Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=kmab20

Download by: [University Rey Juan Carlosi] Date: 21 October 2015, At: 03:32

mAbs

ISSN: 1942-0862 (Print) 1942-0870 (Online) Journal homepage: http://www.tandfonline.com/loi/kmab20

The European antibody network's practical guideto finding and validating suitable antibodies forresearch

Giovanna Roncador, Pablo Engel, Lorena Maestre, Amanda P. Anderson,Jacqueline L. Cordell, Mark S. Cragg, Vladka Č. Šerbec, Margaret Jones, VandaJ. Lisnic, Leonor Kremer, Demin Li, Friedrich Koch-Nolte, Núria Pascual, Jose-Ignacio Rodríguez-Barbosa, Ruurd Torensma, Helen Turley, Karen Pulford &Alison H. Banham

To cite this article: Giovanna Roncador, Pablo Engel, Lorena Maestre, Amanda P. Anderson,Jacqueline L. Cordell, Mark S. Cragg, Vladka Č. Šerbec, Margaret Jones, Vanda J. Lisnic, LeonorKremer, Demin Li, Friedrich Koch-Nolte, Núria Pascual, Jose-Ignacio Rodríguez-Barbosa, RuurdTorensma, Helen Turley, Karen Pulford & Alison H. Banham (2015): The European antibodynetwork's practical guide to finding and validating suitable antibodies for research, mAbs, DOI:10.1080/19420862.2015.1100787

To link to this article: http://dx.doi.org/10.1080/19420862.2015.1100787

© 2015 The Author(s). Published withlicense by Taylor & Francis Group, LLC©Alison Banham

Accepted author version posted online: 29Sep 2015.Published online: 29 Sep 2015.

Submit your article to this journal Article views: 86

View related articles View Crossmark data

The European antibody network’s practical guide to findingand validating suitable antibodies for research

Giovanna Roncador1, Pablo Engel2, Lorena Maestre1, Amanda P. Anderson3, Jacqueline L. Cordell4, Mark S. Cragg5,Vladka �C. �Serbec6, Margaret Jones3, Vanda J. Lisnic7, Leonor Kremer8, Demin Li3, Friedrich Koch-Nolte9, N�uria Pascual10,Jose-Ignacio Rodr�ıguez-Barbosa11, Ruurd Torensma12, Helen Turley3, Karen Pulford3, and Alison H. Banham3,*1Monoclonal Antibody Unit; Spanish National Cancer Research Center; Madrid, Spain; 2Immunology Unit, Department of Cell Biology; Immunology and

Neurosciences; Medical School, University of Barcelona; Spain; 3NDCLS, Radcliffe Department of Medicine; University of Oxford; Oxford, UK; 4OxFabs; University

of Oxford; Oxford UK; 5Antibody and Vaccine Group; Cancer Sciences Unit; University of Southampton; Faculty of Medicine; General Hospital, Southampton, UK;6Center for the Production of Diagnostic Reagents and for Research; Blood Transfusion Center of Slovenia; Ljubljana, Slovenia; 7Center for Proteomics; Faculty of

Medicine; University of Rijeka; Rijeka, Croatia; 8Department of Immunology and Oncology; Spanish National Center for Biotechnology; Consejo Superior de Investi-

gaciones Cient�ıficas (CNB/CSIC); Madrid, Spain; 9Institute of Immunology; University Medical Center; Hamburg, 20246 Germany; 10Custom Antibody Service

(CAbS); IQAC-CSIC/CIBER-BBN; Barcelona, Spain; 11Section of Immunobiology; Institute of Biomedicine; University of Leon; L�eon, Spain; 12Department of

Tumorimmunology; Radboud University Medical Center; Nijmegen, The Netherlands

Antibodies are widely exploited asresearch/diagnostic tools and therapeu-

tics. Despite providing exciting researchopportunities, the multitude of availableantibodies also offers a bewildering array ofchoice. Importantly, not all companiescomply with the highest standards, andthus many reagents fail basic validationtests. The responsibility for antibodiesbeing fit for purpose rests, surprisingly,with their user. This paper condenses theextensive experience of the EuropeanMonoclonal Antibody Network to helpresearchers identify antibodies specific fortheir target antigen. A stepwise strategy isprovided for prioritising antibodies andmaking informed decisions regarding fur-ther essential validation requirements.Web-based antibody validation guides pro-vide practical approaches for testing anti-body activity and specificity. We aim toenable researchers with little or no priorexperience of antibody characterization tounderstand how to determine the suitabil-ity of their antibody for its intended pur-pose, enabling both time and cost effectivegeneration of high quality antibody-baseddata fit for publication.

Introduction

The ability of antibodies to bind specif-ically and with high affinity to their target

antigens has led to the widespread exploi-tation of these reagents within thescientific, clinical and commercial com-munities. Antibodies are commonly usedas research tools to study the expression,localization or function of their targets,and they are one of the most prolific clas-ses of new biologics for clinical therapy,particularly in the fields of oncology, rheu-matology, transplantation and hematol-ogy.1 Given the importance of thesereagents and their impact within the scien-tific and medical communities, it is abso-lutely imperative that all researchers areconfident that any antibodies they use areof the correct specificity. The aim of thisreview is to share the collective experienceof a panel of international experts in anti-body production and characterization toenable the research community not onlyto obtain relevant antibodies but to ensurethe validity of these reagents.

Historically, monoclonal antibody pro-duction and their characterization was per-formed mainly by specialized laboratories.Antibodies were generally raised against mix-tures of proteins and the exact antigen wasoften unknown. In 1982, a numbering sys-tem of unique clusters, each designated acluster of differentiation (CD), for identify-ing antibody reactivity against epitopes onthe surface molecules of leucocytes wasdevised at the first meeting of the HumanLeukocyte Differentiation Antigens (HLDA)

Keywords: BLAST searches, CiteAb,EuroMAbNet, flow cytometry, genesilencing, HLDA workshops, immunohis-tochemistry, monoclonal antibodies, vali-dation studies, western blotting

Abbreviations: BLAST: Basic Local Align-ment Search Tool; CD: Cluster of Differ-entiation; c-FLIP: cellular FLICE-likeinhibitory protein; ELISA: enzyme-linkedimmunoabsorbent assay; FCM: flowcytometry; FOXP1: forkhead box proteinP1; HLDA: Human Leukocyte Differenti-ation Antigens Workshop; IP: immuno-precipitation; NCBI: National Center forBiotechnology Information; SDS-PAGE:sodium dodecyl sulfate polyacrylamide gelelectrophoresis.

© Alison Banham*Correspondence to: Alison Banham;Email: alison.banham@ndcls.ox.ac.uk

Submitted: 08/03/2015

Revised: 09/08/2015

Accepted: 09/22/2015

http://dx.doi.org/10.1080/19420862.2015.1100787

This is an Open Access article distributed under theterms of the Creative Commons Attribution License(http://creativecommons.org/licenses/by/3.0/),which permits unrestricted use, distribution, andreproduction in any medium, provided the originalwork is properly cited. The moral rights of thenamed author(s) have been asserted.

www.tandfonline.com 1mAbs

mAbs 0:0, 1--10; October 2015; Published with license by Taylor & Francis Group, LLCPERSPECTIVE

Dow

nloa

ded

by [

Uni

vers

ity R

ey J

uan

Car

losi

] at

03:

32 2

1 O

ctob

er 2

015

Workshop. This system has served success-fully as a global antibody classificationscheme, and the numbers of CD have risendramatically during the subsequent 9HLDA workshops. It was the exchange ofantibodies among laboratories and theirindependent evaluation by several expertresearch groups using a range of differenttechniques that enabled the utility andreproducibility of these early reagents to beso well characterized.2

During the last decade, the sequencingof the human genome and the success ofantibody-based therapeutics have had aprofound effect on the production andavailability of both polyclonal and mono-clonal antibodies. Many academic groups,including large-scale efforts by the Swed-ish Human Protein Atlas3,4 and the Ger-man Antibody Factory,5 along withcommercial companies have adopted a‘gene-centric’ approach in an effort togenerate antibodies to specific target anti-gens. This has expanded to include anti-bodies to distinguish isoforms of the sameprotein, detect post-translationally modi-fied forms, such as protein phosphoryla-tion and, more importantly, those withthe ability to modulate biological func-tions. There has also been a correspondinggrowth in commercial providers offeringcustomised monoclonal and polyclonalantibody production services. It should benoted that antibodies with functionalactivity against their target antigen, andthose against non-protein targets or post-translational modifications are not consid-ered within the scope of this article. How-ever, many of the principles describedbelow apply equally to these reagents.

High throughput antibody production,and the lucrative returns generated bysales, have led to a flood of antibodiesbecoming widely available to the researchcommunity. There are more than 300antibody suppliers providing >500,000antibodies for the research and clinicalmarkets (www.the-scientist.com/?articles.view/articleNo/32042/title/Antibodies-User-Survey/).

Additionally, many antibodiesdescribed in the scientific literature, andsold commercially, were produced by aca-demic research groups. While some com-panies do make substantial efforts tovalidate their antibodies and provide

detailed product literature showing theresults, not all antibodies from commer-cial sources are fully validated and fit forpurpose. Researchers are frustrated by thedifficulties caused by poorly validatedantibodies and sometimes also by batch-to-batch variability. Even the same mono-clonal antibody from different suppliersmay exhibit variability in performance.This has recently led a coalition of 112researchers to propose that polyclonalantibodies should be phased out and allmonoclonal antibodies should be definedby their antibody gene sequences andexpressed as recombinant proteins toenable standardization of reagents.6 Whilethis will indeed identify individual anti-bodies (something that is often impossiblebetween current suppliers), it will notaddress whether the antibody has thedesired activity or specificity. Further-more, very similar antibodies may beencoded by different sequences (e.g.,humanization variants), while differentpost-translational modifications such aspatterns of glycosylation, which can beregulated by the production methodology,can significantly affect the functionality ofa therapeutic antibody.

Others in the scientific community havestressed the need to improve antibody valida-tion because of the adverse effect that poorquality reagents have on research budgets,the research time wasted on ineffectiveexperiments and the publication of inaccu-rate data in the scientific literature.7-10 Sev-eral academic reviews have discussed how tovalidate antibodies; however, these have beenfor specific fields of research, such as neuro-science;11 for validation of antibodies to asingle protein, e.g., c-FLIP12 or to families ofproteins, including central nervous systemion channel proteins13 and G-protein-cou-pled receptors;14 or validation for specifictechniques such as immunohistochemis-try,15-18 Western blotting (WB),19 reversephase protein arrays20 or chromatin immu-noprecipitation (ChIP).21 There have alsobeen helpful recommendations on how toreport research antibody use to increaseexperimental reproducibility,22 and somejournals are now adopting antibody valida-tion requirements for their publishedarticles.23

In reality, it is not practical to expectantibodies to be validated for all purposes,

and even a partially validated antibody toa new antigen is of potential value to theresearch community. However, thereneeds be more clarity as to exactly whichreagent is being described, its supplier(s)and the validation that has been per-formed. Unaltered original clone namesshould be consistently used for monoclo-nal antibodies and catalog and batch num-bers reliably adopted for polyclonalantibodies. Furthermore, the validationdata (including images) needs to be avail-able for independent review, either in thepublished literature or online, includingantibody databases and commercial prod-uct datasheets. Also, mechanisms must beestablished that: 1) enable meaningful dia-log between antibody suppliers and users;and 2) allow issues about reagents thatmight be a cause for concern to be raisedin a transparent fashion. For example,some companies already encourage andprovide online end user reviews of theirreagents. Some issues may be reagentbased while others may be specific to indi-vidual users. Some companies are willingto independently test off-target activitiesthat are identified for their antibodies. Ina recent case raised by a EuroMAbNetmember, the supplier not only revised theproduct sheet but also contacted all theclients who had previously purchased thereagent. The emergence of independentthird-party test sites that offer antibodyvalidation services indicates that there is ademand for further validation.

EuroMAbNet is the first European net-work of academic laboratories specializedin the production and use of monoclonalantibodies. The motivation behind thewriting of this article is to share our practi-cal experience of antibody production,selection and validation with other mem-bers of the research community. Thisspans several decades, varying experimen-tal techniques and includes the use ofboth undefined antigens (e.g., cell lysates)and gene-centric approaches to generateantibodies. We have worked with poly-clonal and monoclonal antibodies fromour own laboratories and from other sour-ces, including academic groups and com-panies. Furthermore, EuroMAbNetmembers have collaborated with commer-cial antibody suppliers to license theirown antibodies for sale to the research

2 Volume 0 Issue 0mAbs

Dow

nloa

ded

by [

Uni

vers

ity R

ey J

uan

Car

losi

] at

03:

32 2

1 O

ctob

er 2

015

community. This makes us particularlywell positioned to appreciate the view-points of both the academic communityand commercial antibody suppliers.

Step-by-step guide to antibodyselection and validation

Here, we provide our overview to thefundamental principles underlying effec-tive antigen design, antibody selection andvalidation (Fig. 1). This information isprovided in a format to help researchers

who need to source or validate antibodiesacross a broad spectrum of research fieldsand technical applications. However, it isimportant to emphasize that there aremany widely used well-characterized anti-bodies that require no additional valida-tion. This is generally evident from theirfrequent citation in the scientific literatureand their well-populated product informa-tion sheets. CiteAb is particularly helpfulfor identifying these reagents, which areranked by their frequency of literature

citation (www.citeab.com/). However, itshould be noted that citations accrue overtime, thus the initial antibodies generatedagainst an antigen may have accumulatedcitations (sometimes these reagents mayeven be suboptimal but were used due tothe lack of alternatives) while there maybe newer and more effective reagents avail-able that inevitably have fewer citations.However, researchers should always err onthe side of caution and confirm that priorvalidation has been performed, particu-larly if intending to use these antibodies innew techniques, tissues or different spe-cies. Web-based validation guidelines pro-viding additional video tutorials andillustrated examples, including an anti-body catalog with further validation files,are available at www.euromabnet.com.

STEP 1: Define your initialrequirements

When sourcing antibodies, decisionsmust first be made about the: 1) targetantigen, 2) necessary antibody characteris-tics, and 3) proposed technical applica-tions (for a summary and useful websites,see Table 1). Many antibodies are pur-chased purely because the manufacturerrecommends them for a particular applica-tion, while antibodies that might be morespecific but were not extensively tested areignored. Some basic bioinformatics at theoutset can save substantial amounts oftime and money by enabling prioritizationof existing reagents, and the identificationof those with the minimal requirement foradditional downstream validation.

1A. Define the target antigenrequirements

Start by identifying the target antigen,its approved nomenclature and any alter-native names that might also help identifyexisting reagents in the scientific literature.Then, obtain the reference (canonical)protein sequence and identify if variantsexist, e.g., those produced by alternativesplicing, proteolytic cleavage, post-transla-tional modification. Information onknown variants is available on the web,e.g., in Aceview, UniProt and the scientificliterature. Decide whether it is importantto detect all variants or only one in partic-ular, and whether distinguishing domainswith different subcellular localizations

Figure 1. Overview of the fundamental principles involved underlying effective antigen design,antibody selection and validation. A step-by-step guide to defining the target antigen, identifyingrelevant existing antibodies and their specificity to the accurate dissemination of data arising fromtheir use.

www.tandfonline.com 3mAbs

Dow

nloa

ded

by [

Uni

vers

ity R

ey J

uan

Car

losi

] at

03:

32 2

1 O

ctob

er 2

015

may also be critical, e.g., labeling theextracellular versus intracellular portion ofmembrane proteins. For example, theAbnova FOXP1 monoclonal antibody(M01), clone 4E3-G11, (Catalog numberH00027086-M01) product sheetdescribed the immunogen as a full-lengthFOXP1 recombinant protein of 115amino acids, yet the FOXP1 referencesequence was 677 amino acids. Furthervalidation defined the epitope within theunique C-terminus of FOXP1 isoform 2and demonstrated this reagent was unableto detect the full-length FOXP1 protein,which exhibited a significantly differentexpression pattern.24

Perform a search of the non-redundantprotein sequences using the National Centerfor Biotechnology Information (NCBI)Basic Local Alignment Search Tool (BLAST;http://blast.ncbi.nlm.nih.gov/Blast.cgi)using the reference protein sequence. Ourguidelines include an online tutorial forthose with no experience of BLAST(http://www.euromabnet.com/guidelines/example1.php). This simple analysis willenable further identification of proteinvariants, regions of the protein that arehighly conserved in orthologous speciesand regions that share sequence identitywith closely related proteins. This infor-mation is key to using this bioinformat-ics data to help prioritise antibodies forfurther study. The aim of this process isto define whether there are: 1) distinctregions of the chosen protein containinglinear epitopes that are unique to theindividual antigen; and 2) regions of the

target antigen where antibodies arehighly likely to have cross-reactivity withother proteins with which they sharesequence identity. Such reagents may bevery valuable, but are likely to requiremore validation of their specificity.

1B. Define antibody requirementsNext, determine the necessary antibody

requirements, e.g., monoclonal vs. poly-clonal antibody, isotype requirements(further information in Table 2). Whilereported activity in a specific technique isoften one of the main criteria used toidentify antibodies, the intended applica-tion may be less important than specific-ity, particularly when there are no well-characterized antibodies to the target pro-tein. Also, consider whether what isknown about the technical applicationscan help to prioritise reagents for othertechniques. For example, an antibody thatrecognizes a formalin-resistant epitope forimmunohistochemistry may also work inanother technique using formalin fixation,such as ChIP.

Antibodies raised against synthetic pep-tides recognize a linear epitope. Hence,anti-peptide antibodies usually work wellin WB analyses, but not necessarily inassays with native proteins, such as flowcytometry (FCM), ELISA and IP. Con-versely, antibodies raised by cDNA or cellimmunization or by immunizations withnative proteins often work well in FCM,ELISA and IP, but not in WB. With anti-bodies recognizing a native epitope, spe-cial techniques, such as native SDS-

PAGE, refolding on nitrocellulose mem-branes, dot blot, may allow binding onimmunoblots.

STEP 2: Identification and availabilityof existing antibodies

Having defined the fundamentalrequirements for the antibody reagent, avariety of methods can then be used tofind existing antibodies. The publishedscientific literature is the best startingplace for discovering which antibodieshave already been used to study a targetprotein. If little information is available,then try using the antigen name, the term‘antibody’ and the required technique ingeneral search engines (e.g., Google) thatexamine whole manuscripts, commercialdatasheets, theses, patents and abstracts.Bear in mind the importance of antibodyavailability. Despite scientific conventionand journal publishing conditions, not allresearch laboratories will agree to supplyreagents. Collaborations with biotechnol-ogy or pharmaceutical companies mayrequire very restricted terms and a lot ofpaperwork. Some batches of polyclonalantibodies may be in short supply or nolonger in existence.

Examination of online antibody list-ings from commercial suppliers may reveala much wider range of reagents comparedwith thumbing through one or 2 catalogsavailable in the laboratory. Start byemploying web-based resources Biocom-pare (www.biocompare.com/Antibodies/),the Antibody Resource Page (www.antibodyresource.com) and Linscott’s

Table 1. Bioinformatics analysis to define target antigen

Identify the target antigen Be aware that molecules can occasionally share a common name so search for theapproved nomenclature at http://www.genenames.org/

Find alternative gene names Useful for a review of the historical literaturehttps://www.ncbi.nlm.nih.gov/IEB/Research/Acembly/ or http://www.genecards.org/

Obtain ‘canonical’ protein sequence http://www.uniprot.org/ Information also provided on additional isoforms

Further define antigen restrictions E.g. specificity for, or commonality across: isoforms, functional domains, processedforms, domains with different subcellular localization (e.g., extracellular versusintracellular portion), post-translational modifications

Is cross species recognition of the antigen critical? Use BLAST to identify protein sequence identity across orthologous specieshttp://blast.ncbi.nlm.nih.gov/Blast.cgi

Search for related proteins A BLAST search will also identify related proteins that share regionsof identity with the target antigen at http://blast.ncbi.nlm.nih.gov/Blast.cgi

Define ideal epitope(s) Unique regions in the target antigen conferring specificity or sometimesthose regions conferring cross reactivity

4 Volume 0 Issue 0mAbs

Dow

nloa

ded

by [

Uni

vers

ity R

ey J

uan

Car

losi

] at

03:

32 2

1 O

ctob

er 2

015

Directory (www.linscottsdirectory.com),and search tools such as CiteAb that rankantibodies by citations (www.citeab.com)and Antibodypedia (www.antibodypedia.com). Using official nomenclature (www.genenames.org) for the antigen, and alsoother names in common usage, canincrease the number of antibodies identi-fied. Having confidence in the reputationof the supplier may also help when choos-ing antibodies; an Antibodies User Surveyconducted in 2012 by The Scientist andFrost and Sullivan provides some helpfulinformation on the suppliers most fre-quently used and those with the highestcustomer ratings (www.the-scientist.com/?articles.view/articleNo/32042/title/Antibodies-User-Survey/).

Start by identifying individual antibod-ies, bearing in mind that there may bemultiple suppliers providing the samereagent (tips on how to do this are detailedin Table 2). Each antibody should have aunique identifier that distinguishes it fromother antibodies, and this should not bealtered or abbreviated by suppliers orusers. Abbreviations and changes of clonenames make it difficult to link commercialproducts with published validation of the

antibody. Critically, efforts should bemade by researchers, peer reviewers andjournals to ensure that unique identifiersare used in any publications arising fromuse of antibody reagents. Currently, manylaboratories fail to identify the exact anti-body used in their studies and justdescribe whether it was a monoclonal orpolyclonal antibody and the supplier. Thesupplier may at the time, or in the future,license multiple antibodies to the sameantigen making it difficult to sourcereagents or even reproduce published data.

While monoclonal antibodies are usuallyreferenced by their clone name, many poly-clonal antibodies are not adequately identi-fied, and it is possible to unknowingly buythe same reagent from different distributors.There is also rarely any nomenclature usedto differentiate between batches of the poly-clonal antibody produced in different ani-mals. As the same immunogen can generatevery different immune responses in anothermember of the same animal species, whensourcing polyclonal antibodies it is advisableto find up-to-date references to the reagentand possibly to buy several aliquots fromthe same batch to ensure consistency duringa project.

STEP 3: Prioritization of candidateantibodies

When there is a well-characterizedreagent, or very few available antibodies,then little prioritization is needed prior toobtaining the reagent(s) and startingexperimental testing. However, few aca-demic laboratories have the resources totest an extensive panel of antibodies. Becareful to check whether literature referen-ces provided on antibody product sheetsare actually those using the particular anti-body described. Choosing antibodies thatare most likely to meet the specificrequirements defined in STEP 1 is criticalwithin an extensive panel of reagents, par-ticularly if nothing more than the infor-mation provided on the supplier’sdatasheets is available to guide selection.The following section will describe how toproceed when there is no existing pub-lished literature.

Review the product literature to iden-tify the reagents that best match the anti-body requirements devised in STEP 1.Pay particular attention to the quantityand quality of the existing antibody vali-dation data (further information on howto approach this is described in STEP 4)

Table 2. Determine antibody requirements

Monoclonal (mAb) vs. polyclonal antibody(pAb)

mAb - single epitope, single isotype, unlimited supply of identical reagent, identifiable (usually) by clonename. pAb - multiple epitopes, better for some techniques as recognizing a range of different epitopescan increase the number of suitable technical applications and enhance signals by enabling moreantibodies to bind the same antigen molecule and by forming large precipitating lattices. Conversely,there is more risk of cross-reactive epitopes if the immunogen shares identity with other proteins becausea polyclonal antibody will recognize a range of different epitopes. Batch variability, caused by limitedquantity and differences in the immune response during subsequent production in another animal,requires additional validation. Sometimes it can be hard to conclusively identify reagents used in theliterature or distinguish them among those offered by multiple suppliers.

Isotype Different isotypes may be useful for experiments using multiple antibodies e.g. enabling detection usingisotype specific secondary reagents. In both in vitro and in vivo studies, ‘isotype-dependent’ or ‘isotypespecific’ antibody functionality may be important.

Host species As above it can help to have different species for multiple labeling. Also the use of same species antibodiesand tissues needs additional strategies to avoid secondary antibodies detecting endogenousimmunoglobulins, this can be avoided if the antibody is raised in a different species.

Define intended technical applications Consider the intended use in technical applications such as IHC, IHC-P, ICC, IF, FCM, ChIP, WB, ELISA, IP. Absthat recognize linear epitopes (anti-peptide Abs) tend to work well for WB and IHC-P. Abs recognizingnative epitopes tend to work well for IP, ELISA and FCM (Abs raised against native proteins, cDNAimmunization, cell-based immunogens). While antibodies against extracellular epitopes are commonlyused for FCM and for antibody therapeutics, FCM is also increasingly used for antibodies recognizing bothintracellular and intranuclear epitopes.

Identify individual antibodies Refine the list of antibodies using their unique identifiers to remove duplicates such as the same antibodybeing available from different suppliers. Use the clone name to identify monoclonal antibodies. Forpolyclonal antibodies, bear in mind that those with an identical host species, immunogen and/or imagesin validation data may be the same reagent.

Immunohistochemistry (IHC), Immunohistochemistry on formalin fixed and paraffin embedded tissues (IHC-P), immunocytochemistry (ICC), Immunofluores-cence (IF), flow cytometry (FCM), chromatin immunoprecipitation (ChIP), Western blotting (WB), enzyme-linked immunosorbent assay (ELISA), immunopre-cipitation (IP).

www.tandfonline.com 5mAbs

Dow

nloa

ded

by [

Uni

vers

ity R

ey J

uan

Car

losi

] at

03:

32 2

1 O

ctob

er 2

015

and whether it has been performed in thedesired species, tissue or application. Doesthis data fit with what is known about theexpression pattern and potential variantsof the antigen?

Examining the immunogen can, inmany instances, determine which antibod-ies are likely to have the closest match tothe specificity requirements identifiedfrom the bioinformatics analyses in STEP1. Some suppliers do not disclose theimmunogen and consider this to be pro-prietary information. Many researchersfeel that suppliers should be required todisclose detailed information on theimmunogen used for antibody produc-tion, and it has been proposed that with-holding this information preventsindependent scientific replication of thereagent. There have even been suggestionsthat these reagents “are not fit for scientificwork” and that “work done with them willnot be publishable.”8 However, in reality,there is no guarantee that the same immu-nogen will generate a comparable anti-body when used to immunise a differentanimal, and immunization with a mixtureof proteins was used to produce many ofthe widely used CD antibodies.

The key disadvantage of not havingspecific information regarding the immu-nogen is that more extensive validation isoften necessary to confirm the specificityof antigen recognition. While suppliers

may refuse to disclose the exact details oftheir immunogen, some have been willingto provide additional general informationon request. For example, when needing anantibody to target a defined region or toavoid a specific region (e.g., to avoid anti-gens with high sequence homology), anti-body manufacturers are often willing todisclose whether their immunogen mapsto these defined regions. It is in their inter-ests for the research community to suc-cessfully use their reagents and convinceothers to do the same by publishing good-quality data.

Bear in mind that antibodies distinguish-ing protein isoforms might quite legitimatelyhave very different tissue labeling patterns,multiple or variable banding patterns in WBor different subcellular staining. Several pub-lished antibody validation schemes suggestdiscarding reagents that recognize more thana single protein species by WB. Having thisinformation on the protein under study mayhelp to predict the forms of the protein thatare likely to be detected by individualantibodies.

Choosing the region of the proteinwith most sequence conservation (highestidentity) increases the chances of antibod-ies recognizing the target antigen acrossdifferent species. However, species cross-reactivity must be validated experimen-tally, particularly if there are differences inthe protein sequences. The regulatory T-

cell marker FOXP3 is highly conservedacross species, and yet one or 2 aminoacid differences within epitopes were suffi-cient to determine whether antibodies arecross-reactive or species specific.25

At this point, it is also worth investigat-ing additional antibody validation thatmay be available through online antibodydatabases, which are described at the endof STEP 5.

STEP 4: Antibody ValidationBefore using an antibody, without any

further independent characterization, itremains ultimately the responsibility ofthe researcher to review the existing vali-dation data for its specificity for the targetantigen. It is essential that this data relatesto the methodology being used (as anindividual antibody may not be effectiveacross all techniques), the species/tissuetype under study and the technical appli-cation being used. Detailed practicalguides on how to validate antibodies fordifferent technical applications and acrossspecies are provided online (www.euromabnet.com/guidelines/), and an overviewis described below and in Table 4. An iso-type and host species matched controlantibody, without any reactivity in yourexperimental system, should also beobtained and used as a negative experi-mental control for non-specific reactivity.

Table 3. Prioritisation of available antibodies

Validation The ideal situation is to identify a well-characterized antibody that is widely described in the scientificliterature and that already fulfils all the necessary technical requirements. Often a more pragmaticapproach is needed, balancing likely specificity versus suitability for defined techniques.

Review product data sheets Assess the quantity and quality of the validation data and whether this has been performed using thepreferred technical applications. STEP 4 and Table 4 provide more information on how to assess antibodyvalidation.

Immunogen The immunogen is one of the most critical factors determining the specificity and thus likely suitability of anantibody.

Technical applications Established functionality in the desired technical application is desirable. If an otherwise desirable antibody isnot recommended for a technical application it is worth contacting the supplier to find out whether it isknown to be unsuitable or whether the absence of data just means it hasn’t yet been tested in theapplication.

Match antibody data withexisting information

Even when there is no published expression data in the literature there is likely to be information that ispublically available as to where the transcript is expressed. BioGPS provides transcript expression,including meta-analysis of publically available microarray datasets: http://biogps.org/ RNA Seq data fromcancer tissues and cell lines is available through cbioportal: http://www.cbioportal.org/public-portal/Recent drafts of the human proteome provide searchable online resources where protein expressionacross a range of tissues, determined by mass spectrometry, can be searched.http://www.humanproteomemap.org; http://www.proteomicsdb.org

The human protein atlas contains antibody labeling data for many proteins: http://www.proteinatlas.org/index.php

This information can be compared to any data presented in a product data sheet.

6 Volume 0 Issue 0mAbs

Dow

nloa

ded

by [

Uni

vers

ity R

ey J

uan

Car

losi

] at

03:

32 2

1 O

ctob

er 2

015

STEP 4A. Validation of target antigenrecognition

High throughput screening for reactiv-ity with the immunogen using techniquessuch as ELISA, FCM or WB is usually thestarting point for antibody validation.This may be the only form of validation,often in only a single technical applica-tion, provided for many antibodies. Cus-tom antibody services commonlyguarantee the production of antibodiesthat recognize the immunogen (particu-larly by ELISA), as this is achievable fornearly all projects. However, many anti-bodies that recognize the immunogenthen fail to recognize the recombinant orendogenous protein in subsequent testing.This is particularly common if the immu-nogen: 1) is a short linear epitope (e.g., asynthetic peptide), 2) does not adopt thecorrect 3-dimensional native conforma-tion, or 3) is produced in a prokaryoticexpression system that lacks post-transla-tional modifications (e.g., glycosylation).It is also important to note that the expres-sion level of the endogenous protein isoften much lower than that of its recombi-nant counterpart in transfected cells (thelatter commonly being driven by a strongpromoter). Moreover, the endogenousprotein may be engaged in interactionswith other proteins, nucleic acids, ormembranes that may partially or fullymask the binding site of an antibody.

Many groups transfect mammalian celllines that lack or have low levels of thetranscript for the target antigen with epi-tope tagged expression constructs to fur-ther test antibody reactivity with thetarget antigen. VERIFY Antigen

StandardsTM from Origene now providemore than 10,000 tagged overexpressionlysates for antibody validation by WB.This approach can be very helpful as WBenables molecular weight determinationand comparison with that of any endoge-nous proteins that are detected (allowingfor any size difference caused by the tag).Immunolabelling techniques such asimmunohistochemistry also enable a com-parison of the staining by the antibodyunder study with that from an antibody tothe epitope tag to confirm a similar fre-quency of positive cells and comparablesubcellular colocalization of the targetantigen. Detection of a tagged protein orco-expression with fluorescent proteins(such as green fluorescent protein) can beused to identify, sort or purify cellsexpressing the target antigen by FCM,enabling verification that the same popu-lation is detected by the antibody beingcharacterized. Multicolor FCM with apanel of markers can also determinewhether the antigen is expressed on a raresubpopulation of cells. Labeling cells withand without permeabilization for FCMcan also distinguish between antibodiesbinding antigens with an intracellular ver-sus cell surface localization. Conjugationreactions, for generating directly labeledantibodies, need to be carefully evaluatedas they can compromise or change anti-body-antigen binding.26

Importantly, recognition of a recombi-nant protein expressed in the appropriatespecies does not guarantee that the anti-body will recognize the endogenous anti-gen; rare instances where an antibody canrecognize the endogenous protein, but not

the recombinant antigen, have occurred.Antibodies that fail to recognize theendogenous protein may still be usefulresearch tools for experiments usingrecombinant proteins.

A confirmation of antibody reactivitywith the endogenous protein is essen-tial. There are many different ways toachieve this. A comparison with thepublished distribution of the protein inthe literature is helpful, but this is notavailable for all proteins and sometimesthe existing data may be inaccurate. Itmay, for example, have been generatedusing a poor-quality antibody. Thus,combining approaches to test antibodybinding to the endogenous protein withthose to determine the specificity of thisrecognition is advisable, and a com-bined approach is described below.

STEP 4B. Antibody specificity for thetarget antigen

Reactivity with the endogenous proteindoes not in itself guarantee that recogni-tion of the target antigen is specific. Thereare, furthermore, also degrees of specific-ity; some antibodies may only specificallydetect their target antigen in particular tis-sues (e.g., those that lack expression of across-reactive protein) or in particulartechnical applications. While absorptiontests using the immunising peptide toblock antibody binding can be useful tech-nical controls to detect non-specific stain-ing (only of the secondary antibody in thecase of monoclonal antibodies because allbinding by the single epitope will beblocked by the immunogen), such testscannot prove that an antibody is specific

Table 4. Antibody Validation

Reactivity with the immunogen Usually the initial testing performed during high throughput antibody screening e.g., using ELISA, WB,FCM

Reactivity with the target antigen Particularly when linear epitopes are used (e.g., peptides) the antibody reactivity needs to be testedagainst the target protein. Often in an epitope tagged recombinant form.

Reactivity with the endogenous target antigen Reactivity with a recombinant antigen does not guarantee reactivity with the endogenous protein.Specificity for the target antigen Some potential cross reactivity can be predicted based on sequence identity and, if likely, should be

experimentally determined. Whenever possible use more than one antibody to define novelexpression patterns. Further investigation is required if different reagents give discordant data, witheach other or with other sources of information such as transcript expression, or data from otherlaboratories.

Evidence for suitability in intended application(s) Reactivity and specificity for the target antigen in one technique or tissue does not guarantee that thiswill be the same in others. Thus it is important to validate antibodies for additional experimentaltechniques and the range of tissue types being studied.

Validation for use in orthologous species The reactivity and specificity for the target antigen should be validated for each specific species beingstudied.

www.tandfonline.com 7mAbs

Dow

nloa

ded

by [

Uni

vers

ity R

ey J

uan

Car

losi

] at

03:

32 2

1 O

ctob

er 2

015

for target antigens that share commonepitopes. We thus do not recommend thisstep for validating an antibody.

Some degree of likely cross-reactivitywith related proteins can be predictedbased on the sequence homology of thetarget antigens with other molecules.While the ideal situation is to avoidantibodies against these regions, this isnot always possible. When validating anantibody where the immunogen containssignificant homology to other relatedproteins, then it will be necessary to testthe specificity of the reagent against theother proteins. An example of this wasthe BCL11AXL antibody (cloneBCL11A/123, Banham/Pulford Univer-sity of Oxford), which was tested forcross reactivity against BCL11B as thesemolecules shared significant homologywithin the C-terminal immunogen.27

Antibodies can also exhibit cross-reactiv-ity to epitopes (e.g., those that share acommon conformation, epitope mim-icry) that are not predictable based onsequence analysis. An ideal situationwould be testing of several differentantibodies in parallel because commonpatterns of antibody reactivity can sub-stantially strengthen confidence in thevalidation data that are generated. Addi-tional validation should be undertakenwhen using only a single antibody thatshows unexpected patterns of reactivitycompared to expression of its transcriptor other sources of information.

Ideally, start by testing antibody reac-tivity (usually in multiple techniques andpreferably with more than one antibody)with at least 2 positive and 2 negative celllines or tissues (usually transcript C/¡),including the cell type proposed for study.Antibodies will need to be titrated toobtain the optimal dilution to enable sen-sitive detection while minimizing non-specific background binding (compared toan isotype-matched negative control anti-body). Working dilutions commonly varyboth between different laboratories andacross different techniques; a useful guideis available at www.abcam.com/protocols/antibody-dilutions-and-titer. Consider thatpost-transcriptional regulation of proteinexpression, e.g., posttranslational modifi-cations or assembly into multi-proteincomplexes, may occasionally cause failure

to detect protein where transcripts aredetectable. If antibodies work in WB,then this technique is usually routinelyincluded for antibody validation becauseit provides useful data on the molecularweight of the target antigen and whethermultiple isoforms exist.

Silencing the expression of the targetantigen ex vivo in cell lines by siRNA orshRNA, (and more recently usingCRISPR/Cas9 technology), is now morewidely being employed as a method forconfirming the specificity of endogenousantigen recognition, particularly for vali-dating antibodies against human proteinswhere in vivo manipulation cannot be per-formed. However, care must be taken, assilencing a single exon may not be suffi-cient to eliminate all transcripts thatencode the target antigen. If the antibodyrecognizes the murine protein, then knockout (KO) mice provide a useful system totest reactivity with the endogenous pro-tein, as long as this is accompanied by lossof the target antigen/epitope and not justloss of function, e.g., by truncation. How-ever, difficulties can also arise when usingsilencing or KO approaches if the targetantigen also regulates other closely relatedproteins. For example, the circadian regu-lator Bmal1 was found to regulate expres-sion of the related Bmal2 gene,28 resultingin a functionally double Bmal1 andBmal2 KO, which could not have beenused to address antibody specificity fordistinguishing reactivity against Bmal1 vs.Bmal2. Another pitfall arose when charac-terizing an antibody where the immuno-gen was mammalian cells transfected witha transcription factor. Initial antibodyvalidation looked promising, until itbecame apparent that, although thereagent recognized only transfected cellsby immunohistochemistry, the subcellulardistribution of the epitope tagged recom-binant target antigen was different to thatof the endogenous protein recognized bythe monoclonal antibody. Further valida-tion suggested that the antibody recog-nized a protein whose expression wasinduced by its transcriptional regulator(unpublished data).

Antibody validation files available onthe EuroMAbNet website (www.euromabnet.com/monoclonal-antibodies) providefurther examples of how antibodies have

been validated through different technicalapplications and across species.

STEP 5: Reproducibility anddissemination of data

In practice, it is important that anti-bodies are independently evaluated bymultiple users and that the results are dis-seminated throughout the research com-munity. While this activity does occur, itis generally on a fairly small scale, such asindividual research groups reporting expe-riences with a particular reagent, compar-ing the reactivity of a panel of antibodiesto a particular molecule or user feedbackto a particular company or personal com-munication to the academic supplier.There are increasingly suggestions thatthere should be more standardization andseveral platforms have been proposed forsharing antibody validation data.

AbMiner was developed as a relationalweb-based database freely providing infor-mation on more than 600 commercial anti-bodies that were validated by WB forprotein microarray studies.29 While thisoffers information, it does not seem toengage end users in reporting their owndata. The same comment applies to thewealth of data available in the Human Pro-tein Atlas (much of which is generated usingonly partially validated reagents), which doesnot link antibodies with externally publishedvalidation or provide a mechanism for userfeedback. The Human Protein Atlas teamhave released data completing the first draftof the human proteome (www.proteinatlas.org/humanproteome).

Antibodypedia was developed withinthe 6th framework EU program Proteome-Binders and the Human Proteome Organ-ization’s Human Antibody Initiative. Thissite has been launched as a more generalportal for sharing antibody and antigenvalidation data.30 Antibodies available tothe public from commercial or academicproviders can be submitted, along withapplication-specific data that rates themagainst standard validation criteria as sup-portive, uncertain or non-supportive.Users can submit their own validationdata and can also submit comments aboutthe use of a particular antibody, to ensuresharing of both positive and negative find-ings. While this site does contain a largenumber of entries, it appears that these are

8 Volume 0 Issue 0mAbs

Dow

nloa

ded

by [

Uni

vers

ity R

ey J

uan

Car

losi

] at

03:

32 2

1 O

ctob

er 2

015

primarily from a fairly limited number ofcommercial antibody suppliers and theProtein Atlas. The aim of the EuropeanProteomeBinders consortium is to focuson quality-controlled replenishable pro-tein affinity reagents (PARs) for the wholehuman proteome.31 This consortium hasalready proposed a Proteomics StandardsInitiative (PSI)-PAR as a global commu-nity standard format for the representa-tion and exchange of protein affinityreagent data32 that will be adopted byAntibodypedia. Recently another Anti-body Validation Database website wasestablished by a consortium testing >200antibodies raised against 57 different his-tone modifications: more than 25% of theantibodies failed specificity tests by dotblot or WB, and, of the remainder, morethan 20% failed in ChIP experiments.21

Although this site initially only specificallyaddressed histone-modification antibodiestested from the ENCODE (Encyclopediaof DNA Elements) and Roadmap Epige-nomics projects, there are plans to addnew non-histone antibodies from theseprojects. Importantly the site also invitesother researchers to upload their valida-tion data. ProteinSimple is also user-inter-active and reports data from antibodyscreening using simple Western charge-and size-based assays.

Summary and Conclusions

It seems to be fairly universally acceptedthat there is a need for standardization ofantibody validation and for easy publicaccess to validation data to help researchersindependently choose the most specific andeffective reagents. How exactly these goalswill be achieved will probably be deter-mined primarily by the large consortiumsgenerating reagents to characterize thehuman proteome. We hope that our con-tribution will help many researchersthrough the maze of antibody validationand enable them to appreciate the experi-mental nature of these reagents, and toapproach their use with a degree of healthyscepticism. Research to explore the humanproteome and deliver personalised medicinewill continue to fuel the demand forgreater numbers of antibodies with increas-ing specificity and functional activity.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest weredisclosed.

Acknowledgments

Our research has been supported byfunding from Cancer Research UK (Pro-gram A10702 to A.H.B) and Bloodwise(Program 13047 to A.H.B). The researchwas supported by the National Institute forHealth Research (NIHR) Oxford Biomedi-cal Research Center based at Oxford Uni-versity Hospitals NHS Trust and Universityof Oxford. The views expressed are those ofthe author(s) and not necessarily those ofthe NHS, the NIHR or the Department ofHealth. Grant No 310/6 from the DeutscheForschungsgemeinschaft to F.K.-N. Grantsfrom the Instituto de Salud Carlos III(PI14/00703, PN de ICDCI 2013–2016)and the CSIC (201320E109 and201420E109) to L.K. laboratory. Grants ofthe Spanish Ministry of Health (Fondo deInvestigaciones Sanitarias, PI10/01039),Department of Education of Castilla andLeon Regional Government (Grant#LE007A10–2) and Mutua Madrile~na Foun-dation (Basic research grants 2012) to J.I.R.B. This work was supported by a grantfrom the Dutch government to the Nether-lands Institute for Regenerative Medicine(NIRM, grant No. FES0908).

References

1. Reichert JM. Marketed therapeutic antibodies compen-dium. MAbs 2012; 4:413-15; PMID:22531442;http://dx.doi.org/10.4161/mabs.19931

2. Zola H, Swart B. The human leucocyte differentiationantigens (HLDA) workshops: the evolving role of anti-bodies in research, diagnosis and therapy. Cell Res2005; 15:691-94; PMID:16212875; http://dx.doi.org/10.1038/sj.cr.7290338

3. Berglund L, Bjorling E, Oksvold P, Fagerberg L,Asplund A, Szigyarto CA, Persson A, Ottosson J, Wer-nerus H, Nilsson P et al. A genecentric Human ProteinAtlas for expression profiles based on antibodies. MolCell Proteomics 2008; 7:2019-27; PMID:18669619;http://dx.doi.org/10.1074/mcp.R800013-MCP200

4. Asplund A, Edqvist PH, Schwenk JM, Ponten F. Anti-bodies for profiling the human proteome-The HumanProtein Atlas as a resource for cancer research. Proteo-mics 2012; 12:2067-77; PMID:22623277; http://dx.doi.org/10.1002/pmic.201100504

5. Mersmann M, Meier D, Mersmann J, Helmsing S,Nilsson P, Graslund S, Colwill K, Hust M, Dubel S.Towards proteome scale antibody selections usingphage display. N Biotechnol 2010; 27:118-28;PMID:19883803; http://dx.doi.org/10.1016/j.nbt.2009.10.007

6. Bradbury A, Pluckthun A. Reproducibility: Standardizeantibodies used in research. Nature 2015; 518:27-29;PMID:25652980; http://dx.doi.org/10.1038/518027a

7. Kalyuzhny AE. The dark side of the immunohisto-chemical moon: industry. J Histochem Cytochem

2009; 57:1099-101; PMID:19755714; http://dx.doi.org/10.1369/jhc.2009.954867

8. Saper CB. An open letter to our readers on the use ofantibodies. J Comp Neurol 2005; 493:477-78;PMID:16304632; http://dx.doi.org/10.1002/cne.20839

9. Couchman JR. Commercial antibodies: the good, bad,and really ugly. J Histochem Cytochem 2009; 57:7-8;PMID:18854593; http://dx.doi.org/10.1369/jhc.2008.952820

10. Baker M. Reproducibility crisis: Blame it on the anti-bodies. Nature 2015; 521:274-76; PMID:25993940;http://dx.doi.org/10.1038/521274a

11. Rhodes KJ, Trimmer JS. Antibodies as valuable neuro-science research tools versus reagents of mass distrac-tion. J Neurosci 2006; 26:8017-20; PMID:16885215;http://dx.doi.org/10.1523/JNEUROSCI.2728-06.2006

12. Bucur O, Pennarun B, Stancu AL, Nadler M, MuraruMS, Bertomeu T, Khosravi-Far R. Poor antibody vali-dation is a challenge in biomedical research: a case studyfor detection of c-FLIP. Apoptosis 2013; 18:1154-62;PMID:23917691; http://dx.doi.org/10.1007/s10495-013-0880-0

13. Rhodes KJ, Trimmer JS. Antibody-based validation ofCNS ion channel drug targets. J Gen Physiol 2008;131:407-13; PMID:18411328; http://dx.doi.org/10.1085/jgp.200709926

14. Michel MC, Wieland T, Tsujimoto G. How reliableare G-protein-coupled receptor antibodies? NaunynSchmiedebergs Arch Pharmacol 2009; 379:385-58;PMID:19172248; http://dx.doi.org/10.1007/s00210-009-0395-y

15. Pradidarcheep W, Labruyere WT, Dabhoiwala NF,Lamers WH. Lack of specificity of commerciallyavailable antisera: better specifications needed. JHistochem Cytochem 2008; 56:1099-111;PMID:18796405; http://dx.doi.org/10.1369/jhc.2008.952101

16. Ramos-Vara JA, Kiupel M, Baszler T, Bliven L, Bro-dersen B, Chelack B, Czub S, Del Piero F, Dial S, Ehr-hart EJ et al. Suggested guidelines forimmunohistochemical techniques in veterinary diag-nostic laboratories. J Vet Diagn Invest 2008; 20:393-413; PMID:18599844; http://dx.doi.org/10.1177/104063870802000401

17. Bordeaux J, Welsh A, Agarwal S, Killiam E, BaqueroM, Hanna J, Anagnostou V, Rimm D. Antibody valida-tion. Biotechniques 2010; 48:197-209;PMID:20359301; http://dx.doi.org/10.2144/000113382

18. Howat WJ, Lewis A, Jones P, Kampf C, Ponten F, vander Loos CM, Gray N, Womack C, Warford A. Anti-body validation of immunohistochemistry for bio-marker discovery: Recommendations of a consortiumof academic and pharmaceutical based histopathologyresearchers. Methods 2014; 70:34-38;PMID:24525140; http://dx.doi.org/10.1016/j.ymeth.2014.01.018

19. Algenas C, Agaton C, Fagerberg L, Asplund A, BjorlingL, Bjorling E, Kampf C, Lundberg E, Nilsson P, Pers-son A et al. Antibody performance in western blotapplications is context-dependent. Biotechnol J 2014;9:435-45; PMID:24403002; http://dx.doi.org/10.1002/biot.201300341

20. Mannsperger HA, Uhlmann S, Schmidt C, WiemannS, Sahin O, Korf U. RNAi-based validation of antibod-ies for reverse phase protein arrays. Proteome Sci 2010;8:69; PMID:21182776; http://dx.doi.org/10.1186/1477-5956-8-69

21. Egelhofer TA, Minoda A, Klugman S, Lee K, Kolasin-ska-Zwierz P, Alekseyenko AA, Cheung MS, Day DS,Gadel S, Gorchakov AA et al. An assessment of his-tone-modification antibody quality. Nat Struct MolBiol 2011; 18:91-93; PMID:21131980; http://dx.doi.org/10.1038/nsmb.1972

22. Helsby MA, Fenn JR, Chalmers AD. Reportingresearch antibody use: how to increase experimental

www.tandfonline.com 9mAbs

Dow

nloa

ded

by [

Uni

vers

ity R

ey J

uan

Car

losi

] at

03:

32 2

1 O

ctob

er 2

015

reproducibility. F1000Res 2013; 2:153;PMID:24358895

23. Gore AC. Editorial: antibody validation requirementsfor articles published in endocrinology. Endocrinology2013; 154:579-80; PMID:23338708; http://dx.doi.org/10.1210/en.2012-2222

24. Brown PJ, Kagaya R, Banham AH. Characterization ofhuman FOXP1 isoform 2, using monoclonal antibody4E3-G11, and intron retention as a tissue-specificmechanism generating a novel FOXP1 isoform. Histo-pathology 2008; 52:632-37; PMID:18312343; http://dx.doi.org/10.1111/j.1365-2559.2008.02990.x

25. Banham AH, Lyne L, Scase TJ, Blacklaws BA. Mono-clonal antibodies raised to the human FOXP3 proteincan be used effectively for detecting Foxp3(C) T cellsin other mammalian species. Vet Immunol Immunopa-thol 2009; 127:376-81; PMID:19084279; http://dx.doi.org/10.1016/j.vetimm.2008.10.328

26. Vira S, Mekhedov E, Humphrey G, Blank PS. Fluores-cent labelled antibodies – balancing functionality anddegree of labelling. Anal Biochem 2010; 402:146-150;

PMID:20362543; http://dx.doi.org/10.1016/j.ab.2010.03.036

27. Pulford K, Banham AH, Lyne L, Jones M, IppolitoGC, Liu H, Tucker PW, Roncador G, Lucas E, Ashe Set al. The BCL11AXL transcription factor: its distribu-tion in normal and malignant tissues and use as amarker for plasmacytoid dendritic cells. Leukemia2006; 20:1439-41; PMID:16710303; http://dx.doi.org/10.1038/sj.leu.2404260

28. Shi S, Hida A, McGuinness OP, Wasserman DH,Yamazaki S, Johnson CH. Circadian clock gene Bmal1is not essential; functional replacement with its paralog,Bmal2. Curr Biol 2010; 20:316-21; PMID:20153195;http://dx.doi.org/10.1016/j.cub.2009.12.034

29. Major SM, Nishizuka S, Morita D, Rowland R, Sun-shine M, Shankavaram U, Washburn F, Asin D, Kou-ros-Mehr H, Kane D et al. AbMiner: a bioinformaticresource on available monoclonal antibodies and corre-sponding gene identifiers for genomic, proteomic, andimmunologic studies. BMC Bioinformatics 2006;

7:192; PMID:16600027; http://dx.doi.org/10.1186/1471-2105-7-192

30. Bjorling E, Uhlen M. Antibodypedia, a portal for shar-ing antibody and antigen validation data. Mol Cell Pro-teomics 2008; 7:2028-37; PMID:18667413; http://dx.doi.org/10.1074/mcp.M800264-MCP200

31. Taussig MJ, Stoevesandt O, Borrebaeck CA, BradburyAR, Cahill D, Cambillau C, de Daruvar A, Dubel S,Eichler J, Frank R et al. ProteomeBinders: planning aEuropean resource of affinity reagents for analysis of thehuman proteome. Nat Methods 2007; 4:13-7;PMID:17195019; http://dx.doi.org/10.1038/nmeth0107-13

32. Gloriam DE, Orchard S, Bertinetti D, Bjorling E,Bongcam-Rudloff E, Borrebaeck CA, Bourbeillon J,Bradbury AR, de Daruvar A, Dubel S et al. A commu-nity standard format for the representation of proteinaffinity reagents. Mol Cell Proteomics 2009; 9:1-10;PMID:19674966; http://dx.doi.org/10.1074/mcp.M900185-MCP200

10 Volume 0 Issue 0mAbs

Dow

nloa

ded

by [

Uni

vers

ity R

ey J

uan

Car

losi

] at

03:

32 2

1 O

ctob

er 2

015