NEW YORK (GenomeWeb) – Researchers from Uppsala University in Sweden and the University of California, Los Angeles are developing mobile phone microscopy technology to enable in situ mutation analysis and targeted DNA sequencing.
In a study published today in Nature Communications, the team demonstrated proof of concept for the technology by imaging single molecules that had been amplified via rolling circle amplification (RCA) on glass sides using a mobile phone camera.
Malte Kühnemund, a PhD student at Uppsala University and co-lead author on the study with UCLA's Qingshan Wei, said in an interview that the technology could have applications in pathology as well as field-based diagnostic testing for infectious diseases.
The group demonstrated both a targeted sequencing application that could be imaged, as well as in situ mutation analysis of a tumor sample. This "was a nice proof of concept," Kun Zhang, a professor of bioengineering at the University of California, San Diego who was not involved with the study but has worked on developing in situ sequencing technology, said in an interview. "The idea of implementing these sophisticated molecular assays in situ and using a handheld device is novel and impressive," he added.
The researchers first demonstrated that they could design an optomechanical attachment for the mobile phone that contains two laser diodes for fluorescence imaging and a LED for bright-field transmission imaging.
Next, they developed two different methods for molecular analysis — a targeted DNA sequencing method that uses selector probes and sequencing by ligation, and in situ mutation analysis to look at specific KRAS mutations in RNA fragments.
For the targeted DNA sequencing approach, the researchers first extracted DNA from a tumor sample and digested it with a restriction enzyme that created 190-base pair fragments. Next, they used biotinylated selector probes —originally commercialized by Halo Genomics, which Agilent Technologies now owns — to target and circularize DNA fragments containing the KRAS gene.
They then used RCA to amplify the circularized fragments, and sequencing by ligation to selectively sequence just codon 12 of the KRAS gene. In addition, they incorporated fluorescently labeled oligos to bind to either the RCA product with the mutation or the wild-type allele, and used the mobile phone microscope to image the reaction.
The researchers first validated this setup on synthetic KRAS fragments, generating either wild-type or mutant sequences. After verifying that they could distinguish between the two, they attempted to detect single mutants in a background of 1,000 wild-type fragments, successfully detecting nine mutants and 1,552 wild-type alleles.
The authors noted that by sequencing to a great enough depth, they could potentially reach sensitivity levels comparable to PCR-based KRAS tests that have been approved by the US Food and Drug Administration.
Next, the team validated the technology on genomic DNA from a cell line known to be heterogeneous for the codon 12 KRAS mutation, finding 52 percent and 48 percent mutant and wild-type alleles, respectively, representing the expected ratio.
Finally, they analyzed KRAS from DNA extracted from three different colon cancer biopsies, finding that all three were wild type, which was also confirmed via PCR.
Kühnemund said that an eventual assay would likely need to cover more than just codon 12 of the KRAS gene, at least codons 12 and 13. In addition, he said, the researchers are interested in integrating a microfluidics design to cut down on the manual steps on the front end.
Zhang agreed that automation on the front end would be needed. "They've simplified the instrumentation so this can be done on a handheld device, but right now the molecular biology part is still complicated and requires many manual steps," Zhang said.
To demonstrate in situ mutation analysis in a cancer cell line, the researchers used padlock probes that ligated specifically to mutant KRAS cDNA and RCA to amplify the mutant-bound probes. They thenspiked cells with the KRAS codon 12 mutation into normal cells at ratios of 1 in 100 and 1 in 1,000, and were able to detect the mutated cells even in the lower concentrations.
Finally, the team used its setup to successfully detect mutant KRAS directly in colon tumor tissue sections.
Kühnemund said that there are several potential commercialization options for the technology. For instance, the group is interested in commercializing in situ RNA assays for gene expression, but not necessarily for oncology. For DNA sequencing, he said the researchers would be interested in partnering with a sequencing company.
Another area the team is interested in researching further, although not for commercial purposes, is something called "touch imprints," Kühnemund said, which would enable point-of-care analysis of tumor samples. After a tumor is surgically removed, a section is lightly touched to a glass slide, leaving behind a layer of cells, Kühnemund explained. The goal would be to directly analyze that layer of cells, which would also preserve tumor morphology, he said.
Zhang thought this idea was interesting, but said that it would require collaboration with pathologists to make sure it is something they would be interested in. In addition, he said, it might be important to modify the protocol to work for formalin-fixed paraffin-embedded samples, since those are routinely used in clinical pathology labs. "I think it's possible, but would need additional optimization," he said.
Zhang has also worked on developing in situ sequencing technology in collaboration with George Church's group at Harvard. "It's a rapidly evolving field," he said. Much of the current work involves a lot of high-end technical instrumentation, including microfluidics, high-end imaging, and a high-powered computer that can process many samples, so one aspect of this study that stood out was the researchers' focus on simplifying the instrumentation. "Even though it wouldn't let you identify lots of novel events, the technology can identify those that are well-documented, and it is simple for everyone to use," he said.