NEW YORK (GenomeWeb) – A former Arizona State University researcher has launched a diagnostics firm that aims to move a streamlined approach to mass spec-based protein analysis into the clinic.
Dobrin Nedelkov, formerly a group leader at the ASU Biodesign Institute and CEO of clinical proteomics firm Intrinsic Bioprobes (acquired by Thermo Fisher Scientific in 2011), last year launched a new company, Isoformix, focused on identifying clinically relevant protein isoforms and developing tests based on them.
The company plans to concentrate primarily on markers for cardiovascular disease, diabetes, and cancer, Nedelkov said, noting that it won't do extensive discovery work but will, instead, focus on studying isoforms of known protein markers with the expectation that measuring specific isoforms will improve the performance of these markers.
"We may sometimes do a targeted discovery, but we're not going to be discovering proteins that haven't been shown to exist before," he said. Instead, he added, the company plans to analyze known protein biomarkers and then analyze changes in their structure that might arise from post-translational modifications or alternative splicing.
"We're not going to screen 4,000 proteins to see which one has changed," he said. "We're going to look at … maybe a dozen or so proteins that have already been shown to be important in these disease areas [the company is targeting]. And out of these dozen or so proteins, maybe one or two will show [structural] changes that potentially differentiate the disease state."
The notion that protein isoforms could be diagnostically useful is not a new one. In fact, more than half of the cancer protein markers currently approved by the US Food and Drug Administration are glycoproteins. That said, a large proportion of protein biomarker discovery work has consisted of the sorts of experiments Nedelkov alluded to, in which researchers compare the expression of hundreds or thousands of proteins across disease cases and controls, looking for differences that could indicate potential markers.
That approach, which has typically used fairly small sample sets in the discovery stages, has had limited success, with few proteomic tests making it to the clinic in the roughly two decades since the field arose, and even fewer finding commercial success.
This has led researchers and clinicians to rethink these methods, with growing emphasis placed on using large sample cohorts (in the range of hundreds to thousands of samples), throughout the entire discovery and validation process.
"Discovery is important," Nedelkov said, "but to do truly good discovery, you need to work with good clinical cohorts. I'm talking about thousands of samples from patients that have been followed for 10 or more years, where their outcomes are known."
Improvements in mass spec technology have made this more feasible than it once was, but Nedelkov noted that the complexity of current mass spec workflows remains an impediment to running studies of this size. Additionally, this complexity has limited uptake of mass spec as a clinical platform, particularly for protein-based tests.
With Isoformix, Nedelkov aims to address this problem by radically simplifying clinical mass spec workflows. He plans to do this in two main ways: analyzing intact proteins, which will allow clinicians to skip the protein digestion steps that are a major source of complexity for mass spec protein assays; and eliminating liquid chromatography separation upfront of mass spec analysis, which limits throughput while also hampering assay robustness.
Instead, he plans to couple antibody-based enrichment of target proteins to an LC-free form of mass spec like MALDI-TOF, which will enable a simpler, faster analysis. Such a platform would likely offer lower performance than a high-end LC-MS system, but Nedelkov said he believes an emphasis on technology development and performance improvements has in some ways distracted from questions around how to apply the technology's capabilities in the clinic.
"Mass spec has matured in terms of technology, but it hasn't matured in terms of application," he said. He noted that mass spec-based proteomics' greatest clinical success to date has come in microbiology, where the instruments used are typically lower-end MALDI systems, and that Isoformix aims to develop assays that could run on similarly performing equipment.
Nedelkov has experience developing such assays. In 2014, he co-authored a study in PLOS One presenting an assay that combined immunoenrichment with MALDI-TOF mass spec to quantify intact insulin-like growth factor 1 (IGF1) in patient serum. With the method, Nedelkov and his colleagues were able to quantify IGF1 in 1,054 human samples in around nine hours, including forms of the protein featuring point mutations and post-translational modifications.
That assay used the mass spectrometric immunoassay (MSIA) technology developed by Nedelkov's Biodesign Institute colleague Randall Nelson, who founded Intrinsic Bioprobes to commercialize the technology. MSIA uses specialized microcolumns functionalized with antibodies to enrich proteins of interest prior to mass spec analysis. The company promoted the technology for analysis of protein isoforms, similar to what Nedelkov aims to do at Isoformics.
The MSIA microcolumns are available from Thermo Fisher, where Nedelkov worked as an R&D manager for roughly two years after the company purchased Intrinsic Bioprobes. Nedelkov said, however, that he viewed the MSIA columns as still being more of a research technology and that Isoformix aims to develop a more streamlined and less expensive antibody-enrichment method for its assays.
The key question, of course, is whether the company can, in fact, identify clinically useful protein isoform markers that can be measured using the sort of streamlined workflow Nedelkov envisions.
With typical protein markers, "you're looking at amounts of [total] protein and correlating that amount to a specific state of disease," he said. "Now we're trying to open up a second dimension, which is the structural differences, and there we're linking a single protein and [its variants] to a specific state of disease."
"But let's not fool ourselves," he added. "This is not going to be a huge must-do for every protein. Some proteins may not have another dimension. Some proteins may just have a quantitative value that's important, or they may have [isoforms] that are completely unrelated to the disease state."
A second question is, even if Isoformix can identify such markers, will mass spec prove the most effective technology for measuring them?
Nedelkov provided the example of the diabetes marker glycosylated hemoglobin.
"That's one of the best examples of a structural change in a protein that is correlated with disease," he said. "But it took a long time for it to be adopted. And now that it has been adopted, it's not run on mass spec, because mass spec is too expensive. So it's a success story in terms of the protein biomarker based on a structural change, but it's not a success story in terms of mass spec."
At least initially, successful mass spec tests will need to match the performance, price point, and ease of use of immunoassays, while also measuring proteins or protein isoforms not readily detectable by antibody-based methods, Nedelkov said, adding that few markers will likely fit that bill.
"The number of markers that will successfully be run on a mass spec platform initially is not going to be big," he said. "The first one, the first half a dozen will be the hardest, because you need to convince the market that those are important. Then once you have the platforms out there you can convert a lot of the things that are done on immunoassay to mass spec based on the premise that mass spec is more precise and all that."
Nedelkov said Isoformix is about to start on a large clinical cohort study that he said he expects will take several years to complete. He declined to say what disease area it will investigate, but said he has lined up funding for the project and plans to announce it later this year.