OpGen, a startup in Madison, Wis., is betting that its newly launched technology for detecting genetic insertions and deletions can have pharmacogenomics applications from drug metabolism to efficacy predictions and anti-bioterrorism.
“We think that SNP technology is very, very powerful, and some very elegant technologies have been developed to use it,” said Colin Dykes, chief scientific officer at OpGen. “But it’s really looking at part of the story — and in fact it may not even be the most important part. The most important parts may be the insertions and the deletions.”
Dykes was talking about OpGen’s Optical Mapping technology, which was launched last week. Though the platform has only been used to identify and compare microbial isolates, measure haplotypes, and assess the accuracy of gene-sequence assembly, Dykes reckons it has a good shot at becoming a common tool in a variety of drug-discovery laboratories.
To be sure, insertions and deletions — or indels, as they are commonly known — have been chronicled in a number of recent research papers, some of which claim that they may be more important than SNPs as tools for finding variations in evolution and variation. One paper in the Proceedings of the National Academy of Sciences shows that if you take into consideration the indels, humans and chimps are less than 87 percent identical — not the more than 98 percent that SNPs indicate [PNAS, 2003; 100:13].
Another article, this one from a team from Cold Spring Harbor Laboratory, has linked certain genetic deletions and amplifications to breast cancer [see story, page 5].
But will indel-identifying technology like Optical Mapping find a home in pharmaceutical companies, which are inherently averse to investing in new — and, in this case, unproven — technologies?
According OpGen, Optical Mapping aims at creating ordered restriction maps from ensembles of single molecules. The technology enables researchers to score markers across entire genomes in a single manipulation. The technology relies on restriction endonucleases to capture the genome as a collection of single DNA molecules on a derivatized surface. Researchers are then required to add the restriction enzyme in the nuclease, which cleaves at each occurrence of the six-base recognition site. After that, it’s a matter of scoring whether cuts took place or not, Dykes said.
Dykes said it would take 24 hours for researchers to capture the genome and perform the assay, though it is the time to perform the data-analysis that is “a little more difficult to predict.” In fact, people close to the company said OpGen is wrapping up negotiations with Dell and IBM to help it build a supercomputer cluster to aid in this phase of its research. It was not immediately clear whether this collaboration might be with one or both of the computer giants.
Indel-based technology has several advantages over traditional SNP-genotyping platforms, according to Dykes. First, the tool can be applied to a genome without prior sequence information. A second advantage is its ability to haplotype. “Every time we score markers along a molecule, it’s a haplotype. And these haplotypes are very long — they’re run for megabases. And there’s no other technology I’m aware of that can haplotype anywhere near that distance.”
Another advantage is that researchers using Optical Mapping can find things that they weren’t expecting, Dykes said. “With SNP-type assays, you’re basically asking, ‘Is this SNP present in one form or another?’ You can’t find novelty, because you can only screen for what you know about.” OpGen’s technology enables investigators to identify indels or rearrangements that were unsuspected, he said.
To be sure, Dykes said that Optical Mapping was not designed to compete with SNP-genotyping instruments, but rather to be a “strong complement” between the two. For example, researchers can employ indel-identification tools to perform broad genome-wide scans, and use SNP genotyping to interrogate small regions of the genome. “Because you’re not going to get a very high-resolution analysis of a gene-size region using Optical Mapping, [it’s good] for looking at the whole genome.”
Specifically, he said, OpGen’s technology can complement the growing number of drug-metabolism diagnostics on the market today, such as Roche Diagnostics’ AmpliChip or Amersham’s cytochrome P450 chip.
OpGen has so far used the technology to study variations in microbes, though it now hopes to move into plant genomics and, eventually, pharmacogenomics. “The challenge is simply one of scale,” he said. “There are no conceptual barriers.”
To that end, OpGen has been negotiating with a number of clinical centers to use the platform to analyze colorectal tumors, correlating tumor genome rearrangements with a patient’s clinical outcome. “With those data behind us, we’ll be in a stronger position to go to pharma companies with a proof of principle,” said Dykes. “We expect to see different groups of patients will respond to the two or three major colorectal cancer therapies that are available … and we hope to find markers to predict that.”
Dykes said OpGen plans to wrap up in vitro experiments before the end of the year, and move on to these human studies in early 2004.
OpGen currently has a handful of collaborations, but Dykes declined to identify them. However, a person close to the company said current early-stage R&D partners include Baylor College of Medicine, the US Centers for Disease Control and Prevention, and the Washington University Sequencing Center.
Andrew Cameron, a senior research associate at CalTech who co-wrote a paper showing that indels contribute to a majority of divergence between closely related DNA samples, said that for all its promise, the use of indels as research tools is still in the early stages]. “There are not very often good haplotype comparisons in lots of things outside of humans,” he told SNPtech Reporter. “If defects in genes are the basis for making drugs, then yes, [indels can be useful in drug discovery]. But it’s not clear to me how that would work, exactly, since it is the secondary effect that one would have to treat, and whether if it’s a single SNP or a small deletion or insertion. One can therefore engineer a drug to fix that difference.”