FASEB hosts discussion of reproducibility in antibody research

By | September 24, 2015

On September 10, the Federation of American Societies for Experimental Biology (FASEB) hosted a dialogue on ways to improve the reproducibility of scientific studies that use antibodies. The discussion featured presentations by five experts in the field who described different strategies to validate antibodies and the studies that use them.

The roundtable was the second of three FASEB-led discussions on the reproducibility of scientific research. The discussion series will culminate in guidelines to help investigators navigate a new National Institutes of Health (NIH) policy to address rigor and transparency in grant applications that is scheduled to go into effect in January 2016. In addition to the speakers, attendees included delegates from 11 of FASEB’s 27 member societies and the Director of the National Institute of General Medical Sciences (NIGMS), Jon Lorsch, PhD.

Although most biotechnology and pharmaceutical companies perform some quality control on antibodies, the event’s first presenter, Andrew Bradbury, PhD, of Los Alamos National Laboratory’s Biosciences Division, expressed concern about their failure to standardize antibody production and validation. As a solution, Bradbury proposed sequencing and manufacturing antibodies using recombinant technologies. He suggested that stakeholders, including funding agencies and publishers, take an active role in requiring researchers to assure antibodies used in experiments as a condition of funding or publication.

David Rimm, MD, PhD, Yale University Professor of Pathology and Medical Oncology, discussed the need for improved validation of antibodies to ensure rigorous and reproducible research. Rimm said multiple factors, such as sensitivity and specificity, must be evaluated when validating antibodies.  Validation should include tests of replicability with complementary reagents as well as different product lots, laboratories, and experimenters, he said. He recommended use of only well-characterized monoclonal or recombinant antibodies that are specifically validated for each protocol.

Tara Hiltke, PhD, a program manager at the National Cancer Institute’s Office of Cancer Clinical Proteomics Research, described how the Cancer Institute produces and characterizes antibodies for researchers.  Her laboratory develops up to three antibodies for each antigen. Each antibody then undergoes extensive characterization using multiple methods, including western blot, ELISA, Surface Plasma Residence, immuno-mass spectrometry, NCI 60 protein array, and immunohistochemistry (IHC). The results are detailed in an online antibody portal that is available to the public. The program uses a standardized nomenclature and releases detailed experimental procedures to enhance reproducibility.

Sigma-Aldrich’s Aaron Sin, PhD, discussed business models for the manufacture and sale of research antibodies. He began by providing a brief overview of the antibody business and noted the differences between the costs of production for polyclonal, monoclonal, and recombinant antibodies.

Although Sigma-Aldrich typically includes a validated IHC image, titration curve, and western blot analysis for each antibody, Sin agreed with other presenters that the information is often insufficient for validation. To improve the process, Sigma-Aldrich’s development team is focusing on expanded validation of its top-selling antibodies, most of which were developed for research rather than clinical applications. He urged consumers to push companies to provide better validation information.

The final presenter, Michael Okimoto, PhD, chief content officer of the online database BioCompare, described ways in which antibody catalogs can improve rigor. BioCompare, a product aggregator that compiles product information from a variety of sources, features about 10,000 reviews, the majority of which are for antibodies, he said. Users who submit reviews must respond to specific questions about experimental conditions, sample preparation, and applications. Keeping the website up-to-date is a key challenge, Okimoto said. The company uses text mining software to link products to research publications; however, limited access to journal articles and inconsistent reporting of reagents in the materials and methods often slows the process.

Following the presentations, meeting participants developed draft guidelines for investigators to consider and report when describing specific technologies using antibodies in applications for federal funding. FASEB’s final roundtable on scientific rigor will be held October 7. The federation will issue guidelines to enhance reproducibility in January 2016.