximal retention of activity after the labelling procedure.

Acknowledgments Funding support is acknowledged from DSO National Laboratories and the National University of Singapore.

High-Throughput Screening of Metalloproteases Using Small Molecule Microarrays

219

References 1. (2007) The academic pursuit of screening. Nat Chem Biol 3:433 2. Uttamchandani M, Walsh DP, Yao SQ, Chang YT (2005) Small molecule microarrays: recent advances and applications. Curr Opin Chem Biol 9:4–13 3. Hu Y, Uttamchandani M, Yao SQ (2006) Microarray: a versatile platform for highthroughput functional proteomics. Comb Chem High Throughput Screen 9:203–212 4. Duffner JL, Clemons PA, Koehler AN (2007) A pipeline for ligand discovery using smallmolecule microarrays. Curr Opin Chem Biol 11:74–82 5. Uttamchandani M, Wang J, Yao SQ (2006) Protein and small molecule microarrays: powerful tools for high-throughput proteomics. Mol Biosyst 2:58–68 6. Uttamchandani M, Lee WL, Wang J, Yao SQ (2007) Quantitative inhibitor fingerprinting of metalloproteases using small molecule microarrays. J Am Chem Soc 129:13110–13117

7. Lee WL, Li J, Uttamchandani M, Sun H, Yao SQ (2007) Inhibitor fingerprinting of metalloproteases using microplate and microarray platforms: an enabling technology in Catalomics. Nat Protoc 2:2126–2138 8. Wang J, Uttamchandani M, Sun LP and Yao SQ (2006) Activity-based high-throughput profiling of metalloprotease inhibitors using small molecule microarrays. Chem Commun (Camb) 717–719 9. Wang J, Uttamchandani M, Li J, Hu M, Yao SQ (2006) Rapid assembly of matrix metalloprotease inhibitors using click chemistry. Org Lett 8:3821–3824 10. Uttamchandani M, Wang J, Li J, Hu M, Sun H, Chen KY, Liu K, Yao SQ (2007) Inhibitor fingerprinting of matrix metalloproteases using a combinatorial peptide hydroxamate library. J Am Chem Soc 129:7848–7858 11. Chan WC, White PD (2000) Fmoc solid phase peptide synthesis. Oxford University Press, New York

Chapter 14 Metabolic Enzyme Microarray Coupled with Miniaturized Cell-Culture Array Technology for High-Throughput Toxicity Screening Moo-Yeal Lee, Jonathan S. Dordick, and Douglas S. Clark Abstract Due to poor drug candidate safety profiles that are often identified late in the drug development process, the clinical progression of new chemical entities to pharmaceuticals remains hindered, thus resulting in the high cost of drug discovery. To accelerate the identification of safer drug candidates and improve the clinical progression of drug candidates to pharmaceuticals, it is important to develop high-throughput tools that can provide early-stage predictive toxicology data. In particular, in vitro cell-based systems that can accurately mimic the human in vivo response and predict the impact of drug candidates on human toxicology are needed to accelerate the assessment of drug candidate toxicity and human metabolism earlier in the drug development process. The in vitro techniques that provide a high degree of human toxicity prediction will be perhaps more important in cosmetic and chemical industries in Europe, as animal toxicity testing is being phased out entirely in the immediate future. We have developed a metabolic enzyme microarray (the Metabolizing Enzyme Toxicology Assay Chip, or MetaChip) and a miniaturized three-dimensional (3D) cell-culture array (the Data Analysis Toxicology Assay Chip, or DataChip) for high-throughput toxicity screening of target compounds and their metabolic enzyme-generated products. The human or rat MetaChip contains an array of encapsulated metabolic enzymes that is designed to emulate the metabolic reactions in the human or rat liver. The human or rat DataChip contains an array of 3D human or rat cells encapsulated in alginate gels for cell-based toxicity screening. By combining the DataChip with the complementary MetaChip, in vitro toxicity results are obtained that correlate well with in vivo rat data. Key words:  Microarray, Metabolic enzyme array, MetaChip, Cell array, DataChip, High-throughput toxicity screening

1. Introduction The metabolism of chemicals, including drug compounds, in the human body primarily occurs in the liver via a variety of oxidative and conjugative routes. Among the many metabolic enzymes, Sridar V. Chittur (ed.), Microarray Methods for Drug Discovery, Methods in Molecular Biology, vol. 632, DOI 10.1007/978-1-60761-663-4_14, © Humana Press, a part of Springer Science+Business Media, LLC 2010

221

222

Lee, Dordick, and Clark

cytochrome P450 (CYP450) isoforms, which catalyze first-pass (Phase I) functionalization reactions (e.g., hydroxylation, dealkylation, oxidation, deamination, and dehalogenation), are the most important (1, 2). Subsequent conjugation reactions (e.g., glucuronidation, sulfation, acetylation, and addition of amino acids and peptides) are catalyzed by Phase II metabolic enzymes including UDP-glycosyltransferase (UGT), sulfotransferase (SULT), and glutathione S-transferase (GST), resulting in the formation of more soluble compounds and eventually enhancing excretion (3, 4). Thus, understanding the role of these metabolic enzymes is important in drug metabolism and for human toxicology testing. There have been a number of in vitro approaches developed for human metabolism and toxicology screening, including isolated liver slices, primary hepatocytes, transformed cultured human hepatoma cell lines, purified microsomal preparations, or isolated and purified P450s (5, 6). Two-dimensional (2D) hepatocytes cultures in multi-well plates can be adapted to high-throughput screening, and have been considered as the gold standard of in vitro human metabolism and toxicology for replacement of animal testing (7). However, 2D hepatocyte cultures may not emulate the environment and cellular architecture found in vivo (8, 9). Thus, in vitro results obtained from 2D hepatocyte cultures may not provide toxicity information that correlates with in vivo animal data. To address this limitation, we have developed a metabolic enzyme microarray (the Metabolizing Enzyme Toxicology Assay Chip, or MetaChip) and a miniaturized 3D cell-culture array (the Data Analysis Toxicology Assay Chip, or DataChip) for high-throughput toxicity screening of target compounds and their metabolic enzyme-generated products (10–12). The MetaChip can be prepared by spotting human or rat metabolic enzymes including individual CYP450s, a mixture of CYP450s, individual phase II metabolic enzymes, a mixture of phase II metabolic enzymes, a mixture of all metabolic enzymes, liver microsomes, and s9 fractions, all encapsulated in alginate gels (as small as 15 nL) arrayed on a methyltrimethoxysilane (MTMOS)-coated glass slide. The DataChip can be prepared by printing human or rat cells in alginate gels (as small as 30  nL) onto a poly(styrene-co-maleic anhydride) (PS-MA)-coated glass slide for toxicity screening against multiple target cells. Using the human or rat DataChip coupled with the human or rat MetaChip, the toxicity of parent compounds have been compared to that of their products generated by various metabolic enzymes (Fig.  1). We have demonstrated that a single DataChip containing 1,080 individual cell cultures, used in conjunction with the complementary human CYP450containing MetaChip, can simultaneously provide IC50 values for nine compounds and their metabolites from CYP1A2, CYP2D6,

Microarray Coupled with Miniaturized Cell-Culture Array Technology

223

DataChip Spotting of BaCl2-PLL on the PS-MA-coated slide Drying

MetaChip Spotting of compounds in BaCl2-PLL on the MTMOS-coated slide

Spotting alginate containing cells Gelation and incubation

Gasket

Drying

Spotting alginate containing enzymes Removing excess growth medium

Stamping

Store in a - 80°C freezer until used

Washing and incubation

Staining the Data Chip with a live/dead kit Scanning with an array scanner

Cytotoxicity Fig. 1. Schematic of the MetaChip platform coupled with the DataChip for evaluating toxicity of compounds and their enzyme-generated metabolites

and CYP3A4, and an equimolar mixture of the three P450s (12). In addition, we have compared cytotoxicity data from different metabolic enzyme mixtures and cell sources for the validation of the MetaChip/DataChip platform, resulting in better correlations between in  vivo rat data and in  vitro data. We envisage that the MetaChip/DataChip platform represents a promising, highthroughput microscale alternative and creates new opportunities for rapid and inexpensive predictive assessment of human toxicity.

2. Materials 2.1. Slide Treatment

1. Fisherbrand plain microscope slides (Fisher Scientific, now Thermo Fisher Scientific, Rockford, IL). 2. Wheaton glass 20 slide staining dish with removable rack (Fisher).

224

Lee, Dordick, and Clark

3. Methyltrimethoxysilane (Sigma-Aldrich, St. Louis, MO). 4. Poly(styrene-co-maleic anhydride): 14% (w/w) maleic anhydride (Sigma-Aldrich). 5. Potassium phosphate buffer solution: 200  mM, pH 8 (Invitrogen, Carlsbad, CA). Prepare 25  mM potassium phosphate buffer (pH 8) by mixing 200 mM potassium phosphase buffer (pH 8) with de-ionized distilled water. 6. National scientific 20 mL glass sample vials with PTFE-lined solid storage caps (Fisher). 7. Petri dishes with clear lids, 150 × 15 mm (Fisher). 2.2. Microarray Spotting

1. Standard solenoid valve for a normal pin head (Genomic Solutions, Inc., now Digilab, Inc., Ann Arbor, MI). 2. Ceramic dispensing tip, 100 mm orifice (Genomic Solutions). 3. Arrayit hybridization cassette, extra deep chamber, 2.54 mm depth (TeleChem International, Inc., Sunnyvale, CA). 4. Costar clear polystyrene 96-well plates with lid (untreated, round well, sterile) (Corning life sciences, Inc., Lowell, MA). 5. Poly-l-lysine (PLL) solution: 0.01%, molecular weight 70,000–150,000, sterile-filtered, cell culture tested (Sigma). Store the PLL solution in aliquots at 4°C. 6. Barium chloride (Sigma-Aldrich). Prepare a 0.1  M barium chloride (BaCl2) solution by adding barium chloride in sterile de-ionized distilled water and vortex well for complete dissolution. After sterile filtering the solution using a 0.2 mm syringe filter, store the BaCl2 solution in aliquots at 4°C. 7. Alginic acid sodium salt from Macrocystis pyrifera (Kelp), low viscosity (Sigma-Aldrich). Prepare a 3% (w/v) alginic acid solution by dissolving alginic acid sodium salt in sterile de-ionized distilled water in a sterile glass sample vial. Place a small stir bar in the vial and let it mix for at least 3  d for complete dissolution. Store the alginate solution in aliquots at 4°C. 8. Dulbecco’s phosphate-buffered saline (PBS) without CaCl2 and MgCl2 (Invitrogen). Store Dulbecco’s PBS in aliquots at room temperature. 9. Vivid® CYP450 screening kit (Invitrogen). Store CYP450 isoforms in aliquots at −80°C. After sterile filtering the solution using a 0.2 mm syringe filter, store NADP and a regeneration system in aliquots at −80°C. 10. Baculosome® reagents, TNI-WT negative control (Invitrogen). Store TNI-WT negative control in aliquots at −80°C.

Microarray Coupled with Miniaturized Cell-Culture Array Technology

225

11. Supersomes™ human UGT isoforms (BD Biosciences). Store UGT in aliquots at −80°C. 12. UGT reaction mix A and B (BD Biosciences). After sterile filtering the solution using a 0.2 mm syringe filter, store UGT reaction mix in aliquots at −80°C. 13. Human sulfotransferase (SULT) isozymes (Sigma). Store SULT in aliquots at −80°C. 14. Adenosine 3¢-phosphate 5¢-phosphosulfate (PAPS) tetralithium salt (CalBioChem, now EMD Chemicals, Inc.). After sterile filtering the solution using a 0.2 mm syringe filter, store PAPS in aliquots at −80°C. 15. Human glutathione S-transferase (GST) (Sigma). Store GST in aliquots at −80°C. 16. Glutathione (GSH) (Sigma). After sterile filtering the solution using a 0.2 mm syringe filter, store GSH in aliquots at −80°C. 17. Human liver microsome (HLM) (BD Biosciences). Store HLM in aliquots at −80°C. 2.3. Cell Culture

1. Hep3B human hepatoma cells (ATCC, Manassas, VA). 2. RPMI 1640 with l-glutamine, 1× (Mediatech, Manassas, VA). Prepare RPMI supplemented with 10% fetal bovine serum (FBS) and 1% Penicillin-Streptomycin (PS) and store it at 4°C. 3. Fetal bovine serum (FBS) qualified, USA origin, sterile-filtered, cell culture tested (Sigma-Aldrich). Store FBS in aliquots at −20°C. 4. Penicillin-Streptomycin (Invitrogen). Store PS in aliquots at −20°C. 5. Trypsin-EDTA: 0.05% trypsin, 0.53 mM EDTA (Invitrogen). Store Trypsin–EDTA in aliquots at −20°C. 6. BD Falcon tissue culture flasks, 75 cm2 culture area, canted neck, plug seal cap (Fisher). 7. Petri dishes with clear lids, 100 × 15 mm (Fisher).

2.4. Cell Staining

1. Live/dead viability/cytotoxicity kit (Invitrogen). Store the staining dyes at −20°C. 2. BupH phosphate buffered saline (PBS) packs containing 0.1  M phosphate, 0.15  M NaCl, pH 6.9–7.2 when dissolved in 500  mL distilled water (Pierce Biotechnology, now Thermo Fisher Scientific). Store BupH PBS at room temperature. 3. Silicone rubber sheet, 1  mm thick (McMaster-Carr Supply Co., Robbinsville, NJ).

226

Lee, Dordick, and Clark

3. Methods 3.1. Preparation of Functionalized Slides 3.1.1. Acid Cleaning of the Glass Slide

To prepare a uniform coating on a glass surface, it is important to remove all dirt to expose silanol groups (−SiOH) on the slide surface. Acid-cleaning steps are, therefore, essential for standard microscope slides. 1. Wipe the plain microscope slides (25 × 75 × 1 mm) three times with 70% ethanol-soaked paper towels to remove oil and solid particles on the slide surface and then clean ethanol with dry paper towels (see Note 1). 2. Place the slides in a removable glass rack, immerse the rack in a Wheaton glass dish filled with concentrated sulfuric acid (96.5%) for 2  h, and then sonicate the Wheaton glass dish with the slides in acid for 10 min. 3. Rinse the rack containing the slides at least five times with de-ionized distilled water to remove excess sulfuric acid on the slide surface. 4. Immerse the rack containing the slides twice in a Wheaton glass dish filled with acetone to remove excess water on the slide and dry the acid-cleaned glass slides thoroughly under N2 gas stream. 5. Bake the acid-cleaned slides in an oven at 120°C for 10 min to completely remove water on the slide surface. 6. Store the slides in a sterile Petri dish (150  mm diameter, maximum five slides/Petri dish) until use. Never leave the clean slides uncovered for extended times.

3.1.2. Poly (Styrene-co-Maleic Anhydride) (PS-MA) Coating for Cell Printing

1. Prepare a fresh 1% (w/v) solution of PS-MA by dissolving PS-MA in anhydrous toluene in a scintillation vial with a PTFE-coated lid. Shake the solution in an incubating shaker at 50°C, 5  g for 40  min until PS-MA pellets completely dissolve (see Note 2). 2. Prepare a 0.1% (w/v) PS-MA solution by mixing 18 mL of anhydrous toluene with 2  mL of the 1% (w/v) PS-MA solution. Prepare a fresh PS-MA solution and use it immediately. 3. Load 1 mL of the 0.1% (w/v) PS-MA solution onto the acidcleaned slide and spin coating the solution using a spin coater (Model PWM32, Headway Research, Inc., Garland, TX) at 200 g for 30 s. Remove excess PS-MA remaining on the back of the slide by wiping with an acetone-soaked paper towel.

Microarray Coupled with Miniaturized Cell-Culture Array Technology

227

4. Dry the PS-MA-coated slides overnight at room temperature in a sterile Petri dish and store until needed. 3.1.3. Methyltrimethoxy­ silane (MTMOS) Coating for Compound and Enzyme Printing

1. Prepare a fresh MTMOS-HCl sol solution by mixing 8 mL of MTMOS with 3.2 mL of 5 mM HCl, followed by vortex for 2 min and sonication for 10 min (see Note 3). 2. Immediately before spin coating, mix 11.2  mL of the MTMOS–HCl sol solution with 8 mL of potassium phosphate buffer solution (25 mM, pH 8) and use the mixture within 15 min (see Note 4). 3. While spinning the acid-cleaned slide at 200  g for 30  s, place 1.5 mL of the mixture onto the slide and remove any excess MTMOS on the back of the slide using acetone-soaked paper towel. 4. To complete MTMOS gelation, dry the slides overnight at room temperature in a sterile Petri dish and store until needed.

3.1.4. Preparation of a Sol-Gel Gasket on the MTMOS-Coated Slide for Stamping

A sol-gel gasket is fabricated by spotting an MTMOS sol-gel solution around the perimeter of the MTMOS-coated slide and baking the slide to harden the sol-gel drops. The sol-gel gasket on the MTMOS-coated slides is served as a spacer to maintain a suitable distance between the MetaChip and the DataChip and to prevent spot detachment by direct contact of the spots after stamping. For efficient stamping and transfer of compounds onto the cells, the gasket height is optimized as a function of the HCl concentration, the sol-gel volume, and the baking time. 1. Prepare a fresh MTMOS–HCl sol solution by mixing 2 mL of MTMOS with 1  mL of 10  mM HCl, followed by vortex for 2 min and sonication for 10 min. 2. Immediately before spotting the sol solution onto the slides, mix 3  mL of the MTMOS–HCl sol solution with 1  mL of Dulbecco’s PBS without CaCl2 and MgCl2 and use this mixture within 1 h. 3. After aligning the MTMOS-coated slides on the slide deck of a MicroSys™ 5100-4SQ microarray spotter (Genomics Solutions, now Digilab, Inc.), print the mixture onto the periphery of each slide (typically 80 spots/slide, 450 nL per spot, 2 mm spot-to-spot distance). 4. Immediately after printing, bake the MTMOS slides for 5 min in an oven at 100°C to facilitate complete gelation of the sol-gel gasket (see Note 5). The MTMOS slides should be carefully inspected to ensure hemispherical sol-gel gasket spots. 5. Store the MTMOS slides containing the sol-gel gasket in a sterile Petri dish until needed.

Lee, Dordick, and Clark

The MetaChip, a complementary array of encapsulated metabolic enzymes that is designed to emulate the metabolic reactions in the human liver is prepared on the MTMOS-coated slide with a sol-gel gasket. For example, six different compounds can be printed in sections 1–6 of the MetaChip, each region containing a 5 × 9 mini-array (Fig. 2). Within each mini-array, nine different doses of a compound can be assayed for toxicity. For on-chip drug metabolism, human metabolic enzymes can be transversely printed into four regions (a–d), each containing no enzyme and different mixtures of enzymes encapsulated in alginate gel, thereby creating 24 distinct regions on the MetaChip (Table  1). Other human and rat metabolic enzymes can be spotted similarly on the MTMOS-coated slide for drug metabolism. 1. Prepare compound stock solutions by dissolving compounds in DMSO. Typically, higher than 200 mM of compound stock solutions are required to maintain final DMSO content less than 0.5%. DMSO may inhibit many metabolic enzymes. 2. Prepare 200 mL of test compound solutions at 2.5-fold higher concentrations than the desired final concentration (8 dosages plus 1 control, typically 0–1,000 mM) by serial dilutions of test compound stock solutions in DMSO (Fig. 3) and then add an equal volume mixture of 0.1  M BaCl2 and 0.01% Test compounds

2

3

4

5

6

C

D

1

B

3.2. Preparation of a Miniaturized Enzyme Array (the Metabolizing Enzyme Toxicology Assay Chip or MetaChip) on the MTMOS-Coated Slide

A

228

Enzyme # A (45 replicates) C1 C2 C3 C4 C5 C6 C7 C8 C9

9 Concentrations of compound # 4 (5 replicates)

Fig. 2. A layout of the MetaChip (1,080 spots/slide) for in-situ drug metabolism. Specifically, region A contains no enzyme as a test compound only control, region B contains a mixture of human Cytochrome P450 isoforms, region C contains a mixture of human phase II drug-metabolizing enzymes and P450 enzymes (i.e., all enzyme mixture), and region D contains human liver microsome. Each different compound is printed in sections 1–6 of the chip, and the concentrations (C1–C9) vary with each mini-array

Microarray Coupled with Miniaturized Cell-Culture Array Technology

229

Table 1 Typical preparation of enzyme solutions for the MetaChip Region

Enzymes

Co-substrates

A

90 µL PBS containing 5% glycerola

–

B

45 µL human P450 mixture

45 µL P450 cofactor

C

45 µL all enzyme mixturec

45 µL all cofactore

30 µL alginate (2%)

D

45 µL human liver microsome

45 µL all cofactor

30 µL alginate (2%)

b

Matrix 30 µL alginate (2%) d

30 µL alginate (2%)

No enzyme (TNI-WT negative control) is used as a control to determine toxicity of parent compound A mixture of human P450 isoforms contains 0.59 mM CYP3A4, 0.20 mM CYP2D6, 0.08 mM CYP2C9, 0.03 mM CYP2C19, 0.05 mM CYP2E1, 0.04 mM CYP1A2, and 0.01 mM CYP2B6 c All enzyme mixture is composed of 50% of a mixture of human phase II enzymes (containing 1.15 mg/mL UGT1A1, 1.15  mg/mL UGT1A4, 1.15  mg/mL UGT2B4, 1.15  mg/mL UGT2B7, 0.1  mg/mL SULT1A3, 0.1  mg/mL SULT2A1, and 0.2 mg/mL GST) and 50% of the human P450 mixture d A P450 cofactor solution contains 50% of NADP (10 mM) and 50% of a regeneration system from Invitrogen Vivid® CYP450 screening kits e All cofactor solution is composed of 50% of a phase II cofactor solution (containing 50% of 10 mM UDP-GA in 50 mM Tris–HCl buffer, pH 7.5, 25% of 20 mM GSH, and 25% of 20 mM PAPS in 10 mM PBS buffer, pH 8) and 50% of the P450 cofactor solution a

b

X-fold serial dilutions in DMSO

C1 DMSO alone

C2

C3

C4

C5

C6

C7

C8 C9 Compound in DMSO

Mixing with the BaCl2-PLL solution

Fig.  3. Preparation of test compound solutions in the BaCl2-PLL solution after serial dilutions in DMSO

poly- l-lysine (PLL). As a control, prepare the BaCl2-PLL solution without compound. 3. Dispense 200 mL of diluted compound solutions in the round-bottom 96-well plate for printing. Due to potential compound carry-over, always add compound solutions from low to high concentration in the wells. 4. After spotting test compound solutions on the MTMOScoated slides with the sol-gel gasket (typically 1,080 spots/ slide, 30 nL per spot, 1 mm spot-to-spot distance) using the microarray spotter, dry the test compound spots in a sterile Petri dish. The hydrophobic MTMOS coating is used to ensure hemispherical spots on the slide.

230

Lee, Dordick, and Clark

5. Prepare metabolic enzyme solutions in alginate (Table  1). Always prepare fresh enzyme-alginate mixtures on ice and use them immediately. 6. Print 15  nL of 0.5% (w/v) alginate solutions containing metabolic enzymes atop each test compound spot within the microarray spotter chamber under 95% relative humidity. Humidity to maintain enzyme spots hydrated vary depending on the temperature of the slide deck. 7. Immediately after enzyme printing, place the slide in a Petri dish (100 mm diameter) and store in a −80°C freezer until use. Do not dry the enzyme spots, causing enzyme deactivation. 3.3. Preparation of Human Cell Suspension for Spotting

1. Grow Hep3B human hepatoma cells or other mammalian cells in RPMI 1640 or relevant growth media supplemented with 10% FBS and 1% PS in T-75 cell-culture flasks in a humidified 5% CO2 incubator (ThermoForma Electron Co., Marietta, OH) at 37°C. 2. Prepare the cell suspensions by trypsinizing a confluent layer of the cells with 1 mL of 0.05% trypsin-0.53 mM EDTA from the culture flask, and resuspending the cells in 7 mL of 10% FBS-supplemented RPMI. 3. After centrifugation at 100×g for 4 min, remove the super­ natant and resuspend the cells with 10% FBS-supplemented RPMI to a final concentration of 9 × 106 cells/mL. Other mammalian cell suspensions can be prepared similarly.

3.4. Preparation of a Miniaturized 3D Cell-Culture Array (the Data Analysis Toxicology Assay Chip, or DataChip) on the PS-MA-Coated Slide

The DataChip, arrays of 3D cell spots containing Hep3B cells is prepared on the PS-MA-coated slide (Fig. 4). Each 3D cell culture spot on the DataChip contains 6 × 106 cells/mL in 30 nL of 1% (w/v) alginate (180 cells/spot). The 3D cell cultures are arrayed onto the PS-MA-coated slide in 24 regions of 5 × 9 miniarrays. Thus, a single DataChip combined with a single MetaChip can generate 24 sigmoidal dose response curves for 6 compounds and their metabolites generated from human P450 mixture, all enzyme mixture, and human liver microsomes (9 doses and five replicates for each compound). IC50 values (indicating the concentration of test compound required to inhibit 50% of cell growth) can be determined for the parent test compound and its enzyme-generated metabolites against Hep3B human hepatoma cells. Other human and rat cell types can be spotted similarly on the PS-MA-coated slide for toxicology assays. 1. Prepare a sterile BaCl2-PLL mixture by mixing 100 µL of 0.1 M BaCl2 with 200 µL of 0.01% PLL. 2. Spot the BaCl2-PLL mixture onto the PS-MA-coated slide (typically 1,080 spots/slide, 30  nL per spot, 1  mm spot-tospot distance) using the microarray spotter and then dry the

Microarray Coupled with Miniaturized Cell-Culture Array Technology

231

Hep3B cells

5 x 9 Mini-array of Hep3B cells Fig. 4. A layout of the DataChip (1,080 spots/slide) for high-throughput toxicity screening. Each 3D cell culture spot on the DataChip contains 6 × 106 Hep3B cells/mL in 30 nL of 1% alginate (180 cells/spot). The 3D cell cultures are arrayed onto the PS-MA-coated slide as 24 of 5 × 9 mini-arrays. Thus, 24 dose response curves comprising 9 doses (5 replicates each) are obtained from a single DataChip

slide for 10 min in a sterile Petri dish with a lid slightly open to yield flat BaCl2-PLL layers. The hydrophobic PS-MA coating is used to attach PLL with amine groups covalently and to ensure hemispherical spots on the slide. In addition, positively charged PLL has an ionic interaction with negatively charged alginate. 3. Prepare a suspension of the cells in low-viscosity alginate by mixing the Hep3B cell suspension in 10% FBS-supplemented RPMI (9 × 106 cells/mL) with 3% (w/v) alginate solution in distilled water so that the final concentration of cells and alginate were 6 × 106 cells/mL and 1%, respectively. 4. Print the alginate solution containing the cells (1,080 spots/ slide, 30 nL per spot) atop each BaCl2-PLL spot resulting in nearly instantaneous gelation of the alginate matrix (see Note 6). Maintain the microarray spotter chamber at 95% relative humidity to retard evaporation of water during spotting. 5. Immediately after spotting, place the DataChip in 15 mL of 10% FBS-supplemented RPMI in a Petri dish (100  mm diameter) and incubated in the CO2 incubator at 37°C for 18 h prior to use. 3.5. Stamping the DataChip onto the MetaChip

Metabolism-based cytotoxicity assays are performed by stamping the DataChip onto the MetaChip. 1. Take out the MetaChip from the −80°C freezer, place the slide immediately on the slide deck, and then print sterile distilled water onto the periphery of the MetaChip (typically

232

Lee, Dordick, and Clark

80 spots/slide, 150 nL per spot, 2 mm spot-to-spot distance) to retard evaporation during stamping. Maintain the microarray spotter chamber at 95% relative humidity to retard evaporation of water during spotting. 2. Print complete RPMI supplemented with 10% FBS and 1% PS atop the enzyme spots with compounds (1,080 spots/ slide, 30 nL per spot). 3. Simultaneously, take out the DataChip from the growth medium, wipe off the excess RPMI on the back side of the slide with ethanol-soaked Kimwipes, and place the cell slide in a sterile ArrayIt hybridization cassette. 4. Cover a lid of the cassette, slant the cassette with the cell slide at an angle of 45 degree for 5  min to allow excess growth medium on the surface to drain, and wipe off the growth medium with Kimwipes. (see Note 7). 5. Immediately after growth medium spotting, place the MetaChip onto a sterile stamping apparatus (Fig. 5). 6. Immediately after taking out the DataChip from the cassette, stamp the DataChip on top of the MetaChip by aligning the edges of the glass slides in the apparatus, and then tape both edges of the slides. 7. Place the stamped slides, with the cell slide on top, in the hybridization cassette with 100 µL of sterile distilled water, cover the lid tightly, and then incubate the cassette with the slides for 6 h at 37°C. 8. After desired incubation period, lift off the cell slide, clean the back side of the slide with 70% ethanol-soaked paper towels before immersing in the growth medium to prevent contamination, rinse once in a Petri dish with 15 mL of Dulbecco’s PBS to remove any excess compound solution, immerse in 15 mL of the growth medium for 1 h to allow for residual

Stamping apparatus

Sol-gel gasket MetaChip

DataChip

Fig. 5. The stamping apparatus for aligning the MetaChip and the DataChip

Microarray Coupled with Miniaturized Cell-Culture Array Technology

233

compounds to diffuse out from the alginate-gel drops, and transfer to a Petri dish containing 15  mL of fresh medium supplemented with 10% FBS and 1% PS. 9. Culture the cells for 2  d in a CO2 incubator at 37°C for cytotoxicity assays. 3.6. Cell Staining, Scanning, and Data Analysis

Cytotoxicity of test compounds and their metabolites can be determined by staining the DataChip with a live/dead viability/ cytotoxicity kit. Staining dyes such as calcein AM (excitation 495 nm and emission 515 nm) and ethidium homodimer-1 (excitation 495 nm and emission 635 nm) are used to produce a green fluorescent response from living cells and a red fluorescent signal from dead cells. 1. At the end of the 2-d culture period post stamping, wash the DataChip three times by immersing the slide in a 150 mmdiameter Petri dish containing 100 mL of 10 mM BupH PBS (pH 7) with 10 mM CaCl2 for 5 min each. CaCl2 is supplemented to prevent degradation of alginate spots by excess phosphate ions. 2. Prepare a staining dye solution by adding 1.0 µL of calcein AM (4  mM stock) and 4.0 µL of ethidium homodimer-1 (2 mM stock) in 8 mL of Dulbecco’s PBS supplemented with 10 mM CaCl2. 3. Dispense 1.5 mL of dye solution on a glass slide with 1 mmthick perimeter gasket acting as a barrier to prevent loss of the dye solution and evenly spread the dye solution on the gasket slide. 4. Wipe off excess PBS from the cell slide with Kimwipes and place it on top of the gasket slide. 5. Incubate the DataChip in dark for 50  min at room temperature. 6. Wash the cell slide in a Petri dish containing 15 mL of 10 mM BupH PBS (pH 7) with 10 mM CaCl2 on an orbital shaker at 0.5 g for 30 min to remove excess dye in the alginate-gel drops. 7. Blow drying the cell slide gently with N2 gas so as not to detach the cell spots. 8. Detect the location of each cell spot where compounds are added by imaging the entire slide using blue laser (488 nm) and standard blue filter for green dye and blue laser and 645AF75/594 filter for red dye in GenePix® Professional 4200A scanner (Molecular Devices, now MDS Analytical Technologies, Sunnyvale, CA). The PMT gain and power may be adjusted depending upon fluorescent intensity. 9. Save data files as single tiff images for analysis. The green fluorescence intensity is linearly proportional to the total

234

Lee, Dordick, and Clark

number of live cells, and is quantified from the array scanner using GenePix Pro 6.0 (Molecular Devices). 10. Extract fluorescent intensity from each cell spot using GenePix Pro and plot the percentage of live cells against the concentration of the compound tested using Prism 4 (GraphPad Software, Inc., La Jolla, CA). Since the background green fluorescence of completely dead cells (following treatment with 70% methanol for 1 h) was negligible, the percentage of live cells is calculated using the following equation:

a

125

100

75

50

No e nzyme P45 0 mixture All mixture Live r micro so me

25

1

2

3

100 75 50 25

4

5

0

6

Log [Compound A (nM)]

2

3

Live Cell (%)

100 75 50

No enzyme P450 mixture All mixture Liver microsome 1

5

6

2

3

4

5

No enzyme P450 mixture All mixture Liver microsome 1

6

3

4

5

6

4

5

6

125

100 75 50

0

2

Log [Compound C (nM)]

150

100 75 50

No enzyme P450 mixture All mixture Liver microsome

25

25

Log [Compound D (nM)]

50

0

4

No enzyme P450 mixture All mixture Liver microsome

125

25

75

25

150

125

Live Cell (%)

1

100

Log [Compound B (nM)]

150

0

No enzyme P450 mixture All mixture Liver microsome

Live Cell (%)

0

Live Cell (%)

150

125

Live Cell (%)

150

125

Live Cell (%)

b 150

1

2

3

4

5

Log [Compound E (nM)]

6

0

1

2

3

Log [Compound F (nM)]

Fig. 6. (a) Scanning images of the alginate-gel spots containing Hep3B cells (30 nL, 1080 spots/slide) after stamping with 24 regions of 5 × 9 enzyme arrays (from left to right: compound A, B, C, D, E, and F obtained from a collaborator, from top to bottom: control, human P450 mixture, all enzyme mixture, and human liver microsome) and staining. (b) Dose-response curves obtained from the scanning image

Microarray Coupled with Miniaturized Cell-Culture Array Technology

235

F  % Live cells =  Reaction  × 100  FMax  where FReaction is the green fluorescence intensity of the reaction spot and FMax is the green fluorescence intensity of untreated viable cells. 11. To produce a conventional sigmoidal dose-response curve, with response values normalized to span the range from 0 to 100% plotted against the logarithm of test concentration, normalize the green fluorescence intensities of all cell spots with the fluorescence intensity of 100% live cell spot (typically, cell spots contacted with no compound) and convert the test compound concentration to their respective logarithms using Prism 4. The sigmoidal dose-response curves (variable slope) and IC50 values (i.e., concentration of the compound where 50% of cell growth inhibited) are obtained using the following equation:   Top - Bottom Y = Bottom +  ∧  1 + 10 ((Log IC50 - X)* H ) where IC50 is the midpoint of the curve, H is the hill slope, X is the logarithm of test concentration, and Y is the response (% live cells), starting at Bottom and going to Top with a sigmoid shape. An example of the results produced is shown in Fig. 6.

4. Notes 1. Do not use Kimwipes because it generates lots of small paper particles. After wiping with dry paper towels, the microscope slides should not be wet with ethanol. It will leave ethanol stains on the slide surface. 2. After 30 min of shaking at 50°C, 5 g, the solution becomes clear. However, leave the solution for additional 10 min for complete dissolution of PS-MA pellets. The PS-MA solution should be prepared freshly as reactivity of maleic anhydride groups decreases over time. The PTFE-lined scintillation vial has to be used for organic solvents. Never leave the vial with toluene in the incubating shaker unattended due to a concern about an explosion at high temperature. 3. The hydrolysis of MTMOS by HCl is an exothermic reaction. So be careful not to burn skin. The mixture becomes a transparent single phase after vortex mixing (initially two phases).

236

Lee, Dordick, and Clark

Always prepare fresh MTMOS-HCl sol solution and use it immediately. 4. Upon addition of the phosphate buffer solution, gelation of hydrolyzed MTMOS begins. Therefore, use the mixture immediately. Do not use the mixture when the solution becomes turbid, causing uneven coating. 5. Excessive baking (typically longer than 10 min) causes spot detachment. If the sol-gel gasket spots are flattened after baking, it is because of either less hydrophobic MTMOS slide or old MTMOS contaminated with HCl. In the latter case, use fresh MTMOS and HCl to prepare the MTMOS–HCl sol solution. 6. Add freshly suspended alginate solution containing the cells in the round-bottom 96-well plate to make sure well suspension of the cells while spotting. Remind that the cells are gradually precipitated while spotting. For uniform cell spotting, keep re-suspending the cells. In addition, fast spotting and high humidity on the slide surface is critical to maintain high cell viability as spot drying causes cell death. 7. The cell spots should remain hydrated while removing excess growth medium, otherwise the cell would be dead. Particularly pay attention to the cell spots drying on the edge of the slide. Draining time is a function of the spot volume. Optimization of draining time is required to minimize cell death due to drying.

Acknowledgments This work was supported by the National Institutes of Health (ES-012619) and National Science Foundation (0711708). References 1. Furge LL, Guengerich FP (2006) Cytochrome P450 enzymes in drug metabolism and chemical toxicology: an introduction. Biochem Mol Biol Educ 34:66–74 2. Guengerich FP (2006) Cytochrome P450s and other enzymes in drug metabolism and toxicity. AAPS J 8:E101–E111 3. Zamek-Gliszczynski MJ, Hoffmaster KA, Nezasa K, Tallman MN, Brouwer KLR (2006) Integration of hepatic drug transporters and phase II metabolizing enzymes: mechanisms of hepatic excretion of sulfate, glucuronide, and glutathione metabolites. Eur J Pharm Sci 27:447–486

4. Lee MY, Dordick JS (2006) High-throughput human metabolism and toxicity analysis. Curr Opin Biotech 17:619–627 5. Hariparsad N, Sane RS, Strom SC, Desai PB (2006) In vitro methods in human drug biotransformation research: implications for cancer chemotherapy. Toxicol in Vitro 20: 135–153 6. Brandon EFA, Raap CD, Meijerman I, Beijnen JH, Schellens JHM (2003) An update on in  vitro test methods in human hepatic drug biotransformation research: pros and cons. Toxicol Appl Pharmacol 189: 233–246

Microarray Coupled with Miniaturized Cell-Culture Array Technology 7. Gomez-Lechon MJ, Donato MT, Castell JV, Jover R (2004) Human hepatocytes in primary culture: the choice to investigate drug metabolism in man. Curr Drug Metab 5:443–462 8. Sivaraman A, Leach JK, Townsend S, Iida T, Hogan BJ, Stolz DB, Fry R, Samson LD, Tannenbaum SR, Griffith LG (2005) A microscale in vitro physiological model of the liver: predictive screens for drug metabolism and enzyme induction. Curr Drug Metab 6:569–591 9. Hewitt NJ, Lechon MJG, Houston JB, Hallifax D, Brown HS, Maurel P, Kenna JG, Gustavsson L, Lohmann C, Skonberg C, Guillouzo A, Tuschl G, Li AP, LeCluyse E (2007) Primary hepatocytes: current understanding of the regulation of metabolic enzymes and transporter proteins, and phar-

237

maceutical practice for the use of hepatocytes in metabolism, enzyme induction, transporter, clearance, and hepatotoxicity studies. Drug Metab Rev 39:159–234 10. Lee MY, Park CB, Clark DS, Dordick JS (2005) Metabolizing enzyme toxicology assay chip (MetaChip) for high-throughput microscale toxicity analyses. Proc Natl Acad Sci USA 102:983–987 11. Lee MY, Clark DS, Dordick JS (2006) Human P450 microarrays for in vitro metabolite toxicity analysis: toward complete automation of human toxicology screening. J Assoc Lab Automat 11:374–380 12. Lee MY, Ramasubramanian AK, Sukumaran SM, Hogg MG, Clark DS, Dordick JS (2008) Three-dimensional cellular array chip for microscale toxicology assay. Proc Natl Acad Sci USA 105:59–63

Chapter 15 Use of Tissue Microarray to Facilitate Oncology Research Panagiotis Gouveris, Paul M. Weinberger, and Amanda Psyrri Abstract HPV-positive oropharyngeal squamous cell carcinomas (OSCC) represent a distinct disease entity from traditional OSCC. We hypothesized that for HPV DNA-positive cases, p16 expression status differentiates the biologically relevant ones. We determined HPV16DNA viral load in a cohort of 79 oropharyngeal squamous cell cancers by real-time polymerase chain reaction (PCR). We used cervical cancer as a disease model for HPV-initiated epithelial cancer. In cervical cancer, p53 and Rb expression is reduced, while p16 expression is increased. We used TMA technology to facilitate interrogation of this cohort for p53, Rb, and p16 protein expression using a quantitative, in situ method of protein analysis (AQUA analysis). Our results indeed delineate three biologically and clinically distinct types of oropharyngeal squamous cell cancers based on HPV-DNA determination and p16 expression status: one class of HPV-negative/p16-nonexpressing (HPV-negative), one class of HPV-positive/p16-nonexpressing (HPV-inactive), and one class of HPV positive/p16-expressing (HPV-active) oropharyngeal tumors. We demonstrated that only the HPVactive tumors share a similar molecular phenotype to cervical cancers, and are the ones associated with favorable prognosis. Key words: Oropharyngeal squamous cell cancer, HPV16, p16, Real-time polymerase chain reaction (PCR), AQUA, Quantitative immunohistochemistry

1. Introduction The link between HPV16 and oropharyngeal cancer was first proposed in the 1980s (1); since then HPV sequences have been detected in approximately 50% of OSCC (2–4). It is important to point out that mere presence of HPV DNA in tumors does not indicate causal association. Indeed, most experts would argue that the HPV viral oncogenes must be actively transcribed for the virus to contribute to carcinogenesis. The E6 and E7 genes of the oncogenic HPVs encode oncoproteins that bind and

Sridar V. Chittur (ed.), Microarray Methods for Drug Discovery, Methods in Molecular Biology, vol. 632, DOI 10.1007/978-1-60761-663-4_15, © Humana Press, a part of Springer Science+Business Media, LLC 2010

239

240

Gouveris, Weinberger, and Psyrri

degrade p53 and retinoblastoma (Rb) tumor suppressors (5, 6). This results in overexpression of p16 protein due to a negative feedback regulation with Rb. Based on this understanding, we hypothesized that p16 overexpression can define the subgroup of HPV-positive oropharyngeal tumors with biologic similarity to HPV-caused cervical cancer. We used Real-Time PCR to determine HPV16 viral load and interrogated these tumors for expression of p53, Rb, and p16 proteins using a quantitative in-situ method of protein analysis. Our results demonstrated three biologically and clinically distinct types of oropharyngeal squamous cell cancers based on HPV DNA determination and p16 expression status: HPV-negative (HPV negative/p16 nonexpressing), HPV-inactive (HPV positive/p16 nonexpressing), and HPV-active (HPV positive/p16 expressing). Only the HPV positive/p16 expressing tumors were associated with favorable prognosis (7). Tissue microarray (TMA) technology has emerged as a very useful tool in facilitating the process of biomarker validation. TMA technology is a method used to analyze hundreds (or even thousands) of tissue samples on a single slide. The most widely used end-application for TMA is immunohistochemistry, although the technique can be coupled with other molecular biology methods including in-situ hybridization (8, 9) or in-situ PCR (10). Essentially, a TMA is made up of individual cylindrical “cores” taken from representative areas of each sample in a cohort. These cores are then arrayed in an orderly grid into a recipient paraffin block. Subsequent tissue sections are then made of this recipient block and processed identically to traditional wholesection slides. The advantages of TMA over individual whole-slide sections are several. First, all specimens are treated to identical conditions throughout the IHC process. This can become important when one considers the variability that might be introduced if a cohort of 150 tumors is processed 20 at a time – each batch undergoing slightly different incubation times with the final chromagen reaction before quenching, for instance. In a TMA, all spots are being incubated simultaneously, thus greatly reducing “noise” or variability that might obscure true findings. The second and not-insignificant advantage TMA offers is cost and time savings. Consider the experiment described in this work, where 79 tumors (in duplicate) are interrogated for three biomarkers – this would require cutting, deparaffinizing, incubating with primary antibody, staining, and scoring of 474 individual slides (versus 3 slides when using TMA technology). This difference becomes even larger when larger cohorts and more biomarkers are included in a planned experiment.

Use of Tissue Microarray to Facilitate Oncology Research

241

2. Materials 2.1. DNA Extraction from Donor Paraffin Blocks

1. Formalin-fixed Paraffin embedded (FFPE) surgical pathology archival blocks. 2. 18-gauge needles for extraction of tissue cores from the tumor areas (VWR Scientific, West Chester, PA). 3. Chelex-100 (Fisher Bioscience) diluted to 5% in dH2O. 4. Tween-20 (Fisher Biosciences) diluted to 0.5% solution in dH2O. 5. Proteinase K (Fisher Biosciences) diluted to 20 mg/ml stock concentration in dH2O. 6. Chloroform, nucleic acid grade (Fisher Biosciences). 7. Standard PCR thermal cycler with heating block to accommodate 0.6 ml eppendorf tubes.

2.2. Quantitative PCR

1. 96-well thermal cycler equipped with fluorescence detection: BioRad iCycler (Hercules, CA). 2. SYBR Green Supermix (BioRad). 3. HPV16E6 Primers and b-globin primers. 4. Human genomic DNA (Promega). 5. HPV16 plasmid DNA containing full-length HPV16.

2.3. Fluorescent IHC and AQUA Analysis

1. Primary mouse monoclonal antibodies to p53, Rb, and p16. 2. Secondary antibody: Envision Goat anti-mouse (DAKO, Carpinteria CA). 3. Rabbit antipancytokeratin antibody z0622 (DAKO Corp, Carpinteria CA). 4. Goat anti-rabbit antibody conjugated to Alexa546 fluourophore (A11035; Molecular Probes, Eugene, Oregon). 5. Cy-5-tyramide (PerkinElmer Corp, Wellesley, MA). 6. Polyvinyl alcohol-containing aqueous mounting media with antifade reagent (n-propyl gallate; Acros Organics, Geel, Belgium). 7. AQUA system from HistoRx (HistoRx, New Haven CT).

3. Methods The use of automated quantitative analysis (AQUA™) of fluorescent immunohistochemistry (HistoRx, New Haven CT) (11) dramatically improved the implementation of these experiments.

242

Gouveris, Weinberger, and Psyrri

By allowing quantitative determination of protein expression (rather than “by-eye” scoring) we were able to use more robust and sensitive statistical tests. AQUA™ is an automated scoring system for assessing protein expression while preserving in-situ relationships. Fluorescent immunohistochemical staining of either a tissue slide or tissue array is performed for the target of interest plus nuclear and tumor-specific targets. Subsequently a series of florescent images are collected by the PM-2000, a custom robotic fluorescent microscope platform. Each image set is then analyzed by a set of algorithms that provides a reproducible, automated, quantitative analysis of expression of a given marker within a user defined subcompartment (or subcellular locale) in a histospot. This method allows measurements of protein expression within subcellular compartments that results in a number directly proportional to the number of molecules expressed per unit area (12). AQUA analysis allows generation of quantitative protein expression data in a similar manner to western blot or ELISA. Unlike these methods, however, AQUA preserves tissue morphology allowing comparison of protein expression in specific subcellular compartments. Additionally, AQUA analysis can be performed on standard formaldehyde fixed paraffin-embedded specimens, unlike western blot or ELISA, which require fresh or snap-frozen tissue or serum. 3.1. Tissue Microarray Construction and DNA Extraction from Specimens

1. Prepare Hematoxylin and Eosin (H&E) slides of specimen blocks and identify representative areas of tumor (see Notes 1 and 2). 2. Prepare a TMA in standard fashion from the corresponding paraffin-embedded specimens. Cut 5-µm sections and mount on glass slides using an adhesive tape transfer system. 3. Extract four tissue cores from the original FFPE blocks, in the representative areas of SCC using sterile, blunt 18-gauge needles (VWR Scientific). 4. Place core(s) into 0.6  ml eppendorf tubes (see Notes 1 and 2). 5. Add 100 µl of 0.5% Tween20 in dH2O to each tube. Vortex gently. 6. Heat to 90°C for 10  min in thermal cycler, then cool to 55°C (13). 7. Add 5 µl of 20  mg/ml Proteinase K to each tube (final concentration 1 mg/ml). 8. Incubate at 50°C for 3 h, vortexing once each hour. 9. Quick-spin to remove any condensate in cap.

Use of Tissue Microarray to Facilitate Oncology Research

243

10. Add 100 µl of 5% chelex-100 in TE buffer to each tube. Vortex briefly (see Note 3). 11. Heat at 99°C for 10 min in thermal cycler to inactivate the proteinase K. 12. Vortex gently. Centrifuge at 10,500×g for 15 min. 13. Place sample on ice to allow wax to harden. The wax will have formed a cap on top of the solution. Remove wax “cap” with sterile 10 µl pipette tip. 14. Heat to 45°C. Add 100 µl chloroform. 15. Vortex gently. Centrifuge at 10,500×g for 15 min. 16. You will have three phases in the eppendorf tube (Fig. 1) now. Remove top aqueous phase with DNA (~180 µl) to new tubes for storage at −20°C until needed. Discard the rest. 3.2. Quantitative PCR to Determine HPV16 Viral Load

1. Prepare a standard curve for b-globin using serial dilutions of purified high-molecular weight human DNA (Promega Corp, Madison, WI) from 4,000  genomes/µl to 1  genome/µl. These will be used to quantify each qPCR run to allow determination of absolute viral load (see Notes 4–6). 2. Prepare a standard curve for HPV16 E6 using purified wholelength HPV16 plasmid at concentrations from 4,000 copies/ µl to 10 copies/µl. This will allow determination of HPV16 E6 concentration and expression of the final quantity as the number of viral copies present per human genome. 3. Set up the parameters for a qPCR using a standard BioRad iCycler 96-well thermal cycler equipped with fluorescence detection (or equivalent from other manufacturer). A premade Master Mix is recommended for qPCR, to ensure optimal results and minimize batch-to-batch variability. Each reaction vial will consist of 20 µl master mix to which 5 µl target DNA is added for a final reaction volume of 25 µl (see Notes 4–6).

Fig. 1. DNA is isolated from top aqueous layer

244

Gouveris, Weinberger, and Psyrri

4. This assay can be run on the same plate (both ß-globin and HPV16 E6). Divide a 96-well plate in half, with the top being b-globin and the bottom for HPV16E6. Include each patient sample in triplicate, for both b-globin and HPV16E6. Also include a three-log serial dilution to generate standard curves for ß-globin and HPV16 E6, and three negative controls for each. 5. The PCR reaction parameters are: 3 min at 95°C to activate the iTaq DNA polymerase, followed by 40 cycles of 95°C for 15  s followed by 60°C for 30  s. Set fluorescence detection during the combined annealing/extension phase (60°C) of each cycle. Manually set the cycle threshold (Ct) at the point where each sample’s fluorescence crossed 30 units. Following 40 cycles, set up a melt curve analysis by ramping the temperature from 60 to 95°C at a rate of 0.1° per second, while monitoring resulting fluorescence. Following melt curve analysis, insert a final extension step at 60°C for 1  min to reform double-stranded DNA. 6. Using the known standard curves, determine b-globin concentrations and HPV viral load quantities. Viral load is then determined as HPV16E6 concentration/b-globin concentration. For samples with viral load greater than 0.1 (>1 HPV genome copy/10 cells), these should be designated as positive for HPV16 (see Notes 4–6). 3.3. Fluorescent Immunohistochemistry

1. Start with 5-µm sections of the TMA mounted to glass slides. You will need 1 section for each primary antibody studied, plus a negative control slide. 2. Melt excess paraffin off the slides in a 60°C oven for ~45 min. 3. Deparaffinize the slides with 3 washes of xylene followed by graded ethanol washes (100%, 90%, 70%, 50%). Incubate at each step for 5 min. 4. Rehydrate for 5 min in dH2O. From this point on it is essential to keep the slides wet at all times as any drying will introduce artifact and background staining. 5. Perform antigen retrieval (for most targets). We prefer pressure cooking in 0.1 M citrate buffer (pH, 6.0) (see Note 11). 6. Block endogenous peroxidase activity by incubating in 0.3% hydrogen peroxide in methanol for 30 min (10 ml 30% H2O2 in 390 ml methanol). 7. Wash slides in two changes of DDH2O (30 s each). 8. Dry slides carefully around the array edge with a kimwipe, making a meniscus (see Note 12). 9. Block nonspecific antibody binding with 0.3% bovine serum albumin (BSA) in TBS for 30 min at room temperature (see Notes 7–9).

Use of Tissue Microarray to Facilitate Oncology Research

245

10. Decant off BSA/TBS and dry the edge with a kimwipe. 11. Make a cocktail (~300 µl per slide) of each primary antibody in 0.3% BSA/TBS. This cocktail will consist of (1) target antibody at chosen concentration, (2) anti-cytokeratin antibody of a different species than the target (in this example, because the primary antibodies are all mouse, the anti-cytokeratin antibody should be rabbit (anti-pancytokeratin antibody z0622; DAKO Corp)). For this example, you will make 4 cocktails, with primary antibody to p16, p53, Rb, or no antibody for negative control (one antibody per slide). 12. Vortex each briefly to mix, add to slides, and incubate in cold room overnight (see Notes 13, 19, 20). 13. Decant off primary antibody, and wash slides 2 times in 1× TBS/Tween for 10 min each. Do a third wash, but use TBS without detergent (see Note 7). 14. Prepare the secondary antibody cocktail: Envision goat anti-species of target antibody (use as diluent) + Alexa-546 conjugated Goat Anti-Species of cytokeratin antibody @ 1:100 dilution + 4¢, 6-diamidino-2-phenylindole (DAPI) @ 1:100 dilution. In this case you will use Alexa-546 Goat anti-rabbit A11035 (Molecular Probes). 15. Dry slides around the array edge as before and add the secondary antibody cocktail at 300 µl per slide. Incubate 1 h at room temperature in humidity tray (see Note 14). 16. Remove secondary cocktail and wash slides 2× in TBS/Tween (10 min each) Then TBS. Dry slides as before. 17. Dilute the Cy5-Tyramide 1:50 in amplification buffer and add 150 µl to each slide. Incubate slides with the Cy5Tyramide for 10 min at 37°C, keeping slides in the dark (see Notes 15, 18). 18. Wash slides 2× in TBS/Tween for 5 min each, then TBS 1×. Dry edges. 19. Mount the slides in gelvatol by applying a line of ~300 ml to the edge of the slide and then laying the coverslip on pushing the air out as you lay it down (see Notes 10, 16, 17, 21). Allow the slides to dry at room temp (preferably overnight), IN THE DARK! 3.4. Automated Quantitative Protein Expression Analysis (AQUA™)

1. Each slide is subjected to AQUA image capture individually. First, the array layout is recognized by obtaining a low power image of the entire microarray using the DAPI filtercube. The images are stitched together using multiple (~1,500) low-resolution images of the microarray (64 × 64 pixel) at approximately 7-micron resolution. This portion is performed automatically by the AQUA machine.

246

Gouveris, Weinberger, and Psyrri

2. After this, the appropriate wavelength assignments are made. For this experiment, you would assign CY3 (equivalent to the Alexa546 flurophore) to the cytokeratin/tumor mask compartment. Since cytokeratin is expressed by most head and neck cancers (being epithelial in origin) this will allow automatic recognition of tumor versus stroma in each histospot. Your target molecule will be imaged using the CY5 filtercube set. 3. Set appropriate capture times and allow the machine to acquire images of the array. Each histospot is imaged using DAPI, CY3, and CY5 filtercubes, and an in-focus as well as slightly out-offocus image for each is obtained. The latter image is used to perform RESA subtraction to reduce background. 4. Following image capture (which can take several hours for a large array), the data is then subjected to image analysis. Load the dataset into the image analysis software (included with the AQUA platform). 5. Set a threshold for the CY3 image, to create a tumor mask that excludes the stromal elements. 6. Similarly, set a threshold for the DAPI image to create a nuclear mask. 7. Setup a tumor-nuclei mask by combining these two masks. Similarly, you can create a tumor-cytoplasm mask by subtracting the nuclear from the tumor mask. 8. Allow the image analysis program to iteratively perform this analysis for each histospot (again can take several hours). 9. The resulting output will have each histospot assigned an AQUA score from 0 to 255 for each compartment above, for the given protein analyzed.

4. Notes 1. The use of a standard thermal cycler to step through the temperatures needed greatly facilitates this protocol, although another method such as multiple heating blocks at the appropriate temperatures could be substituted. This protocol was adapted from Coombs, et al. Nuc Acid Res 27:16, 1999 (13). The beauty of this is the lack of transfer between multiple vials, so with the tiny amounts of DNA you are working with there is less chance for it to get “lost”. 2. We use black fine-tip permanent marker to do this, under either a 4× objective or a dissecting microscope. Circle the best areas of tumor on the H&E slide. Later, you can compare the grossly apparent shape of the paraffin block to the

Use of Tissue Microarray to Facilitate Oncology Research

247

H&E slide appearance and easily match up the appropriate area of specimen to biopsy. 3. Label only top of tube, leaving the side clear. This is important to allow easy visualization of the 3 layers in subsequent steps. The 5% Chelex-100 quickly falls out of suspension. We found that the best method was to prepare a stock solution in a 100 ml glass jar with screw-top lid (Corning Corporation) and keep a magnetic stir-bar in the jar. When ready to add the Chelex, place the jar on a stir-plate and leave it on. This will keep the Chelex in suspension while you pipette each aliquot into the eppendorf tubes. 4. When diluting the DNA solution, we first created a stock solution of 1 × 1010 genomes/ml, and then performed serial dilutions from this. Prepare a large number of aliquots of each working dilution and keep at −20°C until ready for use. The working dilutions can then be stored at 4°C for up to 30 days; after this discard and thaw a fresh set. 5. Final reaction conditions for our reactions were 1× iQ SYBR Green Supermix (BioRad 50  mM KCl, 20  mM Tris-HCl, 0.2 mM each deoxynucleotide triphosphate, 25 U/ml iTaq hot start DNA polymerase, 3 mM MgCl2, SYBR Green I dye, 10 nM floresein, and stabilizers). 6. Samples with ß-globin values less than 1 human genome/µl indicated lack of amplifiable DNA and were discarded from qPCR analysis. 7. 10× TBS-Tween (10 L, make in 3 batches and combine in a carboy) Batch 1 242.28 g Tris base in 1.8 L water Bring pH to 8.0 with concentrated HCl (will take a lot) Bring final volume up to 2 L (check pH again) Batch 2 438.15 g NaCl in 4 L water Batch 3 438.15 g NaCl in 4 L water 50 ml Tween 20 Combine all batches and verify pH of 8.0 8. 10× TBS (1 L): 87.6 g NaCl, 24.23 g Tris, 800 ml water. pH to 8.0 with concentrated HCl and bring up to 1 L 9. 0.3% BSA in TBS 50 ml 10× TBS, 1.5 g BSA, 450 ml water Sterile filter through a 0.2 micron unit. Store in refrigerator.

248

Gouveris, Weinberger, and Psyrri

10. Gelvatol for mounting coverslips (makes enough for hundreds of slides) 24  g polyvinyl alcohol (m.w. 30,000–70,000) and 60  ml glycerol; vortex well Add 60 ml water, keep at 37°C for 2 h Add 120 ml of 0.2 M Tris (pH 8.5); heat @ 50°C until dissolved (circulating water bath) Centrifuge at 4,500×g for 15  min and collect the supernatant Add n-propyl-galate to 0.6% (weight/volume) Heat to 50°C until dissolved (circulating water bath) Aliquot and store at −20°C When thawed keep at 4°C and protect from light Will set in a few hours or overnight at 4°C 11. We purchased a simple pressure cooker at the local supermarket. Simply fill the pressure cooker one-fourth of the way up with tap water. Place a corningware beaker that will accommodate a slide holder in the pressure cooker. Fill the beaker with the citrate buffer. Close the lid and place on a hotplate. Once the pressure cooker starts emitting steam and whistles, you can quickly quench the pressure cooker by placing in a sink and running cold water over it for several minutes (otherwise you will wait 20–30 min until the pressure runs out). NEVER OPEN THE PRESSURE COOKER when the pressure button is up or you will get severely burned. Recipe for citrate buffer: 3.84 g sodium citrate in 2.0 L distilled water. Bring the pH to 6.0 with 1  M citric acid (21 g Na citrate/100 ml water). Store at room temperature in large carboy. 12. Dry carefully around the array to avoid having cocktail wander to other parts of the slide. We use a small kimwipe to do this. 13. Make sure incubations are done on a flat surface otherwise the cocktails will not cover the array equally. The environment also needs to be moist. One solution that worked well for us was to create an incubation box using an opaque plastic box with a tight fitting lid. Line the bottom with paper towels and moisten them with water. Cut two 10 ml plastic pipettes to fit inside, and place these in the box. Now lay the slides on top of the pipettes (before adding the antibody). We used the following antibodies (all mouse monoclonal): Anti p16 (clone E6H4), Dako corporation (Carpenteria CA) at 1:25 dilution. Anti-p53 (clone DO-7), Dako Corporation (Carpenteria CA) at 1:100 dilution. Anti-Rb (clone 1F8), NeoMarkers – now LabVision (Freemont CA) at 1:50 dilution.

Use of Tissue Microarray to Facilitate Oncology Research

249

14. Prepare the secondary antibody cocktail as a single batch for all slides (if they all have the same species of primary antibody). We found it helpful to vortex the solution and then filter the cocktail using a syringe filter to remove any antibody complexes before incubation. 15. Keep the cy5-tyramide in 10 µl aliquots at −80°C until ready for use. Each aliquot of Cy5-Tyramide contains 10 µl, so add 500 µl of amplification diluent – this is enough to cover 2–3 slides. Cy5 is very photosensitive – make sure you keep this out of the light as much as possible, and perform the incubation in the dark. 16. Gelvatol (see recipe below) is made in large batches and then aliquoted and kept at −20°C until ready for use. To use, the gelvatol must be prepared ahead of time (usually 2 h before you need it!). Place gelvatol in a 100c block until melted, then vortex, and let sit for 1 h. 17. NEVER allow the slide to remain uncovered without fluid for >30 s or so. 18. Keep sensitive fluorescent antibody cocktails in the dark and cold as much as possible while you are working with them. 19. When micropipetting fluorescent antibodies, take your sample from the top of the solution in the vial, not the bottom. 20. Do not vortex the stock vial. Antibody solutions aggregate over time and aggregates fall to the bottom of the vial. Pipetting up antibody aggregates will significantly increase your background. You may want to quick-spin antibody before pipetting. 21. Do not write on the slides with marker – it gives a high background. Use pencil. References 1. de Villiers EM, Weidauer H, Otto H, zur Hausen H (1985) Papillomavirus DNA in human tongue carcinomas. Int J Cancer 36:575–578 2. Franceschi S, Munoz N, Bosch XF, Snijders PJ, Walboomers JM (1996) Human papillomavirus and cancers of the upper aerodigestive tract: a review of epidemiological and experimental evidence. Cancer Epidemiol Biomarkers Prev 5:567–575 3. Snijders PJ, Scholes AG, Hart CA, Jones AS, Vaughan ED, Woolgar JA et  al (1996) Prevalence of mucosotropic human papillomaviruses in squamous-cell carcinoma of the head and neck. Int J Cancer 66: 464–469

4. McKaig RG, Baric RS, Olshan AF (1998) Human papillomavirus and head and neck cancer: epidemiology and molecular biology. Head Neck 20:250–265 5. Psyrri A, DeFilippis RA, Edwards AP, Yates KE, Manuelidis L, DiMaio D (2004) Role of the retinoblastoma pathway in senescence triggered by repression of the human papillomavirus E7 protein in cervical carcinoma cells. Cancer Res 64:3079–3086 6. Alani RM, Munger K (1998) Human papillomaviruses and associated malignancies. J Clin Oncol 16:330–337 7. Weinberger PM, Yu Z, Haffty BG, Kowalski D, Harigopal M, Brandsma J et al (2006) Molecular classification identifies a subset of human

250

Gouveris, Weinberger, and Psyrri

papillomavirus – associated oropharyngeal cancers with favorable prognosis. J Clin Oncol 24:736–747 8. Fischer C, Zlobec I, Stockli E, Probst S, Storck C, Tornillo L et al (2008) Is immunohistochemical epidermal growth factor receptor expression overestimated as a prognostic factor in head-neck squamous cell carcinoma? A retrospective analysis based on a tissue microarray of 365 carcinomas. Hum Pathol 39:1527–1534 9. Graham AD, Faratian D, Rae F, Thomas JS (2008) Tissue microarray technology in the routine assessment of HER-2 status in invasive breast cancer: a prospective study of the use of immunohistochemistry and fluorescence in situ hybridization. Histopathology 52:847–855

10. Goldstine J, Seligson DB, Beizai P, Miyata H, Vinters HV (2002) Tissue microarrays in the study of non-neoplastic disease of the nervous system. J Neuropathol Exp Neurol 61: 653–662 11. Camp RL, Chung GG, Rimm DL (2002) Automated subcellular localization and quantification of protein expression in tissue microarrays. Nat Med 8:1323–1327 12. McCabe A, Dolled-Filhart M, Camp RL, Rimm DL (2005) Automated quantitative analysis (AQUA) of in situ protein expression, antibody concentration, and prognosis. J Natl Cancer Inst 97:1808–1815 13. Coombs NJ, Gough AC, Primrose JN (1999) Optimisation of DNA and RNA extraction from archival formalin-fixed tissue. Nucleic Acids Res 27:15

Chapter 16 Small Molecule Selectivity and Specificity Profiling Using Functional Protein Microarrays Peter R. Kraus, Lihao Meng, and Lisa Freeman-Cook Abstract Small molecules interact with proteins to perturb their functions, a property that has been exploited both for research applications and to produce therapeutic agents for disease treatment. Commonly utilized approaches for identifying the target proteins for a small molecule have limitations in terms of throughput and resource consumption and lack a mechanism to broadly assess the selectivity profile of the small molecule. Here we describe how protein microarray technology can be applied to the study of small molecule-protein interactions using tritiated small molecules. Protein arrays comprising thousands of full-length functional proteins facilitate target identification for those small molecules discovered in cell-based phenotypic assays and both target validation and off-target binding assessment for compounds discovered in target-based screens. The assays are highly reproducible, sensitive, and scalable, and provide an enabling technology for small molecule selectivity profiling in the context of drug development. Key words: Protein-small molecule interaction, Small molecule, Protein microarray, Off target effect, Selectivity profiling, Specificity profiling

1. Introduction The binding of small molecules to proteins is an integral component of biological pathways, including the binding of hormones to hormone receptors and small molecule ligands to G proteincoupled receptors. In addition, many small molecule drugs exert their biological effects, both desired and undesired, through interaction with one or more protein binding partners. A complete understanding of the binding profile of small molecules is rare, and functional protein microarrays provide a powerful approach to the identification and characterization of protein-small molecule interactions. Many aspects of biological pathways have been

Sridar V. Chittur (ed.), Microarray Methods for Drug Discovery, Methods in Molecular Biology, vol. 632, DOI 10.1007/978-1-60761-663-4_16, © Humana Press, a part of Springer Science+Business Media, LLC 2010

251

252

Kraus, Meng, and Freeman-Cook

recapitulated and studied on protein microarrays, including molecular interactions and enzymatic activities such as phosphorylation, ubiquitination, and methylation (1–6). The feasibility of detecting small molecule-protein interactions on functional protein arrays was first demonstrated through the binding of fluorescent molecules to the FK506-binding protein FKBP12 immobilized on chemically derivitized glass slides (7). Interactions were readily observed for molecules with affinities in the nM to mM range. The thyroid hormone receptor has also been shown to interact specifically with its ligand, triiodothyronine, when the receptor is immobilized in an array format (8). Specific binding of small molecule ligands to G protein-coupled receptors has also been demonstrated on arrays, indicating that these transmembrane receptors can retain membrane-like properties when arrayed on a solid support (9). Protein microarrays containing thousands of purified recombinant proteins have established a new paradigm for studying interactions between small molecules and proteins. The binding properties of GTP to GTP-binding proteins on arrays are comparable to those observed in solution (10). High-content protein arrays have been used to identify targets for small molecules that suppress a chemical-induced growth phenotype in the yeast S. cerevisiae (11). Recently, high content human protein arrays were used to profile a compound that is a potential therapeutic agent for the genetic disorder Spinal Muscular Atrophy. The compound was identified in a cell-based screen without knowledge of the molecular target, and the protein array experiments revealed a potential binder that was then validated with additional experiments (12). The sensitivity and ease-of-use support the widespread adoption of protein microarray technology for profiling small molecule-protein interactions. The platform facilitates the rapid identification of binding targets for small molecules and has the potential to dramatically accelerate the pace of drug discovery. The use of tritiated rather than fluorescently labeled or biotinylated small molecules in these assays eliminates any concern that the label may interfere with the binding profile of the small molecule. Here a known hormone–hormone receptor interaction was used to validate the use of tritium for detection of small moleculeprotein interactions using Invitrogen’s ProtoArray Human Protein Microarray version 4.1, which contains over 8,000 functional human proteins. The workflow for profiling tritiated small molecules on protein arrays is described, including a Phase 1 assay in which the optimal blocking conditions are established, and a Phase 2 assay in which proteins are identified that interact with the small molecule probe.

Small Molecule Selectivity and Specificity Profiling Using Functional Protein Microarrays

253

2. Materials 2.1. Materials and Equipment for Tritiated Small Molecule Profiling

1. [3H] Estradiol: [2,4,6,7,16,17-3H(N)]-estradiol (Perkin Elmer, Waltham MA). 2. [3H] experimental Small Molecule (SM): 100 ml at a concentration of 100 mM and a specific activity of ³ 10 Ci/mmol. 3. Protein Microarray Small Molecule Interaction Buffer (SMI Buffer): 50  mM Tris–HCL pH 7.5, 5  mM MgSO4, 0.1% Tween20, 150 mM NaCl (optional), 1% Hammarsten-grade casein (optional) prepared fresh and stored at 4°C. A stock solution of 10% Tween20 should be used to make the buffer. 4. ProtoArray® Human Protein Microarrays and ProtoArray® Control Arrays (Invitrogen, Carlsbad, CA), stored at −20°C until use. See Subheading  2.2 and http://www.invitrogen. com/protoarray for additional information on these protein array products. 5. Four-well quadriPERM incubation tray (Greiner, Monroe, NC). 6. Platform India).

shaker

(Lab-Line

Instruments,

Maharashtra,

7. LifterSlip™ cover slips (Thermo Scientific, Waltham, MA). 8. Incubator (30°C) (VWR, West Chester, PA). 9. Polyacetal slide rack (RA Lamb, Durham, NC) and Eppendorf plate centrifuge (model 5810, Fisher Scientific, Waltham, MA). 10. Tritium Sensitive Phosphor Screen (Perkin Elmer, Waltham, MA). 2.2. Scanning and Data Analysis

1. Cyclone Phosphorimager (Perkin Elmer, Waltham, MA) or any phosphorimager that provides at least 50 mm resolution to acquire the microarray image from the phosphor screen to produce a 16-bit TIFF file. 2. Genepix Pro 6.1 (Molecular Devices, Sunnyvale, CA) suggested, or other microarray analysis software. 3. A “GAL” file or other file containing information about the location of proteins on the microarray (Gal files for Invitrogen products can be downloaded from http://www.invitrogen. com/protoarray). 4. ProtoArray® Prospector provides automated data analysis for Invitrogen microarray products (can be downloaded from http://www.invitrogen.com/protoarray).

2.3. Protein Microarray Manufacturing

Clones used to produce proteins for ProtoArray® Human Protein Microarrays (Invitrogen) were obtained from Invitrogen’s

254

Kraus, Meng, and Freeman-Cook

Ultimate™ ORF (open reading frame) collection or from a collection of kinase clones developed by Protometrix or PanVera. The nucleotide sequence of each clone was verified by full length sequencing. All clones were transferred into a system for expressing recombinant proteins in Sf9 insect cells via baculovirus infection. Using a proprietary high-throughput insect cell expression system, thousands of recombinant human proteins were produced in parallel. Nearly all proteins are expressed as Glutathione-S-Transferase (GST) fusions, which enables highthroughput affinity purification under conditions that retain activity. After purification, every purified protein is quantified to ensure that the protein is present at the predicted molecular weight. ProtoArrays® are manufactured using a contact-type printer equipped with 48 matched quill-type pins. Each protein is printed along with a set of control proteins in duplicate spots on 1″ × 3″ glass slides that have been coated with a thin layer of nitrocellulose. Printing of arrays is carried out in a cold room under dust-free conditions in order to preserve the integrity of

Fig. 1. ProtoArray® Human Protein Microarray. (a) A ProtoArray® Human Protein Microarray was probed with an anti-GST antibody conjugated to Alexa Fluor® 647. The array was dried and scanned at 635 nm on an Axon 4000B scanner. The ProtoArray® contains 48 individual subarrays, each comprising an identical set of negative and positive control elements and variable human protein content. (b) An enlarged image of a single subarray is shown, with the estrogen receptor positional mapping elements highlighted. (c) An enlarged image of one subarray probed with [3H]-estradiol is shown.

Small Molecule Selectivity and Specificity Profiling Using Functional Protein Microarrays

255

both samples and printed microarrays. Before releasing protein microarrays for use, each lot of slides is subjected to a rigorous quality control (QC) procedure, including a gross visual inspection of all the printed slides to check for scratches, fibers, and smearing. Since nearly all of the proteins on the array contain an N-terminal GST tag, a GST-directed antibody detects printed proteins in a second QC assay. The QC assays measure the variability in spot morphology, the number of missing spots, the presence of control spots, and the amount of protein deposited in each spot. For the ProtoArray® Human Protein Microarray, proteins are printed as adjacent pairs in 110 mm spots arrayed in 48 subarrays (4,400-mm2 each) and are equally spaced in vertical and horizontal directions with 22 columns and 20 rows per sub-array. Spots are printed with a 200 µm spot-to-spot spacing. An extra 100-mm gap between adjacent sub-arrays allows quick identification of subarrays (Fig. 1). The proteins printed on the microarray retain functionality even after extensive storage at −20°C, as demonstrated by the auto-phosphorylation of kinases following incubation with 33P-ATP (data not shown).

3. Methods Protein microarrays have been used to study enzyme-substrate modifications, protein binding events including protein – protein interactions, antibody – antigen binding, and for immunological profiling with serum or other antibody-containing biological fluids. This technology is also ideally suited to the study of small molecule-protein binding events, and as such represents an approach for identifying on- and off-target binding partners that is more rapid and less biased than alternative assays that either rely on mass spectrometry or are limited to a specific class of target proteins. The method described here is based on the use of ProtoArray® Human Protein Microarrays (Invitrogen). The workflow for small molecule-protein interaction profiling using protein microarray technology is diagrammed in Fig. 2. 3.1. Tritiated Small Molecule Interaction Profiling on Functional Human Protein Microarrays

Steps 1 through 10 constitute Phase 1 of the assay and are designed to identify the optimal buffer conditions, either with or without casein as a blocking agent. Steps 11 through 19 constitute Phase 2 of the assay, in which interactions of the small molecule(s) with proteins on the high content ProtoArray® are identified. All steps should be carried out at 4°C and samples should be stored on ice unless stated otherwise. Please follow radioactive safety procedures and always check gloves, shoes, and working area for radioactive contamination. Discard radioactive waste into the proper waste container.

256

Kraus, Meng, and Freeman-Cook

Fig. 2. Overview of Tritiated Small Molecule-Protein Interaction Profiling Protocol on ProtoArray® Protein Microarrays. In Phase 1, the optimal blocking conditions are identified using control arrays. After blocking, arrays are incubated with [3H]-estradiol alone or in the presence of the tritiated small molecule of interest. [3H]-estradiol interacts specifically with estrogen receptor alpha (ERa) that is printed in each subarray, and is used for positioning the protein identification grid for subsequent analysis. The arrays are washed to remove unbound small molecule, exposed to a tritium-sensitive phosphor screen, and the images are acquired using a phosphorimager. Optimal blocking conditions are defined as those that provide the lowest levels of background signal across the array. In Phase 2, full content ProtoArray® Human Protein Microarrays are probed with the tritiated small molecule of interest. Assays are performed in duplicate at each of two concentrations, and in the absence and presence of NaCl. A negative control assay containing only [3H]-estradiol is performed in parallel. Arrays are processed as in Phase 1, and significant signals are compared to the negative control assay in order to identify candidate binding proteins.

1. Prior to initiating the small molecule interaction assay, an appropriate number of protein microarrays are obtained (ProtoArray® Human Protein Microarrays and ProtoArray® Control Arrays from Invitrogen are recommended). Two Control Arrays will be needed for each small molecule being profiled. One additional Control Array will be needed for a [3H]-estradiol control. The protein microarrays are stored at −20°C, and must be allowed to equilibrate to 4°C for 10 min prior to initiating the blocking step (see Note 1). To help preserve the protein activity, assays should be carried out at 4°C (see Note 2). 2. Prepare 300  ml SMI Buffer and 100  ml SMI Buffer with Casein for each small molecule that will be profiled. Place slides with the proteins facing up into a 4-Well Tray, 1 slide per well. Carefully add 5 ml SMI Buffer to the array that will be probed with [3H]-estradiol only. For the two control slides to be probed with each small molecule, add 5 ml SMI Buffer to one control slide and 5 ml SMI Buffer with Casein to the second control slide (see Note 3). Ensure that the barcode end of the slide is near the end of the tray with the indented numeral. The indent in the bottom of the tray will be used as the site of buffer exchange. Cover the tray and incubate for 60  min at 4°C with gentle circular shaking (~50 rpm) (see Note 4).

Small Molecule Selectivity and Specificity Profiling Using Functional Protein Microarrays

257

3. While slides are in blocking buffer, prepare the [3H]-estradiol and experimental small molecule solutions in the appropriate SMI Buffer (see Note 5). The final volume of each sample should be 120 ml. For the [3H]-estradiol control slide, add 1 µl [3H]-estradiol to 99 µl SMI Buffer for a 1:100 dilution, which will be used as the working stock solution. For the [3H]-labeled experimental small molecule, prepare two independent solutions. For one solution, add 1.2 ml of the small molecule solution to 117.6 ml SMI Buffer. For the second solution, add 1.2 ml of the small molecule solution to 117.6 ml SMI Buffer with Casein. In a third microfuge tube, aliquot 118.8 ml SMI Buffer for the estradiol-only control. Add 1.2 ml of [3H]-estradiol working stock solution (the 1:100 dilution) to each of the three tubes. Mix the contents of each tube gently by pipet. 4. Remove the array to be probed with [3H]-estradiol alone from the blocking buffer. Tap one edge of the array gently on a laboratory wipe for a few seconds to drain any buffer without allowing the array to dry (see Note 6). Place the slide horizontally on the edge of the 4-well tray. Immediately dispense 120 ml SMI Buffer containing [3H]-estradiol onto the top of the array using a micropipette without touching the array surface. Carefully remove a LifterSlip™ from package and hold the LifterSlip™ from the corner using forceps. Gently lower the LifterSlip™ onto the surface of the slide taking care to avoid trapping bubbles between the slide and the LifterSlip™. This is most effectively achieved by resting one of the short ends of the LifterSlip™ on the slide near the barcode, and slowly lowering (see Note 7). Avoid direct contact of the forceps with the slide surface. Align the LifterSlip™ flush with the top edge of the array to ensure the printed area of the array is completely covered. Lower the slide with a cover slip into a 4-well tray with the printed side (barcode) of the array facing up. Repeat step 4 for each slide that is to be probed with the tritiated SM. Make sure to add the small molecule solution(s) containing casein to only those arrays that were blocked in SMI Buffer with Casein. 5. Cover the 4-well tray(s) and place each 4-well tray on a flat surface in a refrigeration unit set to 4°C such that the printed side of the array is facing up, taking care to ensure that the surface is as level as possible. Incubate the arrays in the trays for 60 min at 4°C without shaking. 6. Remove the 4-well trays containing slides from the refrigerator. Using forceps gently lift the slides from the tray and place into a 50 ml conical tube, one slide per tube. Add 40 ml of SMI Buffer to the [3H]-estradiol control slide and to the slide probed with the experimental SM in SMI Buffer.

258

Kraus, Meng, and Freeman-Cook

Add 40  ml of SMI Buffer with Casein to the slide probed with small molecule in SMI Buffer with Casein. Avoid pouring the solutions directly onto the array surface. Incubate the arrays in the appropriate SMI Buffer for 1  min at room temperature. Gently move the array in the tube to dislodge the LifterSlip™. Using forceps, carefully remove the dislodged LifterSlip™ without touching the array surface (see Note 8). Incubate the arrays in SMI buffer for an additional 1  min. Discard the LifterSlip™ appropriately as radioactive waste. Decant the SMI Buffer solutions into a radioactive waste container. Two additional washes with the appropriate SMI Buffer should be performed for 1 min each. 7. Remove the arrays from the tubes at the end of the probing procedure. Tap one edge of the array gently on a laboratory wipe for a few seconds to drain any buffer. Place each array in a slide holder in a vertical orientation. Ensure the array is properly placed and is secure in the holder to prevent any damage to the array during centrifugation. Centrifuge the array in the slide holder at 800×g for 2 min at room temperature. After centrifugation, allow the arrays to dry at room temperature for 30 min. 8. Obtain a tritium-sensitive phosphor screen. Be sure to erase the phosphor screen by exposure to light prior to exposure to the slides. Place a sheet of filter paper in the bottom of an exposure cassette and lay the slides down with the array side facing up. Tape the top and bottom edges of the slides down onto the filter paper using lab tape to ensure that the slides do not move during exposure. Place the tritium-sensitive phosphor screen down onto the arrays, with the blue side of the phosphor screen facing the slides. Once the tritium-sensitive phosphor screen has been arranged on top of the slides, place a sheet of filter paper on top of the phosphor screen and lower the cover of the exposure cassette. Lock the exposure cassette and expose the arrays for 3 days (see Note 9). 9. Acquire an image of the exposed phosphor screen using a Cyclone phosphorimager or other appropriate phosphorimager (see Note 10). The entire area of the phosphor screen should be scanned with a resolution of at least 50 mm and saved as a 16-bit grayscale TIFF file. 10. Open the saved 16-bit TIFF file in Adobe® Photoshop®. Adjust the brightness and contrast so that each array is visible and evaluate background intensities across each of the arrays. The Control Array probed with [3H]-estradiol alone should have negligible background signal. Evaluate the arrays probed with each small molecule to determine the effect of casein on the profiling. If the background is noticeably higher on the

Small Molecule Selectivity and Specificity Profiling Using Functional Protein Microarrays

259

array incubated with SMI Buffer compared to SMI Buffer with Casein, then the optimal conditions to move forward with in Phase 2 will be with SMI Buffer with Casein. If the background is noticeably higher on the array incubated with SMI Buffer with Casein, then the small molecule likely binds to casein and the optimal conditions to move forward with in Phase 2 will be with SMI Buffer. If there is not a significant difference in background signals between arrays probed with SMI Buffer in the absence and presence of casein, then the optimal conditions to move forward with in Phase 2 will be with SMI Buffer with Casein. As an example, the Phase I experiment results with a [3H] experimental Small Molecule (SM) are shown in Fig. 3. 11. Equilibrate high content ProtoArray® slides at 4°C for 10 min. Eight slides will be needed for each small molecule being profiled: duplicate arrays at each of two SM concentrations. One additional slide will be needed for a [3H]-estradiol-only control. 12. Prepare 1,000  ml SMI Buffer (with or without Casein, as identified as optimal in Phase 1) and 600 ml SMI Buffer with 150  mM NaCl (with or without Casein, as identified as

Fig.  3. Images of Control ProtoArray® Protein Microarrays probed with [3H]-labeled small molecules. Images of ProtoArray® Control Protein Microarrays from Phase 1. Control arrays were incubated with [3H]-estradiol alone or in the presence of [3H]-small molecule (SM1). For arrays probed with [3H]-SM1, blocking and probing steps were carried out in the absence or presence of 1% casein.

260

Kraus, Meng, and Freeman-Cook

optimal in Phase 1) for each small molecule that will be profiled (see Note 11). Place slides with the proteins facing up into 4-well trays, one slide per well. Carefully add 5  ml SMI Buffer to the array that will be probed with [3H]-estradiol only. For the eight slides to be probed with each small molecule, add 5 ml SMI Buffer to four of these slides and 5 ml SMI Buffer with 150  mM NaCl to the other four of these slides. Cover the trays and incubate for 60 min at 4°C with gentle circular shaking (~50 rpm). 13. While slides are in blocking buffer prepare the [3H]-estradiol and experimental small molecule solutions in the appropriate SMI Buffer. The final volume of each sample should be 120 ml. For the [3H]-estradiol-only control slide, add 1 ml [3H]-estradiol to 99 µl SMI Buffer for a 1:100 dilution. For the [3H]-labeled small molecule, prepare four independent solutions. For solution 1, add 2.7 µl of the small molecule solution to 270 µl SMI Buffer. For solution 2, add 2.7 ml of the small molecule solution to 270 µl SMI Buffer with 150 mM NaCl. For solution 3, add 25 ml of solution 1 to 222.6 ml SMI Buffer. For solution 4, add 25 ml of solution 2 to 222.6 ml SMI Buffer with 150 mM NaCl. In a fifth microfuge tube aliquot 118.8 µl SMI Buffer. Add 2.5 ml of the 1:100 dilution of [3H]-estradiol to solutions 1, 2, 3, and 4, and 1.2 µl to the fifth tube that will contain [3H]-estradiol alone. Mix the contents of each tube gently by pipet. 14. Remove the array to be probed with [3H]-estradiol alone from the blocking buffer. Tap one edge of the array gently on a laboratory wipe for a few seconds to drain any buffer without allowing the array to dry. Place the slide horizontally on the edge of the 4-well tray. Immediately dispense 120 ml of solution 5 (SMI Buffer containing [3H]-estradiol only) onto the top of the array using a micropipette without touching the array surface. Carefully remove a LifterSlip™ from package and hold the LifterSlip™ from the corner using forceps. Gently lower the LifterSlip™ onto the surface of the slide taking care to avoid trapping bubbles between the slide and the LifterSlip™. This is most effectively achieved by resting one of the short ends of the LifterSlip™ on the slide near the barcode, and slowly lowering. Avoid direct contact of the forceps with the slide surface. Align the LifterSlip™ flush with the top edge of the array to ensure the printed area of the array is completely covered. Gently adjust the LifterSlip™ to remove any air bubbles, if necessary. Lower the slide with a cover slip into a 4-well tray with the printed side (barcode) of the array facing up. Repeat step 14 for each slide that is to be probed with the tritiated SM, probing two slides with each solution (solution 1, 2, 3,

Small Molecule Selectivity and Specificity Profiling Using Functional Protein Microarrays

261

4, the two dilutions of SM in buffer with and without 150 mM NaCl). 15. Cover the 4-well tray(s) and place each 4-well tray on a flat surface in a refrigeration unit set to 4°C such that the printed side of the array is facing up, taking care to ensure that the surface is as level as possible. Incubate the arrays in the trays for 60 min at 4°C without shaking. 16. Remove the 4-well trays containing slides from the refrigerator. Using forceps gently lift the slides from the tray and place into a 50 ml conical tube, one slide per tube. Add 40 ml of SMI Buffer to the [3H]-estradiol control slide and to the slides probed with the experimental SM in SMI Buffer. Add 40 ml of SMI Buffer with 150 mM NaCl to the slides probed with small molecule in SMI Buffer with NaCl. Avoid pouring the solutions directly onto the array surface. Incubate the arrays in the appropriate SMI Buffer for 1 min at room temperature. Gently move the array in the tube to dislodge the LifterSlip™. Using forceps, carefully remove the dislodged LifterSlip™ without touching the array surface. Incubate the arrays in SMI buffer for an additional 1  min. Discard the LifterSlip™ appropriately as radioactive waste. Decant the SMI Buffer solutions into a radioactive waste container. Two additional washes with the appropriate SMI Buffer should be performed for 1 min each. 17. Remove the arrays from the tubes at the end of the probing procedure. Tap one edge of the array gently on a laboratory wipe for a few seconds to drain any buffer. Place each array in a slide holder in a vertical orientation. Ensure the array is properly placed and is secure in the holder to prevent any damage to the array during centrifugation. Centrifuge the array in the slide holder at 800×g for 2 min at room temperature. After centrifugation, allow the arrays to dry at room temperature for 30 min. 18. Obtain a tritium-sensitive phosphor screen. Be sure to erase the phosphor screen by exposure to light prior to exposure to the slides. Place sheet of filter paper in the bottom of an exposure cassette and lay the slides down with the array side facing up. Tape the top and bottom edges of the slides down onto the filter paper using lab tape to ensure that the slides do not move during exposure. Place the tritium-sensitive phosphor screen down onto the arrays, with the blue side of the phosphor screen facing the slides. Once the tritium-sensitive phosphor screen has been arranged on top of the slides, place a sheet of filter paper on top of the phosphor screen and lower the cover of the exposure cassette. Lock the exposure cassette and expose the arrays for 10–14 days.

262

Kraus, Meng, and Freeman-Cook

1 9. Acquire an image of the exposed phosphor screen using a Cyclone phosphorimager or other appropriate phosphorimager. The entire area of the phosphor screen should be scanned with a resolution of at least 50 mm and saved as a 16-bit TIFF file. 3.2. Data Analysis and Hit Identification

After generating the 16-bit TIFF file, the image must be processed before data acquisition and analysis. This can be performed using Adobe Photoshop using the following procedure. 1. Start Adobe Photoshop and open the TIFF image of the microarrays. 2. Crop a fixed rectangular area (1″×3″) corresponding to each microarray. If the spots are not aligned vertically, rotate the image to correctly align the spots. 3. Invert the data (convert the image from white background with black spots to black background with white spots) and resize the image file to 2,550 × 7,650 pixels (constrained proportions). Do not adjust the image quality (such as contrast or level), which can compress the dynamic range of the data and affect the analysis. 4. Save the cropped and resized images as a 16-bit TIFF file with a new name to a suitable location and proceed with the data acquisition and analysis. 5. The array list file (.gal file) is uploaded to the image analysis software (GenePix 1 from Molecular Devices is recommended). This text file describes the layout of the protein microarray and contains the details of the microarray content, including relevant control elements. For the ProtoArray® Human Protein Microarray, the .gal files can be found by following the Online Tools link on the ProtoArray® Central Portal (www.invitrogen.com/protoarray). The .gal file is used to map the location of each array feature, initially with a fixed feature size based on the diameter of the spotted protein microarray features. To maximize accuracy, a pixel-based segmentation algorithm is recommended for pixel intensity data extraction (Irregular Feature Finding setting, located under the Alignment tab in the Options menu of GenePix 6.1) (see Note 12). Load the .gal file into the software using the Array List button. Use spots corresponding to the positional mapping feature (estrogen receptor, ERa, which binds to the tritiated estradiol) as a reference point to align the grid. Scroll through the image to ensure the grid is in the proper location for each sub-array and make adjustments as necessary. Pixel intensities for each spot on the array are calculated by the software and saved to a text file

Small Molecule Selectivity and Specificity Profiling Using Functional Protein Microarrays

263

formatted for use in GenePix as a GenePix Result file (.gpr filename extension). These files are subsequently opened in other text editing or spreadsheet programs for analysis. 6. Assuming ProtoArray® Human Protein Microarrays are used in the assays, quantitated spot files are processed using the ProtoArray ® Prospector freeware to determine which proteins interact with the small molecule probes. The software incorporates signal scatter compensation, background subtraction, Z-Factor and Z-Score calculations, and replicate spot coefficient of variation (CV) filtering. Z-Factor is a screening window metric that takes into account the signal dynamic range and the variation associated with the control and sample features (13). Z-Score is a value that indicates how far and in what direction a signal from a specific protein feature deviates from the mean of the signals from all of the protein features. The value is expressed in terms of standard deviations. Typically, the threshold criteria for determining interacting proteins are >0.5 for Z-Factor (indicating a signal:noise ratio of >2) and/or a Z-Score >3 (indicating the signal for a given protein is >3 standard deviations above the mean signal for all proteins). These criteria may be relaxed to Z-Factor >0.35 (indicating signal:noise of 1.5) or Z-Score >2 if the stringency of the standard thresholds results in no interacting proteins being identified. Relative Signal Used (backgroundcorrected signal) values should typically be greater than 500 for proteins that interact with the small molecule. Replicate spot CV should be less than 0.5 in order to be considered an interacting protein. 7. Alternately, Microsoft Excel can be used to analyze the raw .gpr data (the .gpr file is a text file which can be open directly in Excel), although the signal scatter correction algorithm used in ProtoArray® Prospector will not be incorporated. The background-subtracted signal values (F635medianB635median) may be used to calculate Z-factors using the following formula: Z-Factor = 1–3 × (stdev feature signal + stdev negative signal)/| (avg feature signal−avg negative signal)| Interacting proteins are defined as those yielding a Z-factor > = 0.5 and a replicate spot CV < 0.5. Coefficient of variation (CV) = standard deviation/mean. As an example, the specificity profile for a tritiated small molecule generated on a high content protein microarray is shown in Figs. 3–5.

264

Kraus, Meng, and Freeman-Cook

Fig.  4. Images of ProtoArray® Protein Microarrays from Phase 2. (a) Images corresponding to ProtoArray® Human Protein Microarrays probed with 115 nM (10 nCi/µl) or 23 nM (2 nCi/µl) [3H]-SM1 in the absence of sodium chloride. The negative control array, probed with 40  pCi/µl [3H]-estradiol, is shown for comparison. (b) Images corresponding to ProtoArray® Human Protein Microarrays probed with 115 nM (10 nCi/µl) or 23 nM (2 nCi/µl) [3H]-SM1 in the presence of 150 mM sodium chloride. The negative control array, probed with 40 pCi/µl [3H]-estradiol in the presence of NaCl, is shown for comparison.

4. Notes 1. Spots may smear or merge if arrays are not equilibrated before use due to the formation of condensation on the array surface.

Small Molecule Selectivity and Specificity Profiling Using Functional Protein Microarrays

265

Fig. 5. Examples of [3H]-SM1 candidate protein interactors. The plots show background subtracted signal (Signal Used) values corresponding to each assay included in Phase 2 of the study.

2. For certain small molecule-protein interactions, such as interactions between ligand and G-protein coupled receptors, the favorable incubation temperature might be higher (ex. room temperature). 3. Assay performance with the optimal blocking reagents is small molecule-specific and can be determined through the pilot experiment described in Phase 1 of this procedure. Buffers containing casein should be heated to 50°C until casein is completely dissolved. The casein used should be Hammarsten grade casein. Do not exceed 60°C and do not microwave the solution. Buffer should be cooled to 4°C before use. 4. Use a shaker that keeps the arrays in one plane during rotation. Nutating or rocking shakers are not recommended because of increased risk of cross-well contamination. 5. The recommended activity range for the final concentration of small molecule probe is 50 pCi/ml-50 nCi/µl, with weaker interactions requiring activity of 10–50 nCi/µl. The tritiated small molecule stock activity should be at least 1 mCi/µl with a specific activity of at least 10 Ci/mmol, and a minimum of 60 mCi should be available to perform each small moleculeprotein interaction experiment. If the tritiated small molecule

266

Kraus, Meng, and Freeman-Cook

is dissolved in an organic solvent such as ethanol or DMSO, the final concentration should be less than 5% ethanol by volume or 1% DMSO by volume. To avoid non-specific interactions and/or high background, the final concentration of the small molecule should not be higher than 1 mM. 6. Do not allow any part of the array surface to dry before adding the next solution as this can cause high and/or uneven background. 7. If air bubbles are trapped under the LifterSlip™, tap the slides gently to drive them out or lift one edge of the LifterSlip™ allowing bubbles to move to the fluid front and then gently lower down again. 8. Do not remove the LifterSlip™ with forceps if the LifterSlip™ is not dislodged from the array. Continue to gently move the array in the tube until the LifterSlip™ floats off. 9. The tritium-sensitive phosphor screen will eventually be damaged due to tritium contamination. Directly washing the screen with methanol can remove some contamination, but for critical experiments we recommend the use of a new screen or a screen that has been verified to be contaminantfree by pre-exposure in an empty cassette followed by scanning and imaging. 10. The following phosphorimagers have been tested with ProtoArray® microarrays: Cyclone® Storage Phosphor System (Perkin Elmer, Inc.) and Typhoon™ Imager (GE Healthcare Life Sciences). 11. High salt concentrations can in some cases modulate binding interactions between small molecules and protein targets through ionic interactions and solvation effects at the protein surface. 12. In general, the use of pixel-based segmentation (irregular feature finding) results in more reproducible Signal Used values. References 1. Predki PF (2004) Functional protein microarrays: ripe for discovery. Curr Opin Chem Biol 8:8–13 2. Boyle SN, Michaud GA, Schweitzer B, Predki PF, Koleske AJ (2007) A critical role for cortactin phosphorylation by Abl-family kinases in PDGF-induced dorsal-wave formation. Curr Biol 17:1–7 3. Gupta R, Kus B, Fladd C, Wasmuth J, Tonikian R, Sidhu S, Krogan NJ, Parkinson J, Rotin D (2007) Ubiquitination screen using

protein microarrays for comprehensive identification of Rsp5 substrates in yeast. Mol Systems Biol 3(116):1–12 4. Hudson ME, Pozdnyakova I, Haines K, Mor G, Snyder M (2007) Identification of differentially expressed proteins in ovarian cancer using high-density protein microarrays. Proc Natl Acad Sci USA 104:17494–17499 5. Satoh J, Obayashi S, Misawa T, Sumiyoshi K, Oosumi K, Abunoki H (2008) Protein microarray analysis identifies human cellular

Small Molecule Selectivity and Specificity Profiling Using Functional Protein Microarrays

6.

7.

8.

9. 10.

prion protein interactors. Neuropathol Appl Neurobiol 35:16–35 Schnack C, Hengerer B, Gillardon F (2008) Identification of novel substrates for Cdk5 and new targets for Cdk5 inhibitors using high-density protein microarrays. Proteomics 8:1980–6 MacBeath G, Schreiber SL (2000) Printing proteins as microarrays for high-throughput function determination. Science 289: 1760–3 Ge H (2000) UPA, a universal protein array system for quantitative detection of protein-protein, protein-DNA, protein-RNA, and protein-ligand interactions. Nucleic Acids Res 28:e3 Fang Y, Frutos AG, Lahiri J (2002) Membrane protein microarrays. J Am Chem Soc 124: 2394–5 Schweitzer B, Predki P, Snyder M (2003) Microarrays to characterize protein interactions

267

on a whole-proteome scale. Proteomics 3: 2190–9 11. Huang J, Zhu H, Haggarty SJ, Spring DR, Hwang H, Jin F, Snyder M, Schreiber SL (2004) Finding new components of the target of rapamycin (TOR) signaling network through chemical genetics and proteome chips. Proc Natl Acad Sci USA 101:16594–9 12. Singh J, Salcius M, Liu S-W, Staker BL, Mishra R, Thurmond J, Michaud G, Mattoon DR, Printen J, Christensen J, Bjornsson JM, Pollok BA, Kiledjian M, Stewart L, Jarecki J, Gurney ME (2008) DcpS as a therapeutic target for Spinal Muscular Atrophy. ACS Chem Biol 3:711–22 13. Zhang J-H, Chung TDY, Oldenburg KR (2000) Confirmation of primary active substances from high-throughput screening of chemical and biological populations: A statistical approach and practical considerations. J Com Chem 2:258–265

Chapter 17 Production and Application of Glycan Microarrays Julia Busch, Ryan McBride, and Steven R. Head Abstract Glycans are vital elements of living organisms and are involved in recognition, communication, cell growth and development, motility, and other significant processes. The interactions of glycans with the proteins that bind them provide valuable information about protein interaction and specificity. By printing glycans on microarrays, investigators are able to effectively determine the binding specificity of certain proteins with an extremely efficient and precise result. Such chips are performed by standard robotic microarray printing. Incubating the slides with various GBP-containing substances not only reveals clear receptor preferences of the proteins, but also detects minute differences in structure specificity. Key words: Glycan, Carbohydrate, Lectin, Microarray, Glycan-binding protein, Chip

1. Introduction The study of interactions between GBPs and glycan ligands has progressed greatly through the development and use of glycan microarrays. The technology of these arrays allows investigators to observe and analyze GBP interactions on individual chips, each of which can contain hundreds of different glycan structures. Arrays can be used to study a number of GBPs including antibodies specific to tumors or HIV, plant lectins, and virus GBPs such as hemagglutinin (HA) (2). Glycan arrays can be very easily customized to contain desired structures, which gives the capability to explore pathogen-specific glycan interactions (3).

Sridar V. Chittur (ed.), Microarray Methods for Drug Discovery, Methods in Molecular Biology, vol. 632, DOI 10.1007/978-1-60761-663-4_17, © Humana Press, a part of Springer Science+Business Media, LLC 2010

269

270

Busch, McBride, and Head

2. Materials 2.1. Plate Setup

1. Printing Buffer: 300 mM phosphate buffer, pH 8.5 containing 0.005% Tween-20. Store at 4°C long term, but must be at room temperature before use to avoid crystallization. 2. 96-well plates: Nunc 96-well conical bottom (Nunc 249944). 3. 96-well plate lids: Nunc 96-well cap, natural rubber, nonsterile (Nunc 276002). 4. Matrix standard clear 384 well small volume microplates. 5. Matrix universal polystyrene lid use with 96, 384, and 1,536 well plates. 6. USA Scientific Sealing Film (2921-0000). 7. Glycans (natural and/or synthetic) with amine derivatized spacers. Store at −20°C long term, but must be at room temperature when printing.

2.2. Printing

1. SCHOTT Nexterion® Slide H, polymer layer activated with N-Hydroxysuccinimide (NHS) esters which covalently binds amine groups. Store sealed in bags at −20°C long term. The slides must be removed from freezer for at least 4 h (usually overnight) before opening bags and printing. 2. ArrayIt Stealth Micro Spotting Pins (SMP4B and SMP4 models were used and have the same printing quality and produce the same spot size) or acceptable equivalent. Stored in the arrayer head or in manufacturers’ box with tip protection intact.

2.3. Humidification/ Immobilizing

1. Large Pyrex dish (5 cm × 30 cm × 15 cm). 2. Rack, such as test tube. 3. Plastic Wrap.

2.4. Numbering, Blocking, and Storage

1. VWR black lab marker. 2. Slide staining rack, 50 slide rack, and dish (Wheaton 90040). 3. Blocking buffer: 50 mM ethanolamine in 50 mM borate buffer, pH 9.2. Stored at 4°C long term but for best results, should be at room temperature before contacting slides. 4. Glass Pyrex loaf dish with fitted glass lid ( 3 1 4 × 7 7 8 × 4 1 8 ). 5. Glass slide staining rack (Wheaton 900200). 6. Drierite Anhydrous Calcium Sulfate, Indicating (Item#23005). Stored at room temperature in tray on bottom of desiccator

Production and Application of Glycan Microarrays

271

boxes. Should be replaced when color indicates-time period varies depending on how often the containers are opened and exposed to humidity. 2.5. Incubation

1. Bioworld PBS 20× solution, pH at 25°C: 7.3–7.5 stored at room temperature. 2. Bioworld PBS 1× solution made by diluting 20× solution 1:20 with ddH2O stored in 1 L quantities at 4°C. 3. Tween® 20, 500  mL (Aldrich Chemical Company, Inc.). Stored at room temperature. 4. Seracare Life Sciences Bovine Serum Albumin standard grade powder. Stored at 4°C. 5. Super PAP PEN hydrophobic slide marker or acceptable alternative. Stored at room temperature lying horizontally on its side. 6. Kimwipes. 7. Invitrogen Molecular Probes Strepdavidin Alexa Fluor 488 conjugate 2  mg/mL in PBS, pH 7.2, 5  mM azide, 5 moles dye/mole. Stored at 4°C. Light sensitive. 8. Plastic Wrap. 9. Aluminum Foil.

3. Methods The process of printing, preparing, and analyzing glycan microarrays, if done properly and carefully, will yield extremely high quality data. The slides used in these prints are very sensitive to a variety of factors including humidity, temperature, and physical contact. In order to draw conclusions based on reliable and reproducible results, special attention must be paid to the storage, handling, and application of all the components of an array. It is assumed that the reader is familiar with the general process of designing an array as well as the access to and experience with a robotic microarrayer and slide scanner. A glycan array is composed of a library of structurally defined sugars that have been modified by the addition of a linker, also referred to as a spacer, containing a terminal amine. These terminal groups bind immediately and irreversibly to the NHS esters on the slide surface during printing and are immobilized before storage to preserve the quality of the chips. By then blocking any additional unwanted bonding to the slide, clear and precise results can be obtained. To be able to analyze the glycan arrays,

272

Busch, McBride, and Head

the GBPs must be either directly or indirectly (through the use of a secondary antibody) fluorescently labeled (1). 3.1. Plate Setup

The design and layout of the glycan array should be completed before this process is started. All of the pipette work in the following method can be performed by hand or with the use of a robot, but great attention to detail and correct placement are extremely important in either case. 1. To prepare a stock of printing buffer, first make about 500  mL of a 300  mM stock solution of dibasic sodium phosphate. Also make about a 50  mL stock of 300  mM monobasic sodium phosphate solution. In a flask using a magnetic stir bar and pH meter, slowly titrate the dibasic sodium phosphate solution with the monobasic solution until a pH of 8.5 is reached. Using a 10% stock solution of Tween-20 in ddH2O, add the appropriate amount to the phosphate buffer to reach an end result of 0.005% Tween20 content. If the buffer has been prepared at a previous date, it might be necessary to place the bottle in a 37°C water bath to remove crystals that may have precipitated over time 2. The sugars are kept at a 1 mM stock concentration in printing buffer stored in 2 mL screw cap microtubes at −20°C. Thaw sugars in warm water bath of 37°C, mix by vortexing, then briefly spin down with centrifuge or small tube spinner to avoid droplets on cap 3. Add 90  mL of printing buffer to each well of the 96-well plates.  10  mL of each sugar should then be added to and thoroughly mixed in each well for a 1:10 dilution, bringing the concentration to a desired 100 mM. Our lab includes an additional concentration of 10  mM sugars on the array. To add these samples to the plates, pipette 10 mL of the 100 mM sugars into wells with 90 mL of printing buffer to perform an additional 1:10 dilution 4. Place sealing lid on 96-well plate and gently mix with vortex. Spin plates in centrifuge briefly, allowing them to reach ~2,000  rpm (1666g) for a moment just to remove any liquid that may have splashed onto the inside of the lid during mixing 5. Transfer 10 mL of each glycan from the 96 well plates to the 384 well plates 6. After the 384 well plates are complete, seal with adhesive film and spin them the same as the 96-well plates to remove any bubbles in the bottom of the wells. Keep all of the plates on ice until it is time to print the arrays

Production and Application of Glycan Microarrays

3.2. Printing

273

The authors make the assumption that the reader has access to a microarrayer and the knowledge of its use. Our lab uses a MicroGrid II 600 by Digilab, but any comparable machine would be acceptable. The slides are sensitive to humidity and as soon as the sealed bags are opened, their surface begins to hydrolyze with any moisture over the course of time. This should be taken into account when planning especially long prints at high humidity. 1. Place arrayer pins in the printing head according to the correct configuration of the desired layout 2. Remove slides from bags, open boxes, and remove any dust particles or small pieces of plastic that might be on their surfaces by blowing ultra-high purity Nitrogen gas on each slide. Anything that is on the printing surface of the slide could potentially get picked up by the pins and inhibit the ability to print. Place the desired number of slides, coated side up, on the arrayer stage (see Note 6) 3. Program the arrayer with the appropriate conditions, load the 384 well plates into the printer, and begin. Periodically check that all the pins are printing by shining a flashlight onto the stage, so the spots are visible with the glare. Relative humidity while printing should be kept between 55 and 65%. A standard hygrometer placed inside the lid of the arrayer is a good measure of chamber humidity.

3.3. Humidification/ Immobilizing

1. Once the slides have finished printing, they must undergo a temporary immobilization step, also called humidifying. This includes the slides being placed in a chamber with 100% humidity immediately after printing for a period of 30 min. The chamber can be constructed by simple placing a few very wet paper towels flat in the bottom of a large Pyrex glass dish, placing the slides on some sort of rack, such as a test tube rack, print side up, and sealing with plastic wrap to trap in the moisture 2. After the humidification step, the slides can be immediately numbered and blocked or stored in a desiccation box.

3.4. Numbering, Blocking, and Storage

Marking, blocking, and storing the slides are very important steps to preparing the slides for incubation. When the slides are bordered and numbered, the orientation of the print must be paid attention to. Desiccating conditions can be achieved by placing a tray of Drierite in the bottom of an air tight container which keeps the slides at 0% humidity for long term storage. 1. Blocking buffer can be prepared in generous quantities since a fairly large amount is used per batch of slides printed.

274

Busch, McBride, and Head

First prepare about 700  mL of a 50  mM boric acid solution. While constantly stirring and monitoring the pH, add the appropriate amount of ethanolamine from a 16.54  M stock solution to reach the desired 50 mM buffer solution. Finally, add concentrated sodium hydroxide to bring the pH up to 9.2 2. After the slides have been humidified, the spots are still visible as small crystals left by the printing buffer. It is necessary to inactivate any unbound groups on the slide surface in order to prevent any nonspecific binding. Once the blocking step has been completed, the crystals will wash away and the grids will no longer be visible on the slide surface, so the print area must be checked and marked prior to the slides being placed in the buffer. Before marking a slide, do a visual inspection to make sure that there are no large parts of grids are missing, smears, or any other major imperfections. This is also when slide numbering should occur. If the barcodes are not the preferred form of identification, a series of numbers can simply be assigned to the slides. This step is also vital to designating the directionality of the array. By numbering each of the slides in an identical fashion, the position of the array can be determined even after the print is no longer visible. This is necessary knowledge to have during the analysis step so if numbering is not used, some other marking to designate an orientation should be used. Using a black VWR lab marker on the back side (not the print surface), mark brackets around each of the four corners of the entire print and number if desired. Be sure not to mark directly behind any portion of the array which could obstruct the scanner from reading a signal (see Fig. 1) (see Note 4) 3. After each slide is marked, carefully place up to 50 into each metal rack. Place the rack inside the Pyrex loaf dish, moving the handle down, so the lid can sit properly on top. Pour enough blocking buffer just to cover the tops of the slides. Place the lid on and shake gently for 1 h 4. Once the slides have finished blocking, remove them from the dish, allowing the excess buffer to drain off. In a separate loaf dish filled about halfway with ddH2O, dip the entire rack of slides 10 times – completely removing and submerging them from the water each time. Dispose of the blocking buffer in the appropriate waste container 5. Transfer the slides to smaller glass dishes and spin dry in centrifuge at 20°C for 5 min at 200 rcf (1,024 rpm) 6. Once the slides are dry, they can be incubated immediately or stored under desiccating conditions for up to 6 months

Production and Application of Glycan Microarrays

275

Fig. 1. The view of the print surface of a slide that has been bordered and numbered on the back

3.5. Incubation

Our lab uses primarily three different protocols for incubating the glycan arrays. On every batch of slides printed, it is important to do some form of quality control. This can be performed by preparing a cocktail of biotinylated plant lectins which are incubated on the slides then fluorescently labeled. Serum can also be incubated on slides and the GBPs can then indirectly be labeled through a secondary antibody. Viruses are another sample that can be incubated on the arrays and require antisera as well as secondary antibodies for the labeling to be successful. All incubations using dangerous of infectious substances should be performed in the hood while non-hazardous samples, such as lectins, can be done on the bench. A sample volume of 1 mL should cover the print surface for most configurations, but large prints using 48 or more pins may need approximately 1.2 mL, and the same logic applies for smaller prints and volumes. For all incubations involving fluorescence, cover the humidification chamber with foil during the labeling step to avoid any loss of signal. The first four steps are universal for all glycan array incubations.

3.5.1. For All Array Incubations

1. Draw two lines, one on either side of the print area, with hydrophobic marker on outside of print area and let it dry for several minutes. Be sure to mark the slide on the print surface

276

Busch, McBride, and Head

and to go all the way to the edge of the slide to create a true barrier. The purpose of this step is to limit the area the liquid will cover on the array during incubation and also helps preserve sample volume 2. Place a couple of paper towels in the bottom of a Pyrex dish (size depending on how many slides are being incubated at a time) and soak them generously with water. Use a test tube rack or lid of some sort inside the dish to put the slides on during the incubation to raise the slides up off the wet towels. Place the dish, which is now the humidification chamber, on a rotating shaker 3. Soak the slides in PBS for 2 min to hydrate the print surface. This is an important step to eliminate nonspecific binding, which can cause high background 4. Remove the slides one at a time and dry off the backside with a Kimwipe, be careful not to wipe the print surface. Return each slide in the humidification chamber and pipette 1 mL of the sample in between the hydrophobic marker barrier (see Note 5) 3.5.2. For Plant Lectins

1. Dilute the lectin(s) in incubation buffer composed of 1× PBS and 0.05% Tween-20 to a final concentration of 10 mg/mL 2. Pipette 1 mL of the sample onto the print surface and incubate in the sealed humidification chamber while rotating gently for 1 h. Take care that the speed is not too fast and the liquid on the slide surface does not spill 3. At the end of the first incubation period, discard sample by simply pouring it off the slide into the bottom of the humidification chamber 4. Wash the slides one at a time by holding the edges and dipping them four times in a mixture of PBS and 0.05% Tween, then four times in 1× PBS, and finally four times in ddH2O 5. Dry off the backside of the slide with a Kimwipe and place back in the humidification chamber 6. Add 1 mL of Strepdavidin at 0.4 mg/mL concentration to the print surface and incubate in the sealed humidification chamber covered with foil while rotating gently for 1 h 7. After the labeling step, wash the slides with four dips in PBS/0.05% Tween-20, four dips in 1× PBS, and lastly a 3 × 3 ddH2O wash. This final wash step includes three separate dishes with ddH2O, and the slides are dipped three times in each dish 8. Spin dry in centrifuge at 20°C for 5  min at 200 rcf (same drying protocol as after blocking step) or with a gentle stream of ultra high purity Nitrogen gas

Production and Application of Glycan Microarrays 3.5.3. For Serum

277

1. Dilute the serum in incubation buffer composed of 1× PBS, 3% BSA, and 0.01% Tween-20. Various dilutions can be incubated on the arrays, or even pure serum with no incubation buffer can be used. However, more concentrated serum has a lesser specificity than that which is diluted. This is something that can be tested with a series of dilutions on a group of slides 2. Pipette 1  mL of the sample onto the print surface and incubate in the sealed humidification chamber while rotating gently for 1 h. Take care that the speed is not too fast and the liquid on the slide surface does not spill 3. At the end of the first incubation period, discard sample by simply pouring it off the slide into the bottom of the humidification chamber 4. Wash the slides one at a time by holding the edges and dipping them four times in a mixture of PBS and 0.05% Tween, then four times in 1× PBS, and finally four times in ddH2O 5. Dry off the backside of the slide with a Kimwipe and place back in the humidification chamber 6. Pipette 1 mL of a biotinylated secondary antibody onto the print surface and incubate in the sealed humidification chamber while rotating gently for 1 h. The secondary antibody is usually diluted to 10  mg/mL from its stock solution in PBS/3% BSA/0.05% Tween-20 incubation buffer, but this is also something that can be varied and tested for optimal results. Take care that the speed is not too fast and the liquid on the slide surface does not spill 7. At the end of the second incubation period, discard sample by again simply pouring it off the slide into the bottom of the humidification chamber 8. Add 1  mL of Strepdavidin at a 2  mg/mL concentration in PBS/3% BSA/0.05% Tween-20 incubation buffer to the print surface and incubate in the sealed humidification chamber covered with foil while rotating gently for 1 h 9. After the labeling step, wash the slides with four dips in PBS/0.05% Tween-20, four dips in 1× PBS, and lastly a 3 × 3 ddH2O wash. This final wash step includes three separate dishes with ddH2O, and the slides are dipped three times in each dish 10. Spin dry in centrifuge at 20°C for 5 min at 200 rcf (same drying protocol as after blocking step) or with a gentle stream of ultra high purity Nitrogen gas

3.5.4. For Viruses

1. Dilute the virus in incubation buffer composed of 1× PBS and 3% BSA. Various dilutions resulting in different HA

278

Busch, McBride, and Head

concentrations can be used on the arrays. This is something that can be tested with a series of dilutions on a group of slides (see Note 8) 2. Pipette 1 mL of the sample onto the print surface and incubate in the sealed humidification chamber while rotating gently for 1 h. Take care that the speed is not too fast and the liquid on the slide surface does not spill 3. At the end of the first incubation period, discard sample by simply pouring it off the slide into the bottom of the humidification chamber 4. Wash the slides one at a time by holding the edges and dipping them four times in a mixture of PBS and 0.05% Tween and then four times in 1× PBS 5. Dry off the backside of the slide with a Kimwipe and return it to the humidification chamber 6. Pipette 1  mL of antisera specific to the virus diluted in the PBS/3% BSA incubation buffer onto the print surface and incubate in the sealed humidification chamber while rotating gently for 1 h. Take care that the speed is not too fast and the liquid on the slide surface does not spill. The antisera is typically incubated at a 1:1,000 dilution, however, this can be varied and tested for optimal results. Take care that the speed is not too fast and the liquid on the slide surface does not spill 7. At the end of the second incubation period, discard sample by again simply pouring it off the slide into the bottom of the humidification chamber 8. Wash the slides one at a time by holding the edges and dipping them four times in a mixture of PBS and 0.05% Tween and then four times in 1× PBS 9. Pipette 1  mL of a biotinylated secondary antibody onto the print surface and incubate in the sealed humidification chamber while rotating gently for 30  min. The secondary antibody is usually diluted to 10  mg/mL from its stock solution in PBS/3% BSA incubation buffer, but this is also something that can be varied and tested for optimal results. Take care that the speed is not too fast and the liquid on the slide surface does not spill 10. At the end of the third incubation period, discard sample by again simply pouring it off the slide into the bottom of the humidification chamber 11. Add 1  mL of Streptavidin at a 2  mg/mL concentration in PBS/3% BSA incubation buffer to the print surface and incubate in the sealed humidification chamber covered with foil while rotating gently for 30 min

Production and Application of Glycan Microarrays

279

12. After the labeling step, wash the slides with four dips in PBS/0.05% Tween-20, four dips in 1× PBS, and lastly a 3 × 3 ddH2O wash. This final wash step includes three separate dishes with ddH2O, and the slides are dipped three times in each dish 13. Spin dry in centrifuge at 20°C for 5 min at 200 rcf (same drying protocol as after blocking step) or with a gentle stream of ultra high purity Nitrogen gas. 3.6. Scanning and Data Analysis

The glycan arrays can be scanned at different qualities and powers. By varying the PMT, laser power, and resolution, the instrument can produce an image with as many signals as possible with its dynamic range. Our lab uses a Perkin Elmer slide scanner with a 20-slide auto loader and normally scan at a fixed laser power while varying the PMT. It is assumed that the reader is familiar with creating a GeneID or map file, made by the arrayer or by hand in a tab-delimited spreadsheet. 1. After the arrays have been incubated, they are ready to be immediately scanned. Regardless of what type of scanner being used, take care to make sure the print surface is facing the laser source 2. Using the scanner software, set up the instrument to scan with the desired parameters. The settings can be tested and optimized, but keep in mind that each scan lowers the fluorescence of the signals. The images should be saved to a known location in the form of a TIFF file (see Fig. 2) (see Note 7) 3. Once the slides have finished scanning, transfer the images to a computer with the image processing program installed. ImaGene (Biodiscovery Inc.) is the software most often used in our lab;however, there are a variety of other acceptable programs available. Using a previously constructed GeneID or map file and customized grids, measure the array and save the data results. The placement of a sample marker known to react upon analysis is important in order to accurately place the grid file. Sample markers might include: a fluorophore like GFP; fluorescently labeled antibody; or a detectable tag like His. Our lab prints NHS-biotin, diluted to 100  nM in printing buffer, at the corner of each subarray. This allows the correct placement of the analysis grid by visualization with labeled streptavidin 4. The output file, usually in a text format, can then be opened in Microsoft Excel. With the use of a multistep macro, the data can be quickly sorted, organized, and graphed as desired for all slides of the same print format (see Fig. 3)

280

Busch, McBride, and Head

Fig. 2. An image, as it appears in ImaGene version 6.1 software, of a glycan array printed for the CFG that was incubated for quality control according to the given procedure with a cocktail of plant lectins. The slide was scanned at 60% PMT and 80% Laser Power at 10 mm resolution. The plant lectins used are from Vector and include AAL, ACL, BPL, ConA, GS-I, Jac, LEL, LTL, MAA, PTL-I, RCA-I, SBA, SJA, SNA, STL, WFA, and WGA

Production and Application of Glycan Microarrays

281

Fig. 3. The output file produced by a macro run in Excel for a quality control slide incubated with plant lectin cocktail

4. Notes 1. Only ultra high purity Nitrogen gas must be used on slides, pins, and any other pieces of equipment involved in printing. If a lower quality gas is used, the oils and debris will cause endless problems with pins sticking and printing to be defective 2. Always use gloves when handling slides, pins, and any part of the arrayer. Oils from the skin can have the same negative effect as dirty gas tanks when in contact with printing materials. It is even recommended to wash gloved hands with soap and water before handling or cleaning the pins to get rid of any powder or particles that could clog the tips 3. Even if the pins in the arrayer are clean, be sure to add at least 2 wash cycles before the first pick up of a print. This ensures no dust is in the tips, and it has been found that wetting and drying the pins briefly right before printing improved the quality of the first set of spots printed 4. Do not use Sharpie© or other regular permanent marker when marking and numbering slides. While the product description

282

Busch, McBride, and Head

may claim permanence, the ink will gradually soak off in the blocking buffer resulting in a mess and wasted slides. VWR lab markers have shown to withstand the buffer for such an extended amount of time 5. If during incubation, the wrong side of the slide is accidently wiped, the array is not necessarily destroyed. The stability of slide surface and the covalent bond formed between the slide surface and the glycans is sufficiently robust, and the print will most likely be fine except for a few minor smears 6. Attach a piece of plastic hose to the Nitrogen gas tank and insert a 1,000 mL pipette tip inside the end of tubing. Snip the tip off to create a larger opening, creating a precise and easy to use way of cleaning off slides and drying pins 7. If slides cannot be scanned immediately after incubation, they should be stored in a dark place such as a cabinet. This is not recommended because of the possible loss of fluorescence, however, sometimes it is necessary if the scanner is being used or breaks down 8. Do not use Tween-20 or other detergents in incubation buffers for viruses. The virus will break down and not bind as effectively causing less than desirable results

Acknowledgments The authors would like to thank Ola Blixt, James Paulson, Nahid Razi, and Julia Hoffmann for all of their help and guidance. Andrew Hemingway and all of Schott/Nexterion for outstanding product support. Special thanks to the Consortium for Funtional Glycomics (CFG) for making everything possible. References 1. Blixt O, Head S, Mondala T, Scanlan C, Huflejt ME, Alvarez R, Bryan MC, Fazio F, Calarese D, Stevens J, Razi N, Stevens DJ, Skehel JJ, van Die I, Burton DR, Wilson IA, Cummings R, Bovin N, Wong C-H, Paulson JC (2004) Printed covalent glycan array for ligand profiling of diverse glycan binding proteins. Proc Natl Acad Sci USA 101:17033–17038 2. Stevens J, Blixt O, Paulson JC, Wilson IA (2006) Glycan microarray technologies: tools

to survey host specificity of influenza viruses. Nat Rev Microbiol 4:857–864 3. Stevens J, Blixt O, Glaser L, Taubenberger JK, Palese P, Paulson JC, Wilson IA (2006) Glycan microarray analysis of the hemagglutinins from modern and pandemic influenza viruses reveals different receptor specificities. J Mol Biol 355:1143–1155

Index A

F

Absorption, distribution, metabolism, and elimination (ADME)...................100–102, 118 Adverse drug reaction..................................................... 100 Affymetrix GeneChip............... 46, 48, 58, 64–65, 115, 161 Agilent bioanalyzer.................................................... 13, 15, 17, 19, 29, 33, 37–39, 53–54, 57–59, 64, 66, 71, 161, 166 Allele-specific extension (ASE)............................ 75, 77–78 Alternate splicing....................................................... 63–72 Apurinic/apyrimidinic endonuclease........................ 68, 167 Automated quantitative protein expression (AQUA) analysis.........................241–242, 245–246

Feeder-independent ESC culture......................... 47–51, 60 Flourescence activated cell sorting.............................. 27–43 Flow cytometer................................................28, 30, 32, 39 Fragmentation.......................................... 38, 48, 54, 57, 59, 68, 105, 111–115, 122, 130–131, 167, 175–176, 178–180, 182, 191–200

G

Bacterial artificial chromosomes (BACs)........................ 125

Gene regulation...................................................... 102, 174 Genomics..................................... 1–3, 9, 16–17, 23, 38, 47, 69, 87, 102–103, 105–110, 118, 121, 126–132, 137, 161, 174, 192–193, 195, 198, 224, 227, 241 Genotyping......................... 9, 100, 105–106, 116–121, 123 Globin reduction.................................................. 13–15, 25 Glycan arrays........................... 269, 271–272, 275, 279–280

C

H

Carbohydrate.................................................................... 71 Cell differentiation........................................................... 74 Cells-to-CT.......................................................... 91–92, 94 cell surface markers..................................................... 28, 39 Chromatin immunoprecipitation............179, 181, 186–188 Combinatorial synthesis......................................... 204–211 Comparative genomic hybridization.............................. 126 Copy variation................................................................ 118 CpG island..............................................142, 174, 188, 192 Cytogenetics........................................................... 125–139

Haplotype............................................................... 118–119 Health Insurance Portability and Accountability Act (HIPAA)...................................................... 4–5 High throughput screening.....................203–218, 221–236 HpaII tiny fragment enrichment by ligation-mediated PCR (HELP) assay.................................... 191–200 Hypermethylation.......................................................... 192 Hypomethylation........................................................... 192

D

Illumina beadchip....................................................... 73–86 Immunoprecipitation......................................161, 163–167, 169–170, 176–188 In-situ hybridization....................................................... 240 Integrins..................................................................... 63–64 Isotype-control................................................................. 30

B

DASL assay...................................................................... 75 DataChip.........................................222–223, 227, 230–233 DNA methyaltion...................................173–174, 191–200 Drug metabolism.....................................100–101, 222, 228 Drug metabolizing enzymes and transporters (DMET) assay.................................................... 119

E Embryoid bodies........................................................ 45–60 Embryonic stem cells.................................................. 45–60 Endogenous control................................................... 90–91 Epigenomics................................................................... 192 Exon splicing.............................................................. 70–71 5¢-Exonuclease assay......................................................... 88 Expression profiling........................... 27, 46, 69, 73–86, 192

I

L Lectin..............................................269, 275–276, 280–281 Leukemia........................................................................ 192 Ligase detection reaction (LDR)............................ 141–156 LM-PCR....................................................................... 192

M Melanoma.................................................................. 27–43 b-Mercaptoethanol...................6–7, 16, 19, 29, 47, 186, 194

283

Microarray Methods for Drug Discovery 284  Index

  

Metabolic enzyme.................................................. 221–236 MetaChip................................................222–223, 227–233 Metalloprotease...................................................... 203–218 Microarrays.................................... 1–25, 27, 32, 42, 45–60, 64–69, 71, 73–74, 83, 87–89, 99–123, 125–139, 142, 144–145, 148, 160–162, 167–168, 174, 177–178, 182, 185–186, 192, 197–198, 203–218, 221–236, 239–249, 253, 255, 262, 266, 269–282 microRNA.................................................................. 73–86 Molecular inversion probe (MIP)............105, 108–111, 122 Mononuclear cells............................... 6–7, 9, 15–18, 23–24 Multicenter clinical study................................................... 2 Multiplexed detection..................................................... 142 Multiplex polymerase chain reaction (mPCR)................ 95, 102, 105–109, 121, 142, 155–156

Ribominus.......................................................... 65–66, 170 Ribonomics.................................................................... 161 Ribonucleases (RNases).............................................. 28, 41 RIP-Chip............................................................... 159–170 RNA-binding protein (RBP)..................160–167, 169–170 RNaseAlert................................................................. 29, 39 RNaseZap.......................................................28–29, 34, 39 RNA viruses........................................................... 144–145

S

NanoDrop.................................... 13, 15, 17, 19–20, 23–24, 29, 33, 36–38, 43, 64, 66, 71, 128–129, 161, 166 Nucleic acid extraction................................................... 149

Selectivity profiling................................................. 251–266 shRNA............................................................................. 64 Single nucleotide polymorphisms (SNPs)................... 9, 47, 101, 105, 109–110, 117–120, 193 Slide printing..................................................147, 213–214, 218, 227, 255, 271, 273, 275–276 Small molecule arrays......................................252, 256, 259 Specificity profiling................................................ 251–266 Streptavidin magnetic beads....................................... 13–14 Streptavidin-phycoerythrin.................................49, 69, 113

O

T

Oligonucleotide-based................................................... 126 Oropharyngeal squamous cell cancer.............................. 240

TaqMan Array Card................................................... 87–96 Tissue microarray................................................... 239–249 Toxicity............................................................100, 221–236 Transcription factors........................173–174, 178, 186–187 Transcriptome.....................................................65, 87, 160 Transfection, Transplantation.................................................1–5, 8–9, 23

N

P PAXgene Blood RNA tube...........................6–8, 11–15, 25 PCR amplification............................................90, 102, 105, 142, 144, 146, 149, 155, 182–183, 187 Personalized medicine.............................................. 99–123 Pharmacogenetics............................................. 99–100, 119 Pharmacogenomics............................................. 91, 99–123 Poly-A polymerase (PAP).................................. 75–77, 271 Post-transcriptional........................................................ 159 Protein extraction................................................. 6–7, 9–24 Protein microarray.................................................. 251–266

U Uracil DNA glycolase (UDG).......................68, 75, 79, 167

W Whole blood.......................... 3, 6–9, 11, 13–15, 21–23, 127 Whole transcript (WT).................................................. 159

Q

Z

QPCR............................32, 42, 66, 87–90, 93–95, 243, 247 immunohistochemistry................................................... 239

Zip-codes oligonucleotides.............. 142, 144, 147, 151–155

R Radiofrequency....................................................... 206, 209 Reverse transcription...............................68, 71, 92–95, 145