NEW YORK (GenomeWeb) – Harnessing a patented technology to map open chromatin regions across the genome, Stanford University spinout Epinomics is working on clinical applications of epigenomic signatures, including a diagnostic test to predict response to immunotherapy.
The company was founded to exploit ATAC-seq, which stands for "assay for transposase accessible chromatin by sequencing," for use in biopharmaceutical research and in clinical applications. The epigenomic profiling technology was developed by researchers at Stanford University School of Medicine and published in Nature Methods in 2013.
Earlier this week, the Stanford team, led by Howard Chang, a professor of dermatology at Stanford, and William Greenleaf, an assistant professor of genetics at Stanford, both scientific co-founders of Epinomics, published a follow-up paper in Nature Methods in which they describe an improved ATAC-seq protocol, called Omni-ATAC, that reduces the background noise and allows the method to be applied to sections of archival frozen tissue.
For ATAC-seq, a transposase carrying sequencing adapters is added to nuclei prepared from cells, where it integrates the adapters preferentially into regions of open chromatin. Those regions can then be PCR-amplified and analyzed by next-generation sequencing.
The method's main advantage over other approaches for profiling accessible chromatin regions, such as DNase-seq or FAIRE-seq, is that it only involves two simple steps, and that it requires hundreds to thousands of cells instead of millions.
As a result, ATAC-seq could be used for profiling the epigenomes of individuals from a blood draw, on a timescale that is compatible with clinical testing, whereas "you would have to exsanguinate someone to assay their blood with the old techniques," Greenleaf explained.
"The ATAC-seq method is very simple, it can be executed by a first year grad student," he said, and it can be applied to primary tissue and small clinical samples. As such, "it allows the translation into the clinical setting."
In their original paper, the researchers demonstrated that they could use ATAC-seq to generate sequencing libraries from clinical blood samples in just under five hours, and Chang said they have shown proof of concept that the entire assay, including sequencing, can be performed in 10.5 hours.
According to Greenleaf, the epigenetic changes that ATAC-seq measures are both highly dynamic and large in number. A cell usually expresses on the order of 5,000 genes, "but we see hundreds of thousands of different open chromatin sites that are extremely dynamic," he explained. "So there is much more cell-type specific information in the open chromatin landscape, many more potential biomarker observations."
In their original paper, the authors cautioned that deeper data analysis "will take additional time" and said they anticipated "that bioinformatic analyses — not the molecular biology or sequencing — will become the bottleneck of epigenomic studies in the future."
Stanford holds intellectual property on the ATAC-seq method, which Epinomics licensed exclusively.
In this week's publication, the researchers described further improvements to the approach that result in better data quality but do not change the simplicity of the method. For example, they employed different types of detergents, included another wash step, and used a new buffer to access different cell types, remove mitochondria, enhance the complexity of the sequencing library, and increase the signal-to-noise ratio.
Those changes "extend the capabilities for ATAC-seq, primarily for more clinical questions," Chang said. "This method now works even better for very small numbers of cells, and it works very well for frozen archival tissue, very small pieces, even sections of tissues, which is what a pathologist would generate when they take a biopsy to diagnose disease," he said. "That makes the entire workflow very compatible with the translational path."
In their paper, the researchers tested the new Omni-ATAC protocol on frozen tissues from human thyroid cancer and brain, and on a variety of frozen mouse tissues. They also applied it to 50-micrometer sections of frozen human brain tissue and compared the results with those from adjacent 5-micrometer pathology sections, finding that the open chromatin profiles correlated well with histopathological staining. For example, regions with a high content of neurons as measured by immunohistochemistry also showed high chromatin accessibility near neuron-specific genes.
"Thus, the Omni-ATAC protocol enables the application of epigenomics to clinically relevant specimens, paving the way for assays and diagnostics that leverage the highly informative and cell-type specific signals of open chromatin landscape," the authors wrote.
"Some of the work from our groups in the last couple of years has shown that, for example, in cancer samples, you can actually determine the prognosis of the patient based on the open chromatin information," Chang said.
This summer, for example, Chang's group published a paper in Cancer Cell in which they used ATAC-seq to profile chromatin accessibility in cutaneous T cell lymphoma and found that the clinical response to histone deacetylase inhibitors was strongly associated with a gain in chromatin accessibility.
Epinomics wants to harness ATAC-seq for clinical applications, including diagnostics. According to Fergus Chan, the company's CEO and another co-founder — the fourth co-founder is Paul Giresi, a former postdoc in Chang's lab and a co-inventor of the ATAC-seq method — the firm is both offering ATAC-seq as a service to biopharmaceutical partners and is developing it as a clinical tool in conjunction with its data analytics platform.
"A core clinical application that we're looking at now is for cancer immunotherapy diagnostics," Chan said. The idea is to use ATAC-seq on blood samples to read out a variety of metrics related to the immune system. The company has already developed analytics modules that allow it to obtain a readout of various immune functions, he explained. That readout is comparable to existing assays used in immunology, for example to measure cellular composition or cytokine levels. "We have been focusing on development of this capability for use in the cancer immunotherapy space, in collaboration with the Parker Institute for Cancer Immunotherapy, to aid prediction of treatment response and adverse events," he said.
Earlier this year, Epinomics announced a collaboration with the Parker Institute that aims to use biomarkers discovered by Epinomics to improve outcomes and reduce adverse events in immunotherapy clinical trials. Chan said several clinical studies are currently ongoing, some of which are expected to be completed later this year.
Going forward, the company plans to exploit epigenomics to determine the health status of cells and tissues in the context of other diseases. "Having all this data is like building a map of the health status of a human," he explained. "And given that epigenomics is dynamic, we are able to use our technology almost like a GPS to guide personalized medicine treatment."
So far, Epigenomics, which counts Stanford's Mike Snyder and the Salk Institute's Joe Ecker among its scientific advisors, has raised an undisclosed amount of funding from a number of venture capital firms, including Felicis Ventures, SVAngel, Lightspeed Venture Partners, Founders Fund, and from Stanford's StartX business incubator.
The company currently has about a dozen employees and is located in Menlo Park, California.