Is your antibody binding the right target?
Detect gene expression, knockdown, or knockout with confidence
Only 30-50% of commercial antibodies demonstrate specificity for their targets. Furthermore, the correlation in the protein binding of the same antibody from different lots can be as low as r2=0.038. Antibodies have been shown to detect proteins of incorrect molecular weight, no protein at all, or worst, detection of a protein of the correct molecular weight which is not the intended target. The program director of the Human Protein Atlas consortium, Matthias Ulhen, said in 2014 that after testing more than 25,000 antibodies from 50 plus suppliers the vast majority did not function as intended. There is an interesting simile described in Bordeaux, et al.; when buying a DNA isolation kit, there is generally the security it will isolate DNA and not proteins, but the same is not true for antibodies.
The antibody crisis has many consequences such as false conclusions, uninterpretable or misinterpreted results, wasted samples, inability to replicate published results, paper retractions, stronger reviewers’ challenges, and researchers’ time and money wasted. Commercial antibodies are often screened and optimized for narrow experimental conditions and may not work as advertised causing misleading or irreproducible results.
What is testing vs validation?
There is a difference between testing an antibody and validating it. In testing, a positive result is the acceptance criteria, but the signal must be in the right place in a relevant sample. Validation goes beyond a positive signal, it tests the boundaries/truth of the positive signal. Validation is testing that the reagent is specific, selective and reproducible. To validate, the reagent must be at maximal dilution:
- Bind the intended protein (specificity),
- React minimally with other proteins (selectivity), and
- Consistently and repeatedly exhibit the same result (reproducibility).
It means a very good signal in the correct sample and hardly any signal in a negative control using the same dilution. Validation is also different from conformity, the latter implies that the antibody should perform as described by the provider (same cell line including treatment, the protein preparation, dilution, technique, and detection). Validation can be defined as the demonstration, using specific laboratory research tools, that the performance characteristics of any analytic reagent are suitable for its intended use. There is no gold standard validation strategy to assess antibodies across all applications because expected performance in one application does not necessarily predict applicability in another one. Therefore, each researcher should validate the analyte in the specific assay and biological sample of interest, using appropriate loading controls, and negative controls based on cell or tissue-specific expression.
Is there a link between QC standards and reproducibility, and why should I worry about it?
Antibodies are the only extensively used reagent in molecular biology uncharacterized at the molecular level and hard to validate by users. Similar to the qPCR guidelines (MIQE guidelines), the implementation of the Western-blot minimal reporting standards (WBMRS) has been proposed to improve Western-blot reproducibility. Retraction Watch reports a substantial amount of papers retracted, and academic reputations damaged as result of poor Western-blot practices.
To reproduce antibody results there is a minimal information needed:
- Antibody provider catalog and lot numbers
- Specific experimental design
- Sample information
- Antibody usage conditions
However, in order to replicate published results and obtain reliable results, antibody validation regarding specificity, selectivity and reproducibility must be carried out for the singular experimental setup. The validation data using a particular cell line or tissue cannot necessarily be used to prove the antibody to be functional in a different one. Additionally, reagent aging, instability, inappropriate storage conditions and handling, incorrect dilution, among other factors can also lead to low performance.
How can I make sure my antibody is working correctly?
To set up standards and implement the creation of a catalog of validated antibodies, different researchers have participated in initiatives such as Human Antibody Initiative (HAI) and International Working Group for Antibody Validation (IWGAV). These groups have defined 5 criteria for antibody validation in an application- and experimental condition- specific manner:
- Genetic strategies using tissues with differential protein expression, and/or exogenous: overexpression, reduction (knockdown), or elimination (knock out) of the target protein. This requires the use of more than one reagent (eg: at least 2 different siRNA molecules) or negative cell lines, and includes the verification by RT-qPCR to establish a direct link between mRNA reduction and protein detection by the antibody.
- Orthogonal strategies carrying out protein quantification in diverse cell samples using an antibody-independent approach such as mass spectrometry-based targeted proteomics and comparing it with the antibody signal.
- Expression of tagged proteins, employing affinity or fluorescently tagged proteins to perform a parallel detection. The signal must be comparable between the tag detection and the specific antibody, otherwise, there is a potential cross-reactivity.
- Independent antibodies strategy testing at least 2 different antibodies with no overlapping epitopes in a panel of different samples. This minimizes the possibility to obtain the same off-target binding signals. Ideally, samples with variable protein expression should be used including the knockdown and knock out cells.
- Immunocapture and mass spectrometry (IMS), it involves the determination of protein abundance by MS after immunocapture, the three top peptides must be derived from the expected target.
Minimum one strategy of validation must be applied, and although some of them are not at hand in all laboratories, genetic and independent antibodies strategies can be applied without incurring in excessive investments, and certainly, the benefit at long run is higher, regarding the quality, and accuracy, of the results and the derived conclusions.
- Baker M (2015) Reproducibility crisis: Blame it on the antibodies. Nature 521: 274-276
- Baker M (2015) Antibody anarchy: A call to order. Nature 527: 545–551
- Bordeaux J et al (2010) Antibody validation. BioTechniques 48: 197-209
- Bradbury A, Pluckthun A (2015) Reproducibility: Standardize antibodies used in research. Nature 518: 27-29
- Couchman, J.R. 2009. Commercial antibodies: the good, bad, and really ugly. J. Histochem. Cytochem. 57:7-8.
- Freedman LP et al (2016) [Letter to the Editor] The need for improved education and training in research antibody usage and validation practices. BioTechniques 61: 16-18
- Gilda JE (2015) Western Blotting Inaccuracies with Unverified Antibodies: Need for a Western Blotting Minimal Reporting Standard (WBMRS). PLoS ONE 10(8): e0135392.
- Helsby MA et al (2014) The F1000Research Antibody Validation Article Collection. F1000Res. 3:241-242
- Perkel, J.M. 2014. The antibody challenge. Biotechniques 56:111-114.
- Polakiewicz R (2015) Antibodies: The solution is validation. Nature 518: 483
- Schonbrunn, A. 2014. Editorial: Antibody can get it right: confronting problems of antibody specificity and irreproducibility. Mol. Endocrinol. 28:1403-1407.
- Uhlen M et al (2016) A proposal for validation of antibodies. Nat Meth 13:823-827
- Voskuil JL (2017) The challenges with the validation of research antibodies. F1000Res. 6:161-165
- Weller MG (2016) Quality Issues of Research Antibodies. Analytical chemistry insights 11: 21-27
Author: Johanna De Castro Arce, Ph.D. Field Application Scientist