GE Global Research has developed a fluorescence microscopy-based protein detection system capable of measuring more than 60 analytes in a single sample and visualizing them with subcellular resolution.
Beginning midyear, the company, through its Clarient cancer diagnostics unit, plans to offer the platform as a research service called MultiOmyx, Fiona Ginty, high value diagnostics biology platform leader at GE Global Research, told ProteoMonitor. Longer term, the company is considering developing the technology as a device for selling into clinical pathology settings, she said.
The company has partnered with Vanderbilt University to continue development of the platform as part of work on colon cancer led by Vanderbilt researcher Robert Coffey. That work is supported by a five-year, $3.75 million grant from the National Institutes of Health.
The research will focus on colon cancer stem cells, in particular, Coffey told ProteoMonitor, investigating events driving pathogenesis and progression of the disease.
The platform uses antibodies conjugated to fluorescent dyes to stain proteins of interest in batches of two to four analytes at a time. Researchers then image the stained tissue and deactivate the fluorescent dyes via a proprietary process. They can then stain the tissue with the next round of antibodies, multiplexing in an iterative fashion.
Thus far, GE researchers have measured as many as 65 proteins in a single sample, Michael Gerdes, a GE Global Research biologist and leader of the platform development effort, told ProteoMonitor. In theory, the system could allow for even higher multiplexing, he added.
"There's not really a theoretical limit that we know of," he said, suggesting that, ultimately, physical degradation of samples due to the handling involved might prove the limiting factor.
The platform can analyze in the range of 3,000 to 5,000 cells in a single sample and can "in each individual cell quantify biomarker expression down to the subcellular level," Gerdes said. "So you end up with a very powerful dataset to really explore what's happening within the [cell] populations that we're examining."
"In one 6-micron section you can look at [more than 60] antibodies and get the spatial compartmentalization of these markers," Coffey said. "So in the case of colon cancer, we know it's a heterogeneous mix of elements – fibroblasts, lymphocytes, blood vessels – and more and more we're recognizing that the microenvironment plays a critical role in tumor formation and progression.
"So now what you can do is in one section look at the relationship of the different cells in that section, and [that] will allow you to draw what we think are very strong inferences about the communication that is going on between different cell types," he said.
Initially Coffey plans to use the platform to investigate two different colon stem cell populations – one identified by the well-known stem cell marker LGR5, and the other a distinct population of stem cells recently discovered by his lab that is identified by the marker Lrig1.
Using the GE platform, the researchers plan to characterize these two normal stem cell populations and then create tumors in mouse models and analyze the stem cell populations of these tumors, Coffey said.
"We think it's going to be incredibly important to give insight into the pathogenesis and progression of colon cancer," he added. "We'll be able to tell what cells in the stroma – what B cells, what T cells – are now entering this tumor microenvironment and begin to see what proteins they are producing, and I think it's going to give us some unique insights into both the pathogenesis of colon cancer and the events involved in progression."
Beyond his own research interests, Coffey suggested the platform could also prove useful in a clinical setting, particularly for predicting and tracking response and resistance to targeted cancer therapies – much as proteomics researchers and companies like Theranostics Health have been using reverse phase protein arrays (PM 6/15/2012). The GE platform, Coffey added, would have the added advantage of providing high-resolution spatial information in addition to data on protein signaling and activation.
Although GE is bringing the platform to market as a research service, the development team has been keeping such clinical applications in mind, Ginty said. She noted that by offering the service out of Clarient they would be "working side-by-side with an operation that delivers clinical workflows.
"We recognize – and this is a motivation for developing this technology – that in pathology there is a need for the ability to do multiple markers on the same sample," she said. "Especially where samples are small or where complex signatures are required to determine prognosis or outcome, and where morphology and histology are needed in conjunction with protein information.
"Our goal would certainly be to work within the specifications of clinical requirements," she added. "We've really been working hard to make the workflow compatible with the expected turnaround time within a clinical environment."
Currently, each staining cycle takes roughly an hour, with imaging times running from ten minutes to an hour, Ginty said. She noted that quantification of the markers is automated, potentially reducing quality and reproducibility issues associated with manual reading.
According to Gerdes, GE currently has antibodies to more than 175 markers ready for use with the platform. He added that in experiments to test for background resulting from multiple staining cycles the researchers have found no added background after as many as 100 cycles.
Gerdes did note, though, that antibody competition could prove problematic in certain instances – particularly when two epitopes sit relatively close to each other on a single protein.
"If you have two antibodies to, say, an intracellular kinase and one of the antibodies is to the native form of the protein and one is to the phospho-form and those epitopes are really close together, we have seen evidence of some competition at that point," he said.
Given the large amounts of data collected by the platform, analysis is a significant challenge, Gerdes said, noting that working out and optimizing such methods would also be a key component of the Vanderbilt collaboration.
"We've gone up to 65 markers on single samples… and the emphasis of this grant is really going to be on how we analyze that data," he said. "We can do things like clustering algorithms to look at what sort of patterns different cell populations have relative to one another, and it really helps to get a lot of insight into what's happening in this spectrum of disease heterogeneity."