McGill University researchers have developed an antibody array that could significantly mitigate the cross-reactivity and interference problems that have troubled such devices, allowing improved multiplexing of immunoassay-based protein detection methods.
Described in a paper published in the current edition of Molecular & Cellular Proteomics, the array platform – named an antibody co-localization microarray, or ACM – consists of individually spotted sandwich ELISAs, an approach that David Juncker, a McGill scientist and leader of the effort, said allows for a much higher level of multiplexing.
In conventional multiplex sandwich arrays, the capture antibodies are first arrayed on a slide and then the detection antibodies are typically applied as a mixture, meaning they interact not only with their intended protein target, but also with each other, the capture antibodies, and other proteins immobilized on the array. Such cross-reactivity leads to false positives and background noise.
To get around this problem, the McGill team spotted not only the capture antibodies, but also the individual detection antibodies, thereby eliminating the mixing of reagents and the potential for cross-reactivity.
"The trick [behind the ACM] is that we spot the capture antibodies, then we incubate the sample, and then we [spot] each individual detection antibody specifically to a single spot so it's co-localized [with the protein and capture antibody]," Juncker told ProteoMonitor this week. "So in this way we get rid of the interactions that cause cross-reactivity."
Using this approach, the researchers built a 50-plex array that they used in a proof-of-concept experiment investigating protein biomarkers for breast cancer. In an analysis of serum samples from 15 ER-positive breast cancer patients and 11 age-matched controls, they identified six candidate markers that could be useful for detecting the disease.
Additionally, Juncker noted, the study attempted to quantify levels of cross-reactivity in conventional multiplexed sandwich assays, establishing a scaling law for MSAs vulnerability to cross-reactivity. The researchers found that this vulnerability increased quadratically with the number of targets, making substantial scaling of MSAs impractical.
"People know there is cross-reactivity [in MSAs]," Juncker said, "but it hadn't really been rationalized in any way. By putting numbers on [the problem], it makes it really jump out at you."
The ACM platform, on the other hand, should, in theory, be able to multiplex thousands of assays on a single array, he said. However, he noted, even without cross-reactivity concerns, the expense of antibody pairs remains a limitation.
"What limits most people [using conventional MSAs] is that their antibodies cross-react. What limits us is the cost of pairs," Juncker said. "I could go out and buy 300 or 400 pairs of antibodies, but 300 or 400 pairs at $1,000 each – that's a heavy investment."
The elimination of cross-reactivity concerns could, however, allow the researchers to use less expensive reagents that might not work well in a conventional MSA, he noted.
"That's a great strength of this approach," he said, adding that the small amount of reagent used in the microarrays should also help keep costs down. "It costs a lot, but when you calibrate how much antibody we use in the spotting, it's very economical."
The platform might also aid in areas of research where relatively few good antibodies are available, Juncker said, pointing to the team's proof-of-concept breast cancer work.
"There are a lot of cancer markers where people want to multiplex [assays], but it can be very difficult because there are relatively few antibodies against some of these targets and so when they cross-react you don't have an alternative," he said. "So that's an important advantage."
The researchers have patented the technique and are interested in commercializing it, Juncker said, but he added, "we haven't gone very far yet in pursuing that."
The team has another study, detailed in a paper currently under review, which could help push the platform toward commercialization. In this work, the researchers spotted two slides – one with capture antibodies and the other with the corresponding detection antibodies – which can then be shipped as a kit or stored for the future. To use the array, researchers incubate their sample on the slide with capture antibody and then transfer the detection antibodies from the slide they were originally spotted on to the sample slide.
Production of the arrays wasn't conceptually difficult, Juncker said, but required tending carefully to practical details.
"The spots are only around 100 micrometers in diameter, so we need very precise mechanical alignments when we make the [capture antibody] spots, take the slide out [to incubate with sample], and then put it back on the spotter [to spot the detection antibody]," he said. "
Juncker added that the researchers used a custom-built printer to spot the arrays – essentially an inkjet with a pin spotter head mounted on it.
Moving forward, he and his collaborator Morag Park – also an author on the MCP paper – hope to use the platform to expand on the breast cancer biomarker work they presented in the study, he said.
Park "has an ongoing sample collection [project] and a blood bank that goes back several years, and so we have access to that and we'd like to run those samples," he said. "We're thinking about doing a pilot study over the course of this year, using about 100 patients."
"In parallel, I would also like to increase the range of targets on the chip," Juncker said, noting that at 50 antibodies, the chip is currently more of a validation platform than one suitable for biomarker discovery work.
"One of the main goals of setting this [platform] up was to look for blood-based biomarkers, and the reason we [developed] the platform was to address this limitation in multiplexing," he said. "One thing that struck me is that DNA microarrays went from a few hundred to a few million pairs and have had tremendous success in research and now impact clinically. Why don't antibody arrays scale up in a similar way? From my perspective these things just [haven't had] the required performance for the biomarker applications that we want to use them for."
Have topics you'd like to see covered in ProteoMonitor? Contact the editor at abonislawski [at] genomeweb [.] com.