NEW YORK (GenomeWeb) – NanoString Technologies has unveiled its plans for a new technology to provide spatial data on protein and RNA in tumor samples. Formerly known as "digital immunohistochemistry" (IHC), the Digital Spatial Profiler (DSP) is already drawing interest from the firm's partners in pharmaceutical and immuno-oncology research.
On a conference call following the release of the firm's third quarter financial results last week, NanoString CEO Brad Gray announced the name of the new technology and provided initial details of an early-access program that already has several pharma companies enrolled.
It's an idea built around two existing NanoString products, Gray told GenomeWeb: the nCounter analyzer and the photocleavable antibody-driven 3D biology assays the firm has launched in the last year.
At the moment, NanoString has only a few prototype devices, but eventually the DSP will take the form of a sample processing instrument that will feed plates of DNA barcodes to an nCounter instrument.
Using antibodies to target specific proteins, the instrument will lay them down on tissue samples while a focused UV beam will "knock off" the barcodes in a region of interest, defined by the researcher. "Those barcodes are sipped up, collected, put on a plate, and analyzed the same way as any other sample," he said. "That way you can come back and assign [protein or RNA] expression information to any particular region."
Gray said NanoString's R&D team has been able to focus light down to the single-cell level to look at one cell at a time, but that's not the driving purpose of the instrument. It's the ability to go back to different regions of the same sample that makes it an exciting technology, he said.
"Each region is a separate sample for analysis," he said. "One slide yields a larger number of samples for processing." And proof-of-concept data suggests the technology is highly multiplexable. "It's just as good for twelve [regions] as it is for one," Gray said.
Added to the company's existing multiplexing capabilities — which include over 800 barcodes — providing information about different areas of the sample "has the potential to make an unparalleled discovery tool." That's according to David Rimm, a pathologist at Yale Medical School who has collaborated with NanoString by providing samples for early work on the DSP. "Spatial info is orders of magnitude more information than can be obtained from a ground-up specimen."
The DSP can be traced back to a different NanoString collaboration with an academic researcher. In 2014, researchers at Massachusetts General Hospital led by Ralph Weissleder developed a way to perform protein quantitation using antibodies linked to photocleavable DNA barcodes.
Though developed at MGH, NanoString chose to partner with the University of Texas MD Anderson Cancer Center to further develop the technology.
"It is necessary to have a CLIA-compliant, FDA-compliant approach to move multiplex RNA and protein analysis to the clinic," MD Anderson's Gordon Mills told GenomeWeb. Mills is a NanoString-sponsored researcher and has invented and licensed patents to the firm that contribute to the 3D Biology portfolio.
Already approved under both regulatory regimes for the Prosigna breast cancer assay, the NanoString nCounter platform was attractive to Mills as his sprawling research outfit looked for a single assay that provided quantitative information on both RNA and proteins.
IHC is a major part of the research he directs, but existing technologies use up large amounts of ever shrinking samples. "To go in now with a single slide to analyze up to 800 RNAs and proteins at once is an exciting step forward that we couldn't do previously," Mills said.
"Barcoded antibodies are the basis for our proteomics and 3D Biology assays," Gray said, adding that the firm offers a library containing 30 barcode-linked antibodies at once. "A natural question was, 'Could you do that in situ on a tissue [to] see how protein expression varies in space?'"
NanoString began working on proof-of-concept experiments in 2015, providing the first description of 'digital IHC' at the meeting of the American Association for Cancer Research in April 2015, Gray said. "We proved it could work," he said.
Since then, adding RNA quantification has been the biggest technology advancement. "We think it's a really strongly differentiating feature," Gray said. "It's challenging to develop good antibodies and [researchers] are limited to looking at targets where they have an antibody. With RNA, they can look at targets they have not yet developed antibodies for."
This week, at the meeting of the Association for Molecular Pathology, NanoString is presenting a poster on the abilities of DSP to provide spatially resolved, multiplex characterization of both protein and RNA in formalin-fixed, paraffin-embedded samples. One experiment shows a colorectal cancer tumor slide, analyzed by three-color staining but also using DSP for 12 regions of interest. NanoString scientists were able to analyze more than a dozen proteins in each of the 12 regions, far more than traditional imaging would allow, Gray said. "We are not limited by colors," he said.
Stained colorectal tumor slide and circled regions analyzed by Digital Spatial Profiler. Image credit: NanoString Technologies
Moreover, the spatial data allowed the researchers to present a heatmap of protein expression, based on whether the region of interest was in the stroma, the tumor, or the immune-enriched area. "That's the type of information you see in mRNA expression [studies]," Gray said. Using traditional IHC, each row would have had to be a separate stain, slide, and color. Moreover, it demonstrates the new ways that researchers will be able to visualize their data.
"This is very new spatial data," Gray said. "No one's ever had so much before, so we need new software tools that probably don't exist today."
It's one of the factors NanoString will be considering under its tech access program, announced last week. For an undisclosed fee, NanoString will provide participants a standard DSP analysis package, looking at 20 protein targets in 24 regions of interest in 20 samples.
"One thing we hope to get out of our tech access program is to see how [the partners] would like to visualize this data," Gray said. "They'll send us 20 slides, we'll show them a few ways to visualize data and see what works for them," he said. That process will inform a data visualization software package that NanoString plans to develop alongside the instrument, which is still in the early phases of development.
So far, NanoString has signed up three pharmaceutical firms, but declined to disclose which ones. "The biopharma companies are the most eager to get their hands on this," Gray said. "They're the ones with too little tissue and too much information they want to get out of it."
Under the early access program, NanoString will process samples on behalf of its customers to get more experience with DSP. In return, it will receive feedback on what data the partners value most and which features they'll want to see on the final instrument.
While NanoString is open to proposals, which can be submitted via its website, Gray said the firm would be choosy at the beginning. "We only have so much capacity. We plan to start with a handful of select customers and then we'll see what level of demand is," he said.
Whenever NanoString launches the DSP, it will likely be available as an add-on option for existing customers or in a package deal with nCounter analyzers.
Spatial profiling could be interesting to any number of researchers, especially in the immunoncology space. MD Anderson's Mills said that while his labs are evaluating a number of other technologies for similar purposes, including cyclic immunofluorescence and mass cytometry, he expects NanoString's DSP to offer certain abilities that will complement the other techniques.
"The ability to do spatially oriented, single-cell research with RNA, DNA, and protein analysis offers an incredible research platform," he said. Quantitative measurement and sample conservation are two important qualities that could enable a new level of investigation into breast cancer markers for several aspects of disease.
Initial data from breast cancer models suggests that measuring the level of functioning HER2 protein, instead of total protein, could refine the ability to determine benefit from targeted therapies. A multiplex assay could provide information on post-translational modifications and transcriptional programs related to HER2 function. "The combination of measuring HER2 levels accurately and at the same time measure copy number, RNA levels, and functional activation in a single assay on a small amount of tissue, in a quantitative rather than qualitative manner, has the potential to be a major step forward," Mills said.
The multiplex aspect could also normalize protein assays, Rimm, the Yale pathologist, said. "Historically IHC has been essentially impossible to normalize," unlike reverse transcriptase PCR assays.
"Because there are so many slots for multiplexing, you can build in normalization factors like housekeeping protein controls, which could be run alongside query targets to assess tissue quality," he said. "IHC to date has never been able to do that."