Stringent validation of antibodies is now required in science. It takes longer and it’s more expensive, but our recent experiences showed us that if you want to trust your results, you can’t live without it.

Our lab was looking at the impact of a cellular protein on influenza severity. The protein, Beta 6 (β6) integrin, is found in many cells and is incredibly important in wound repair and development. In fact, it plays a significant role not only in flu severity in preclinical models, it is also increased during cancer. While conducting our study, we began to have problems reproducing our results.  Surely, we were doing something wrong?

We kept checking.

Beta 6 integrin teaches us a valuable lesson

While studying the β6 integrin, we found the antibodies that we were using to detect this protein didn’t always work properly. We noticed extra bands on our Western blots. We saw background fluorescence in our imaging experiments. Sometimes the antibodies didn’t work at all. Of greatest concern, we detected protein in cells where the β6 integrin had been removed. So we ended up taking a close look at every commercial antibody for the β6 integrin, and we found many of them were extensively cross-reacting with other proteins.

For a long time, we questioned our methods and thought we’d done something wrong. As it turned out, it wasn’t us. The antibodies were not as specific as advertised.

It takes a long time

Taking the time to validate our antibodies significantly slowed our research. Our investigations took about two years longer and cost us thousands of dollars. Although validation was costly and time-consuming, we decided we weren’t going to publish our work until we were confident that the reagents we used were specific. Separately, we also published our findings on the antibodies as a word of warning to both the scientific community and to clinicians using these antibodies for diagnostic purposes.

It’s not just time … it’s more serious than that

We were surprised at how many commercial antibodies we tested had problems with specificity, and I recommend taking a close look at the antibodies in your lab. Non-specific commercial antibodies not only contribute to a lack of reproducibility, they have far more serious implications. As a researcher, your reputation is on the line with every publication. Others questioning the reproducibility of your data could lead to serious professional integrity issues.

In addition, β6 antibodies are commonly used by cell biologists, pulmonologists and others who study lung repair and development. They’re also used to test diagnostically for metastatic potential of tumors. How can clinicians use them in diagnostic testing if the results can’t be trusted?

Caveat emptor

Unfortunately, the responsibility to validate these reagents currently lands on the principal investigator. Don’t trust something because you bought it from a company. Always include those controls in your experiments. If we didn’t have cells with and without the β6 protein, we may not have known there was a potential problem. On the plus side, the importance of reproducibility is a hot topic in the scientific community these days, and this just highlights one more way we can work to ensure our results are trustworthy and reproducible.

Because we’re in biomedical research that has direct implications on human health, we want it to be right. My message to the graduate student, postdoctoral fellows, and researchers everywhere — be rigorous, police your own work, and question everything.