Protein Microarrays for Validation of Affinity Binders

Home , Antibody microarray, Protein microarray

Protein microarrays for validation of affinity binders

MÅRTEN SUNDBERG

KTH - Royal Institute of Technology

School of Biotechnology

Stockholm

KTH - Royal Institute of Technology School of Biotechnology Science for Life Laboratory SE-171 21 Solna Sweden

Printed By: AJ E-print AB Oxtorgsgatan 9 111 57 Stockholm Sweden

ISBN 978-91-7501-149-3 TRITA-BIO Report 2011:23 ISSN 1654-2312

Mårten Sundberg (2011): Protein microarrays for validation of affinity binders, School of Biotechnology, Royal Institute of Technology (KTH), Stockholm, Sweden. ISBN 978-91-7501-149-3

Abstract

Is specificity an important issue regarding affinity reagents? What about the validation of affinity reagents today, is it good enough? This depends on the application and the producer of the reagent. Validation should be the most important marketing argument that can be found. Today there is a continuous growth of both the number of affinity reagents that are produced and the different types of affinity reagents that are developed. In proteomics they become more and more important in exploring the human proteome. Therefore, validated affinity reagents should be on top of every proteomic researcher’s list. How should this be accomplished? Better international agreements on how affinity reagents should be tested to be regarded as functional reagents are needed. One of the most important issues is the specificity of the affinity reagent. An international standard for which specific validation that is needed for different kinds of applications would be very useful. In this thesis, it is shown that the protein microarray platform that was established within the HPA project at KTH is a very good tool to determine the specificity of different affinity binders. In the first study, the production of mono-specific antibodies for tissue profiling in the Human Protein Atlas (HPA) project is presented. The section describing the use of protein microarrays for validation of the antibodies is relevant for this thesis. The implementation of protein microarrays in the HPA workflow was an important addition, because a deeper insight of the specificity of all the antibodies produced were now available. In a second study, bead based arrays were compared to planar protein microarrays used in the HPA project. In this study, 100 different bead identities were coupled with 100 different antigens and mixed together to generate an array. The correlation between the two types of assays was very high and the conclusion was that the methods can be used as backup to each other. A third study was a part of an international initiative to produce renewable affinity binders against proteins containing SH2 domain. Here, the HPA protein microarrays were modified to analyze different types of reagents produced at six laboratories around the world. Monoclonal antibodies, single chain fragment and fibronectin scaffolds were tested as well as mono-specific antibodies. It was shown to be possible to adapt protein microarrays used in the HPA project to validate other kinds of affinity reagents.

Keywords: Microarray, protein, antibody, antigen, affinity, validation

List of publications

This thesis is based upon the following three papers. The papers are included in the Appendix.

I. Nilsson, P., Paavilainen, L., Larsson, K., Odling, J., Sundberg, M., Andersson, A. C., Kampf, C., Persson, A., Al-Khalili Szigyarto, C., Ottosson, J., Bjorling, E., Hober, S., Wernerus, H., Wester, K., Ponten, F., and Uhlen, M. (2005) Towards a human proteome atlas: high-throughput generation of mono-specific antibodies for tissue profiling, Proteomics 5, 4327-4337. II. Schwenk, J. M., Lindberg, J., Sundberg, M., Uhlen, M., and Nilsson, P. (2007) Determination of binding specificities in highly multiplexed bead-based assays for antibody proteomics, Mol Cell Proteomics 6, 125-13. III. Sjoberg, R*., Sundberg, M*., Gundberg, A., Sivertsson, A., Schwenk, J, M., Uhlen, M., Nilsson, P (2011) Validation of various affinity reagents using protein microarrays. Manuscript submitted to New Biotechnology.

All papers are reproduced with permission from the copyright holders

*These authors contributed equally to this work.

Related publications

Colwill, K., Persson, H., Jarvik, N. E., Wyrzucki, A., Wojcik, J., Koide, A., Kossiakoff, A. A., Koide, S., Sidhu, S., Dyson, M. R., Pershad, K., Pavlovic, J. D., Karatt-Vellatt, A., Schofield, D. J., Kay, B. K., McCafferty, J., Mersmann, M., Meier, D., Mersmann, J., Helmsing, S., Hust, M., Dubel, S., Berkowicz, S., Freemantle, A., Spiegel, M., Sawyer, A., Layton, D., Nice, E., Dai, A., Rocks, O., Williton, K., Fellouse, F. A., Hersi, K., Pawson, T., Nilsson, P., Sundberg, M., Sjoberg, R., Sivertsson, A., Schwenk, J. M., Takanen, J. O., Hober, S., Uhlen, M., Dahlgren, L. G., Flores, A., Johansson, I., Weigelt, J., Crombet, L., Loppnau, P., Kozieradzki, I., Cossar, D., Arrowsmith, C. H., Edwards, A. M., and Graslund, S. (2011) A roadmap to generate renewable protein binders to the human proteome, Nat Methods 8, 551- 558.

Rimini, R., Schwenk, J. M., Sundberg, M., Sjoberg, R., Klevebring, D., Gry, M., Uhlen, M., and Nilsson, P. (2009) Validation of serum protein profiles by a dual antibody array approach, J Proteomics 73, 252-266.

Lundberg, E., Sundberg, M., Graslund, T., Uhlen, M., and Svahn, H. A. (2007) A novel method for reproducible fluorescent labeling of small amounts of antibodies on solid phase, J Immunol Methods 322, 40-49.

. Eriksson, C., Agaton, C., Kange, R., Sundberg, M., Nilsson, P., Ek, B., Uhlen, M., Gustafsson, M., and Hober, S. (2006) Microfluidic analysis of antibody specificity in a compact disk format, J Proteome Res 5, 1568-1574.

Uhlen, M., Bjorling, E., Agaton, C., Szigyarto, C. A., Amini, B., Andersen, E., Andersson, A. C., Angelidou, P., Asplund, A., Asplund, C., Berglund, L., Bergstrom, K., Brumer, H., Cerjan, D., Ekstrom, M., Elobeid, A., Eriksson, C., Fagerberg, L., Falk, R., Fall, J., Forsberg, M., Bjorklund, M. G., Gumbel, K., Halimi, A., Hallin, I., Hamsten, C., Hansson, M., Hedhammar, M., Hercules, G., Kampf, C., Larsson, K., Lindskog, M., Lodewyckx, W., Lund, J., Lundeberg, J., Magnusson, K., Malm, E., Nilsson, P., Odling, J., Oksvold, P., Olsson, I., Oster, E., Ottosson, J., Paavilainen, L., Persson, A., Rimini, R., Rockberg, J., Runeson, M., Sivertsson, A., Skollermo, A., Steen, J., Stenvall, M., Sterky, F., Stromberg, S., Sundberg, M., Tegel, H., Tourle, S., Wahlund, E., Walden, A., Wan, J., Wernerus, H., Westberg, J., Wester, K., Wrethagen, U., Xu, L. L., Hober, S., and Ponten, F. (2005) A human protein atlas for normal and cancer tissues based on antibody proteomics, Mol Cell Proteomics 4, 1920-1932.

Janzi, M., Odling, J., Pan-Hammarstrom, Q., Sundberg, M., Lundeberg, J., Uhlen, M., Hammarstrom, L., and Nilsson, P. (2005) Serum microarrays for large scale screening of protein levels, Mol Cell Proteomics 4, 1942-1947.

Contents

Introduction ...... 1 Human Protein Atlas project ...... 2 Affinity reagents used in protein microarrays ...... 3 Antigens ...... 3 Antibodies ...... 4 Polyclonal antibodies ...... 4 Monoclonal antibodies ...... 5 Antibody fragments ...... 6 Other kinds of affinity reagents ...... 6 Protein microarrays - production and application ...... 7 Different approaches of protein microarrays ...... 7 Antigen microarray ...... 7 Antibody microarray ...... 8 Peptide microarray ...... 8 Reverse-phase microarrays ...... 8 Production of protein microarrays ...... 9 Printing robots ...... 9 Surfaces and immobilization ...... 10 Blocking of substrate ...... 11 Sample labeling and detection systems ...... 12 Incubation, scanning and data evaluation ...... 12 Bead based protein microarrays ...... 13 Present investigation ...... 14 Validated affinity reagents – an important factor in today’s biotechnology . 15 Paper I ...... 15 Paper II ...... 16 Paper III ...... 18 Conclusions ...... 19 Abbreviations ...... 20 Acknowledgments ...... 21 References ...... 23

Mårten Sundberg

Introduction

The human body is built by billions of cells. The genetic code is the basis of all living cells and the information stored in the code controls all activities of the living cells. The genetic code contain billions of nucleotides folding into a double helix spiral, the deoxyribonucleic acid, DNA (1). DNA is the template for proteins, which are the building blocks in the human body and involved in most of its processes. Today it is approximated that the human proteome is based on about 21.000 protein coding genes (2). Many of these are yet of unknown function and some of them might be key targets in the medicinal products that will be produced tomorrow. One of the challenges is to characterize these proteins to generate more knowledge about their involvement both in the processes in a healthy human body and in one that suffer from a disease. Proteomics is a fast growing field and the concept proteome was defined by Marc Wilkins during his doctoral studies in the 1990s (3, 4) as a way to classify the study of the proteins of humans and other species. This area of research includes a large number of different techniques that can be used to find proteins in different samples. One of the major problems in the proteomic research field is the large dynamic range that is found in e.g. the plasma proteome (5). There is a need to establish sensitive techniques that can detect low abundant proteins. Regarding affinity based methods, there must be affinity reagents available that can be relied on when it comes to their specificity. One effort to make it easier for researchers to find the reagents they need and to provide as much validation data as possible about antibodies and antigens, is an internet based portal, called Antibodypedia.org (6). This database is publically available for anyone who search for a specific reagent or wants to contribute with additional validation information to commercially available reagents. There are two major areas within proteomic. First of all there are techniques that do not use affinity reagents for the detection such as mass spectrometry (MS) and gel electrophoresis (7). Then there are techniques that use affinity reagents for the detection of proteins. This group includes e.g. ELISA, Western blot, RIA, SPR, IHC, IF and different kinds of protein microarrays. All these methods have in common that they have been established to obtain more information about the proteins that e.g. the human body is built of (7, 8). Some of them are used both as research tools and for routine analysis at hospitals, e.g. ELISA. Two of the most promising methods to acquire information about the human proteome would be MS and antibody arrays. If used complementary they can provide new insight in the world of proteins, but there is a lot more work that has to be done both concerning the sensitivity and the throughput before a whole proteome can be covered. The Human Protein Atlas (HPA) project, founded 2003, is a large scale effort to produce antibodies against all human proteins (2, 9, 10). The aim of the project is to produce at least one mono-specific antibody directed against each and every one of the 21.000 proteins. This project would not have been possible without the huge effort that was first presented some ten years ago, by the sequencing of the human genome (11, 12). All the work

Protein microarrays for validation of affinity binder done in the project is based on the human genome sequence and the predictions that are made on which part of the genome actually will be expressed as a protein. All this information is now accessible through web-based databases such as, UniProt and Ensembl (13, 14). Without this information, it would not be possible to create the proteins that are templates for the antibodies, which are used to generate protein profiles in human tissues and cells. When the proteins are produced, affinity binders that recognize those specific proteins can be manufactured as in the HPA project. Today, this is done at large scale and there are more than 500.000 antibodies (15) produced commercially that can be used by researchers. This is a fantastic resource for researchers as tools to increase the knowledge of the human body, if the reagents are well characterized, which is not always the case (16). This is one of the key problems of today’s proteomic research. What kind of validation is carried out on the antibody and is the quality of that antibody good with the validation the company provide? In contrast to the situation for approval of new medicinal products there is today no generally accepted standard for the validation of antibodies and there is no authority that has the responsibility to ensure the quality of antibodies on the market (8, 17).

Human Protein Atlas project

To be able to reach the goal to produce antibodies against all proteins in the human body, a large scale production has been set up to produce recombinant protein fragments that are named Protein Signature Epitope Tag (PrEST) in the HPA project. To select a region of the full length protein, that is unique for that protein, an in house developed software is used. This web based program for antigen selection and visualization is based on data from Ensembl. In addition there is also built in functions for prediction of transmembrane regions and predictions of solubility (13, 18). In the software a unique fragment of the protein, consisting of about 25-150 amino acids, selected to have as low homology as possible to other protein sequences in the human body, to avoid cross-reactivity. With this as a template, primers are designed in the software and the PrEST is produced with RT-PCR with RNA as template. The sequence is ligated into a vector that contains a hexa-histidine albumin binding protein (His6ABP) gene sequence. The resulting His6 tag on the protein is used for the purification of the PrESTs before immunization (19) and the ABP is used to enhance the immune response (20). The vector construct is transformed into E. coli. The sequence of the clones is verified by Sanger sequencing (21). After that, the E.coli colonies are cultured, the protein expression is induced and the resulting PrESTs are collected and purified (22, 23). The PrESTs are immunized in rabbits and the polyclonal serum is collected. The first step in the purification procedure of the serum is a depletion column used to remove antibodies directed against the His6-ABP tag. The serum is affinity purified on a second column with the corresponding PrEST coupled (24). This process results in a mono-specific antibody mix, directed against different epitopes on the PrEST (21). The specificity of the antibodies is then verified by microarrays were the PrEST is used as target agent (24). The antibodies are

Mårten Sundberg also tested on Western blot. In Immunohistochemistry (IHC) the localization of the corresponding protein is explored (24). To get an even better view of the localization immunofluorescence (IF) is used for the subcellular localization (25, 26).

Affinity reagents used in protein microarrays

There are many types of molecules that can be used for protein microarrays such as antibodies, recombinant protein, antibody fragment, different kinds of scaffolds and others (27). In what context the protein microarray is going to be used determine what kind of capture reagents that are used. In this thesis the focus is put on the antigen protein microarray as a tool for validation of produced antibodies or other affinity reagents. In antigen microarrays the capture reagent is the antigen that the affinity reagent is produced against. Antigen microarrays can also be used for screening for autoantibodies in different body fluids. Another way to use the affinity reagents is to print the reagents and use them as capturing agents, such as antibody arrays used for protein detection in different kinds of body fluids.

Antigens

An antigen is usually described as a substance foreign to the host organism. The substance will trigger an immune response that produces antibodies targeting that substance. The region the antibody recognizes on an antigen is called an epitope. Epitopes can be either linear or conformational and it is important that the antibody can reach the epitope to be able to bind. This can change due to environmental changes causing e.g. denaturation. Depending on the ability of the affinity reagent to recognize the epitope in a native or denatured state, this can either be an advantage or a disadvantage. Antigens can also be polysaccharides, lipids, nucleic acids or proteins, which are the most common to produce antibodies used for research. When antibodies are produced in host animals there are usually three different types of antigen used: recombinant full-length proteins, other recombinant proteins and small peptides. Working with recombinant full-length proteins can be both laborious and time consuming due to the fact that it is more complicated to produce these kinds of proteins. To be able to ensure a native folding, they should be expressed in a system which makes it feasible e.g. mammalian cells (28). Native proteins can also be use but it can be laborious to extract them from human tissue or body fluids. If peptides are used as antigens, a length of about 10-20 amino acids are usually used to have functional linear epitopes. It is assumed that a peptide shorter than approximately 7 amino acids is insufficient to function as an epitope and if they are longer they could fold into a structure that is not found in the native protein (29). Peptides have also successfully been used for epitope mapping of proteins. It was shown that a 10 amino acids long peptide

Protein microarrays for validation of affinity binder was not enough to map the epitope but when 15 amino acids long peptides were used the epitope mapping was successful (30). Recombinant proteins that are not full length proteins are a commonly used as antigens because they can be produced in large amounts, in e.g. E.coli., which can be used in the downstream process of the affinity reagent production. There are examples of both monoclonal, polyclonal and antibody fragments produced with recombinant proteins as antigen (24, 31-34).

Antibodies

Antibodies are important components of the adaptive immune system. Antibodies circulates the body and have a function to eliminate microbes in the extra cellular fluids, but also to activate other components of the adaptive immune system (35). Antibodies are glycoproteins secreted by specialized B lymphocytes, called plasma cells. They are also referred to as immunoglobulins (Ig) because they contain a common structural domain. Antibodies are built-up by four polypeptides, two identical copies of a heavy chain and two identical copies of a light chain which are hold together with disulfide and noncovalent bonds. This structure is usually schematically presented like a Y-shaped molecule. Based on functional differences, antibodies can be further divided in two parts, the Fc–region and the Fab-region. The Fab domain is the arms of the Y-shape and contains the antigen binding sites of the antibody, that region is also referred to as the Fv region. The base of the Y is called the Fc region and the constant region of the Fc region is involved in binding between effector molecules in the complement system and more important Fc receptors that are present on phagocytes and other immune system molecules (36). Each variable region of the heavy chain and the light chain contains hyper variable regions, called CDRs. These regions make it possible for the antibodies to adapt to almost any antigen that can appear in the body. In mammals there are five subclasses of Ig that are classified depending on the conformation of the heavy chains (IgG , IgA, IgE, IgM and IgD) which all have their function in the immune system and can be used for different purposes in research. For research there are different types of antibodies to choose from e.g. polyclonal, monoclonal and mono-specific antibodies but there are also antibody fragments. Each of them has its strength and weakness but it can be useful in different applications.

Polyclonal antibodies

Polyclonal antibodies are generally produced by immunizing an antigen to a host animal, usually mouse, rabbit or goat. The antigen is recognized as a foreign substance by the immune system of the host animal and the immune system starts to produce antibodies by specific B-cells. An antigen normally has several different epitopes that are recognized by the immune system and each specific B-cell will produce an antibody directed to one of the epitopes. This means that when polyclonal antibodies are produced there will be a whole panel of different antibodies that recognize different parts of the antigen with different

Mårten Sundberg affinity. This is both an advantage and a disadvantage for the polyclonal immune response when antibodies are used as a capturing tool. It is an advantage that there are different epitopes that are recognized by polyclonal antibodies. This can be used in protein microarrays due to the fact that the apparent affinity for polyclonal antibodies is higher because there are more binding sites and this will enhance the limit of detection (37). The pool of polyclonal antibodies might be less sensitive to changes in the environment because if one of the epitopes are destroyed there are others that could be detected by other antibody fractions of the pool (38). The potential that lies in the polyclonal antibodies to recognize different types of epitope, both linear and structural epitopes (39) makes them useful in many kinds of application (40-46). Another advantage that makes the polyclonal antibodies so useful is that they are relatively easy to produce compared to other types of binders. Already after a couple of months after the first immunization, it is often possible to collect the antibodies from the animal (38). The recognition of different epitopes could in some cases be a disadvantage because it is not known exactly which part of e.g. a protein that is the target and it could be hard to distinguish unspecific binding from the correct ones. This means that it is currently not possible to use polyclonal antibodies for in vivo diagnostics, but they are used much in in vitro diagnostics. There is also a reproducibility issue because there will be different pools of antibodies produced between different immunizations even in the same animal. This means that there has to be more batch to batch validation done on polyclonal antibodies than on a renewable resource of affinity reagent. There are strategies that can be used to pull out the specific antibodies from a polyclonal serum, because typically these antibodies is found in concentrations from 50 to 200 µg/mL (38). This is done by purification of the polyclonal serum on a column coupled with the antigen that was used for the immunization. Compared to monoclonal antibodies there will still be problems with reproducibility, since there is a need to produce new antibodies and their specificity may differ.

Monoclonal antibodies

Monoclonal antibodies are often the first choice when it comes to clinical diagnostics and research today due to the fact that they bind with one specificity to their targets. The principle is that an antigen is immunized into a host animal, usually a mouse. So far, it is the same as the production of polyclonal antibodies, but in 1975 Köhler and Milstein published a paper where they describe how they fused activated, antibody producing cells with myeloma cells (47). The hybridoma that is the outcome of this fusion is an immortal cell line that produces identical monoclonal antibodies. One advantages of this technology is that there will be identical antibodies that can be produced rapidly and reproducibly recognizing the same epitope. This can however be a disadvantage if the epitope that the antibodies are directed against is altered in a changing environment. Compared to polyclonal antibodies, the production of monoclonal antibodies is labor-intense and it can take more than a year to produce a functional monoclonal antibody. On the other hand there is a renewable resource when the cell line is established. More about the differences between monoclonal and

Protein microarrays for validation of affinity binder polyclonal antibodies can be found in Lipman et al. (38). There has been efforts to enhance the throughput of monoclonal antibodies by using robotics and antigen microarrays for screening (32).

Antibody fragments

Instead of using the whole antibodies, fragments, like single-chain variable (scFv) and fragment antigen binding (Fab), can be used. The Fab fragment was found as a product after enzymatic cleavage of the whole antibody with papain and pepsin and got the name simply because it was the antigen binging fragment (48). Single-chain variables simply consist of the variable regions of the heavy and light chain and are coupled tighter with a linker (49). In 1990 McCafferty et al published an article where they used phage display to display and select antibody fragments (50). Since then, there has been a lot effort on developing phage display and there are a growing number of large libraries that consist of random combination of antibody fragment genes (33, 34, 51, 52). In parallel with the further development of the phage display, other systems like ribosome display, cell surfaces display on Staphylococci and yeast displays have been developed and implemented in the generation of different kinds of affinity reagents (53, 54). Antibody fragments are also produced in E.coli and already in 1988 there were two articles published that descried how functional antibody fragments could be produced in E.coli (55, 56). This technology has since then been developed and is today widespread (57).

Other kinds of affinity reagents

There are an increasing number of affinity reagents without immunoglobulin as template. One of them is scaffolds called Affibody molecules. The origin of Affibody molecules is the B domain of protein A which is a Stapylococcus aureus protein and consists of 58 amino acids (58). By randomization in 13 of the surface exposed amino acids on two α-helices that are involved in the Fc-binding activity, libraries of binders can be produced and these can be used for screening of desired binding properties (59). Another one is DARPins that stands for Designed Ankyrin Repeat Proteins and is another example of a biotechnical designed scaffold. The original ankyrin repeat proteins are natural receptor proteins located mostly in cellular cytoplasm but are also present in the extracellular compartments (60, 61). Binders developed from the 10th fibronectin type III domain are named monobodies/ AdNectins. The fibronectin is a 94 amino acid large molecule with a β-sandwich structure that resembles the Ig structure. (62, 63) Anticalins are molecules that have been developed from lipocalins with combinatorial techniques. Lipocalins are a family of transport and storage proteins for vitamins, hormones and secondary metabolites with a central highly conserved β-barrel (64). Aptamers is another kind of scaffold that can be used as affinity reagent. Aptamers are synthetic single stranded DNA or RNA molecules with a unique 3D-structure that makes it

Mårten Sundberg possible to specifically bind target molecules (65). They can be amplified by PCR and through repeated cycles of amplification and screening be affinity matured (66). More about different scaffolds and their usability can be found in two review articles written by Nygren and Skerra (67, 68).

Protein microarrays - production and application

A microarray is an ordered structure of samples such as DNA, proteins or other sample of interest. This technique took off during the 1990s with the evolvement of the DNA microarrays (69, 70). The principles of miniaturized and parallelized immunoassay where already described in the late 1980s by Roger Ekins (71). Although, it was not until the end of the 1990s that the protein microarrays developed (72-75). At that time DNA microarrays and the tools needed to produce microarrays in a high-throughput fashion were established. The principle is to miniaturize and parallelize the assay to get as much information as possible from one run. Although, it was challenging to adapt to protein microarrays, since proteins are far more complex to work with than DNA. One of the biggest limitations was and still is the production of proteins used for the production of the protein arrays. Today protein microarrays are established and used in many different applications (53, 76). A lot of expectations are put into the technique as a diagnostic tool and a screening tool for new biomarkers and targets for new drugs.

Different approaches of protein microarrays

Antigen microarray

An antigen microarray consists of immobilized antigens on a solid support. The antigen can be recombinant proteins used for large scale screening or it can be allergens used for diagnosis of allergy (77). Antigen microarrays can be used as a validation tool of polyclonal antibodies (24, 78) or as a screening assay for human autoantibodies (79, 80). In Ayoglu et al. more about different antigen microarray applications can be found (53). Antigen microarrays can be a valuable tool for validation when new affinity reagents are produced. The specificity of an affinity binder is important, in order to be sure that the molecule that is detected is the wanted target. There is a growing need to get a deeper understanding of different autoimmune diseases and here antigen microarrays could be an important tool to give new information and new targets that could be used in both diagnosis and therapy of these diseases. There have been studies with antigen microarrays as a tool for identification of the immunodominant pathogen in different infectious diseases (81, 82).

Protein microarrays for validation of affinity binder

Antibody microarray

Antibody arrays can be produced with polyclonal, monoclonal, single chain variable (scFv), Fab, Affibody molecules or other kinds of affinity binders. Antibody microarrays are together with antigen microarrays the most widly used protein microarrays. There is an example were small size arrays are used in a diagnostic application. This application is used for diagnosis of acute coronary syndrome (ACS) (83). A limiting factor is the expense and lack of available antibodies and this is indicated when looking at the size of the produced and published studies with antibody arrays. It is usually from ten and up to a couple of hundred antibodies that are included. One of the larger antibody arrays were produced by Schröder et al. and contained 810 antibodies directed against 741 different proteins (84). There are already initiatives taken to use the arrays for large scale protein profiling, even if the one referred here is based on a bead based assay (85). This kind of initiative will open up a new field that can be used in parallel with e.g. mass spectrometry protein profiling. This will hopefully give many new biomarker candidates and potential therapeutic targets. There is a lot of potential in using antibody arrays both as a diagnostic tool, but also as a biotechnological tool.

Peptide microarray

Peptide microarrays are similar to antigen microarrays and could possibly be used more or less in the same applications. In the last years these kind of microarrays have been improved a lot and there is a potential to synthesize millions of peptides directly on a glass slide (86). This gives a new opportunity to have a whole proteome on the same slide. There are limitations in using peptide microarrays due to lower affinity, structural issue compared to the native or recombinant protein and that the binding regions might not be fully covered. Because of this, peptide microarrays are today used as a complement to antigen microarrays (87). Another promising application for peptide microarrays is epitope mapping of proteins and this might be a breakthrough for the peptide microarrays (30, 88).

Reverse-phase microarrays

A reverse-phase microarray is constructed by spotting a complex sample e.g. human tissue lysate or serum/plasma on a solid support. The strength of doing this is that thousands of patient samples can be screened in a single run at equal conditions (89). Due to the fact that the amount of sample deposit is small, there is not much patient sample needed for the analysis. On the other hand this is one of the drawbacks with these kinds of microarrays because the number of molecules of each protein in the sample will be very low, it might be below 100 molecules (53). In a study done by Janzi et al. with a serum microarray targeting IgA, the sensitivity was in the lower µg/ml range (89). This shows that the sensitivity today is

Mårten Sundberg not good enough for the low abundant proteins (concentrations below µg/ml) in these kinds of studies. Although it can still be of interest due to the fact that thousands of samples can be tested at the same time and for medium to high abundant proteins this could be useful. Still there are possibilities to enhance the signal and there have been some studies to do this (90). In the future there might be new amplification techniques that could be used to increase the sensitivity for reverse-phase microarrays and then they could be a very good tool for large scale screenings of large patient cohorts.

Production of protein microarrays

In protein microarrays, proteins are printed with a robot and immobilized on a solid surface. Microscope slides are usually used, with an adsorbent or a chemically activated surface. This implies that there is a small amount of protein solution that is deposited either by a drop or by physical contact with a steel pin. The amount of material that is deposited is usually 100- 400 pL. This will result in a spot size of about 100-300 micrometer on the solid support. Tens of thousands of spots can be placed within one slide and the possibility to print many different proteins in one slide is one of the advantages with planar protein microarrays. If the purpose is screening a lot of proteins with many different patient sample it is achievable, but the throughput will be rather low. The throughput can be improved if there is not more than about 1000 proteins/samples that are being included in the study. Then the slide can be divided in different sub arrays with up to 1000 proteins in each.

Printing robots

A very important component when producing protein microarrays is the printing robot, which is a device that very precisely delivers a small amount of the sample onto a solid support. There are two main methods of printing microarrays and those are, contact and non-contact printing. If contact printing is used there is a robot equipped with a steel pin that deposits the sample. The system can differ, either there is only a solid steel pin that prints the amount of sample that sticks to the pin or the sample can be aspirated by capillary forces into a compartment in a split pin. There is also one concept with a pin and a ring, where the ring is immersed into the sample. During a print run the printing steel pin pass through the ring and deliver sample down on the surface. The idea with the later ones is to pick up more volume to improve the amount of spots that can be produced. The contact printing has the advantage that there is always very good alignment of the different spots, but there are some limitations, especially when proteins are printed. When printing with a contact printer there are often more problems with doughnut effects, which means that there is a ring of the sample in the edge of the spot but not much sample in the middle. One limitation with the system is the fact that there cannot be any systems to control if sample is delivered

Protein microarrays for validation of affinity binder or not. One of the biggest problems for proteins is that there is quite high risk of contamination of other proteins if more than one is printed due to the stickiness of proteins. Proteins can be difficult to wash off due to adherence to the surface, especially if a split pin is used. The proteins might be negatively affected by the fact that the pin touches the slide surface. Non-contact printing is done by deposits of droplets. With this technique a well- defined drop of usually between 100-400 pL is dispensed on the surface. This means that the surfaces are left untouched and there is a possibility to calculate how much of the protein that is deposit on the surfaces and that the sample is delivered to the surface. There are better opportunities to wash out the protein solution that are not printed before the next protein will be printed. There is usually quite large volumes that can be aspirated which means that a lot of spots can be spotted with only one aspiration. One drawback is that they are more sensitive to disturbance during printing which can result in bad alignment of the spots. One key issue with printers is the flexibility to make different patterns. Other issues are the speed of the printing and how many slides that could be produced in a single run. The most advanced non-contact printers today work with higher speed than previous printers and they are equipped with more dispensers which means that more sample can be printed in parallel. Additional equipment such as control of the temperature and humidity can be installed. There can be stackers for both sample plates and slides added for large scale production.

Surfaces and immobilization

There is a large diversity of solid supports that can be used for protein microarrays on the market (75, 91, 92). In table 1, different surfaces are listed that can be used and the binding mode that is used for the immobilization. Unfortunately, there is no universal surface that could be used for all kinds of protein arrays. Every new application has to start with an optimizing step were the optimal solid support and other important parameters has to be established. One effect of the immobilization can be that the epitope is hidden or that the protein can be degraded. As can be seen in table 1 there are many different kinds of surfaces such as covalent coupling of the proteins or adsorption. There could be a risk of using surfaces that include hydrophobic interactions due to the risk of exposure of hidden residues that are needed for the stability of the protein which could cause denaturation of the protein (93, 94). Hydrophilic surfaces can be favorable in this respect, nitrocellulose-coated slides can be stored for many months (91). Unfortunately nitrocellulose-coated slides are usually associated with high unspecific binding, especially when fluorescent detection system is used. Then different kinds of hydrogel slides could be an alternative. At least for antibodies there has been shown that the sensitivity has been affected positively when covalent immobilization has been used (95, 96). If oriented immobilization is used it is not necessarily advantageous to random immobilization, but it could be easier to interpret the result (96).

Mårten Sundberg

There are a lot of studies done on different surfaces to investigate the best properties for protein arrays which is a very important factor when producing protein microarrays. Thereby not said that there is a general guideline of what should be the best solid support.

Table 1. Some of the surfaces that can be used in protein microarrays.

Surface chemistry Binding mode/ reactive groups Epoxy Covalent/ amino, thiol or hydroxyl groups Aldehyde Covalent/ amino groups Poly-L-Lysine Adsorption, ionic interaction Aminosilane Adsorption, ionic interaction Amine Covalent/ amino groups NHS Covalent/ amino groups PVDF Adsorption Nitrocellulose Adsorption Polystyrene Adsorption/polar groups Polyacrylamide gel Adsorption, interactions with the 3-D matrix Biotin immobilized in polycarboxylate Streptavidin/avidin/ neutravidin hydrogel Polycarboxylate hydrogel, NHS activated Adsorption, amino groups and interactions with the 3-D matrix Ni²+ ions complexed in polycarboxylate Poly histidine tag Hydrogel Protein A polycarboxylate hydrogel Human immunoglobulin proteins Streptavidin polycarboxylate hydrogel Biotin

Blocking of substrate

Blocking the surfaces that are not immobilized with protein is an important issue when working with protein microarrays. This can be tricky and a lot of testing might be needed to find a good blocking solution. Usually this is done with some kind of protein blocking. The most common used is bovine serum albumin (BSA) or milk or a combination of them. There is a growing commercial market of blocking solutions. Different kinds of chemical blocking e.g. ethanolamine can be used for some of the activated surfaces. It is important to find a proper blocking solution for the assay that is developed to get the background signal as low as possible to reach a good signal to noise ratio. One problem could be that the blocking solution decreases the interaction signal from the spotted protein due to steric hindrance. If the assay setup is not that complex with antigens and antibodies there is usually not a big effort to find a proper blocking solution. The situation is totally different if crude body fluids like, serum/plasma, CSF or urine are used. Then it can be a challenge to find something that work satisfactory in the assay.

Protein microarrays for validation of affinity binder

Sample labeling and detection systems

There are many different detection systems that can be used, but in protein microarrays fluorescence based are most frequent used (97). Some research groups use chemiluminescent detection (98). There are two different ways of labeling the samples that are being tested: direct labeling or indirect labeling. Indirect labeling, called sandwich assay, is an approach where the proteins in a complex sample like plasma are captured on an antibody array, then a second complementary fluorophore conjugated antibody is introduced for detection (99). This system is usually more sensitive, but the limitation is that there have to be two antibodies against the same target. It can be difficult to find two validated antibodies against the same target and it will be expensive if many targets are tested. Direct labeling means that the sample is labeled with e.g. a fluorophore and then incubated on the slide. Han et al. used an approach where they labeled the sample with two different fluorophores (84, 100, 101). This assay format is used a lot in DNA microarrays. The idea is to get a ratio between these two fluorophores. A problem with an approach like this is that the different fluorophores can give rather different signals and one have to be aware of that and try to avoid differences that are related to technical reasons (102). There is a risk that low abundant proteins can be missed due to the fact that the limit of detection is increased when the two different labeled samples compete for the same target (103). Haab et al. has improved the two color assay by including a rolling circle amplification (104, 105). It is also called direct labeling when the sample is labeled with e.g. biotin. In these types of assays the biotinylated samples are incubated on an antibody array and then the detection is done with a fluorophore conjugated streptavidin. In antigen microarrays the detection is usually done with a fluorophore conjugated secondary antibody which is specific for the host animal in which the primary antibody is produced. There have been studies to compare different labeling strategies to find out advantages and disadvantages with the different approaches (103, 106).

Incubation, scanning and data evaluation

When the microarray surface is printed and blocked the slides are ready for use. How this is done is very much depending on what kind of protein microarray that is produced and what kind of questions that are to be to answered with the assay. First of all the sample of interest, which can be antibodies or human serum/plasma or other kind of body fluid, depending on the setup, are put on the slide and incubated for 1h up to 24 h. After that, unbound material is washed off. If the sample is not directly labeled with a fluorophore a secondary antibody e.g. an anti-rabbit/anti-human antibody label with a fluorophore or other kind of detectable reagent is put on and incubated, usually 1 h. When

Mårten Sundberg this is done the slide is washed again to wash away unbound secondary antibody and after that the slide is dried and scanned. The evaluation of the scanned slide is done with an image analyzing software, which translates the fluorescent signals into numbers that can be analyzed using different kinds of statistical methods often in a statistical program like R or equal.

Bead based protein microarrays

In a bead based system the solid support is color-coded microspheres that can be identified with a flow cytometry based system (53, 74). Today there are 500 unique color-coded microspheres available, which means that this is the upper limitation of how many analytes can be tested. Each of these beads can be coupled to an antibody or antigen in suspension and after coupling the different beads can be mixed together. This is very flexible due to the fact that when the beads are coupled there are no limitations how they can be mixed together. In contrast to planar arrays the whole procedure is done in suspension which means that there is no drying that can make the coupled molecule inactivated or denatured. The suspension setup means that there are better possibilities to make the whole process automated by robots. Compared to planar arrays the bead arrays are faster when analyzing many different patient samples due to the fact that it is only adding more wells in the microplate and there is no image analysis. If slides are used, many slides have to be included and there is also a scanning and image analysis. The limitation compared to planar arrays is that there are only 500 unique bead identities instead of, in theory, theoretically tens of thousands on planar microarrays if one slide is used for one sample. The procedure starts with coupling the beads, where a specified amount of antibodies or antigens is added to a specified amount of beads. The coupling efficiency is checked with a flow cytometer based system. The beads are mixed together to create a microarray in suspension. Then sample is incubated with the beads and as in planar microarrays it can be done by direct labeled samples or indirect labeled samples. In bead based system the detection is done by fluorescent dyes. More about specific details for bead coupling and sample labeling can be found in two articles by Schwenk et al (78, 85). When the incubation is done the incubated beads are detected in a flow cytometry based system equipped with two lasers. One detects the identity of the bead and the other the intensity of the signal from the coupled sample. In the output file from the system, the mean intensity of all the beads with the same identity is reported as well as how many of that specific bead identity that are counted. It is always more than one bead that is counted, usually it is set to count at least 50 beads/well of the different bead identities included. This means that the output data is a mean value based on at least 50 different beads and the signal generated from each of them. It is more convenient when no external image software has to be used for the analysis of the result. The image analysis is usually one of the most labor intensive steps when working with planar microarrays. The bead based protein microarrays are a quite easy system to work with, but it has its limitations due to the number of unique beads that are available. Looking into a user perspective, this system is much more user-friendly and it is much more flexible because

Protein microarrays for validation of affinity binder the different beads can be mixed in any combination while the planar array have a fixed pattern even if you can include many spots but then the cost can be an issue.

Present investigation

When the HPA project started 2003, the standard method to validate the specificity of the produced antibodies was dot blot (Figure 1). If dot blot is compared with one of the sub arrays produced today incubated with a HPA antibody there is a substantial difference, as can be seen in Figure 1. The development of the protein array accelerated during 2004. There was a thorough discussion about the robustness of the microarray and if there should be any cut off values when an antibody should not be approved. The conclusion was that it worked and the dot blots were replaced by antigen microarrays produced with the PrEST as the method for specificity validation of the produced antibodies. Up until today more than 32.000 antibodies have been tested on the platform and almost 29.000 of them were approved. The PrEST arrays have been the foundation of my work at KTH and continuous development of both the equipment and the assay itself has been done. Spin off projects that started with the serum arrays in 2005 (89), have been on the schedule. The work has also included many different projects with other PhD students such as printing of Affibody molecules in different kinds of applications, lateral flow applications and a collaboration concerning labeling of antibodies. In the latter, my part was to check that the labeling worked as it should (107). To be able to compare different strategies, three different assays were set up to validate the specificity of antibodies produced in the HPA project. The microarray assay was compared with an assay in a compact disc format (108) and a bead based array (78).

Figure 1. Dot blot compared with the antigen microarray used today in the HPA project.

Many hours have been spent to be able to produce antibody arrays based on the HPA produced antibodies. Many different types of slides, blocking buffers and assay buffers have been tested through the years (109).

Mårten Sundberg

Validated affinity reagents – an important factor in today’s biotechnology

Why is it important to have validated affinity reagents to work with? First of all, it is a lot of money involved in affinity reagents and since they are valuable and crucial for the assays the purchaser needs to be sure that the reagent works. If there is an affinity reagent that does not work at all, a lot of money and time is wasted. If the affinity reagent that is bought is unspecific, much time is spent on investigating something that in the end shows to be of low value. Therefore, it is in the researcher’s interest that better and more similar validation strategies are set up for the affinity reagents that are commercially accessible. If there is no or limited information of the validation of the affinity reagent, the researcher has to put valuable time to investigate that the product that is bought work as it should. To speed up the growing market of affinity based diagnostics, there is a need for a more generally approved validation strategy for affinity reagents. It is important that this validation is not too time consuming and that the cost is not too high which will limit the reagent to reach the market. Today there are many different ways to test the antibodies such as western blot, immune precipitation and many more. Unfortunately there is no general agreement on what should be done and what kind of cut off that should be used to be able to state that an affinity reagent fulfills the criteria of being a validated and fully functional reagent. A first step could be if all producers of affinity reagents published all available data of the reagents.

Paper I

In the first paper, the structure of the production line in the HPA project 2005 is presented. The focus of the paper lies on how the antibodies are purified from rabbit sera and how these mono-specific antibodies are characterized. There are both similarities and differences compared to the structure of the project today. First of all, the size of the project is much larger today and more methods are involved to further characterize the antibodies. Today the antibodies are also tested with Western blot and they are used in immunofluorescence to get a picture of the subcellular localization of the proteins. There are projects ongoing that deal with plasma profiling and epitope mapping of the antibodies that are produced. Despite of these changes, the set up for the other methods is similar to how it is today. The affinity purification is done in the same way but more streamlined and the tissue microarrays that are used have similar appearance today as they had in 2005. A microarray containing 72 different antigens (PrESTs) that were printed in 12 different sub arrays on an epoxy slide is presented. This was a great progress since the previous method was dot blot with only 4 spots, three different antigens and one negative control. Today the immobilization surface is still the same but the size has been increased so that one slide contains 384 different antigens in 21 sub arrays. In 2005 much of the

Protein microarrays for validation of affinity binder incubation on the microarrays was made manually but today a lot of that procedure is semi- automated. There is no presentation of the development and all the tests that were done with different surfaces, printing robots, blocking buffers, incubation times, washing and different masks to separate the sub arrays. In 2005 there was a lot of DNA microarrays produced at KTH with very high quality. Therefore, there was a lot of knowledge in house how to produce microarrays. However it is much more difficult to produce protein microarrays than DNA microarrays since proteins are more complex than DNA. Much time was spent to find the most appropriate surface, blocking solution, washing protocol and so on. Although some advice can be found in published articles, there is still a lot of trial and error that has to be done in the laboratory to find out the best solution for the assay. In the paper, 96 % of the 464 tested antibodies were approved for further use, 19 % were approved with the comment low specificity and 2 % were failed due to low specificity. Today these numbers are 88 % approved and of these 17 % were approved with low specificity and 6 % were failed due to low specificity. It is interesting to see that today there are fewer antibodies that are approved with low specificity but there are more antibodies failed with that comment. This can partly be explained by the fact that there are now 384 PrESTs on each array instead of 72. Maybe the lower percent of antibodies that are approved with low specificity could be explained by better purification routines. Another consideration to keep in mind is that the antibodies today are more diluted than they were in 2005 which will result in what might appear as less unspecific binding. This can also be indicated by the fact that 1.1 % of the antibodies in the paper were failed due to high background compared to 0.3 % today. This is due to insufficient depletion of tag-specific antibodies and the problem is eliminated with better purification routines. When the antigen microarray was implemented in the HPA project it was shown to work very well. Today it is obvious that it has worked well as a validation tool for the produced antibodies. Much of the effort that has been invested in the antigen microarray has been valuable when other kinds of microarrays have been developed. The antigen array has proven to be a helpful tool to generate antibodies that function well in both immunohistochemistry and in the immunofluorescence.

Paper II

The second paper is based on the set-up of a bead based assay for specificity determination of mono-specific antibodies. The set-up is similar to what was done on the planar microarrays in the first paper. The bead based assay is presented as an alternative to the planar microarrays. In the study, 84 mono-specific antibodies were run in parallel on both the planar arrays and the bead based arrays and the results were compared. Another purpose with the study was to show that it was possible to set up a multiplexed assay with up to 100 different bead identities. Until then, no set-up of this complexity had been published. It was found that there was no need for any washing steps during the run which simplified the handling throughout the run. Most of the 84 tested mono-specific antibodies showed similar

Mårten Sundberg results in both assays. When there was unspecific binding in the planar microarray, the same unspecific binding could be seen with the bead array. There was one case with a mono- specific antibody that did not show any interaction in the bead based assay. The corresponding PrEST was short and one explanation could have been that the epitope was masked on that specific antibody. This is due to different chemistries used to immobilize the PrESTs in the two assays. In the planar array, epoxy chemistry was used, while in the bead based assay N-hydroxysuccinimide ester chemistry was used. There could have been some kind of masking effect, due to the fact that out of the 55 amino acid long PrEST in total 11 were either lysine or arginine. One suggestion to solve this issue is to use direct labeling by biotin-avidin or through a His6-chelator. This was not tested due to the fact that it will be a non-covalent interaction, which raised some new concerns about the stability. The bead based assay is a multiplex set-up and the use of a non-covalent strategy could lead to that antigenic information could be transferred between different bead identities. Therefore it was suggested that more testing has to be done before a strategy like this could be implemented. It was shown that the bead based assay was comparable with the standard planar microarray set-up used in the HPA project. The parallel run that was performed with 84 mono-specific antibodies showed similar results even if there were differences in signal intensities. This can be explained by differences in what dyes that are used and differences in the readout system. There are fewer steps involved in the bead based assay compared to planar arrays which makes it easier and faster. Due to the fact that there is no scanner or image analyzing software needed for the bead based system it might be a better choice to the laboratory that want to set up such an assay if this equipment is not already available in the laboratory. At the time of the study, the limiting factor for the bead based system was that there were only 100 different bead identities that could be used. Today that number has been raised to 500. Still, there are no such limitations when planar microarrays are used. In this context, where a limited number of antibodies will be tested in each batch, the planar microarrays should be considered as a better choice due to the cost of the beads and the storage of the beads. The bead based assay is a much more flexible system and if a lot of different antibodies or biological samples are to be tested, it is a cheaper and faster system. The problem is the limitation in how many targets that can be tested in the same run. This can partly be solved by multiple runs with the same sample but different target but then there will be an issue with plate to plate variations. Even if decided that the antigen microarray was the assay to be used in the HPA project the bead based array has been used in many other applications such as the plasma profiling and the epitope mapping. The development and assays set-up that was done for the validation of antibodies have been useful for the further work with the bead based array. This is one of the advantages with large projects as the HPA project. There are large quantities of material to work with and there are always different kinds of development projects ongoing. Therefore, something that shows to not work in one part could be useful in another part of the project. This has been done in the HPA project in order to find out as much about the human proteome as possible and it will probably result in many useful products for both diagnostics and therapeutics in the future.

Protein microarrays for validation of affinity binder

Paper III

In the previous papers, it was shown that PrESTs could be immobilized on both microscope slides and microspheres and these two different assays were used as a tool to determine the specificity of the antibodies produced in the HPA project. In the study presented in paper III, the planar antigen microarray was further developed to be used as a validation tool for other kinds of affinity regents produced externally. The article is part of a bigger international initiative to produce well characterized renewable affinity binders. The affinity binders included in the study were directed against SH2-containing human proteins. In the study, four different kinds of affinity reagents were included; mono-specific HPA antibodies, mouse monoclonal antibodies, recombinant single-chain variable fragment (scFv) and fibronectin-based binders that were produced by six different academic laboratories. A protein microarray was constructed including 432 antigens, whereof 384 PrESTs and the others produced by the Structural Genomics Consortium (SGC) at Karolinska Institute. Of the PrESTs that were included, 85 were SH2 domain containing proteins and all of the SGC produced proteins contained SH2 domains. The functionality of the PrESTs was known, while the behavior of the SGC proteins was unknown. These proteins were delivered with varying concentrations and it was decided to print them both as they were delivered but also diluted. When the affinity reagents were tested, it was shown that the SGC proteins could be printed in the same way as the PrESTs and their functionality was equal to the PrESTs. All in all, 401 affinity reagents were tested on the produced protein microarrays, some of them where tested more than once due to unsatisfactory results. All included participating laboratories had affinity reagents that were approved and at least one affinity reagent targeting the included SH2-containing proteins was approved. Most of the proteins had multiple affinity reagents that were approved. As the case is in the HPA project in whole, there is anything from one spike of binding to the right target to binders that bind all of the antigens. This can be adjusted by dilutions or can in some cases have a possible explanation of targets that have similarities in their amino acid chain. This is truly a big challenge for the future since very few proteins are expressed specifically in one organ. There are a lot of regulation mechanisms in the body to fine-tune the amount of proteins in different parts of the body. This kind of regulation mechanism cannot be included in a microarray and it has to be investigated what are normal levels of a specific protein and what indicates a disease state. Therefore we might in the future see another definition of how a pattern for an affinity reagent should look like to be used as a biomarker. One suggestion could be that the results from an assay are presented with a number of different concentrations which could be similar or very different. More investigations with different kinds of assays are needed to show that the affinity reagent is specific enough so it could be used for diagnostic purposes. The study showed that it is possible to print proteins produced by other laboratories on the protein microarray platform that is used in the HPA project. Furthermore it was

Mårten Sundberg proven that the platform can be used as a validation tool of the specificity not only for the HPA produced mono-specific antibodies but also for other kinds of affinity reagents. This is important because if the assay that is used for the validation can be used for all different kinds of affinity reagents it will be easier to compare the different behavior and in that sense find the most appropriate reagent to proceed with.

Conclusions

Today there are no general agreements about how to validate affinity reagents and this cost a lot of money and restricts the growing market of affinity based diagnostics. Therefore, better international agreements on how affinity reagents should be tested to be regarded as functional reagents are needed. One of the most important issues is the specificity of the affinity reagent. There has to be agreement in what kind of application specific validation that is needed, because different assays need different validation of the antibodies that are used. It would be too expensive to have the same validation for all antibodies that are available but there has to be a basic level. Then the suppliers should specify in what kind of assays they recommend their affinity reagents and have appropriate validation for that. In this thesis, it is shown that the protein microarray platform that was established within the HPA project at KTH is a very good tool to determine the specificity of different affinity binders. The protein microarray could be used in future collaboration as a tool to produce specific affinity binders against different targets. In the presented papers, it has been shown that the assay is not only usable with the antigens and affinity binders that are produced in the HPA project but also with reagents produced by other research groups around the world. The methods have to be able to handle different kinds of affinity reagents produced at different laboratories. The methods should be so easy to use and so robust that there is no difference by whom and where the assay is carried out. There is still much work needed before such generally accepted and standardized methods are established. There are promising indications that shows that we are heading in that direction, e.g. by researchers publishing articles where they demand better and more standardized methods for affinity reagent validation.

Protein microarrays for validation of affinity binder

Abbreviations

ACS Acute coronary syndrome BSA Bovine serum albumin CDR Complementarity determining regions CSF Cerebrospinal fluid ELISA Enzyme-Linked Immunosorbent assay Fab Fragment, antigen binding Fc Fragment, crystallizable Fv Fragment, variable HPA Human Protein Atlas IHC Immunohistochemistry IF Immunofluorescence Ig Immunoglobulin NHS N-Hydroxysuccinimide MS Mass spectrometry PrEST Protein Epitope Signature Tag PVDF Polyvinylidene fluoride RIA Radioimmunoassay scFv Single-chain variable fragment SPR Surface Plasmon Resonance

Mårten Sundberg

Acknowledgments

Det är många som jag skulle vilja tacka för att jag har tagit mig till denna punkt. Nedan nämns några särskilt men det är betydligt fler som varit med och bidragit till den jag är idag.

Först och främst är det min handledare Peter och bihandledare Jochen som bidragit till att jag rott detta projekt i hamn. Det har varit en lång resa från det att jag steg in genom dörren på Albanova för första gången. Det har definitivt inte varit rätlinjigt och inte alltid självklart. Men det har varit många spännande och givande år med många lärdomar på vägen.

Alla trevliga array människor, till att börja med Annelie W som var okrönt DNA microarray drottning då jag kom in som färsking i microarrayvärlden. Tack för alla roliga stunder vi hade med diskussioner om allt från krånglande printrar till att bli förälder. Sedan har vi Ronald som kom in precis när vi skulle sjösätta LIMS för PrEST array. Jag var nog inte så pedagogisk men det gick ju alldeles lysande ändå. Passar även på att rikta ett särskilt tack för det arbete du lagt ner med att fixa med kommentarerna vi fick på SH2 peket. Kimi min avbytare under två pappaledigheter och nu en fena på att kartlägga plasma proteomet. Rebecca, why didn´t we manage to get the antibody microarrays to work? With all our work it should have been done. But we had a lot of fun during that time. Burcu, Maja, Ulrika, och Anna H, när ni kom in började det verkligen arta sig till en riktig grupp det här med protein array. Det har varit en ynnest att arbeta med er. Anna G, det är sällan man träffar någon med så mycket livsglädje och sprudlande energi. Tack för att du delar med dig av ditt glada humör. Utan din arbetsinsats och smittande entusiasm skulle jag aldrig ha skrivit den här boken utan klivit av tåget någon station tidigare. Stort tack även till alla andra som varit med på proteinarraytåget. Jag hoppas det tuffar mot en lysande framtid.

Mitt första hem i HPA projektet har också en särskild plats bland mina minnen från KTH. Min instruktör Hanna T under den första tiden i Molbio får ett extra tack. Lycka till med din avhandling! Marica utan dig skulle jag inte ha kommit in i HPA projektet alls. Ett stort tack även till Bahram, Anna S, Sam och Anja för utmärkt introduktion i projektet.

Jag vill även tacka alla härliga människor som jobbar och har jobbat inom HPA projektet. Särskilt tack till Johanna S, Pia A, Anna K, Cajsa, Erik M, Ulla W, Jenny O, Henrik och det glada LIMS gänget.

Kicki och Annelie W, mina mesta rumskompisar. Tack för alla lätta på trycket samtal och prat om allt som inte har med forskning att göra. Stisse får även vara med på ett hörn här eftersom ni tre hade skapat en så bra stämmning i Kanelen som även jag fick ta del av.

Snyggrummet: Maja, Johanna, Helena, Basia, Anna P och Feifan. Det blev såå tyst när jag flyttade till Scilife, men det störde mig betydligt mer. Tack för att ni livade upp min tillvaro trots att ni hade svårt att förstå hur jag så fullständigt kunde koppla bort era diskussioner.

Protein microarrays for validation of affinity binder

Alla doktorander och ex-jobbare som jag fått äran att hjälpa i era projekt: Torun, Björn R, Emma L, Jinghong, Cilla, Magda, Johan R, Holger, Karin L, Linda Paavilainen, Calle, Johan L, Jesper, Tarek, Nima och Viktor.

Tack även alla trevliga Uppsala HPA människor för alla roliga konferenser och ett särskilt tack till Caroline, Figge och Kenneth för många givande diskussioner under åren.

Mathias, det är ett häftigt projekt som du drog igång och jag har sagt det många gånger tidigare: jag kommer aldrig att jobba i ett projekt av den här digniteten igen. Tack för att jag fick möjlighet att bidra med min kunskap.

Sophia, det har varit så skönt att veta att jag kunnat gå in till dig och prata om vad som helst och du lyssnar. Sedan har vi kanske inte alltid varit överens men det är en helt annan sak. Fortsätt att lyssna, det behövs fler som dig i dagens värld. Per-Åke, jag upphör aldrig att förundras över hur fort tiden går när jag sitter inne hos dig för att diskutera det jag trodde var ETT problem som när jag går ut därifrån visade sig vara en hel djungel av möjligheter som nog borde undersökas snarast. Helt underbart att jag fått möjligheten att diskutera med dig.

Roger, Mia, Lisbeth, Peter, Christoffer, Christine och Frida med familjer som funnits där med goda middagar eller bara en massa prat om barn, jobb och livet i största allmänhet.

Mamma och pappa som alltid finns där när det behövs som mest. Tack för att ni alltid har ställt upp för mig. Det känns tryggt att ha er som jag kan luta mig emot. Simon, min bror, vad skulle jag göra om jag inte kunde ringa dig när bilen krånglar eller när jag behöver prata av mig.

Tack även till mina svärföräldrar Agneta och Hasse för att ni har en sån bedårande dotter och för att ni alltid är beredda att ställa upp om det behövs hjälp vad än det kan gälla.

Min älskade Katarina, tack för att du stått ut med mig och mitt hetsiga humör inte minst under denna höst då jag förutom att få ihop denna skrift, varit väldigt frånvarande p.g.a. av mitt nya jobb. Skönt att veta att jag kan komma hem och få känna din värme. Sist men inte minst mina två underbara killar Viggo och Joel. Jag kan inte tänka mig ett liv utan er och den enorma upptäckarglädje och livsbejakande som finns inom er båda. Tänk om jag bara hade hälften av den inlärningskapacitet och den förmåga till problemlösning (och problemskapande) som ni besitter. Då skulle vi kunna börja snacka forskning på riktigt hög nivå.

Mårten Sundberg

References

1. Watson, J. D., and Crick, F. H. (1953) The structure of DNA, Cold Spring Harb Symp Quant Biol 18, 123-131. 2. Berglund, L., Bjorling, E., Oksvold, P., Fagerberg, L., Asplund, A., Szigyarto, C. A., Persson, A., Ottosson, J., Wernerus, H., Nilsson, P., Lundberg, E., Sivertsson, A., Navani, S., Wester, K., Kampf, C., Hober, S., Ponten, F., and Uhlen, M. (2008) A genecentric Human Protein Atlas for expression profiles based on antibodies, Mol Cell Proteomics 7, 2019-2027. 3. Wilkins, M. (2009) Proteomics data mining, Expert Rev Proteomics 6, 599-603. 4. Wilkins, M. R., Pasquali, C., Appel, R. D., Ou, K., Golaz, O., Sanchez, J. C., Yan, J. X., Gooley, A. A., Hughes, G., Humphery-Smith, I., Williams, K. L., and Hochstrasser, D. F. (1996) From proteins to proteomes: large scale protein identification by two-dimensional electrophoresis and amino acid analysis, Biotechnology (N Y) 14, 61-65. 5. Anderson, N. L., and Anderson, N. G. (2002) The human plasma proteome: history, character, and diagnostic prospects, Mol Cell Proteomics 1, 845-867. 6. Bjorling, E., and Uhlen, M. (2008) Antibodypedia, a portal for sharing antibody and antigen validation data, Mol Cell Proteomics 7, 2028-2037. 7. Tyers, M., and Mann, M. (2003) From genomics to proteomics, Nature 422, 193- 197. 8. Uhlen, M. (2008) Affinity as a tool in life science, Biotechniques 44, 649-654. 9. Uhlen, M., Oksvold, P., Fagerberg, L., Lundberg, E., Jonasson, K., Forsberg, M., Zwahlen, M., Kampf, C., Wester, K., Hober, S., Wernerus, H., Bjorling, L., and Ponten, F. (2010) Towards a knowledge-based Human Protein Atlas, Nat Biotechnol 28, 1248-1250. 10. Uhlen, M., Bjorling, E., Agaton, C., Szigyarto, C. A., Amini, B., Andersen, E., Andersson, A. C., Angelidou, P., Asplund, A., Asplund, C., Berglund, L., Bergstrom, K., Brumer, H., Cerjan, D., Ekstrom, M., Elobeid, A., Eriksson, C., Fagerberg, L., Falk, R., Fall, J., Forsberg, M., Bjorklund, M. G., Gumbel, K., Halimi, A., Hallin, I., Hamsten, C., Hansson, M., Hedhammar, M., Hercules, G., Kampf, C., Larsson, K., Lindskog, M., Lodewyckx, W., Lund, J., Lundeberg, J., Magnusson, K., Malm, E., Nilsson, P., Odling, J., Oksvold, P., Olsson, I., Oster, E., Ottosson, J., Paavilainen, L., Persson, A., Rimini, R., Rockberg, J., Runeson, M., Sivertsson, A., Skollermo, A., Steen, J., Stenvall, M., Sterky, F., Stromberg, S., Sundberg, M., Tegel, H., Tourle, S., Wahlund, E., Walden, A., Wan, J., Wernerus, H., Westberg, J., Wester, K., Wrethagen, U., Xu, L. L., Hober, S., and Ponten, F. (2005) A human protein atlas for normal and cancer tissues based on antibody proteomics, Mol Cell Proteomics 4, 1920-1932. 11. Lander, E. S., Linton, L. M., Birren, B., Nusbaum, C., Zody, M. C., Baldwin, J., Devon, K., Dewar, K., Doyle, M., FitzHugh, W., Funke, R., Gage, D., Harris, K., Heaford, A., Howland, J., Kann, L., Lehoczky, J., LeVine, R., McEwan, P., McKernan, K., Meldrim, J., Mesirov, J. P., Miranda, C., Morris, W., Naylor, J., Raymond, C., Rosetti, M., Santos, R., Sheridan, A., Sougnez, C., Stange- Thomann, N., Stojanovic, N., Subramanian, A., Wyman, D., Rogers, J., Sulston, J., Ainscough, R., Beck, S., Bentley, D., Burton, J., Clee, C., Carter, N., Coulson, A., Deadman, R., Deloukas, P., Dunham, A., Dunham, I., Durbin, R., French, L., Grafham, D., Gregory, S., Hubbard, T., Humphray, S., Hunt, A., Jones, M., Lloyd, C., McMurray, A., Matthews, L., Mercer, S., Milne, S., Mullikin, J. C., Mungall,

Protein microarrays for validation of affinity binder

A., Plumb, R., Ross, M., Shownkeen, R., Sims, S., Waterston, R. H., Wilson, R. K., Hillier, L. W., McPherson, J. D., Marra, M. A., Mardis, E. R., Fulton, L. A., Chinwalla, A. T., Pepin, K. H., Gish, W. R., Chissoe, S. L., Wendl, M. C., Delehaunty, K. D., Miner, T. L., Delehaunty, A., Kramer, J. B., Cook, L. L., Fulton, R. S., Johnson, D. L., Minx, P. J., Clifton, S. W., Hawkins, T., Branscomb, E., Predki, P., Richardson, P., Wenning, S., Slezak, T., Doggett, N., Cheng, J. F., Olsen, A., Lucas, S., Elkin, C., Uberbacher, E., Frazier, M., Gibbs, R. A., Muzny, D. M., Scherer, S. E., Bouck, J. B., Sodergren, E. J., Worley, K. C., Rives, C. M., Gorrell, J. H., Metzker, M. L., Naylor, S. L., Kucherlapati, R. S., Nelson, D. L., Weinstock, G. M., Sakaki, Y., Fujiyama, A., Hattori, M., Yada, T., Toyoda, A., Itoh, T., Kawagoe, C., Watanabe, H., Totoki, Y., Taylor, T., Weissenbach, J., Heilig, R., Saurin, W., Artiguenave, F., Brottier, P., Bruls, T., Pelletier, E., Robert, C., Wincker, P., Smith, D. R., Doucette-Stamm, L., Rubenfield, M., Weinstock, K., Lee, H. M., Dubois, J., Rosenthal, A., Platzer, M., Nyakatura, G., Taudien, S., Rump, A., Yang, H., Yu, J., Wang, J., Huang, G., Gu, J., Hood, L., Rowen, L., Madan, A., Qin, S., Davis, R. W., Federspiel, N. A., Abola, A. P., Proctor, M. J., Myers, R. M., Schmutz, J., Dickson, M., Grimwood, J., Cox, D. R., Olson, M. V., Kaul, R., Shimizu, N., Kawasaki, K., Minoshima, S., Evans, G. A., Athanasiou, M., Schultz, R., Roe, B. A., Chen, F., Pan, H., Ramser, J., Lehrach, H., Reinhardt, R., McCombie, W. R., de la Bastide, M., Dedhia, N., Blocker, H., Hornischer, K., Nordsiek, G., Agarwala, R., Aravind, L., Bailey, J. A., Bateman, A., Batzoglou, S., Birney, E., Bork, P., Brown, D. G., Burge, C. B., Cerutti, L., Chen, H. C., Church, D., Clamp, M., Copley, R. R., Doerks, T., Eddy, S. R., Eichler, E. E., Furey, T. S., Galagan, J., Gilbert, J. G., Harmon, C., Hayashizaki, Y., Haussler, D., Hermjakob, H., Hokamp, K., Jang, W., Johnson, L. S., Jones, T. A., Kasif, S., Kaspryzk, A., Kennedy, S., Kent, W. J., Kitts, P., Koonin, E. V., Korf, I., Kulp, D., Lancet, D., Lowe, T. M., McLysaght, A., Mikkelsen, T., Moran, J. V., Mulder, N., Pollara, V. J., Ponting, C. P., Schuler, G., Schultz, J., Slater, G., Smit, A. F., Stupka, E., Szustakowski, J., Thierry-Mieg, D., Thierry-Mieg, J., Wagner, L., Wallis, J., Wheeler, R., Williams, A., Wolf, Y. I., Wolfe, K. H., Yang, S. P., Yeh, R. F., Collins, F., Guyer, M. S., Peterson, J., Felsenfeld, A., Wetterstrand, K. A., Patrinos, A., Morgan, M. J., de Jong, P., Catanese, J. J., Osoegawa, K., Shizuya, H., Choi, S., and Chen, Y. J. (2001) Initial sequencing and analysis of the human genome, Nature 409, 860-921. 12. Venter, J. C., Adams, M. D., Myers, E. W., Li, P. W., Mural, R. J., Sutton, G. G., Smith, H. O., Yandell, M., Evans, C. A., Holt, R. A., Gocayne, J. D., Amanatides, P., Ballew, R. M., Huson, D. H., Wortman, J. R., Zhang, Q., Kodira, C. D., Zheng, X. H., Chen, L., Skupski, M., Subramanian, G., Thomas, P. D., Zhang, J., Gabor Miklos, G. L., Nelson, C., Broder, S., Clark, A. G., Nadeau, J., McKusick, V. A., Zinder, N., Levine, A. J., Roberts, R. J., Simon, M., Slayman, C., Hunkapiller, M., Bolanos, R., Delcher, A., Dew, I., Fasulo, D., Flanigan, M., Florea, L., Halpern, A., Hannenhalli, S., Kravitz, S., Levy, S., Mobarry, C., Reinert, K., Remington, K., Abu-Threideh, J., Beasley, E., Biddick, K., Bonazzi, V., Brandon, R., Cargill, M., Chandramouliswaran, I., Charlab, R., Chaturvedi, K., Deng, Z., Di Francesco, V., Dunn, P., Eilbeck, K., Evangelista, C., Gabrielian, A. E., Gan, W., Ge, W., Gong, F., Gu, Z., Guan, P., Heiman, T. J., Higgins, M. E., Ji, R. R., Ke, Z., Ketchum, K. A., Lai, Z., Lei, Y., Li, Z., Li, J., Liang, Y., Lin, X., Lu, F., Merkulov, G. V., Milshina, N., Moore, H. M., Naik, A. K., Narayan, V. A., Neelam, B., Nusskern, D., Rusch, D. B., Salzberg, S., Shao, W., Shue, B., Sun, J., Wang, Z., Wang, A., Wang, X., Wang, J., Wei, M., Wides, R., Xiao, C., Yan, C.,

Mårten Sundberg

Yao, A., Ye, J., Zhan, M., Zhang, W., Zhang, H., Zhao, Q., Zheng, L., Zhong, F., Zhong, W., Zhu, S., Zhao, S., Gilbert, D., Baumhueter, S., Spier, G., Carter, C., Cravchik, A., Woodage, T., Ali, F., An, H., Awe, A., Baldwin, D., Baden, H., Barnstead, M., Barrow, I., Beeson, K., Busam, D., Carver, A., Center, A., Cheng, M. L., Curry, L., Danaher, S., Davenport, L., Desilets, R., Dietz, S., Dodson, K., Doup, L., Ferriera, S., Garg, N., Gluecksmann, A., Hart, B., Haynes, J., Haynes, C., Heiner, C., Hladun, S., Hostin, D., Houck, J., Howland, T., Ibegwam, C., Johnson, J., Kalush, F., Kline, L., Koduru, S., Love, A., Mann, F., May, D., McCawley, S., McIntosh, T., McMullen, I., Moy, M., Moy, L., Murphy, B., Nelson, K., Pfannkoch, C., Pratts, E., Puri, V., Qureshi, H., Reardon, M., Rodriguez, R., Rogers, Y. H., Romblad, D., Ruhfel, B., Scott, R., Sitter, C., Smallwood, M., Stewart, E., Strong, R., Suh, E., Thomas, R., Tint, N. N., Tse, S., Vech, C., Wang, G., Wetter, J., Williams, S., Williams, M., Windsor, S., Winn- Deen, E., Wolfe, K., Zaveri, J., Zaveri, K., Abril, J. F., Guigo, R., Campbell, M. J., Sjolander, K. V., Karlak, B., Kejariwal, A., Mi, H., Lazareva, B., Hatton, T., Narechania, A., Diemer, K., Muruganujan, A., Guo, N., Sato, S., Bafna, V., Istrail, S., Lippert, R., Schwartz, R., Walenz, B., Yooseph, S., Allen, D., Basu, A., Baxendale, J., Blick, L., Caminha, M., Carnes-Stine, J., Caulk, P., Chiang, Y. H., Coyne, M., Dahlke, C., Mays, A., Dombroski, M., Donnelly, M., Ely, D., Esparham, S., Fosler, C., Gire, H., Glanowski, S., Glasser, K., Glodek, A., Gorokhov, M., Graham, K., Gropman, B., Harris, M., Heil, J., Henderson, S., Hoover, J., Jennings, D., Jordan, C., Jordan, J., Kasha, J., Kagan, L., Kraft, C., Levitsky, A., Lewis, M., Liu, X., Lopez, J., Ma, D., Majoros, W., McDaniel, J., Murphy, S., Newman, M., Nguyen, T., Nguyen, N., Nodell, M., Pan, S., Peck, J., Peterson, M., Rowe, W., Sanders, R., Scott, J., Simpson, M., Smith, T., Sprague, A., Stockwell, T., Turner, R., Venter, E., Wang, M., Wen, M., Wu, D., Wu, M., Xia, A., Zandieh, A., and Zhu, X. (2001) The sequence of the human genome, Science 291, 1304-1351. 13. Flicek, P., Amode, M. R., Barrell, D., Beal, K., Brent, S., Chen, Y., Clapham, P., Coates, G., Fairley, S., Fitzgerald, S., Gordon, L., Hendrix, M., Hourlier, T., Johnson, N., Kahari, A., Keefe, D., Keenan, S., Kinsella, R., Kokocinski, F., Kulesha, E., Larsson, P., Longden, I., McLaren, W., Overduin, B., Pritchard, B., Riat, H. S., Rios, D., Ritchie, G. R., Ruffier, M., Schuster, M., Sobral, D., Spudich, G., Tang, Y. A., Trevanion, S., Vandrovcova, J., Vilella, A. J., White, S., Wilder, S. P., Zadissa, A., Zamora, J., Aken, B. L., Birney, E., Cunningham, F., Dunham, I., Durbin, R., Fernandez-Suarez, X. M., Herrero, J., Hubbard, T. J., Parker, A., Proctor, G., Vogel, J., and Searle, S. M. (2011) Ensembl 2011, Nucleic Acids Res 39, D800-806. 14. Consortium, U. (2011) Ongoing and future developments at the Universal Protein Resource, Nucleic Acids Res 39, D214-219. 15. www.antibodybeyond.com//abfinder/abfinder.htm. 16. Couchman, J. R. (2009) Commercial antibodies: the good, bad, and really ugly, J Histochem Cytochem 57, 7-8. 17. Bordeaux, J., Welsh, A., Agarwal, S., Killiam, E., Baquero, M., Hanna, J., Anagnostou, V., and Rimm, D. (2010) Antibody validation, Biotechniques 48, 197-209. 18. Berglund, L., Bjorling, E., Jonasson, K., Rockberg, J., Fagerberg, L., Al-Khalili Szigyarto, C., Sivertsson, A., and Uhlen, M. (2008) A whole-genome bioinformatics approach to selection of antigens for systematic antibody generation, Proteomics 8, 2832-2839.

Protein microarrays for validation of affinity binder

19. Porath, J., Carlsson, J., Olsson, I., and Belfrage, G. (1975) Metal chelate affinity chromatography, a new approach to protein fractionation, Nature 258, 598-599. 20. Sjolander, A., Nygren, P. A., Stahl, S., Berzins, K., Uhlen, M., Perlmann, P., and Andersson, R. (1997) The serum albumin-binding region of streptococcal protein G: a bacterial fusion partner with carrier-related properties, J Immunol Methods 201, 115-123. 21. Agaton, C., Galli, J., Hoiden Guthenberg, I., Janzon, L., Hansson, M., Asplund, A., Brundell, E., Lindberg, S., Ruthberg, I., Wester, K., Wurtz, D., Hoog, C., Lundeberg, J., Stahl, S., Ponten, F., and Uhlen, M. (2003) Affinity proteomics for systematic protein profiling of chromosome 21 gene products in human tissues, Mol Cell Proteomics 2, 405-414. 22. Tegel, H., Steen, J., Konrad, A., Nikdin, H., Pettersson, K., Stenvall, M., Tourle, S., Wrethagen, U., Xu, L., Yderland, L., Uhlen, M., Hober, S., and Ottosson, J. (2009) High-throughput protein production--lessons from scaling up from 10 to 288 recombinant proteins per week, Biotechnol J 4, 51-57. 23. Steen, J., Uhlen, M., Hober, S., and Ottosson, J. (2006) High-throughput protein purification using an automated set-up for high-yield affinity chromatography, Protein Expr Purif 46, 173-178. 24. Nilsson, P., Paavilainen, L., Larsson, K., Odling, J., Sundberg, M., Andersson, A. C., Kampf, C., Persson, A., Al-Khalili Szigyarto, C., Ottosson, J., Bjorling, E., Hober, S., Wernerus, H., Wester, K., Ponten, F., and Uhlen, M. (2005) Towards a human proteome atlas: high-throughput generation of mono-specific antibodies for tissue profiling, Proteomics 5, 4327-4337. 25. Barbe, L., Lundberg, E., Oksvold, P., Stenius, A., Lewin, E., Bjorling, E., Asplund, A., Ponten, F., Brismar, H., Uhlen, M., and Andersson-Svahn, H. (2008) Toward a confocal subcellular atlas of the human proteome, Mol Cell Proteomics 7, 499-508. 26. Lundberg, E., and Uhlen, M. (2010) Creation of an antibody-based subcellular protein atlas, Proteomics 10, 3984-3996. 27. Phelan, M. L., and Nock, S. (2003) Generation of bioreagents for protein chips, Proteomics 3, 2123-2134. 28. Geisse, S., Gram, H., Kleuser, B., and Kocher, H. P. (1996) Eukaryotic expression systems: a comparison, Protein Expr Purif 8, 271-282. 29. Hancock, D. C., and O'Reilly, N. J. (2005) Synthetic peptides as antigens for antibody production, Methods Mol Biol 295, 13-26. 30. Hjelm, B., Fernandez, C. D., Lofblom, J., Stahl, S., Johannesson, H., Rockberg, J., and Uhlen, M. (2010) Exploring epitopes of antibodies toward the human tryptophanyl-tRNA synthetase, N Biotechnol 27, 129-137. 31. Colwill, K., Persson, H., Jarvik, N. E., Wyrzucki, A., Wojcik, J., Koide, A., Kossiakoff, A. A., Koide, S., Sidhu, S., Dyson, M. R., Pershad, K., Pavlovic, J. D., Karatt-Vellatt, A., Schofield, D. J., Kay, B. K., McCafferty, J., Mersmann, M., Meier, D., Mersmann, J., Helmsing, S., Hust, M., Dubel, S., Berkowicz, S., Freemantle, A., Spiegel, M., Sawyer, A., Layton, D., Nice, E., Dai, A., Rocks, O., Williton, K., Fellouse, F. A., Hersi, K., Pawson, T., Nilsson, P., Sundberg, M., Sjoberg, R., Sivertsson, A., Schwenk, J. M., Takanen, J. O., Hober, S., Uhlen, M., Dahlgren, L. G., Flores, A., Johansson, I., Weigelt, J., Crombet, L., Loppnau, P., Kozieradzki, I., Cossar, D., Arrowsmith, C. H., Edwards, A. M., and Graslund, S. (2011) A roadmap to generate renewable protein binders to the human proteome, Nat Methods.

Mårten Sundberg

32. De Masi, F., Chiarella, P., Wilhelm, H., Massimi, M., Bullard, B., Ansorge, W., and Sawyer, A. (2005) High throughput production of mouse monoclonal antibodies using antigen microarrays, Proteomics 5, 4070-4081. 33. Mersmann, M., Meier, D., Mersmann, J., Helmsing, S., Nilsson, P., Graslund, S., Colwill, K., Hust, M., and Dubel, S. (2010) Towards proteome scale antibody selections using phage display, N Biotechnol 27, 118-128. 34. Pershad, K., Pavlovic, J. D., Graslund, S., Nilsson, P., Colwill, K., Karatt-Vellatt, A., Schofield, D. J., Dyson, M. R., Pawson, T., Kay, B. K., and McCafferty, J. (2010) Generating a panel of highly specific antibodies to 20 human SH2 domains by phage display, Protein Eng Des Sel 23, 279-288. 35. Abbas, A. K., and Lichtman, A. H. (2011) Basic immunology : functions and disorders of the immune system, Saunders, Philadelphia, Pa. ;. 36. Woof, J. M., and Burton, D. R. (2004) Human antibody-Fc receptor interactions illuminated by crystal structures, Nat Rev Immunol 4, 89-99. 37. Wingren, C., Steinhauer, C., Ingvarsson, J., Persson, E., Larsson, K., and Borrebaeck, C. A. (2005) Microarrays based on affinity-tagged single-chain Fv antibodies: sensitive detection of analyte in complex proteomes, Proteomics 5, 1281-1291. 38. Lipman, N. S., Jackson, L. R., Trudel, L. J., and Weis-Garcia, F. (2005) Monoclonal versus polyclonal antibodies: distinguishing characteristics, applications, and information resources, ILAR J 46, 258-268. 39. Rockberg, J., Schwenk, J. M., and Uhlen, M. (2009) Discovery of epitopes for targeting the human epidermal growth factor receptor 2 (HER2) with antibodies, Mol Oncol. 40. Angenendt, P., Glokler, J., Konthur, Z., Lehrach, H., and Cahill, D. J. (2003) 3D protein microarrays: performing multiplex immunoassays on a single chip, Anal Chem 75, 4368-4372. 41. Engvall, E., and Perlmann, P. (1971) Enzyme-linked immunosorbent assay (ELISA). Quantitative assay of immunoglobulin G, Immunochemistry 8, 871-874. 42. Uhlen, M., and Ponten, F. (2005) Antibody-based proteomics for human tissue profiling, Mol Cell Proteomics 4, 384-393. 43. Wide, L., Bennich, H., and Johansson, S. G. (1967) Diagnosis of allergy by an in- vitro test for allergen antibodies, Lancet 2, 1105-1107. 44. Yalow, R. S., and Berson, S. A. (1959) Assay of plasma insulin in human subjects by immunological methods, Nature 184 (Suppl 21), 1648-1649. 45. Burnette, W. N. (1981) "Western blotting": electrophoretic transfer of proteins from sodium dodecyl sulfate--polyacrylamide gels to unmodified nitrocellulose and radiographic detection with antibody and radioiodinated protein A, Anal Biochem 112, 195-203. 46. Warford, A., Howat, W., and McCafferty, J. (2004) Expression profiling by high- throughput immunohistochemistry, J Immunol Methods 290, 81-92. 47. Kohler, G., and Milstein, C. (1975) Continuous cultures of fused cells secreting antibody of predefined specificity. 1975, J Immunol 174, 2453-2455. 48. Porter, R. R. (1959) The hydrolysis of rabbit y-globulin and antibodies with crystalline papain, Biochem J 73, 119-126. 49. Bird, R. E., Hardman, K. D., Jacobson, J. W., Johnson, S., Kaufman, B. M., Lee, S. M., Lee, T., Pope, S. H., Riordan, G. S., and Whitlow, M. (1988) Single-chain antigen-binding proteins, Science 242, 423-426.

Protein microarrays for validation of affinity binder

50. McCafferty, J., Griffiths, A. D., Winter, G., and Chiswell, D. J. (1990) Phage antibodies: filamentous phage displaying antibody variable domains, Nature 348, 552-554. 51. Pansri, P., Jaruseranee, N., Rangnoi, K., Kristensen, P., and Yamabhai, M. (2009) A compact phage display human scFv library for selection of antibodies to a wide variety of antigens, BMC Biotechnol 9, 6. 52. Kwong, K. Y., Baskar, S., Zhang, H., Mackall, C. L., and Rader, C. (2008) Generation, affinity maturation, and characterization of a human anti-human NKG2D monoclonal antibody with dual antagonistic and agonistic activity, J Mol Biol 384, 1143-1156. 53. Ayoglu, B., Haggmark, A., Neiman, M., Igel, U., Uhlen, M., Schwenk, J. M., and Nilsson, P. (2011) Systematic antibody and antigen-based proteomic profiling with microarrays, Expert Rev Mol Diagn 11, 219-234. 54. Lin, H., and Cornish, V. W. (2002) Screening and selection methods for large- scale analysis of protein function, Angew Chem Int Ed Engl 41, 4402-4425. 55. Better, M., Chang, C. P., Robinson, R. R., and Horwitz, A. H. (1988) Escherichia coli secretion of an active chimeric antibody fragment, Science 240, 1041-1043. 56. Skerra, A., and Pluckthun, A. (1988) Assembly of a functional immunoglobulin Fv fragment in Escherichia coli, Science 240, 1038-1041. 57. Fernandez, L. A. (2004) Prokaryotic expression of antibodies and affibodies, Curr Opin Biotechnol 15, 364-373. 58. Nord, K., Gunneriusson, E., Ringdahl, J., Stahl, S., Uhlen, M., and Nygren, P. A. (1997) Binding proteins selected from combinatorial libraries of an alpha-helical bacterial receptor domain, Nat Biotechnol 15, 772-777. 59. Lofblom, J., Feldwisch, J., Tolmachev, V., Carlsson, J., Stahl, S., and Frejd, F. Y. (2010) Affibody molecules: engineered proteins for therapeutic, diagnostic and biotechnological applications, FEBS Lett 584, 2670-2680. 60. Stumpp, M. T., Binz, H. K., and Amstutz, P. (2008) DARPins: a new generation of protein therapeutics, Drug Discov Today 13, 695-701. 61. Zahnd, C., Wyler, E., Schwenk, J. M., Steiner, D., Lawrence, M. C., McKern, N. M., Pecorari, F., Ward, C. W., Joos, T. O., and Pluckthun, A. (2007) A designed ankyrin repeat protein evolved to picomolar affinity to Her2, J Mol Biol 369, 1015-1028. 62. Koide, A., Bailey, C. W., Huang, X., and Koide, S. (1998) The fibronectin type III domain as a scaffold for novel binding proteins, J Mol Biol 284, 1141-1151. 63. Koide, A., and Koide, S. (2007) Monobodies: antibody mimics based on the scaffold of the fibronectin type III domain, Methods Mol Biol 352, 95-109. 64. Schlehuber, S., and Skerra, A. (2005) Lipocalins in drug discovery: from natural ligand-binding proteins to "anticalins", Drug Discov Today 10, 23-33. 65. Proske, D., Blank, M., Buhmann, R., and Resch, A. (2005) Aptamers--basic research, drug development, and clinical applications, Appl Microbiol Biotechnol 69, 367-374. 66. Tuerk, C., and Gold, L. (1990) Systematic evolution of ligands by exponential enrichment: RNA ligands to bacteriophage T4 DNA polymerase, Science 249, 505-510. 67. Skerra, A. (2007) Alternative non-antibody scaffolds for molecular recognition, Curr Opin Biotechnol 18, 295-304. 68. Nygren, P. A., and Skerra, A. (2004) Binding proteins from alternative scaffolds, J Immunol Methods 290, 3-28.

Mårten Sundberg

69. Fodor, S. P., Read, J. L., Pirrung, M. C., Stryer, L., Lu, A. T., and Solas, D. (1991) Light-directed, spatially addressable parallel chemical synthesis, Science 251, 767-773. 70. Schena, M., Shalon, D., Davis, R. W., and Brown, P. O. (1995) Quantitative monitoring of gene expression patterns with a complementary DNA microarray, Science 270, 467-470. 71. Ekins, R. P. (1989) Multi-analyte immunoassay, J Pharm Biomed Anal 7, 155- 168. 72. Ekins, R. P. (1998) Ligand assays: from electrophoresis to miniaturized microarrays, Clin Chem 44, 2015-2030. 73. MacBeath, G., and Schreiber, S. L. (2000) Printing proteins as microarrays for high-throughput function determination, Science 289, 1760-1763. 74. Templin, M. F., Stoll, D., Schwenk, J. M., Potz, O., Kramer, S., and Joos, T. O. (2003) Protein microarrays: promising tools for proteomic research, Proteomics 3, 2155-2166. 75. Seong, S. Y., and Choi, C. Y. (2003) Current status of protein chip development in terms of fabrication and application, Proteomics 3, 2176-2189. 76. Chandra, H., Reddy, P. J., and Srivastava, S. (2011) Protein microarrays and novel detection platforms, Expert Rev Proteomics 8, 61-79. 77. Melioli, G., Bonifazi, F., Bonini, S., Maggi, E., Mussap, M., Passalacqua, G., Rossi, E. R., Vacca, A., and Canonica, G. W. (2011) The ImmunoCAP ISAC molecular allergology approach in adult multi-sensitized Italian patients with respiratory symptoms, Clin Biochem. 78. Schwenk, J. M., Lindberg, J., Sundberg, M., Uhlen, M., and Nilsson, P. (2007) Determination of binding specificities in highly multiplexed bead-based assays for antibody proteomics, Mol Cell Proteomics 6, 125-132. 79. Balboni, I., Chan, S. M., Kattah, M., Tenenbaum, J. D., Butte, A. J., and Utz, P. J. (2006) Multiplexed protein array platforms for analysis of autoimmune diseases, Annu Rev Immunol 24, 391-418. 80. Hueber, W., Tomooka, B. H., Batliwalla, F., Li, W., Monach, P. A., Tibshirani, R. J., Van Vollenhoven, R. F., Lampa, J., Saito, K., Tanaka, Y., Genovese, M. C., Klareskog, L., Gregersen, P. K., and Robinson, W. H. (2009) Blood autoantibody and cytokine profiles predict response to anti-tumor necrosis factor therapy in rheumatoid arthritis, Arthritis Res Ther 11, R76. 81. Crompton, P. D., Kayala, M. A., Traore, B., Kayentao, K., Ongoiba, A., Weiss, G. E., Molina, D. M., Burk, C. R., Waisberg, M., Jasinskas, A., Tan, X., Doumbo, S., Doumtabe, D., Kone, Y., Narum, D. L., Liang, X., Doumbo, O. K., Miller, L. H., Doolan, D. L., Baldi, P., Felgner, P. L., and Pierce, S. K. (2010) A prospective analysis of the Ab response to Plasmodium falciparum before and after a malaria season by protein microarray, Proc Natl Acad Sci U S A 107, 6958-6963. 82. Xu, Y., Bruno, J. F., and Luft, B. J. (2008) Profiling the humoral immune response to Borrelia burgdorferi infection with protein microarrays, Microb Pathog 45, 403-407. 83. Viswanathan, K., Kilcullen, N., Morrell, C., Thistlethwaite, S. J., Sivananthan, M. U., Hassan, T. B., Barth, J. H., and Hall, A. S. (2010) Heart-type fatty acid- binding protein predicts long-term mortality and re-infarction in consecutive patients with suspected acute coronary syndrome who are troponin-negative, J Am Coll Cardiol 55, 2590-2598. 84. Schroder, C., Jacob, A., Tonack, S., Radon, T. P., Sill, M., Zucknick, M., Ruffer, S., Costello, E., Neoptolemos, J. P., Crnogorac-Jurcevic, T., Bauer, A.,

Protein microarrays for validation of affinity binder

Fellenberg, K., and Hoheisel, J. D. (2010) Dual-color proteomic profiling of complex samples with a microarray of 810 cancer-related antibodies, Mol Cell Proteomics 9, 1271-1280. 85. Schwenk, J. M., Igel, U., Neiman, M., Langen, H., Becker, C., Bjartell, A., Ponten, F., Wiklund, F., Gronberg, H., Nilsson, P., and Uhlen, M. (2010) Toward next generation plasma profiling via heat-induced epitope retrieval and array- based assays, Mol Cell Proteomics 9, 2497-2507. 86. Breitling, F., Nesterov, A., Stadler, V., Felgenhauer, T., and Bischoff, F. R. (2009) High-density peptide arrays, Mol Biosyst 5, 224-234. 87. Cretich, M., Damin, F., Pirri, G., and Chiari, M. (2006) Protein and peptide arrays: recent trends and new directions, Biomol Eng 23, 77-88. 88. Chiari, M., Cretich, M., Corti, A., Damin, F., Pirri, G., and Longhi, R. (2005) Peptide microarrays for the characterization of antigenic regions of human chromogranin A, Proteomics 5, 3600-3603. 89. Janzi, M., Odling, J., Pan-Hammarstrom, Q., Sundberg, M., Lundeberg, J., Uhlen, M., Hammarstrom, L., and Nilsson, P. (2005) Serum microarrays for large scale screening of protein levels, Mol Cell Proteomics 4, 1942-1947. 90. Pawlak, M., Schick, E., Bopp, M. A., Schneider, M. J., Oroszlan, P., and Ehrat, M. (2002) Zeptosens' protein microarrays: a novel high performance microarray platform for low abundance protein analysis, Proteomics 2, 383-393. 91. Kusnezow, W., and Hoheisel, J. D. (2003) Solid supports for microarray immunoassays, J Mol Recognit 16, 165-176. 92. Angenendt, P. (2005) Progress in protein and antibody microarray technology, Drug Discov Today 10, 503-511. 93. Butler, J. E. (2000) Solid supports in enzyme-linked immunosorbent assay and other solid-phase immunoassays, Methods 22, 4-23. 94. Butler, J. E., Ni, L., Brown, W. R., Joshi, K. S., Chang, J., Rosenberg, B., and Voss, E. W., Jr. (1993) The immunochemistry of sandwich ELISAs--VI. Greater than 90% of monoclonal and 75% of polyclonal anti-fluorescyl capture antibodies (CAbs) are denatured by passive adsorption, Mol Immunol 30, 1165-1175. 95. Vijayendran, R. A., and Leckband, D. E. (2001) A quantitative assessment of heterogeneity for surface-immobilized proteins, Anal Chem 73, 471-480. 96. Wilchek, M., and Miron, T. (2003) Oriented versus random protein immobilization, J Biochem Biophys Methods 55, 67-70. 97. Sanchez-Carbayo, M. (2006) Antibody arrays: technical considerations and clinical applications in cancer, Clin Chem 52, 1651-1659. 98. Usui-Aoki, K., Shimada, K., and Koga, H. (2007) A novel antibody microarray format using non-covalent antibody immobilization with chemiluminescent detection, Mol Biosyst 3, 36-42. 99. Haab, B. B. (2003) Methods and applications of antibody microarrays in cancer research, Proteomics 3, 2116-2122. 100. Han, M. K., Oh, Y. H., Kang, J., Kim, Y. P., Seo, S., Kim, J., Park, K., and Kim, H. S. (2009) Protein profiling in human sera for identification of potential lung cancer biomarkers using antibody microarray, Proteomics 9, 5544-5552. 101. Haab, B. B., Dunham, M. J., and Brown, P. O. (2001) Protein microarrays for highly parallel detection and quantitation of specific proteins and antibodies in complex solutions, Genome Biol 2, RESEARCH0004. 102. Wingren, C., and Borrebaeck, C. A. (2008) Antibody microarray analysis of directly labelled complex proteomes, Curr Opin Biotechnol 19, 55-61.

Mårten Sundberg

103. Wingren, C., Ingvarsson, J., Dexlin, L., Szul, D., and Borrebaeck, C. A. (2007) Design of recombinant antibody microarrays for complex proteome analysis: choice of sample labeling-tag and solid support, Proteomics 7, 3055-3065. 104. Orchekowski, R., Hamelinck, D., Li, L., Gliwa, E., vanBrocklin, M., Marrero, J. A., Vande Woude, G. F., Feng, Z., Brand, R., and Haab, B. B. (2005) Antibody microarray profiling reveals individual and combined serum proteins associated with pancreatic cancer, Cancer Res 65, 11193-11202. 105. Sanchez-Carbayo, M., Socci, N. D., Lozano, J. J., Haab, B. B., and Cordon-Cardo, C. (2006) Profiling bladder cancer using targeted antibody arrays, Am J Pathol 168, 93-103. 106. Kusnezow, W., Banzon, V., Schroder, C., Schaal, R., Hoheisel, J. D., Ruffer, S., Luft, P., Duschl, A., and Syagailo, Y. V. (2007) Antibody microarray-based profiling of complex specimens: systematic evaluation of labeling strategies, Proteomics 7, 1786-1799. 107. Lundberg, E., Sundberg, M., Graslund, T., Uhlen, M., and Svahn, H. A. (2007) A novel method for reproducible fluorescent labeling of small amounts of antibodies on solid phase, J Immunol Methods 322, 40-49. 108. Eriksson, C., Agaton, C., Kange, R., Sundberg, M., Nilsson, P., Ek, B., Uhlen, M., Gustafsson, M., and Hober, S. (2006) Microfluidic analysis of antibody specificity in a compact disk format, J Proteome Res 5, 1568-1574. 109. Rimini, R., Schwenk, J. M., Sundberg, M., Sjoberg, R., Klevebring, D., Gry, M., Uhlen, M., and Nilsson, P. (2009) Validation of serum protein profiles by a dual antibody array approach, J Proteomics 73, 252-266.