NEW YORK (GenomeWeb) – In another sign of the life science community's increasing concern about antibody quality, a team led by Mathias Uhlén of the Royal Institute of Technology in Stockholm has formed a new working group to develop guidelines for validation and use of these reagents.
The International Working Group on Antibody Validation held its first meeting last month in Vancouver, Canada, coinciding with the Human Proteome Organization's annual meeting. According to Uhlén, it aims to have a set of guidelines it can present in a white paper at next year's HUPO meeting.
Thermo Fisher Scientific is providing support for the group, but the company is not present at its meetings and will not have input into any guidelines it develops, Uhlén said.
Antibody quality has been a challenge in life sciences research since such reagents were introduced decades ago, but discussion of the matter has intensified recently with, for instance, the publication of editorials in a variety of high-profile journals.
As leader of the Human Protein Atlas project, which aims to collect antibody-based data on the full human proteome, Uhlén has particular incentive to take up the matter.
Indeed, he told GenomeWeb that in his work on the HPA he spends around half his time dealing with validation of antibodies used in the project.
"So it is very timely now to try to get together guidelines both for providers and users ... [as well as] funders and journals," he said.
Journals, in particular, he noted, could serve as key motivators for users and vendors to make sure that their antibodies are properly validated.
For instance, he said, if journals were to require certain levels of validation for antibodies used in published papers, this would essentially force vendors to meet these requirements as researchers would be reluctant to purchase reagents that would not readily pass muster with the standards of journals in which they hoped to publish.
Uhlén said that the group hopes this spring to engage a number of journals in discussions about guidelines, adding that the group has "very good support from some of the major publishing houses."
While interest in questions of antibody quality is intensifying, the emergence of genomic techniques like CRISPR gene-editing have recently made more stringent levels of antibody validation feasible, Uhlén noted.
"During the last few years there have [been developed] some new ways of validating antibodies which are very good and much better than what we had before," he said, adding that one of the goals of the working group was to make the community more aware of these validation methods and to push vendors to adopt them.
"The most dramatic [improvements] are genetic methods — siRNA and CRISPR," he said. "So, interestingly, the new ways to validate antibodies are coming from the genomics community."
He added that "mass spectrometry" is an "obvious" choice for additional independent validation of antibodies.
There are signs of uptake of such methods among vendors. For instance, last month Abcam announced that it had begun using CRISPR-based knockout validation to test its reagents. And, last year, antibody seller Proteintech began offering antibodies validated via siRNA knockdown experiments. Uhlén also noted that Novus Biologicals and Atlas Antibodies, a company he founded to commercialize the antibodies used in the HPA project, also use knockdown and knockout validation.
Knockdown and knockout experiments are key in that they allow vendors to test their reagents for non-specific binding, which, Uhlén noted, is the primary challenge facing researchers using antibodies.
While cases of outright fraudulent antibodies have been reported, these are relatively rare, he said.
"I'm sure there are some antibodies that are just crap, but in the vast majority of cases vendors sell antibodies that bind to the target," he said. "But in some cases the cross-reactivity in some applications makes them more or less useless."
The lack of broadly accepted guidelines for how antibodies should be validated for various purposes exacerbates this problem, he said, noted that the working group hopes for this reason to develop guidelines that will be adopted by the field.
Given the vast number of antibodies researchers have to choose from, he added, clearer guidelines could help force vendors to provide the necessary validation to attract customers.
For instance, Uhlén said, the Antibodypedia database established by his group has information on 1.7 million antibodies to human proteins.
"If there were clear guidelines how you validate an antibody for a specific application, people could look at it and the vendors would be forced into doing that [validation]," he said. "Since we have 1.7 million antibodies to choose from, I am pretty sure we will find good antibodies [among them]."
Even with adoption of new validation standards, though, antibody quality will remain a challenge for life sciences research, Uhlén noted.
For instance, the question remains of how specific a knockout experiment must be to validate an antibody for a particular use.
"Just because a vendor has shown that an antibody works fine for Western blot in liver it doesn't mean that you can take kidney samples and run it on a Western," he said. Additionally, he noted, the performance of an antibody depends not only on the antibody itself but will also vary with, for instance, the relative abundance of its target compared to the abundance of other potentially cross-reacting analytes in a sample.
"I don't think that this will be foolproof," he said. "There will always be grey zones and there will always be wrong use with antibodies. It's not very easy. This is why we have been working with antibodies for 40 years and we still haven't sorted out completely the cross-reactivity and background of different assays, and we will have to live with that."