NEW YORK (GenomeWeb) – Startup Genetics Research is developing a sample prep approach that's turning the traditional concept of finding a needle in a haystack on its head.
"We protect the needle, and then burn the haystack down," said Anthony Shuber, the company's president and chief operating officer.
Essentially, the company's negative enrichment technology uses CRISPR nucleases to protect a region of interest on a strand of DNA and then degrades the surrounding material, until the target fragments are all that are left to be sequenced and analyzed.
The technology is Shuber's brainchild and came from his experience as a scientific cofounder and the former chief technology officer of colorectal cancer diagnostic firm Exact Sciences. His approach to sample prep is very simple: "Garbage in, garbage out." In other words, if the sample isn't prepared correctly, the data that comes from it probably won't be worth much.
When he cofounded Exact Sciences, Shuber found that no one had yet developed an at-home diagnostic for colorectal cancer because it was too hard to extract human DNA from stool samples. Using regular DNA enrichment methods at the time resulted in predominantly bacterial DNA sequences, which aren't helpful in determining whether a person has or is developing colorectal cancer. To develop Exact Sciences' Cologuard diagnostic — which identifies cancer DNA in human stool samples — Shuber took a different approach, using sequence-specific oligonucleotides to enrich for human sequences he was interested in amplifying.
Shuber's experience with Cologuard showed him how powerful the right technology could be for getting specific data out of a sample as efficiently as possible. And when CRISPR came along, he saw how that technology could be used in this niche.
"We are applying this to a space that no one really thinks is important," he said. "The sample prep is paramount to developing good, viable clinical information."
Lighting the match
Genetics Research's negative enrichment technology does exactly what its name suggests — rather than enriching a region of DNA by targeting it with RNA baits and trying to pull it out of a larger sample, it simply locks on to the region of interest and gets rid of everything surrounding it.
"[When we began,] we were most interested in long DNA enrichment. And using baits to get long DNA fragments is really difficult," Shuber said. "So I said, 'Let's take the Cas9-guide RNA complexes, we will target specific sequences at various distances apart from each other, and we'll let the Cas9 do its job and cut those locations.' Once we protect the ends of the region that contain the fragments, we intentionally throw exonuclease into the reaction and it chews up all the background."
Using this first version of the technology, Shuber and his team demonstrated that they could target a single copy of sequence within the human genome with two Cas9-guide RNA ribonucleoprotein (RNP) complexes, expose the genome to an exonuclease, and make a library out of the resulting sample, whether it was 10 kilobases, 12 kilobases, or 30 kilobases long. Further, when they sequenced one of the negatively enriched libraries on both an Illumina HiSeq 2500 and an Oxford Nanopore MinIon, they only obtained sequence data for the targeted 10-kilobase region.
In a subsequent version of the technology, additional endonucleases were attached to the Cas9 RNP in order to target even longer fragments of DNA, Shuber noted, and a third generation of the method went in the opposite direction, targeting much smaller fragments.
"There's a lot of work being done on looking for signatures or biomarkers in FFPE samples, and those fragments are all small — we're talking about 180 to 200 base pair fragments," he said. "So we've developed a number of procedures to also do sequence-specific enrichment for those types of targets or those samples."
Positives of the negative approach
The method's ability to target DNA of various lengths isn't its only advantage over the traditional bait-type sample prep, Shuber added. The incubation step associated with positive enrichment, for example, can take several hours, whereas Cas9 works far more quickly.
"We saw binding efficiencies in one minute [with Cas9], equivalent to binding efficiencies in one hour [with oligonucleotides]," Shuber said. "You can imagine, when you start to look at a multiplex sequence-specific enrichment technique that can go for a minute versus overnight, that would be pretty beneficial."
Another benefit is efficiency. Enriching multiple targets with baits can be time consuming as the baits can sometimes interfere with each other. However, several CRISPR RNPs can be combined to target several DNA fragments at once and don't interfere with each other in the same way that the baits do, he said.
"Any one RNP has the same binding efficiency, whether there are very few molecules [or] up to billions. We did a titration of 1,000-molecule targets up to 400 million and it had the same binding efficiencies regardless of the number of target molecules," Shuber said. "And those binding efficiencies are not like hybridization efficiency. They have up to 70 percent to 90 percent binding efficiency regardless of the number of targets."
This also allows the use of multiple Cas enzymes at once, or even the use of dead Cas9, to create different types of sample libraries.
The technology's greatest advantage, Shuber noted, is its versatility, allowing it to be used in a large number of applications. "If it's a researcher that is interested in querying a certain number of sequences, it can be applicable to that. If it's a clinical diagnostic laboratory that wants to query a large number of regions or biomarkers to get clinical information, it can be applied to that as well," he said.
Perhaps one of the most obvious applications is in preparing samples for liquid biopsy diagnostics. "If you're developing clinical assays, you have to make sure that the whole population is informative, whether it's positive or negative," Shuber said. "I think we're seeing the same thing in liquid biopsy [that we saw with finding human DNA in stool samples]. You'll notice that even though people undergo different liquid biopsy-associated analyses, not all of them have [sufficient]DNA in their circulation. We believe that our ability to capture DNA straight out of plasma may very well lead to a higher yield [of circulating tumor DNA], and therefore a higher clinical informativeness for the population."
Genetics Research has also developed a negative enrichment application for hyper-methylation analysis that obviates the need for treating DNA with bisulfite, Shuber noted.
"There are methods that can target methylated sites using monoclonal antibodies. But one can't do sequence-specific targeting of those regions," he said. "What we normally do is get the DNA, bisulfite-treat it, amplify it, and we look for the sequence changes based on the bisulfite treatment. But it would be nice to be able to analyze DNA in its native state as opposed to its amplified state. And that is one particular application we're pretty excited about — applying this method of enrichment to then be able to do analysis of methylated sequences without having to go through the bisulfite treatment."
The firm has also developed a diagnostic based on negative enrichment of circulating tumor DNA for recurrence monitoring in cancer patients, specifically by looking for passenger mutations. "If you're looking for mutations in liquid biopsy, if someone's negative, you don't know if there was any ctDNA there or not," Shuber said. "But if one uses it in conjunction with a targeted analysis of a biomarker that is specific to the tumor — for example, the presence or absence of passenger mutations — that would allow confidence that a negative is a true negative."
Genetic Research has received a patent for the underlying technology and is amassing a patent portfolio for different applications of the approach. Shuber said the firm is in talks with other companies that have expressed an interest in collaborating with it on further developing the technology, or licensing it.
Chairman and CEO Thomas Shields is currently the principal investor, but Shuber is also speaking with other potential investors. Though he couldn't reveal who he's talking to or how much money he's looking to raise, he said he's targeting strategic partners who can help the Wakefield, Massachusetts-based firm develop negative enrichment technology and its applications.
But if he's looking for partnerships with other firms, he may have some luck with instrument developers. Oxford Nanopore, for example, announced in May that it is is developing a method for targeted sequencing using deactivated Cas9 and a guide RNA to pull out regions of interest and transport them close to the pore. The company had planned to make this available as a kit "later in the year," according to Chief Technology Officer Clive Brown. And Pacific Biosciences has been working on its own CRISPR-based approach for target enrichment, using CRISPR-Cas9 in combination with its Sequel instrument.