Antibody quality has long loomed as a significant concern for proteomics research, with much discussion in the field of what validation steps are needed and what data vendors should provide to ensure the performance of a given affinity agent.
According to Nader Rifai, director of clinical chemistry at Children's Hospital Boston and the editor-in-chief of the journal Clinical Chemistry, these concerns have recently taken on a new urgency due to the appearance on the market of a large number of faulty antibodies.
Rifai raised the issue in an editorial he authored for last month's issue of Clinical Chemistry along with Greg Miller, president of the American Association for Clinical Chemistry, and Ian Watson, president of the European Federation of Clinical Chemistry. In that article, the trio warned researchers that while proteomics and protein biomarker researchers might welcome commercial immunoassay kits as a convenient and time-saving research tool, "users of these kits are advised to proceed with great caution."
The editorial spoke in general terms about the potential hazards of using commercial immunoassay kits and provided guidelines for the minimum levels of validation and reference information such kits should include.
However, in an interview with ProteoMonitor this week, Rifai suggested that the article was written primarily in response to the behavior of one firm – a Chinese life sciences company that, he said, is selling a large number of faulty antibodies and has achieved wide distribution in the US and European markets.
Rifai declined to identify this company on the record, saying that it would be inappropriate given his role as a journal editor. He said, though, that he considered its practices "fraudulent" and noted that while antibody quality is a concern generally in protein research, he has "never seen anything like this."
Despite his decision not to name the company, Rifai said that there are several things about its products and practices that should tip off researchers to potential problems – most notably the lack of meaningful reference material accompanying the antibodies and the large number of antibodies, including to quite obscure proteins, offered by the firm. In fact, Rifai notes, this latter factor was a main cause of his suspicions in the first place.
"We have a lot of investigators [at Children's Hospital] who come to us because they just identified a marker [and need an assay for it]," he said. "Many of them will have to raise their own antibodies, especially if it is a very novel marker, and that is very tedious and costly work. And so it's wonderful to be able to go on the [Internet] and find that actually there is a [commercial] assay for whatever obscure protein you are working with."
His group was approached by a researcher wanting to measure several "very obscure" proteins," Rifai said. "And to our surprise, we found that there were kits for them out there" available from the Chinese company.
"So we purchased those kits," he said. "They were priced between $800 and $1,200 for a kit with 38 measurements, and they came with absolutely no instructions. There was absolutely no description of the reagents, no description of the performance, no demonstration that they were actually measuring what they said they were measuring."
In instances where some information was provided, Rifai said, "in many cases it was totally incorrect information."
Concerned, Rifai and his team ordered more antibody kits from the company, this time for more common analytes they were familiar with. These kits, he said, "were just total garbage."
"We are a very experienced lab," Rifai said. "We know what we are doing [with immunoassays], and these things just didn't work."
More worryingly, he noted, he has since discovered these antibodies being sold under different names and through reputable US and European distributors.
"When these kits came, there was absolutely no mention of where the kits were made or who produced them, but if you look at the vials, you will find the insignia of the initial company," he said. "So they have infiltrated very effectively the American and European markets."
These findings, Rifai said, led him to write the Clinical Chemistry editorial, recruiting his two co-authors to lend the stature of their positions to the piece. He said he has also contacted several noted proteomics researchers including the National Cancer Institute's Henry Rodriguez and SISCAPA Assay Technologies CEO Leigh Anderson to make them aware of the problem.
"This is particularly a problem in the proteomics community, and I'm very worried about that," he said. "People are buying these kits over the web and they have no idea what they are measuring."
In an email, Anderson told ProteoMonitor that he did not "have a very broad experience of this [problem]" himself, but that he and his colleagues "have been working on SISCAPA [immunoassays] for two proteins for which the commercial immunoassays don't work correctly."
"The problem," Anderson added, is that while a better estimation of the frequency of bad immunoassays would no doubt be useful for the field, "this is an expensive and thankless thing to find out."
Of course, the concern that some affinity reagents may not perform as expected is not a new one for the proteomics field. For instance, Mathias Uhlen, professor of microbiology at Stockholm's Royal Institute of Technology, has tested a vast number of antibodies from vendors around the world as part of his Human Protein Atlas project, and in the course of this work has discovered many commercial antibodies that don't perform as advertised (PM 11/19/2010).
And the Proteome Standards Initiative of the Human Proteome Organization in 2010 established criteria via its Minimum Information About a Proteomics Experiment guidelines for what information must be included regarding affinity agents used in a proteomics experiment (PM 7/16/2010).
Rifai noted that while the Chinese firm's products are a problem, he believes it is unique in its behavior. "It is one rogue [company]," he said, adding that he didn't know of any others with similar practices.
Rifai called for individual researchers to take more seriously the task of verifying their reagents' validity. "If you look on the web and find no information – because now every manufacturer has the insert sheet on the web – if it is incomplete or the information is not there then you should either request it or you should not buy it," he said. "And if you do end up buying it, you should not just buy a kit and use it; you have to validate it first."
He added that researchers should be suspicious of outfits offering antibodies to a very wide range of proteins, including uncommon analytes.
"It was strange," he said, citing the case of the Chinese firm, "no matter what highly unusual protein we were looking for, it was available. And not only for humans but for about 15 other species as well. They have like 2,000 proteins [covered]."
He compared that to the menu of high-quality antibody producers like R&D Diagnostics, noting that, if you look at [R&D's] menu it is miniscule compared to what [the Chinese firm] is trying to offer."