NEW YORK – Bio-Rad has commercially launched its single-cell chromatin profiling assay, which is based on transposase accessible chromatin by sequencing (ATAC-seq) technology, and has demonstrated its utility in a collaboration with researchers at the Broad Institute and elsewhere.
The launch follows Bio-Rad's announcement of the assay last year, when the company said that it was making it available to early-access users. The SureCell ATAC-Seq library prep kit runs on Bio-Rad's droplet-based ddSeq Single-Cell Isolator instrument and comes with an analysis toolkit.
The commercial availability of the assay coincides with a new paper published by Bio-Rad in Nature Biotechnology this week in collaboration with ATAC-seq pioneer Jason Buenrostro, a professor at Harvard University, and others. The paper describes some of the platform's technical specifications and shows off product refinements and avenues of research made possible by looking at open chromatin.
"Having peer-reviewed technology is of critical importance to the adoption and advancement of science," Buenrostro said. "In that way, this is a big milestone."
According to the paper, the Bio-Rad platform provides 100,000 nuclear reads per cell and a cell capture efficiency between 65 to 90 percent. The paper showed a transcription start site (TSS) enrichment value of 30.
Bio-Rad said its platform can analyze 5,000 cells per sample, or 20,000 cells per experiment, but Buenrostro noted that the system could be pushed even harder. "We developed a new approach to get more throughput that is not part of the standard product," he said. By using a barcoded transposase, for example, his team was able to get approximately 10 times more cells per experiment than the off-the-shelf assay.
"The paper by Buenrostro and colleagues highlights the rapid progress in scaling up single-cell epigenomics," Stanford University Professor Howard Chang said in an email. Chang's lab is collaborating with 10x Genomics on a study to highlight that firm's ATAC-seq assay. "The authors combine two powerful technologies: droplet-based microfluidics and combinatorial indexing. The data are very high quality and I think this would open the door for many applications.”
Carolyn Reifsnyder, director of global marketing at Bio-Rad's digital biology group, added that Bio-Rad has developed novel bead overloading and sequence deconvoluting technology that enabled the high end of the cell capture efficiency range, and has filed for patents on the biochemistry and bioinformatics of that. "We've got our dropletsworking a lot harder," she said.
The ATAC-seq method, developed by Buenrostro while he was a graduate student at Stanford University, together with Stanford professors Howard Chang and William Greenleaf, measures open chromatin using a Tn5 transposase enzyme that inserts sequencing adapters into accessible regions of the genome. The single-cell implementation of this technique uses PCR to amply DNA from cells with barcoded primers prior to next-generation sequencing and analysis.
Early on, Buenrostro used Fluidigm's C1 platform for single-cell ATAC-seq experiments. But recently, Bio-Rad and 10x Genomics have both developed platforms offering much higher throughput, from hundred to tens of thousands of cells per experiment. These firms have also developed end-to-end analysis pipelines producing biologically interpretable data and visualizations. The 10x Genomics offering has been available since October 2018, while Bio-Rad made its assay available as part of an early-access program last September.
And while Bio-Rad's showcase study came out first, 10x Genomics is not far behind. Chang said his lab collaborated with the firm on a similar a paper, which is currently in press. They posted it as a BioRxiv preprint April 18 — the same day that the Bio-Rad collaborators posted their preprint.
According to a 10x Genomics webinar, its ATAC-seq platform offers the ability to analyze 10,000 nuclei per sample and up to 80,000 cell nuclei per chip, with a capture efficiency of 65 percent. The company recommended sequencing coverage of 25,000 read-pairs per cell, but "more reads per cell is not necessarily better if the reads are duplicates and are therefore filtered out for downstream analysis," said Laura DeMare, associate product manager of epigenomics at 10x Genomics. She suggested that unique fragments per cell is a better metric, which can be used to assess library complexity. In the preprint, Chang's team reported a per-cell average of 23,000 unique fragments mapped to the nuclear genome, with approximately 40 percent of insertions at sites present in aggregated ATAC-seq profiles from all cells and TSS enrichment values up to 20. 10x Genomics later added that the preprint's supplemental data suggest up to 62 percent of insertion sites are present in aggregated ATAC-seq profiles from all cells and that TSS enrichment values can be as high as 30.
Both new platforms "give you an opportunity to do experiments that you couldn't do before," Buenrostro said. "Transcriptomics has held the spotlight for so long and chromatin assays had lagged, relatively speaking. But with this paper, we're seeing rapid acceleration and the single-cell research community will converge on those opportunities."
DeMare said 10x Genomics has data to back up the rise of ATAC-seq. "In the field of epigenetics, the ATAC-seq method has been the fastest growing as measured by the number of publications," she said.
Fergus Chan, CEO of Epinomics, which was acquired by 10x in 2018, added that "people not historically looking at epigenetics are starting to look at it." Chan and Chang cofounded Epinomics and spun it out of Stanford to commercialize ATAC-seq. Chang is also a paid advisor to 10x Genomics.
As Buenrostro put it, RNA sequencing and chromatin profiling are "two sides of a coin, both wholly distinct" and complementary.
Reifsnyder said the new assay is fully commercialized now, although she declined to provide pricing information or early sales numbers. She noted that the early-access program brought in both existing users of the ddSeq instrument as well as new ones.
She added that she thought the market was waiting for both a peer-reviewed article in a major journal and a "plug-and-play product with software that takes you all the way through."
"We saw it in the adoption of digital PCR, [those elements were] a huge part of the adoption of that technology, especially in the early days," she said.
As for the market for its new assay, Reifsnyder said Bio-Rad is "excited to bring a tool like this and see what people can actually solve" with it. She suggested that oncology and immuno-oncology were two prime areas of research, with ATAC-seq available as an "orthogonal validation" of gene expression studies, and perhaps even as a substitute for classifying cells.
Buenrostro said the extra throughput gained with the barcoded transposases allowed the study to describe more than 500,000 single cells, making it the largest published ATAC-seq study to date. This enabled his team to create an atlas of brain cell diversity in mice and find new clusters of cells. In another experiment, they were able to do "tissue-scale perturbations" of bone marrow cells. "We put them together, added a stimulant, and watched the whole population react at the same time," he said.
"It's one thing to say you have high throughput, but it's another to say you can investigate cellular dynamics and how [cells] interact with each other."
That kind of study is just one of the opportunities that didn't exist until now.
"When you have something that just works and gives you good data quality, so you can trust what you see, it opens up new territories like tumor samples where you have no idea of what you expect to find," he said.