NEW YORK (GenomeWeb) – Scientists at MIT and the Broad Institute have developed an inexpensive "virtual microfluidics" method to compartmentalize samples for amplification reactions that could be applied to single-cell sequencing and digital PCR.
The technique, which was described recently in Nature Methods, uses a thin layer of poly ethylene glycol (PEG) hydrogel to immobilize single cells or molecules while still enabling some reagent flow. It can be used to amplify DNA from single cells for sequencing in place of expensive instruments or proprietary consumables and reagents, Paul Blainey, a member of the Broad Institute and MIT's Department of Biological Engineering and the corresponding author on the article, told GenomeWeb.
Interestingly, the number of chimeric sequences generated when using the method along with multiple displacement amplification (MDA) is dramatically reduced. This is likely due to immobilization preventing cross-hybridization and reducing chimera formation, Blainey said, which could in turn reduce genome misassemblies and false-positive signals in horizontal gene transfer studies.
The method was inspired by a 2003 PNAS study from George Church's group at Harvard Medical School. The researchers happened to be reviewing some of Church's early "polony" cloning efforts — work that ultimately led to massively parallel clonal amplification technologies and next-generation sequencing.
"We were looking back at some of that old work and saw that it was all PCR in polyacrylamide gels, and we saw that there was an opportunity to update the gels to more highly-engineered PEG gels, and the chemistry from PCR to MDA, and shift the application from molecular cloning to single-cell sequencing," Blainey said.
The gels are "super low-tech" slabs, similar to those used for DNA separations, he said. There are no real compartments in the gel, Blainey explained, but rather "virtual compartments, dynamically forming whereever the molecules are."
The PEG gels are commercially available as soluble pre-polymers with complementary reactive groups, such that a simple covalent chemical reaction solidifies the gel over a period of minutes at room temperature. The method is also quite mild and does not require radical polymerization chemistry like acrylamide gels do, thus minimizing damage to nucleic acids.
The gels are also available with a variety of terminal functional groups. The researchers selected one that was stable with heat under neutral pH, allowing them to use heat for cell lysis and to do PCR in situ in the gel. However, other forms of the PEG gel could provide reversibility, tranforming back to a liquid state under certain conditions, Blainey noted.
The gels can also be functionalized, with molecules of interest directly linked to the gel matrix. "I think that is a big opportunity as well," Blainey said. For the Nature Methods study, the group was looking at DNA with a high molecular weight, so it was immobilized well in the gel. "If people are thinking about applications with RNA and cDNA — smaller molecules that may not be as well immobilized — they might want to graft the cDNA primer onto the gel, so that the cDNA ends up anchored," he said.
Because the hydrogel method uses standard reagents and the gels are commercially available, Blainey said, the cost is quite low. "In terms of single-cell workflow, if somebody wants to compare to other platforms out there, there is no special equipment, and only a small quantity of standard reagents per cell, similar to microfluidic platforms," he said.
Specifically, the cost would be "a few dollars in reagents and supplies to run the hydrogel reactions, plus the cost of library construction and sequencing," Blainey said.
The study also suggested that virtual microfluidics might serve as a simple platform for digital quantification assays, such as digital PCR.
At this point in the initial conceptual demonstration, the implementation of the method is still a little clunky, said Blainey, but if another academic lab wished to try it out, "it is as low-cost as it gets — they probably already have a lot of the reagents and just need to put somebody on the project to test it out."
Blainey suggested that the hydrogel approach could be industrialized, using an instrument with high-throughput mechanical picking, or barcoding, so that all the library construction and processing could be done on a large scale. The thin hydrogel also allows easy access for subsampling approaches like punches and pickers, as well as for imaging.
Although throughput was limited in the study by use of a 60 nanoliter punch volume, the researchers suggest using a thinner gel with more surface area, or a smaller punch, would yield more subsamples from a single hydrogel. Using imaging data to guide product retrieval could also improve the fraction of samples that contain single-cell whole genome amplification reaction product.
In the Nature Methods study, Blainey and his colleagues used in-gel digital MDA, or dMDA, to analyze a unique collection of saliva, fecal, and environmental microbiome samples from the Fiji Community Microbiome Project. That population has not been exposed to a lot of the features of the Western lifestyle, Blainey said, and thus might reflect a more ancestral microbiome that could be more representative of humanity's natural state. Research published earlier this year in Nature described mobile gene elements in these microbiota and utilized hydrogel-based single-cell amplification.
Two of Blainey's co-authors on the study, Ilana Brito and Eric Alm, have also developed a method for high-throughput investigation of conserved genetic traits called emulsion, paired isolation, and concatenation PCR, or epic-PCR. Alm was also involved in work that tracked daily changes in the human microbiome over a period of two years.
The goal of the hydrogel study was ultimately "to make a complete technical description ... and encourage other people to try it out," Blainey said, noting that the method could be particularly useful for labs that are hesitant to invest in the existing commercial platforms. Now, Blainey's group is pursuing ways to further improve the data quality and applying the method to human cells, and there are also several other groups currently evaluating the approach, he said.