NEW YORK (GenomeWeb) –University of Toronto researchers have devised a workflow using mass spectrometry to provide a quantitative measure of antibody quality.
Detailed in a study published last month in Nature Methods, the approach is intended to be specific to antibodies used in immunoprecipitation experiments but appears to have relevance to other types of assays, as well, Aled Edwards, a University of Toronto researcher and senior author on the paper, told GenomeWeb.
Antibody quality is a vexing issue in proteomics and life science research more broadly. While antibody-based assays are overwhelmingly the tool of choice for measuring proteins, antibody reagents are notorious for their poor performance and lack of validation.
A comment by Stockholm Royal Institute of Technology (KTH) researcher Mathias Uhlén, who has been compiling data on antibody quality as part of his work leading the Human Protein Atlas, suggests the large scope of the problem. In an interview last year with GenomeWeb, he noted that his group in the course of its work has tested more than 25,000 antibodies from roughly 50 different suppliers, and that the majority of these reagents have not worked in their hands.
Adding to the complexity of the problem is the fact that antibody performance is highly dependent on context. For instance, an antibody may perform well in one application —say, Western blotting — but poorly in another, like immunohistochemistry. Factors like concentration and the type of sample used can also have significant effects.
"There are, of course, thousands and thousands of publications where people have used antibodies, and the results they are showing are actually not what they think they are looking at," Uhlén said last year. "So there are quite a lot of question marks around the use of antibodies."
Indeed, Edwards said this week, "There are millions of antibodies on sale, none of which have a mechanism by which you can compare their quality quantitatively. So where do you start? It quickly becomes a problem where you just want to throw your hands up."
Rather than throw up their hands, Edwards and his colleagues looked for an antibody-based approach where they could potentially generate quantitative data on reagent performance.
"We looked at the techniques that are commonly used, like immunofluorescence and immunohistochemistry and Western blotting and immunoprecipitation, and we realized that only one of them – immunoprecipitation, is amenable to some sort of quantitative analysis," he said.
In an IP experiment, it is possible to obtain solid quantitative data on how well an antibody captures its target antigen compared to other off-target antigens by measuring the levels of the various antigens pulled down by the antibody in question. Starting from this notion, Edwards and his colleagues developed a series of standard operating procedures using mass spec to evaluate the quality of antibodies used in IP assays.
Antibodies for which the most abundantly captured protein was either the antibody target or a known member of that target's protein complex were designated the highest quality, or, "IP gold standard."
"The general thesis was that here is at least one small slice of the antibody world in which we can apply standard operating procedures and quantitative biology," Edwards said. "It had all the ingredients [needed] to be able to create a robust platform where you put [the antibodies] through the platform, and you get numbers out on the other side."
Edwards' group spent around five years working out the details of the SOPs, after which they enlisted five other labs around the world to test the process using their SOPs and blinded antibodies.
"We let them analyze it to show that it wasn't a lab-specific phenomenon, and it proved very reproducible," he said. In all, the researchers looked at the performance of 1,124 recombinant antibodies to 152 chromatin-related human proteins in HEK293 cells. Of those, 1,124, 452 detected their target antigen.
Of the 152 antigens looked at in the study, 98 had antibodies that could detect them. Of those 98, 71 had had at least one antibody that met the researchers' "IP gold standard" designation, while for the other 27 antigens, the best performing antibodies were "inconclusive," the authors wrote.
The failure of the majority of the antibodies studied to effectively detect their target antigen indicates the challenges involved in using these reagents. The Nature Methods study, Edwards said, gives researchers "at least a starting point of a platform that the community can use to assess antibody quality."
The approach is aimed at evaluating antibodies for use in IP experiments, but, Edwards noted, the researchers found that it also appeared to provide some indication of how well a reagent would perform in other types of assays like immunofluorescence and ChIP.
"We were able to show that once an antibody was given a gold standard in the [IP] method we used, then proportionally, it worked better than average in other applications," he said, "So, it looks like a good first pass standardization" for these other methods.
Generating solid quantitative data for antibody performance in methods like immunofluorescence, however, would be quite challenging, Edwards said.
"How do you quantify immunofluorescence?" he asked, noting that such an effort would likely require comparison of antibody staining in target cells to staining in the same cells with the antigen knocked out.
"But you would need to make knockouts in every sample you tried, and that is hard to do" he said. "Say you take a human muscle biopsy from a patient. The control for that is that patient biopsy [with the target protein knocked out]."
Perhaps an even bigger problem than the technical challenges, however, is getting researchers to take the issue seriously, Edwards suggested.
Two things that could have a significant impact, he said, would be if journals and granting agencies demanded more proof from researchers that they were using high-quality reagents in their work.
"Every scientist responds to those two pressure points," he said. "If funders insisted that one used high-quality antibodies in grant applications or insisted that all antibodies have to pass quality controls or some metric like that, that would be one mechanism to make scientists, as it were, behave. [And] if journals insisted that antibodies used in their journals have gone through some sort of standardized characterization, then I think that would change behavior."
"It is a lot of work to use a good reagent," Edwards said. "It's expensive. People don't have the time. People don't have the reagents or mass spec in house, and so I can acknowledge the difficulty. But it is super important because a lot of taxpayer dollars are being used."