(This story has been updated to correct a description of the company's technology.)
Prognosys Biosciences has developed a cDNA-based system for coding customized peptide libraries and has begun offering it commercially as a research tool.
The platform, which uses next-generation sequencing for its read-out, is part of an ongoing effort by the company to parlay its expertise in sequencing and nucleic acid research into proteomics applications, Jill Dombrauckas, Prognosys' director of marketing, told ProteoMonitor.
Headed by Mark Chee — a co-founder of Illumina and former director of genetics research at Affymetrix — Prognosys has to date focused largely on services and technology for nucleic acid research, Dombrauckas noted.
However, the company also has a number of proteomics R&D projects underway, including five active National Institutes of Health Small Business Innovation Research grants funding work such as development of assays for protease and kinase activity profiling and multiplexed protein biomarker quantification.
"There are a lot of exciting things that we're developing in the proteomics space," Dombrauckas said. "To date a lot of our services have been centralized more in nucleic acid technology, but this [peptide assay platform] is one that is starting to bridge [the firm's nucleic acid experience] into proteomics."
Proteomics "is the next logical step," she said. "The genomics space has been studied and has advanced tremendously over the last several years … but the next logical step is to understand the proteomics side and the functional side of what's encoded within the genes. The proteomics space hasn't advanced at quite the same rate due to the complexities and challenges of developing proteomic assays … so some of the technologies [under development at Prognosys] are addressing that."
Detailed in a paper published this week in PLoS One, the company's peptide assay platform uses pools of DNA templates generated via microarray-based synthesis to create collections of peptide-cDNA fusions. The peptides can then be assayed with analytes of interest and the read-out performed via next-generation sequencing.
In the PLoS One study, the researchers built two custom libraries: one containing peptides comprising the hepatitis C proteome, which they used to identify HCV NS3/4A protease cleavage sites; the other a 3,243-peptide library that they screened to identify substrates of Abl kinase.
In the first case, the platform identified all three known trans cleavage sites in HCV and in the second, detected Abl kinase phosphorylation events that were specific to and consistent with observed patterns of Abl kinase activity.
A number of other technologies, including the DNA synbodies developed by Arizona State Biodesign researcher John Chaput (PM 8/19/2011) and Integrated Diagnostics' Protein-Catalyzed Capture Agents (PM 10/28/2011), make use of large peptide libraries for developing proteomics reagents in proteomics research.
Many of these techniques, however, rely on large combinatorial libraries, using random assemblies of peptides in their screens.
Prognosys, on the other hand, has opted for a more targeted approach, using peptide libraries tailored to particular research interests.
"Combinatorial approaches can be valuable in identifying novel sequences," Dombrauckas said, "but a lot of times, especially within maybe the drug development community, [researchers] have a target of interest and so want to explore a specific or customized set [of peptides]. This allows you to really target various organisms or substrates and more specifically address the biological questions you're interested in."
Prognosys has begun offering the platform to customers, including collaborators at the Scripps Research Institute and Sanford-Burnham Medical Research Institute, the latter of which published a paper in April in PLoS One using the system in a study of the HCV NS3/4A protease.
"What we're looking to do is work with customers in any space where they feel this could be a beneficial tool for them, and then we can expand the applications from there," Dombrauckas said.
The company plans in the near-term to offer the system solely on a service basis, she added, noting that "if there's a custom region of interest or a collection of peptides, customers can contact us to work on that."
Eventually, Prognosys hopes to offer the platform both as a service and as a kit that researchers could order to run on their own, Dombrauckas said. "You would be able to bring us the custom content and we would manufacture the [cDNA-peptide] pools and give those back to you."
"We can forsee [the technology] being applicable to a number of different fields in the proteomic space," she said. "We've validated it [in the PLoS One work] for proteases and kinases, so there may be more extensions in those domains, but there is also potential for binding assays for antibodies, autoimmune [research] – a lot of different fields where this may be an excellent research tool."
Prognosys developed the technology on Illumina sequencers, using a Genome Analyzer IIx machine for the PLoS One work. However, Dombrauckas said, the company has not collaborated on the platform with Illumina and it should be easily transferrable to any brand of sequencer.
"We've validated it on an Illumina platform," she said, "but if there are any changes necessary to optimize, say, the barcoding or the sample prep [in order to] use it on another platform, that shouldn't be problematic."
Currently, Prognosys is offering customers and collaborators assays of up to 1,536 peptides, Dombrauckas said, as the company works to develop new applications for the platform. However, she said, that number is not a hard limit and should expand considerably as the technology develops.
She noted that the company might also be interested in adding post-translational modifications to the libraries, an option made potentially possible by recent synthetic biology efforts (PM 10/7/2011).
"We haven't validated [assays featuring PTMs]," she said. "If there's a customer who has that interest, it's something we'd be interested in discussing."
The company may also adjust the length of the peptides in its libraries as more data on the platform comes in. One potential issue with peptide-based screens is that these short sequencings may not contain all the relevant structural information of a full-length protein and so may bind differently than the molecule does in vivo.
In the PLoS One paper the researchers used 10-mer peptides and plan to offer 15-mer peptides to their collaborators going forward. These lengths, though, Dombrauckas said, aren't set in stone.
"Peptides are great tools to explore the proteome, and I do think you can get a lot of information out of a small portion as opposed to having to represent the whole protein," she said. "[But] our technology doesn't have a limitation at 15[-mer]. So if the data reveals that maybe longer peptides are more interesting, then we would probably adjust accordingly."