Reproducibility remains a problem in life sciences research. A major cause of the lingering issue is poorly characterized reagents—especially antibodies—that can vary wildly in composition and yield different results in different labs.
A group of scientists has issued a call for the community to improve the quality of antibodies and other protein-binding reagents. In a comment in Nature, Andrew Bradbury of Los Alamos National Laboratory, Andreas Plückthun of the University of Zurich, and 110 other life scientists propose that all antibodies used as reagents be sequenced and that the sequences obtained be used to produce standardized versions (2015, DOI: 10.1038/518027a).
“People have been complaining about the quality of antibodies used in research for a long time, but no real solutions have been forthcoming,” Bradbury says. “Antibodies are the only widely used reagents in biological research that are not defined at the sequence level.” Bradbury and colleagues thought it was time to finally address this issue.
For decades, scientists have used polyclonal antibodies. Researchers inject an animal—often a rabbit—with an antigen and then collect the resulting antibodies. Only some of the antibodies bind the intended target, and the mix of antibodies varies from batch to batch.
“Polyclonal antibodies are completely undefined,” Bradbury says. “The polyclonal I use and order from company A will be different from the polyclonal you order from company B.” The researchers are calling for phasing out the use of polyclonal antibodies as reagents.
Monoclonal antibodies are an improvement. Unlike polyclonal antibodies, which are produced by different types of B cells in an animal, monoclonal antibodies are made by just one, a hybridoma. This cell is a fusion of a cancer cell and a B-lymphocyte cell. But even this process can result in mixtures of multiple antibodies.
If antibodies are defined by their DNA sequences, then researchers or reagent sellers can enlist viruses and microbes to synthesize the proteins to ensure consistency and eliminate mixtures. Also, with a published sequence, researchers can be sure they’re using the right reagent when they try to replicate others’ experiments. “The sequence of an antibody or binding reagent is the ultimate bar code,” Bradbury says.
Bradbury and his cosignatories are calling for an international collaboration and funding initiative to underwrite the necessary characterization. This effort would cost about $1 billion and take about 10 years to characterize reagents for all human genes, the scientists estimate. They also ask that the scientific leaders of the U.S. National Institutes of Health and the European Union’s funding agencies convene various stakeholders to establish a realistic timeline for a transition to well-characterized reagents.
As part of these efforts, the scientists want NIH to expand its Protein Capture Reagents program, which was launched in 2010 to develop new methods of producing and validating reagents such as antibodies. Over the past four years, NIH has invested more than $42 million in the program.
Hundreds of monoclonal and recombinant antibodies have been produced as a result of the reagent program, an NIH spokeswoman says. But those reagents still need to be validated, and users need to be convinced that they work better than existing ones. Reviewers of the program have told NIH that scientists will start using the new reagents once there are data validating their effectiveness, the spokesman says.
This validation is the current focus of the reagents program. “Since different investigator communities use protein-capture reagents in different ways, there is no single set of validation criteria,” the NIH spokesman says. “Professional societies may need to develop new standards that are most useful for their communities.” This would enhance reproducibility and adoption of new reagents across communities.
In a separate effort, the Global Biological Standards Institute has convened a task force “to focus on the need for standards and best practices to advance the credibility, reproducibility, and translatability of preclinical research using antibodies,” Leonard P. Freedman, GBSI president, told C&EN. GBSI is developing an online survey to identify current practices and needs for antibodies in biomedical research. The results of that survey should be available in the summer or fall of 2015.