NEW YORK (GenomeWeb) – Researchers at Albert Einstein College of Medicine have developed a method that eliminates critical sources of artifacts in single-cell sequencing sample preparation – and they have founded a company, called SingulOmics, to offer the method as a service.
The technique was published last month in Nature Methods. Jan Vijg, leader of the research team at Einstein, studies the way biological aging might be the result of accrual of DNA mutations in somatic cells, a hypothesis supported by the relationship between age and cancer or neurodegenerative diseases. Vijg looks at cells from young and old people to see if mutations accumulate, but bulk tissue is not a particularly useful to way to understand mutation patterns in single cells, he said, since they occur at such low abundance.
Unfortunately, most whole-genome amplification techniques were never intended for single-cell examination. In fact, Vijg noted that a 2015 review in Science on somatic mutations in cancer and normal cells highlighted that it is now easy to find mutations, but hard to know if they are artifacts. This may be because "the error rate of single-cell sequencing remains too high for accurate detection of de novo mutations, and only clonally expanded mutations can be reliably detected with current technologies," the authors wrote.
Vijg's own research had also been hampered by artifacts caused by multiple displacement amplification, or MDA, a non-PCR based DNA amplification technique that typically uses psi 29 DNA polymerase. MDA was designed for small amounts of DNA, Vijg said. "It is important to realize MDA was never designed for single cells, so most people [started to use it], as we originally did, but then you find out that either you have many artifacts or else it is very inefficient," he said.
Furthermore, the workarounds developed to apply MDA to single cells, which tend to involve elevating temperatures, can themselves cause artifacts, he said. This may be because lysis using high temperature causes cytosine deamination and DNA denaturation that leads to CG -> AT artifacts.
Vijg's group instead developed a method using alkaline denaturation with sodium hydroxide and low temperature. This was, of course, not very efficient, he said, so they also modified the workflow further to improve efficiency by, for example, switching around the point when primers are added.
Finally, the group developed a novel computational algorithm, dubbed the single-cell variant caller, or SCcaller, that verifies each mutation to see if it is "real." This approach examines heterozygous SNPs. "You can use those to recognize where you are — so suppose we see a somatic mutation, but now we look at a heterozygous SNP close by, and we can can determine the allelic bias in the amplification reaction," Vijg said. A suspect site is then flagged, considered an artifact, and left out from the final analysis.
To test the overall method, called single-cell multiple displacement amplification, or SCMDA, Vijg and his colleagues performed the amplification prep on primary fibroblasts followed by whole-genome sequencing to identify mutations, as well as on clonal cell populations, which are expected to have nearly identical mutations as the parent cell.
The researchers found about the same frequency and spectrum of mutations in clones using their method. To whit, a commercial elevated-temperature procedure led to an average of about 23,000 candidate somatic mutations per cell in the Nature Methods study, while the SCMDA method had about 930 mutations on average, similar to the 860 or so average mutations in unamplified clones.
They also validated the method on published single-cell data sets, comparing their SCcaller to MuTect, VarScan, and Monovar variant callers, which showed the superior performance of SCcaller. Interestingly, the Monovar caller was designed specifically for single-cell analyses, but the SCMDA method still showed higher specificity, or lower false discovery rate, in their hands, he said.
Vijg said this showed that the group had developed "the first reliable, accurate method for looking at mutations in single cells." He noted that Sunney Xie and his colleagues at Harvard and Peking University have also published a method recently, called linear amplification via transposon insertion (LIANTI), that was focused on other types of mutations, and that group's study also "shows very clearly that MDA, the way it is usually done, is full of errors." Beyond temperature-related C to T mutations, the Xie group also found other, less dramatic, sources of artifacts, such as oxidation.
Vijg and his colleagues launched SingulOmics a few months ago with support from Einstein. The company will offer the SCMDA prep as a service for a price of $300 per cell.
Einstein has also filed patent applications for the method, and is supporting the fledgling firm by renting it space, providing infrastructure, and helping with the commercialization process. "Our main customers are not going to be the people who work in labs and can easily do this; they will do it themselves. But, if you deal with an epidemiologist who is interested in mutations in blood cells, for example, they usually cannot do this themselves," Vijg said.
The firm is also considering developing clinically applicable tests using the technology. One will be a liquid biopsy technique evaluating mutation levels in circulating tumor cells. Somatic mutation frequency can also be used for radiation exposure testing, Vijg said. And, since the BRCA-1 gene is a DNA repair gene, Vijg reasons that changes in mutation rates in carriers of that breast cancer risk gene might be correlated to age of onset, and this could be another commercially viable test as well.
The company is now promoting its service at conferences, such as the American Association for Cancer Research annual meeting held earlier this month in Washington. So far, the firm is self-funded, but it is seeking investment to further develop the technology and potential assays, and it is also considering collaborations with industry partners, such as pharma companies interested in single-cell biology.