Setting standards for antibodies

We all know that some antibodies work better than others, and some don’t work very well at all. Commercially available antibodies may not be target-specific with the information on the supplied datasheet revealing little as to their function in specific assays. Antibodies from one lab may work superbly in their hands only to completely fail in yours.

These discrepancies in antibody quality have called the antibody-based data from whole research areas into question, driven the demand for improved antibody validation and led to the development of alternative tools, such as aptamers and our own Affimer® reagents. To address this issue a consortium of scientists have established a protocol for quantitative antibody (link is external) validation and a framework for creating a community resource that allows scientists to identify which antibodies are the best for a particular application.

Antibody

The team from the University of Toronto assessed a total of 1124 different IgG antibodies and Fab fragments targeting 152 chromatin-related proteins by immunoprecipitation from HEK293 cells followed by mass spectrometry analysis: antibodies that captured their target antigens or a member of its known protein complex as the most abundant were considered ‘gold standard’. A subset of the Fab fragments of varying performance were sent ‘blindly’ to four different labs for validation. Although an SOP was used between the five different labs for the IP protocol, each lab used their own methods for trypsin digestion and mass spec. The identification and relative abundance of each of the Fab fragment-captured proteins could then be compared between labs.
Of the original 1124 antibodies only 354 were able to pull-down their target antigen from HEK293 lysates as one of the three most abundant proteins, achieving ‘gold standard’. These represented 71 of the 152 targets. In the validation stage across the five laboratories, of the 20 Fab fragments that were assessed just 9 showed results across all the labs that were consistent with the initial screen. These inconsistencies may be due to misclassification of the Fab fragment in the initial screen, instability of the Fab fragment or some of the Fab fragments may be susceptible to small changes in the IP conditions. Whilst developing a validated list of reproducible affinity reagents may result in the loss of many currently available antibodies, analysis across multiple labs in this manner can establish reproducibility.

The group also tested a subset of their antibodies in immunofluorescence, western blot and ChIP experiments. In general the gold standard antibodies performed well in immunofluorescence and ChIP applications, but not western blot, which may be attributed to the fact that these antibodies were selected to bind folded proteins.

Analysis of all commercially available and in-house natural and synthetic antibodies through a system such as this would undoubtedly provide the much needed confidence to antibody users within the life science community and an increased level of robustness to the data produced. However, the question remains of who should take up this challenge. The study authors suggest that giving individual researchers the responsibility of validating their own antibodies involves too much risk, and state that instead this should be a public effort involving the whole antibody-using scientific community.