Scientists from the University of California, San Francisco, and the California Institute for Quantitative Biosciences, or QB3, have published a study demonstrating how picoinjection can be used to precisely and efficiently add reagents to individual digital PCR reaction volumes without compromising assay quality or detection rates.
Although the picoinjection technique may not have immediate application to digital PCR experiments, per se, the study serves as an important proof of principle for the use of picoinjection in droplet-based biological assays.
In addition, the researchers have founded a startup company called Torrent Bio to further develop their method as a tool to enable single-cell PCR-based gene expression studies, Adam Abate, principal investigator on the study, told PCR Insider this week.
Abate helped develop the picoinjection technique while he was a student in the laboratory of David Weitz, a physics professor at Harvard University. Abate, Weitz, and others first described the method in a paper published in PNAS in 2010.
Weitz also co-founded nascent sequencing company GnuBio, which uses the picoinjection method as part of its droplet-based sequencing technology (see PCR Insider sister publication In Sequence, 10/4/2011).
"The picoinjector is basically a microfluidic pipettor," Abate told PCR Insider this week. "Essentially the droplets are test tubes, and the microfluidic device is the automation that allows you to manipulate the test tubes, put reagents in them, sort them — all stuff that has to be done with microfluidics, because these droplets are tiny, on the order of 10 microns to 50 microns in diameter. You could never manipulate them manually, and especially if you want to [process] them in large numbers, you need some kind of very high-throughput technique."
Abate said that prior to the development of picoinjection, the most reliable way to add reagents to droplets was a method called droplet merger, or electrocoalescence, which essentially involves controlling multiple droplets so that certain ones "merge" together inside a device.
"It's an effective way to add reagents, but one of the downsides is that … it's sometimes technically challenging to get it to work well — particularly when you have what are called re-injected emulsions, drops that you formed in a previous process and then want to add another reagent to," he said. "It's very hard to flow drops at a very periodic and controlled rate through a device. It just leads to more inefficiency."
Enter picoinjection, which Abate, Weitz, and others had previously shown to be an effective method to add reagents to drops in a sequential fashion for a number of applications. However, the method's compatibility with biological reactions had never been thoroughly demonstrated — until Abate and colleagues at UCSF and QB3 published their study earlier this month in PLoS One.
"A classic reaction that people have been doing in picodroplets for a while, and has recently become even more popular, is digital PCR," Abate said. "We thought this was a nice reaction to test our picoinjector. Basically we just … rigorously tested how adding reagents to the drops with picoinjection affects the digital PCR."
In droplet-based digital PCR — which is enabled by instrumentation platforms offered by companies like Bio-Rad and RainDance Technologies — a biological sample and all necessary PCR reagents are added at the beginning of a reaction. Then, the mixture is divided into a large number of uniform picoliter-scale droplets, subjected to thermal cycling conditions, and then analyzed to see which droplets contain an amplification product of target nucleic acid.
"Our variation was to separate the reagents from the digital PCR so you are picoinjecting some of them," Abate said. "So your reactions won't occur unless you picoinject the necessary reagents."
In their study, Abate and colleagues created their own digital PCR setup composed of a microfluidic T-junction and carrier oil, creating monodisperse 30-µm (about 14-picoliter) drops, then compared TaqMan-based digital PCR with and without picoinjection of various reagents to droplets.
For instance, they compared the detection of two cancer-relevant human transcripts, EpCAM and CD44, in both picoinjected and non-picoinjected drops. In one case, they separated RT-PCR reagents into two solutions added at different times. The first contained total RNA, RT-PCR buffer, primers, probes, and DNA polymerase dispersed into 30-µm droplets. These droplets were not capable of digital PCR since they lacked reverse transcriptase. For the second solution, they used picoinjection to add reverse transcriptase to the droplets and were able to successfully amplify and detect the target transcripts.
The group also conducted experiments demonstrating that picoinjection of reaction components at different times afforded transcription detection rates equivalent to that of standard digital RT-PCR; and that picoinjection did not cause any cross-contamination between different reaction volumes.
In short, the researchers found that using picoinjection to conduct digital PCR was essentially as effective as standard digital PCR with all reagents mixed together at the outset. Abate told PCR Insider that the research provided important proof of principle that picoinjection works for biological reactions, but that its actual applicability to digital PCR specifically might not be the most important takeaway.
"Up until now, most people who have done digital PCR have added everything up front, and that's true for all the commercial instruments and, I think, all of the papers that have been written on digital PCR," Abate said.
"That's fine for digital PCR," he added. "Basically the picoinjector allows you to change the conditions in the drops … [but] for digital PCR it turns out that's not really important, because you don't need to change the conditions. You add everything up front and then you're done. But there are other applications where that's not going to be the case."
One of those applications is single-cell PCR, a research area that has recently garnered attention from researchers and tool vendors alike because of its potential to elucidate heterogeneity in large populations of cells. To wit, early digital PCR innovator Fluidigm has recently made single-cell genomics the main focus of its business, introducing an automated single-cell sample prep platform to complement its chamber-based (i.e., non-droplet) digital PCR instrument systems.
"The chamber-based people … have been the most successful in getting down to single-cell PCR," Abate said. "There are a couple of papers using droplets, but they haven't made much of an impact for many reasons. Probably the biggest [reason] is the throughput has been limited … because of problems such as: How do you digest cells, break them open, and access their genomes and transcriptomes without destroying your PCR reagents? If you add everything up front, you'll have to add detergents, maybe proteases, heat, and you can imagine this can wreak havoc on your PCR enzyme."
Abate said that this application was a "deeper motivation" for his group's recent paper, and that the team is currently preparing another manuscript exploring single-cell PCR for peer-reviewed publication.
In addition, Abate and others at QB3 recently founded Torrent Bio to commercialize single-cell PCR and other downstream applications using picoinjection and other microfluidic techniques. Abate said he could not disclose many more details about the company since he is not an official representative, and because the firm is in the very nascent stages. He did note that those involved with the startup, along with UCSF's tech-transfer office, have filed a provisional patent on certain technologies. The picoinjection technology, he added, is owned by Harvard University where it was first developed.
"We've done a lot of scientific work," Abate said. "The company formally exists, it's an incorporated entity. Stay tuned, because we have a much more interesting paper that has been submitted [demonstrating] massively parallel single-cell PCR."