Review
Review Antibody val validatio idation n Jennier Bordeaux, Allison Al lison W. Welsh, Welsh, Seema Se ema Agarwal, Agar wal, Elizabeth Eli zabeth Killiam, Kil liam, Maria . Baquero, Jason A. Hanna, Valsamo K. Anagnostou, and David L. Rimm Department of Pathology, Pathology, Yale Yale University School School of Medicine, Medicine, New New Haven, CT, CT, USA BioT Bio Techniques 48:197-209 (March 2010) doi 10.2144/000113382
Keywords: antibody; validation; immunohistochemistry; immunoluorescence
Antibodies are among the most requently used tools in basic science research and in clinical assays, but there are no universally accepted guidelines guidel ines or standardized methods or determining the validity o these reagents. Furthermore, or commercially available antibodies, it is clear that what is on the label does not necessarily correspond to what is in the tube. o o validate an antibody, it must be shown to be specific, spe cific, selective, and reproducible in the context or which it is to be used. In this review, we highlight highlig ht the common pitalls when working with antibodies, common practices or validating valida ting ant antibodies, ibodies, and levels o comm commerci ercial al antibody antibody valida validation tion or seven ven vendors. dors. Fin Finally ally,, we we share share our algorith algorithm m or antibody validation or immu immunohistochemistry nohistochemistry and quantita quantitative tive immunofluorescence. immunofluorescence.
Introduction
Antibodies are among the most commonly used research tools, routinely used or Western W estern blot blot (WB), immunop immunoprecip recipitati itation on (IP), enzyme-linked immunosorbent assays (ELISA), quantitative immunofluorescence (QIF), and immunohistochemistry (IHC). Tey are also important tools in clinical management with extensive use in both laboratory medicine (ELISA assays and flow cytometry) and anatomic pathology (IHC). In anatomic pathology, IHC serves as a diagnostic, prognostic, and predictive method and IHC readings directly influence i nfluence the management o patients in the clinical setting. For example, the assessment o estrogen receptor α (ER-α), and human epidermal growth actor receptor 2 (HER2) by IHC in breast cancer patients is the definitive test to determine whether or not a patient patie nt will receive therapies therapies that can cost as much as $100,000 per year. Tus, in the clinic, as well as in the research laboratory, careul accurate validation o antibody reagents is critical or correct results. Te influence o antibody-based tests on clinical decisions has led to a number o publications that have highlighted the unmet need or standardization o such assays and development development o antibody validat val idation ion gu idel ideline iness (1–8). A ltho lthoug ughh many groups have enunciated theguideneed, there are no universally accepted lines or best practice in antibody-based tests. Tere are a number o books on the topics by world leaders such as Clive
aylor and David Dabbs, and recently recently,, an ad hoc group published a set o “recommendations” (2). However, these works ocus on the clinical aspects o IHC, ofen using subjective criteria and ofen not taking advantage o recent scientific advances that allow more quantitative evaluation o antibodies. antibodies . Conversely, there there are other groups that have done biologically rigorous evaluatio e valuationn o antibodies using surace plasmon resonance (9) and even X-ray crystallization o antibodies bound to their antigens anti gens (10), (10), methods that are unachievable in a routine research or clinical setting. Te wide range o rigor and methodology in what is used or validat val idation ion is proba probably bly resp responsi onsible ble or a lack o consensus on a single method or antibody validation. Here we present an overview o antibody validation approaches and the pitalls pitall s associated with the ailures o validation. val idation. Tis work speciically ocuses ocuses on assessment o prognostic and predictive cancercancer-related related biomarkers on ormalin-ixed parain embedded (FFPE) tissue.
What is antibody a ntibody validation? Te FDA defines validation as “the process o demonstrating, demonstrating, through t hrough the use o specific laboratory investigations, that the t he perormance o an methodcharacteristics are suitable or its analytical intended analytical use” (www.da.gov/downloads/ Drugs/GuidanceComplianceRegulatoryInormation/Guidances/UCM070107.pd).
For antibodies, one must demonstrate that they are specific, selective, and reproducible in the context or which they are used. When it comes to IHC, standardization can be quite challenging due to the number o pre-analytical, analytical, and postanalytical actors known to inluence staining in FFPE tissue. Variable time to fixation, inadequate fixation period, differ di ffer-ences in fixative used, and a nd tissue processing can all aect tissue antigenicity (5,11). Antibody clone and dilution, antigen retrieval, detection system, and interpretation o results using different di fferent cutoff points are also important variables that regulate IHC measurements (3,12) (V.K. (V.K.A., A., unpublished data). Here we ocus on analytical actors and highlight the importance o proper antibody validation, especially especial ly or IHC or QIF use.
Common pitfalls in antibody studies Nonspecific antibodies
A recent editorial by Michel et al. (13) emphasizes the lack o target specificity or 49 antibodies against 19 subtypes o G protein–coupled receptors calling or more stringent antibody validation criteria. Examples highlighted by the authors included double-knockout lacking the M2 and M 3 subtypes omice muscarinic receptors still staining positive or M 2 and M3 receptor antibodies (14), and tripleknockout mice or the three α1-adrenoceptor
Vol. | No. |
www.Bio www.Bio echniques.com
Review
subtypes demonstrating staining patterns similar simil ar to wild-type wi ld-type mice m ice (15). (15). Determining the speciicity o an antibody is in part dependent on the type o the immunogens: synthetic peptides or purified proteins. Synthetic peptides provi pro vide de th thee ad adva vant ntag ag e o kn know owin ingg the amino acid sequence to which the antibody binds; however, these peptides do not necessarily structure or post-translational recapitulatemodificathe 3-D tions o the native protein (16). (16). As a result, antibodies generated against a synthetic peptide may not work work well when a protein protein is in its native conormation with intact 3-D structure. Such antibodies may not be useul or IP or IHC experiments, but may bind the protein o interest afer it is ully denatured when running SDS WB. Te opposite could also be the case, especially i the immunogen was the purified protein, where the antibody antibody works works well well or prote proteins ins in their native conormation, but not when denatured. Tus WB cannot be an absolute standardization or antibody binding in IHC or other assays where the antigen is in its native conormation. Te native versus denatured confirmation is urther ur ther complicated by methods used to fix tissue. Epitopes that are not exposed in the native proteins can be exposed in fixed tissue and vice versa, even though they may not be truly denatured. Tus, an antibody could recognize one epitope in resh tissue, but when applied to fixed tissue recognize another epitope (17,18). A representative example is an epitope on BCL-2 (41–54 amino acids) that is exposed when BCL-2 is present in the cytoplasm but is inaccessible in the nuclear compartment most likely due to interaction with other proteins (19). (19). However, However, when BCL-2 is phosphorylated at sitesavailable, close to the epitope, the epitope becomes as seen when the protein is extracted extract ed rom cells or denatured by SDS (20). he issue o epitope speciicity is urther complicated by the choice o monoclonal versus polyclonal antibodies. Polyclonal antibodies represent a pool o antibodies against the immunogen and typically show a higher probability or detection in a range o different conditions, whereas a monoclonal antibody is more likely to work in only one set o conditions. Polyclonal antibodies might nevertheless contain superior ainity clones. Perhaps the best example o this is the FDA-approved FDA-approved Herceptest antibody rom Dako, which is the industry standard sta ndard or HER2 testing in breast cancer (21). Monoclonal antibodies are more pure, since they are produced rom single clones o used cells producing immune
A
B
C
D
Figure 1. Examples of non-specific antibodies f or HoxA1 and phospho-4EBP1. (A) Cell lysates were denatured and separated by SDS PAGE, transferred to nitrocellulose and blotted with a mouse polyclonal antibody against HoxA1. A band of the expected molecular weight is seen in CaCo2 lysate (marked by the ar row). Note the numerous bands in all cell lines at unexpected molecular weights. (B) A representative example of IHC staining on FFPE breas t carcinoma tissue for HoxA1. Note the cytoplasmic staining for a nuclear transcription factor. (C) Western blot for phospho-4EBP1. Arrow denotes band at the expected molecular weight; again, all lysates show numerous bands at unexpected molecular weights. (D) A representative example of IHC staining on FFPE lung carcinoma tissue for p-4EBP1. Note the predominantly nuclear localization for a protein expected to localize to the cytoplasm and nucleus.
globulins. However, these clones may be grown in host animals where the ascites
lysates (denoted by the arrow). All o the lysates show several bands above and below
fluid containing secreted antibody is collected (16,18); the thereore, these antibody preparatio prepa rations ns may be conta contamin minated ated by antibodies other than the monoclonal antibody o interest. A study by Spicer et al. demonstrated that seven o 20 monoclonal antibody preparations (35%) they analyzed had staining patterns localized to the Golgi cisternae unrelated to the antigenic specificity o the antibody, and that five o these cross-reactive antibodies ailed to even stain the antigen o interest (22). Antibody validation efforts in our lab have shown similar lack o specificity, even in monoclonal antibodies. Figure 1 provides two examples o nonspecific antibodies tested in our lab. In Figure 1A, a mouse polyclonal poly clonal antibody against HoxA1 HoxA1 (B0 (B01P; 1P; Abnova, aipei City, aiwan) aiwan) was used to probe pro be lysates rom ten cell lines by WB. Te Te expected molecular weight is 37 kD and a band o this size was seen in the CaCo2
this expected molecular weight at a lower signal level, such that i this a ntibody antibody were validated only against a cell line transected to overexpress HoxA1, HoxA1, these bands ba nds would have been missed. Tis level o noise as seen by WB raises concerns about nonspecificity. Figure 1B provides a representative example o the staining pattern seen with this HoxA1 antibody a ntibody on breast carcinoma tissue. Te predominately cytoplasmic staining or a homeobox transcription actor indicates that this antibody should not be used or IHC or IF methods. Figure 1C provides an example o a monoclonal antibody or phospho-4EBP1 (rabbit MAb, clone 236B4; Cell Signaling echnology, Beverley, MA, USA) showing a band o the expected molecular weight (denoted by the arrow) as well as numerous additional bands at higher molecular weights. Figure 1D is a representative example o the nuclear staining observed when using this
Vol. | No. |
www.Bio www .Bio echniques.com
Review
A
C
could represent represent the same sa me target at different post-tra nslationa post-translat ionall mod modiic iication ation stat status, us, breakdown products, or splice variants. However, such observations should raise concerns or using this antibody or urther experiments. Major et al. (2006) published their database AbMiner as a resource or inormation including the immunogen, im munogen, vendor,, and antigen on over 600 commer vendor commer-
ciallythe that available group validated monoclonal via WB antibodies against
pooledd cell poole cell lysates lysates rom rom each each o the NCI-60
B
D
Figure 2. Lack of reproducibility is seen with the VEGF VG-1 clone antibody. (A) IHC on FFPE FF PE lung carcinoma tissue with VG-1 (1:50 dilution) demonstrates specific staining of expected localization. (B) Serial sections of our lung TMA stained with VG-1 (1:50) do not correlate with each other. (C,D) IHC staining of serial sections of the same patient tissue showing a high level of positivity in Run 1 (C) is negative for VEGF in Run 2 (D). Inset represents cytokeratin staining of the tumor for reference.
antibody or IHC on FFPE lung carcinoma
tissue when the expected localization or p-4E BP1 is cytoplasmi p-4EBP1 cy toplasmic. c. Non-reproducible antibodies
Poor correlations between antibody lots are cause or concern, as demonstrated by a recent study rom our lab on antibodies against the Met tyrosine kinase receptor (23). Strikingly, different lotsstained o the monoclonal 3D4two Met antibody on an array o 688 breast cancer cases showed opposite staining patterns—one nuclear and one membranous membranous and cytoplascy toplasmic—with a regression between the two lots having an R2 value o 0.038 (23 (23). ). Another example o non-reproducible antibodies was demonstrat demonstrated ed by by Grimsey et al. (2008 (2008)) with antibodies a ntibodies supposedly targeting the cannabinoid CB1 receptor. Te authors tested multiple lots o antibodies rom various commercial sources by both WB WB and IF against HEK cells transected with HA-tagged H A-tagged CB1 receptor. Only two o the six antibodies tested showed specific membranous staining that co-localized with detection detection using an anti-HA antibody antibody.. Additionally, WB analysis on HEK cell lysates demonstrated that the antibodies with poor IF specificity specificit y detected proteins o incorrect molecular weight, no proteins at all, or proteins present in wild-type
HEK cells which do not even express the CB1 receptor (24). Some examples o non-reproducibility can be very subtle. Figure 2 shows two examples o what appears to be specific staining with expected localization. However, staining with the same conditions on a serial section s ection o the MA (representing the same patient tissues) shows extremely pooractor correlation. is aansecreted growth and as VEGF such, we expected cytoplasmic staining on patient tissues with the VG-1 clone (Figure 2A). Staining on a serial section o the same tissue microarray with the same lot o antibody was not reprodu reproducible, cible, with an R2 value o 0.016 (Figure 2B).
Current practice in antibody validation Antibody specificity speci ficity
WB is wid widely ely use usedd to det deter ermi mine ne an antibody’s speciicity and is an appro priate first validat validation ion step i the t he antibody recognizes the denatured antigen. Te first indication that the antibody is specific or the selected target would be observing a single band at the known molecular weight or the target. Presence o multiple bands or bands not at the proper molecular weight
cell lines. An antibody was considered validated i it produced a band (or bands) o the expected molecular weight(s) or the target protein (25). (25). While Whi le this type o antibody validation is a useul first step, it only guarantees that a given antibody will provide accurate results or WB analysis. I the goal is to use the antibody or IHC or IF, then the user must demonstrate that the antibody is also able to specifically recognize its target when used or those applications. Antibody pedia is also an a n online onli ne portal or sharing antibody validation data or WB, IHC, and IF. Only antibodies commercially available are submitted, and in addition a ddition to a validation score, the inclusion o original experimental data is required. (26). Antibody speciicity has also been evaluated by using blocking peptides especially or IHC (27). Tese peptides are the sequences used to generate the antibody and are incubated with the antibody in great excess. Te antibody with and without the blocking peptide is then used to stain tissue known to express the target o interest. I the antibody is specific, the addition o the blocking peptide will result in loss o staining on the tissue. An example using this technique was recently published or the validation o phosphophospho-specific specific(27) antibodies or ERα. Although this method demonstrates that the antibody is specific or the immunogen rom which it was generated, it does not prove sele selectiv ctivity ity o the antib antibody ody sinc sincee off-target binding activity o the antibody willl also wil a lso be inhibited i nhibited by pre-adsorption pre-ad sorption withh the blo wit block ckin ingg pept peptide ide.. So whi le blocking peptides can prove that an antibody is bad, when nonspecific staining is seen in the presence o the peptide, they cannot prove that an antibody is good. Blocking peptides had been used with a number o the G protein–coupled protein– coupled receptor antibodies and have been ound not to be selective upon more stringent validation (13). Tus, we do not typically include blocking peptides as part o our antibody validation validat ion process. Te key to proving antibody specificity is ofen the correct use o controls. A negative control, including no primary
Vol. | No. |
www.Bio www .Bio echniques.com
Review
Figure 3. The Rimm L ab Algorithm for antibody validation of IHC/QIF. Step 1 of antibody validation is using cell lines in vitro to test antibody spe cificity. Step 2 (below the dotted line) line) involves further validation of antibody antibody on tissue microarray (TMA) for expected tar get localization and reproducibility between assay runs and different antibody lots.
antibody in an IHC I HC or QIF experiment, is valuable but insuicient. A better negative control is a cell line or tissue that is known not to express the protein o interest. Knockout cells thus provide the best negative controls. Similarly, non-expressing cells, transected with the protein o interest, provide the best positive controls. controls . Since these reagents reagent s are ofen beyond the reachthat o many labs, there are other approaches can be used to obtain comparable control results. Ofen
there are readily available cell lines that have been biologically proven not to express a specific speci fic protein o interest. For example, H1650 H1650 cells are PEN-null and a nd thereby make a good negative control or PEN antibodies. Similarly, overex pression pres sion ca cann be va valuab luable le as a pos positive itive control.l. For example, contro ex ample, A431 cells overex overex- press wild-type EGFR and can be used as a positive contro controll or anti-EGFR antibodies. An alternative alternat ive to knockout cell controls controls is siRNA or shRNA knockdown controls, as will be discussed with our Rimm Lab Algorithm o antibody validation. Finally, another type o quality control slide has been developed by Sompuram et al. that couples peptides to glass slides that mimic the epitope o the native antigen and are thereore specific controls or a given
monoclonal antibody (28). Te drawback to this type ty pe o control is that it only works or monoclonal antibodies, as the antibody must interact with a known epitope in order to generate the appropriate peptide controls. Antibody reproducibi lity An important criterion or validation and standardization standardizat ion is antibody reproducibility. reproducibility.
et al. compared HER2 IHC staining o 283 breast adenocarcinomas on tissue microarrays with the 4B5 rabbit monoclonal antibody to the previously established CB11 mouse monoclonal antibody. Additionally, both antibodies were also compared to each other and to the corresponding FISH scores or each case. Te authors ound no significant differences in sensitivity, speci-
Tis staining o thelots same over means time with different on antibody different days, as well as comparison o a new antibody to either a previously validated antibody or to a second independent means o measuring the target that would yield similar results (29). Tere are not many studies on reproducibility o antibodies, since this is ofen assumed a ssumed when using new lots o the same antibody or new aliquots o an antibody used previously by the same sa me lab. Our lab published one such example o this type o evaluation (23), and a similar work to show assay reproducibi lity has been done or HER2 by Gustavson and colleagues (30). A second issue related to reproducibility is when a new reagent is introduced and there is a desire to compare the new antibody to previously validated standards. Tis sort o study is seen more commonly in the literature. For example, van der Vegt
ficity, or predictive values between the two antibodies, validating that 4B5 could be used to assess HER2 expression (31). A second example example by Zarani Za rani et al. a l. compared antibodies or ER and PR with IHC to biochemical determination using enzyme immunoassayy and ound significant links immunoassa between the two methods (32). A similar study by Sasano et al. validated the use o an aromatase monoclonal antibody by correlating the IHC score with biochemical activity as a s determined by product isolation assays (33). Sometimes this is done in an attempt to prove superiority superiority o a new reagent. In those cases, investigators must compare the reagents in the setting o intended use with ou outco tcome me or res respon ponse se to ther therapy apy,, not the previous previo us assay, as the criterio criterionn standard. A good example o this is i s the work o Cheang et al. that showed the ER antibody SP-1 to be more prognostic and predictive than the current standard 1D5 (34), although this
Vol. | No. |
www.Bio www .Bio echniques.com
Review
A
B
site-directed mutants. It also strived to demonstrate lot-to-lot consistency aided by combination o two or more pre-qualified individual lots to create each batch to minimize batch-to-batch variability. Company 7 describes their stringent validation validat ion protocol protoco l or al alll antibodies a ntibodies to include testing o each antibody by WB, IP, IHC, IF, flow cytometry, and ELISA. ELISA . It
It is estimated that there are over 180 antibody companies that produce over 350,000 antibodies or the research and clinical markets (www.antibodyresource.com/onlinecomp.html). So it is not surprising that there are commercially available antibodies against a huge collection o target proteins. When one buys a kit to isolate DNA, it is generally sae to assume it will not end up puriying protein. However, However, the same is not true or antibodies. Just because the manuacturer claims an antibody is specific or protein Z does not necessarily mean that it will only bind protein Z and not a range o other proteins (29,36). In 2005, RamosVara noted that the inormation about antibodies targeting the same antigen varied var ied sig signi ni ica icantly ntly depe dependi nding ng on the manuacturer (16). In act, it is now clear
observed or Company 1’s phosphoAMPKβ1 polyclonal antibody. No inormation was readily available on its web site or datasheet describing any level o antibody validation criteria. Te datasheet itsel contained minimal inormation about the antibody including a brie background on the target with corresponding reerences, recommended applications and starting dilutions or WB, IF, and ELISA. However, no examples o successul use in any o these applications are provided and none o the reerences provided had utilized this antibody. Te datasheet also includes the animal host and immunogen source as a synthetic peptide, but does not include exact peptide pepti de sequence, sequence, although the peptide is available or purchase or use as a blocking peptide. Te datasheet also cautions that this antibody “may” cross-react with correspondingly phosphorylated phosphorylated AMPK β2 . A moderate level o validation was observed or Companies 2–5. hese companies also did not provide any in-depth descriptions o antibody validation procedures. Datasheets all included background on the target, as well as inormation i normation on the
verifies the specificity specificit reproducibility o antibodies by usingy and appropriate kinasespeciic activators or inhibitors when available; testing against a large panel o cell lines with known target expression; phosphatase treatment or phospho-speciicity; comparison to isotype control; verification in transected cells, cel ls, knockout cells, and siRNA-treated cells; utilizing blocking peptides to eliminate all signal; veriying veriyi ng correct subcellular localization or treatment-induced translocation; comparing new antibody lots to previous lots; providing optimal dilutions and buffers, as well as a s speciying both positive positive and negative control cell lines. For IHC, Company 7’s antibodies are tested on paraffin-embedded cell pellets including cell pellets created afer the cell lines are subjected to treatments known to induce signaling changes or treated with siRNA to block expression o the target. issue is treated with phosphatase to additionally test phospho-specificity phospho-specificity on FFPE tissues. Xenografs with cell lines o known target expression or treatments to modulate expression are parain-embedded and then stained. Te datasheets or these companies include everything seen with Companies 2–5 plus additional examples o the successul use o their antibody. For example, Company 6 includes represen-
that responsibility or not proo specificitythe is with the purchaser, theovendor. Different vendors provide different levels o validation, depending on their approach to the balance between making a profit and providing high quality. For For this review, review, we sought to compare the level o validation and amount o inormation provided by seven dierent companies (reerred to herein simply as Companies 1–7) to the randomly selected molecule AMPKβ1. he companies are not revealed since only 7 o over 180 were chosen at random; ra ndom; urthermore, the level o validation at some companies varies based on the product. When availa available, ble, we compared the validation protocol on the datasheet or phospho-specific antibodies, as these require another level o specificity and validation validat ion (37 (37,38). In general, we ound three levels o validation. he least validation was
immunogen. out o sequence our companies included theTree complete used and the ourth identiied the region surrounding the phosphorylation site as the sequence used. Tese companies also provided recommended applications with starting star ting dilutions, and a nd all al l provided at least one example o the antibody successully identiying its target in one o the recommended applications. applications. WBs— W Bs— either with transected cells expressing the target or pre-incubation o the antibody with block blocking ing peptide peptide—were —were the most commonly shown antibody validation examples. Te highest levels o validation were seen or Companies 6 and 7. Company 6 describes its validation procedures or phospho-speci phosphospeci fic antib antibodie odiess to includ includee WB analysis in (i) multiple cell lines, (ii) peptide pept ide and phosp phospho-p ho-pepti eptide de comp compeetition experiments, and (iii) analysis o
tative data or all anded ll validated va lidated applicati applications ons wit with h rec recomme ommended antibo ant ibody dy conc concenentrations and includes low cytometry, immunocytochemistry on a positive control control cell line li ne (including (including incubation with either the phospho-peptide or the non–phospho peptide pept ide demon demonstra stratin tingg sig signa nall abs absence ence only with the phospho-peptide), IHC o a positive control tissue, and a WB o a positive control cell line lysate also including detection afer pre-incubation withh the phos wit phosphopho- or non– non–phos phospho pho-specific peptides demonstrating a single band o the expected molecular weight with signal si gnal absent only on ly afer a fer incubation with the phospho-specific peptide. Antibodies purchased rom Company 1 (which provided the minimal amount o inormation and no examples o successul use o the product) would would require extensive ext ensive validation by a researcher to demonstrate demonstrate that the results describe only the target
Figure 4. Stathmin expression is reduced by siRNA. (A) WB of BT-20 lysates 24 h post-mock; control scrambled siRNA or Stathmin siRNA transfection demonstrates specific reduction in Stathmin levels. GAPDH serves as a loading control. (B) BT-20 cells were fixed in 4% PFA 24 h post-transfe ction with scrambled control or Stathmin siRNA, and IF with Stathmin demonstrates a reduction in Stathmin expression.
conclusion was later disputed by Brock and colleagues (35).
Vendor-based Vendor -based validation valid ation of commercial antibodies
Vol. | No. |
www.Bio www .Bio echniques.com
Review
o interest; the company provides little in the way o assurances that it will work. Companies 2–5 all provide more inormation on their respective antibodies and include at least one example o the product successully being used or at least one o their recommended applications with corresponding positive controls. While this is clearly preerable to Company 1’s
A
approach, antibodies still need to bethese validated by thewould researcher or target specificity in the application o interest and or lot-to-lot reproducibility. Company 6 describes the validation validat ion steps it uses to determine that its phospho-specific antibodies are both specific speci fic and reproducible. Furthermore, or each recommended application, an example o successul use is provided with corresponding inhibition by phospho-speciic blocking peptides. Company 7 provides what we consider the gold standard or antibody validation. Te extensive in-house testing described provides pro vides a high level o o confidence confidence that the antibodies it provides provides will work or all the applications that are recommended or the particular parti cular antib antibody ody.. It also pr provid ovides es a grea greater ter level o confidence that results obtained will be specific or the described target and will be reproducible among antibody lots. Even though Companies 6 and a nd 7 provide extensive validation, the researcher is still obliged to confirm that the product gives specific and reproducible reproducible results in the cell lines or tissues o interest in the lab. Indeed, even the best companies cannot control what happens afer the product leaves their door. Issues during shipping, shipping , inappropriate storage storage on or afer arrival in the laboratory, and antibody contamination contamination during usage are all potential pote ntial sour sources ces o o error error in antibod antibody-b y-based ased testing. Tus vigilant and comprehensive
B
C
Figure 5. ER- α antibody Clone 1D5 as an example of strong correlation between quantitative WB and IHC methods. (A) FFPE MCF7 cells stained with 1D5 (1:50 dilution) demonstrate expected nuclear localization for ER- α. (B) WB with 1D5 of a panel of breast cancer cell lines demonstrating varying expression of ER-a. Puro cells are MCF7 cells stably expressing an inducible ER- α. β-tubulin serves as a loading control. (C) ER- α expression by WB was quantified using ImageJ and correlated to AQUA scores corresponding to the cell lysates demonstrating a strong correlation.
association, and thus we are not so bold as to call this a “recommendation.” Instead, we provide what we eel is a compr compromise omise between rigor and lab economics that results in a level o evidence sufficient or data dissemination. Our algorithm or
siRNA were also ixed and decreased Stathmin expression was visualized with IF I F (Figure 4B). Additionally, when known target activators and inhibitors are available, lysates rom treated cells can also be included when validating antibodies
antibodyor validation (Figure 3) is especially ocused the end use application applicatio n o IHC
specific orIdentification a particular state (e.g., phosphorylation). o a single band (or multiple bands i more than one isoorm o the protein is expected to interact with the t he
here are no uniorm or enorceable standards or antibody validation. Unlike drugs—whose sale is prohibited without FDA approval—there is no ederal agency governing what can be sold into the antibody-based antibody-based assay market. In the uture, we may see urther FDA clearance o antibodies or more rigorous labeling and regulation o reagents to be used in
or QIF on paraffin-embedded tissues, but could be equally valid va lid or modified or other antibody-based assays. Our irst line o evidence that an antibody is specific or the target o interest is done by WB. A variety o cell line l ine lysates (orr tissue homogenates) are selected, ideally (o with known levels o target expressio expression n such such that both positive and negative cell lines are analyzed. Practically speaking, we ofen have no idea o target levels and select cell line series at random. Lines are ofen selected in such a way that we suspect some o the lines to be b e completely negative or the target protein (e.g., using fibroblast lines or epithelial targets). When true negative lines are not available, or when levels o target protein are highly dependent on growth conditions, we produce lysates o cell lines li nes where the target protein has been knocked down using RNAi or lysates rom cell lines typically null or the target that have been transected or overexpression constructs. In a recent example, we have used inducible overexpression (A.W.W., unpublished data). An example o using siRNA to aid in validating an antibody is shown in Figure 4 or Stathmin (rabbit monoclonal; Epitomics, Burlingame, CA, USA), a microtubule-destabilizing microtubule-destabilizing protein. Lysates rom B B-20 -20 cells cel ls either e ither
antibody) o the correct molecular weight only in cell lines expressing the target encourages urther validation or use in IHC/QIF. IHC/Q IF. An antibody demonstrating no binding on WB may still be specific or its intended target when in its native conormation and an IP experiment can be the next step in determining the specificity o the antibody i the goal is or use with IHC/QIF. Te second step in the validation o an antibody or IHC/QIF is to titer the antibody on a tissue microarray (MA) comprised o FFPE cell line pellets corresponding to the identical cell lines li nes used to first validate valid ate by WB in addition to patient tissue expected to express the target (or more inormation on the useulness o MAs, MA s, see Reerences 39 and 40). A good antibody will have the ollowing charactercharacteristics: (i) it will stain only the cell pellets expressing the target, (ii) the level o staining willl decrease with increasing wil i ncreasing dilutions di lutions o the antibody, and (iii) it will demonstrate an expression pattern that is consistent with biological biological and mechanistic mechanistic data in the published literature. Te target expression expression seen with these first two steps should be quantified—we use ImageJ or quantifi-
clinical testing. standard. However, We to date, there is no universal thereore present our lab’s approach (the Rim Rimm m Lab Algorithm) to the validation issue. Our approach is not sanctioned or approved by any governing body or by any trade
mock-transected transected with scrambled siRNA orcontrol or siRNA specific or Stathmin were analyzed by WB dem demons onstra tratin tingg los losss o Stat Stathm hmin in expression (Figure 4A). B-20 B-20 cells transtrans ected with either scrambled or Stathmin
cation o WBs and AUA or measuring the level o protein expression in each MA spot—and then the results should show a strong correlation to each other. An example o an antibody meeting these conditions is shown in Figure 5 or ER-α
controls should be done with each assay.
Te Rimm Lab Algorithm Al gorithm for antibody validation for IHC/QIF IHC/QIF
Vol. | No. |
www.Bio www .Bio echniques.com
Working with primary cells ?
A
B
Figure 6. Examples of reproducible antibodies between assay repeats and antibody lots. (A) Lot-to-lot reproducibility is shown using a MAP-tau monoclonal antibody on a breast index TMA. (B) MENA demonstr ating a high level has an R 2 value of 0.932 when stained on serial sections of our lung TMA demonstrating of reproducibility among assays run on different days using the same lot and dilution of antibody.
Cardiac Cell Promotion www.promocell.com/promotion
Go for PromoCell PromoCe ll ! Your partner for primary cell cultivation
Huge variety of human primary cells e.g. Cardiac Myocytes & Fibroblasts Cardiac Microvascular Endothelial Cells
www.promocell.com
PromoCell GmbH
North America: 1 - 866 - 251 - 2860 2860 (toll-free) Deutschland: 0800 - 776 66 23 (gebührenfrei)
(1D5 clone; Dako, Glostrup, Denmark). As expected rom rom published literature, the staining on patient breast tissue is predominately nuclear (Figure 5A). Breast cell lines with varying var ying ER expression ex pression levels (Figure 5B), 5B ), as well as a calibration curve o recom recom-binant ER protein, were analyzed by WB and the expression o ER protein (pg per μg o loaded lysate protein) against again st thecell expression level o was ER asplotted determined by IHC and quantified by AUA demonstrating a strong correlation ( R 2 value o 0.91 0.91)) between bet ween the two methods met hods (Figure 5C). he inal step in validation o an antibody or any application is to demonstrate that the antibody is reproducible between assay runs and between lots. We accomplish this by using a MA containing ~120 spots that is stained with each lot o the antibody. Te target expression is quantified and a regression between the t he two scores or each spot is perormed. A high correlation among multiple lots o
(13,18,36,41). hese authors point out that one o the best controls or antibody validation is tissue tissue rom rom a knockout knockout animal lacking the antigen altogether. When available, this tissue tis sue would be an excellent addition to any control control MA, MA , but this is not available or every protein o interest and would be a time-consuming time-consuming and expensive expensive investmentt or the sole purpose o validating investmen an antibody. Ideally, as suggested sugge sted in a recent perspective by Alexander A lexander Kalyuzhny, Ka lyuzhny, we will one day have have guidelines on the gener gener-ation and use o antibodies (42) where all companies comp anies will wil l be held to high standards or antibody validation and researchers can have greater confidence that their precious grant or clinical revenue dollars are not wasted on inaccurately labeled “vials o PBS” (36). For the meantime, however, the responsibility ultimately lies with the researcher or laboratory director to ensure that the antibodies used in their labs are validated or specificity and reproducibility.
the antibody our finalorequirement validation. Anisexample this is shown shown or or
Acknowledgments
an antibody to MAP-tau (United States Biological, Swampscott, MA, USA), where two different lots o the antibody were quantified with AUA demonstrating a high correlation (Figure 6A). Each time the antibody is used on a different MA, sections rom this control MA are stained in parallel as an additional staining control. control. Te control control MA M A slide should always a lways have a high correlation to previous experiments experi ments or a specific and reproducible antibody Te example in Figure 6B shows inter-assay reproducibility o MENA (mouse MAb, clone 21; BD ransduction Laboratories, Franklin Lakes, NJ, N J, USA). USA). Tese validation steps are relatively ast and inexpensive, but most importantly, they are comprehensive. Tis algorithm or antibody validation is somewhat similar to many other previously published approaches
Te authors thank Elaine Alarid o the University o Wisconsin or provision o cells lines with inducible expression o ER-α. Tis work was supported by the National Institutes o Health (NIH; grant nos. CA139431, CA 114277, CA 110511, and CA 106709, to D.L.R.); the Susan G. Komen Foundation (grant no. KG090562, to D.L.R.); D.L.R .); and the Department o Deense Breast Cancer Research Program [grant nos. 1W81XWH-06-1-0746 1W81XW H-06-1-0746 (to M. M ..B.), .B.), 1W81XWH-08-1-0404 (to J.B.), and 1W81XWH-08-1-0784 (to A.W.W.)]. Tis paper is subject to the NIH Public Access Policy.
Competing interests he authors declare no competing interests.
United Kingdom: 0800 - 96 03 33 (toll free) France: 0800 90 93 32 (ligne verte) Other countries: +49 6221 - 649 34 0
www.Bio www .Bio echniques.com
Review
References 1. Dowsett, M., W.M. Hanna, M. Kockx, F. Penault-Llorca, J. Rüschoff, T. Gutjahr, K. Habben, and M.J. van van de Vijv Vijver. er. 2007. Standardization of HER2 testing: results of an international proficiency-testing ring study. Mod. Pathol. 20 :584-591. :584-591. 2. Goldstein, N.S., S.M. Hewitt, C.R. Taylor, H. Yaziji, D.G. Hicks, and Members of the Ad-Hoc AdHoc Comm Committee itteeon Immun Immunohisto ohistochem chemistry istry Standardization. 2007. Recommendations for improved standardization of immunohistochem-
istry. Appl. Immunohistochem. Mol. Morphol. 15:124-133. 3. Shi, S.-R., C. Liu, and C.R. Taylor. 2007. Standardization of immunohistochemistry for formalin-fixed, paraffin-embedded tissue sections based on the antigen-retrieval technique:
from experiments to hypothesis. J. Histochem. Cytochem. 55:105-109. 4. W Wolff, olff, A.C., A.C., M.E.H. Hammond, J.N. J.N. Schwartz, K.L. Hagerty, D.C. Allred , R.J. Cote, M. Dowsett, P.L. Fitzgibbons, et al. 2007. American
Society of Clinical Oncology/College of American Patholo Pathologists gists guideline guideline recommen recommendadations for human epidermal growth factor receptor 2 testing in breast cancer. J. Clin. Cli n. Oncol. 25:118145. 5. Hicks, D.G. and S. Kulkarni. 2008. HER2+
breast cancer: of biologic and optimal use ofreview diagnostic tools.relevance Am. J. Clin. Pathol. 129 :263-273. :263-273. 6. Yaziji, H., C.R. Taylor, N.S. Goldstein, D.J. Dabbs, E.H. Hammond, B. Hewlett, A.D. A.D. Floyd, T.S. Barry, et al. 2008. Consensus recommendations on estrogen receptor testing in breast cancer
by immunohistochemistry. Appl. Immunohistochem. Mol. Morphol. 16 :513-520. :513-520. 7. Deutsch, E.W., C.A. Ball, J.J. J.J. Berman, G.S. Bova, A. A. Brazma, R.E. Bumgarner, D. Campbell, H.C. Causton, et al. 2008. Minimum infor-
mation specification for in situ hybridization and immunohistochemistry experiments (MISFISHIE). Nat. Biotechnol. 26 :305-312. :305-312. 8. McShane, L.M., D.G. Altman, Altman, W. W. Sauerbrei, S.E. Taube, M. Gion, and G.M. Clark. 2006. REporting recommendations for tumor MARKer
prognostic studies (REMARK). Breast Cancer Res. Treat. 100 :229-235. :229-235. 9. Mullett, W.M., W.M., E.P. Lai, and J.M. J.M. Yeung. 2000. Surface plasmon resonance-based immunoassays. Methods 22:77-91. 10. Franklin, M.C., K.D. Carey, F.F. Vajdos, Vajdos, D.J. Leahy, A.M. A.M. de Vos, and M.X. Sliwkowski. 2004. Insights into ErbB signaling from the structure of the ErbB2-pertuzumab complex. Cancer Cell 5:317-328. 11. Hsi, E.D. 2001. A practical approach for evaluating new antibodies in the clinical immunohistochemistry laboratory. Arch. Pathol. Lab. Med. 125:289-294. 12. Leong, A. S. 2004. uantitation in
immunohistology: immuno histology: fact or fiction? A discussion of variables that influence results. Appl. Immunohistochem. Mol. Morphol. 12:1-7. 13. Michel, M.C., T. Wieland, Wieland, and G. Tsujimoto.
2009. How reliable are G-protein-coupled receptor antibodies? Naunyn Schmiedebergs Arch. Pharmacol. Pharmacol. 379 :385-388. :385-388.
Jositsch, tsch, G., T. Papadakis, R.V. Haberberger, M. 14. Josi Wolff, W olff, J. J. Wess, Wess, and W. W. Kummer. 2009. Suitability of muscarinic acetylcholine receptor antibodies
for immunohistochemistry evaluated on tissue
sections of receptor gene-deficient mice. Naunyn Schmiedebergs Arch. Pharmacol. 379 :389-395. :389-395. 15. Jensen Jensen,, B.C., P.M. Swigart, and P.C. Simpson.
2009. Ten commercial antibodies for alpha-1adrenergic receptor receptor subtypes are nonspecific. Naunyn Schmiedebergs Arch. Pharmacol. :409-412. 379 :409-412. 16. Ramos-Vara, J. A. 2005. Technical
30. Gustavson, M.D., B. Bourke-Martin, D. Reilly, M. Cregger, C. Williams, Williams, J. J. Mayotte, M. Zerkowski, G. Tedeschi, et al. 2009. Standardization of HER2 immunohistochemistry in breast cancer by automated quantitative analysis. Arch. Pathol. Pathol. Lab. Med. 13 1333:1413-1419. Vegt, B., G.H. de Bock, J. J. Bart, N.G. 31. van der Vegt, J. Wesseling. Wesseling. 2009. Validation of Zwartjes, and J.
aspects of immunohistochemistry. Vet. Pathol. 42 :405-426. 17. Willingham, M.C. 1999. Conditional epitopes.
the 4B5 rabbit monoclonal antibody in deter-
is your antibody always specific? J. Histochem. Cytochem. 47 :1233-1236. :1233-1236.
32. Zafrani, B., M.H. Aubrio Aubriot, t, E. Mouret, P. De Crémoux, Y. De Rycke, A. A. Nicolas, E.
18. Saper, C.B. and P.E. Sawchenko. 2003. Magic
Boudou, A. A. Vincent-S Vincent-Salomon alomon,, et al. 2000. High sensitivity and specificity of immunohistochemistry for the detection of hormone receptors in breast carcinoma: comparison with biochemical determination in a prospective study of 793 cases. ca ses. Histopathology 37 :536-545. :536-545. Anderson, S.G. Silverberg, R.J. 33. Sasano, H., T.J. Anderson, Santen, M. Conway, D.P. Edwards, A. A. Krause, A.S. Bhatnagar, et al. 2005. Te validation of new new aromatase monoclonal antibodies for immunohistochemistry--a correlation with biochemical activities in 46 cases of breast cancer. J. Steroid Biochem. Mol. Biol. 95:35-39. 34. Cheang, M.C., D.O. Treaba, C.H. Speers, I.A. Olivotto, C.D. Bajdik, S.K. Chia, L.C. Goldstein, K.A. Gelmon, et al. 2006. Immunohistochemical detection using the new rabbit
peptides, magic antibodies: guidelines for appro-
priate controls for immunohistochemistry. J. Comp. Neurol. 465:161-163. A.G. Tse, J.L. J.L. Cordell, K.A. Pulford, 19. Pezzella, F., A.G. K.C. Gatter, and D.Y. Mason. 1990. Expression of the bcl-2 oncogene protein is not specific for fo r the 14;18 chromosomal translocation. Am. J. Pathol. 137 :225-232. 20. Wi l li ng ha m, M.C. an d K. Bhalla.
1994. Transient mitotic phase localization of bcl-2 oncoprotein in human carcinoma cells and its possible role in i n prevention of apoptosis. J. Histochem. Histochem. Cytoche Cytochem. m. 42:441-450. 21. Jaco Jacobs, bs, T.W., A.M. A.M. Gown, H. Yaziji, M.J. Barnes, and S.J. Schnitt. 1999. Specificity of HercepTest
in determining HER-2/neu status of breast cancers using the United and Administr Administratio ation-a n-appr pprove oved dStates scoring scor ingFood system. syste m. J.Drug Clin. Oncol. 17 :1983-1987. :1983-1987. 22. Spicer, S.S., M.A. Spivey, M. Ito, and 22. B.A. Schulte. 1994. Some ascites monoclonal
antibody preparations contain contaminants that bind to selected Golgi zones or mast cells. J. Histochem. Histochem. Cytoche Cytochem. m. 42:213-221. 23. Pozner-Moulis, S., M. Cregger, R.L. Camp, and D.L. Rimm. 2007. Antibody validation by quantitative analysis of protein expression using
expression of Met in breast cancer as a model. Lab. Invest. 87 :251-260. :251-260. 24. Grimsey, N.L., C.E. Goodfellow, E.L. 24. Scotter, M.J. Dowie, M. Glass, and E.S. Graham.
2008. Specific detection of CB1 receptors; cannabinoid CB1 receptor antibodies are not all created equal! J. Neurosci. Methods 17 1711:78-86. 25. Major, S.M., S. Nishizuka, D. Morita, R.
Rowland, M. U. 2006. Shankavaram, Washburn, D. Sunshine, Asin, Asin, et al. AbMiner: F. a bioinformatic resource on available monoclonal antibodies and corresponding gene identifiers for genomic, proteomic, and immunologic studies. BMC Bioinformatics 7 :192. :192. 26. Björling, E. an d M. Uhlén. 2008. 26. Antibodypedia, Antibod ypedia,a portal for sharin sharingg antib antibody ody and antigen validation data. Mol. Cell. Proteomics :2028-2037. 7 :2028-2037. Al-Dhaheri, C. 27. Skliris, G.P., B.G. Rowan, M. Al-Dhaheri, Williams, S. Troup, S. Begic, M. Parisien, P.H. Watson, and L.C. Murphy. 2009. Immunohistochemical validation of multiple phosphospecific epitopes for estrogen receptor alpha (ERalpha) in tissue microarrays of ERalpha positive human breast carcinomas. Breast Cancer Res. Treat. 11 :443-453. 118 8 :443-453. 28. Sompuram, S.R., V. Kodela, K. Zhang, 28. H. Ramanathan, G. Radcliffe, P. Falb, and S.A. Bogen. 2002. A novel quality control slide for quantitative immunohistochemistry testing. J. Histochem. Cytochem. 50 :1425-1434. :1425-1434. 29. Saper, C.B. 200 2005. 5. An open letter to our readers on the use of antibodies. J. Comp. Neurol. 493:477-478.
mining Her2/neu status in breast cancer. Mod. Pathol. 22:879-886.
monoclonal antibody SP1to ofmouse estrogen receptor in breast cancer i s superior is monoclonal antibody 1D5 in predicting survival. J. Clin. Oncol. 24 :5637-5644. :5637-5644. 3 5. B roc k, J J.. E . , J J.. L . Horn i c k, A . L . Richardson, D.A. Dillon, and S.C. Lester. 2009. A comparison of estrogen receptor SP1
and 1D5 monoclonal antibodies in routine clinical use reveals similar staining results. Am. J. Cl in. Pathol . 132:396-401. 36. Couchman, J. R . 2009. Commercial antibodies: the good, bad, and really ugly. J. Histochem. Cytochem. 57 :7-8. :7-8. 37. Mandell, J. J.W. W. 2003. Phosphorylation statespecific antibodies: applications in investigative and diagnostic pathology. Am. J. Pathol. 1633:1687-1698. 16 38. Mandell, J.W. 2008. Immunoh is-
tochemical assessment of protein phospho phosphoryrylation state: the dream and the reality. Histochem. Cell Biol. 130 :465-471. :465-471. V. Neumeister, and D.L. Rimm. 39. Camp, R.L., V.
2008. A decade of tissue microarrays: progress in the discovery and validation of cancer biomarkers. J. Clin. Oncol. 26 :5630-5637. :5630-5637. 40. Kay, E., A. O’Grady, J.M . Morgan, S. 40. Wozniak, and B. Jasani. Jasani. 2004. Use of tissue microarray for interlaboratory validation of HER2 immunocytoche immunocytochemical mical and FISH testing. J. Clin. Pathol. 57 :1140-1144. :1140-1144. J.S. Trimmer. 2008. Antibody 41. Rhodes, K.J. and J.S. based validation of CNS ion channel drug targets. J. Gen. Gen. Physiol. Physiol. 13 1311:407-413. A.E. E. 2009. The dark side 42. Kalyuzhny, A. 42. of the immunohistochemical moon: industry. J. Histochem. Cytochem. 57 :1099-1101. :1099-1101. Received 26 January 2010; accepted 28 January 2010. Address correspondence to David L. Rim m, Department of Pathology, Yale University School of Medicine, 310 Cedar St., PO Box 208023, New Haven, CT, 06520 USA. e-mail:
[email protected]