Skip to main content
Premium Trial:

Request an Annual Quote

MGH Team Develops DNA-Based Assay for Single Cell Proteomics; NanoString Secures Option to IP


Researchers at Massachusetts General Hospital have developed a proteomic assay using DNA-barcoded antibodies that can simultaneously measure in the range of 100 proteins at single-cell sensitivity.

Detailed in a study published this week in Science Translational Medicine, the method uses NanoString Technologies' nCounter Analysis System as its readout platform. NanoString has an exclusive option from MGH to license the underlying IP.

The technique's low sample requirements and high multiplexing capabilities make it particularly amenable to analysis of clinical samples, study author Cesar Castro, an oncologist and director of the cancer program at MGH's Center for Systems Biology, told ProteoMonitor. He said that the researchers hoped to develop it for applications including guiding personalized cancer therapies and designing clinical trials.

The MGH method uses antibodies linked to photocleavable DNA barcodes. These DNA-linked antibodies detect the proteins of interest, upon which the DNA barcode is released and can then be quantified by various means – NanoString's nCounter platform in the case of the STM paper.

The nCounter system quantifies nucleic acids using fluorescence hybridization, an approach that Castro said proved advantageous due to its multiplexing capacity, quick turnaround time, and the fact that it requires no amplification, a process that, he noted, can introduce error.

In the study, the researchers measured more than 90 cancer-relevant proteins in a human breast cancer cell line profiling both bulk samples and single cells. They followed this up with analyses of fine-needle aspirates taken from patients with lung adenocarcinoma, using these samples to investigate questions of tumor heterogeneity and finding that levels of correlation between individual cells and bulk tumor measurement varied widely – with the highest correlation being 0.79 and the lowest 0.43.

Cesar noted that while it is difficult to say what such single-cell heterogeneity data means from a clinical perspective, it represents an important challenge for researchers working to develop an understanding of cancer biology.

"It could be that you have 100 cells and 90 are similar and 10 are disparate, and it's those 10 that drive the clinical picture," he said.

The researchers also used the platform to study the response of protein pathways to various therapies, both in cell lines and in FNA's from four cancer patients before and after PI3 kinase inhibitor treatment. In both cases, they were able to observe protein changes across a number of pathways and, in the case of the cancer patients, distinguish responders from non-responders via their protein profiles.

The results suggest that the method could prove useful for monitoring patient response to molecularly targeted therapy, Castro said, noting that, from his perspective as an oncologist, better tools for targeting and monitoring such drugs are badly needed.

"What we're seeing is that with targeted therapies, the measurements on the back end [to determine their effectiveness] are very crude," he said. "We've learned the hard way that using single therapies you may get a nice [initial] response, but you will invariably get a relapse because of redundant pathways, compensatory pathways, that get activated."

Such pathways are often not captured in the handful of traditional immunohistochemistry markers used for assessing whether a patient will respond or is responding to a targeted therapy, Castro said.

"But by looking at a hundred [proteins], you can look at these other pathways to see if they get modulated as well, if there is crosstalk," he said. "So there can be more parity between the novelty of the [molecularly targeted] drugs and the novelty of the metrics used to assess the drug response."

Because of its extremely low sample requirements, the method could also prove useful for determining the best way to administer combination therapies to patients, Castro said. For instance, with the cells from an FNA, researchers could do several passes, exploring different combinations or different sequences for each individual patient.

"You could incubate the [biopsied cells] with the drugs and test different combinations to see [which] for that particular patient shows the most change in a pathway," he said. "Because with a lot of these targeted therapies, we're still in the dark ages with respect to [whether we] should give drug A before drug B; or drug B before drug A; or drug A on Monday, Wednesday, and Thursday; or drug B at night."

The MGH researchers are currently planning a follow-up study using the method to look at patient response to PI3 kinase inhibitors, Cesar said.

In many respects, their work and aims are similar to clinical research being done using reverse phase protein arrays, an approach that similarly offers high multiplexing capacity and low sample requirements. George Mason University researchers Emanuel Petricoin and Lance Liotta, for instance, recently demonstrated the use of RPPA for developing protein profiles for guiding personalized therapies in breast cancer as part of a study run by the breast cancer charity the Side-Out Foundation.

Cesar suggested that assay turnaround time could be one potential advantage of the MGH method compared to RPPA. While RPPA analyses like those used in the Side-Out trial can take on the order of weeks, Cesar said the MGH researchers were able to deliver results within 24 to 48 hours of receiving a sample.

In addition to protein measurements, the researchers also aim to perform DNA and RNA analyses simultaneously using the nCounter system, Cesar noted. He said that they planned in the near future to submit a paper for publication detailing a workflow for "achieving true DNA, RNA, and protein [data] on the same sample simultaneously."

Based on his work as an oncologist, though, Cesar said he expected protein data would ultimately prove most informative in guiding patient therapy.

"The field is looking more at proteins — whether certain pathways are up- or down-modulated," he said. "Certainly in my book the proteins are the better approximation of the tumor biology at the time of [patient] enrollment, which is really the most relevant thing in terms of therapeutics."

In an email to ProteoMonitor, NanoString Senior Vice President of R&D Joseph Beechem said that the STM study "demonstrates the flexibility of the nCounter platform in current and emerging areas of research and provides a compelling case for its advanced capabilities in protein targeting." He added that the company is "committed to expanding the range of applications enabled on the nCounter Analysis System."

The firm is the latest of several genomics-focused outfits to dip its toe into the proteomics space. In August, Fluidigm entered a co-marketing deal with Olink Bioscience to offer Olink's Proseek Multiplex protein assays on Fluidigm's BioMark HD real-time PCR platform.

And in 2012, Cepheid CEO John Bishop said during the company's investor day that it planned to pursue protein detection on its GeneXpert system. The company, he said, had "done all the feasibility work" including developing in-house assays for thrombin on the platform. Since Bishop's investor day remarks, however, Cepheid has declined to comment further on these plans.