Skip to main content
Premium Trial:

Request an Annual Quote

RainDance Debuts PCR-Based Sequence Enrichment Platform


RainDance Technologies launched the first application for its microdroplet-based technology this month, designed to enable customers to amplify many genomic loci in parallel by polymerase chain reaction for targeted resequencing.

The company claims that compared to other targeted sequence-enrichment methods, its approach leads to more uniform coverage, or less enrichment bias. This would allow researchers to obtain the same results with less sequencing and therefore lower sequencing costs, according to RainDance.

During a workshop at the Advances in Genome Biology and Technology conference in Marco Island, Fla., earlier this month, the company showed in-house data on amplifying up to 4,000 target regions in picoliter-sized PCRs and sequencing the products on the 454 Genome Sequencer and the Illumina Genome Analyzer.

A researcher from the Genome Center at Washington University — one of the company's early-access customers — talked about the center's experience with an early version of the instrument, and about modifications it has made to sample-prep and clean-up protocols.

RainDance's sequence-enrichment technology sets up millions of PCRs in picoliter-sized droplets — "the world's smallest test tube" — and combines them in a single vial for PCR amplification.

Using PCR has a number of advantages, being a "very robust, very simple procedure" that has been optimized over time, said Jeff Olson, who manages RainDance's nucleic acids applications group, during the company's workshop. Olson also claimed that PCR covers regions with repetitive elements better than enrichment methods that rely on hybridization capture probes.

The main drawback of PCR has been that it is difficult to scale, he said, but RainDance promises to change that. "Instead of setting up hundreds of thousands of wells on different [microtiter] plates, you just mix together droplets of different primer pairs in a single tube," Olson said.

Customers wanting to use the technology need to purchase the company's RDT 1000 instrument, which can produce 10 million microdroplets per hour and has a list price of $225,000. The instrument uses disposable chips to process the droplets, thus preventing contamination between samples.

Researchers start by sending the company the DNA regions they wish to analyze — up to 1,536 at the moment. RainDance then designs primer pairs to amplify these sequences and has them synthesized by an outside manufacturer. The company fills primer pairs into microdroplets, generating a primer pair library where each primer pair is represented equally, and sends this library to the customer.

RainDance declined to provide pricing information for the custom library offering. In the coming year, the company plans to launch catalog libraries, according to Jeremy Lambert, a RainDance product manager. Starting in the third quarter, customers will also be able to order custom libraries with up to 4,000 primer pairs.

Currently, RainDance supports targeted sequencing of the human genome, but libraries for other organisms "will be managed on a case-by-case basis," according to Lambert, as long as a well-annotated reference genome sequence is available for primer design.

After receiving the primer library, customers use their RDT 1000 instrument to mix, on-chip, the primer library with template and reagents for the amplification reaction. During this process, each primer pair droplet merges with a droplet that contains genomic DNA, fragmented to 2 to 4 kilobases, and reagents. The process takes about 45 minutes, generating approximately 1.5 million droplets that are collected in a single tube and can go directly into a thermocycler.

Following the amplification, users break the emulsion, a "quick, five-minute process," according to the company, purify the PCR products, and sequence them on a second-generation sequencer.

[ pagebreak ]

To test their technology, RainDance researchers amplified and sequenced a set of 384 exons from human cancer genes that covered 176 kilobases in total, using between 3 and 7 micrograms of genomic DNA as input. These exons represented amplicon sizes from 300 to 600 bases, a range of primer melting temperatures, and GC content from 30 to 60 percent.

They sequenced the amplification products on both the 454 Genome Sequencer FLX and the Illumina Genome Analyzer, though they believe this will "work just as well" on the Applied Biosystems SOLiD, which they are planning to use "in the near future," according to Olson.

The researchers found that with both sequencing platforms, all but one of the 384 exons were covered by sequence reads. Coverage was slightly more uniform with the Illumina platform, where 89 percent of amplicons had coverage that differed by fivefold or less, than with the 454, where 83 percent of amplicons had coverage within that range.

The specificity — the percentage of reads that mapped uniquely to the target — also differed between platforms, reaching about 84 percent for 454 but only about 40 percent for the Illumina GA. The likely reason is genomic DNA from the PCR that gets onto the sequencer, Olson explained, and the problem might be alleviated with better cleanup methods.

To demonstrate that their technology can handle larger numbers of reactions, in a second proof-of-concept, RainDance scaled up to 4,000 amplicons. These were chosen from genes involved in pancreatic cancer and covered 1.63 megabases in total. After sequencing the amplicons on the GS FLX and the GA, the scientists found that 99.7 percent of the regions were covered by Illumina reads, and 97 percent by 454 reads. For Illumina, 81 percent of amplicons fell within a fivefold range of coverage, and for 454, 89 percent of amplicons did.

According to Olson, though the company is offering customers up to 4,000 amplicons this year, but "there is no reason we can't go higher than that." Ultimately, he said, the process is limited by the amount of genomic DNA required — right now, he said, the system needs between 2 and 4 micrograms of DNA to go up to 4,000 amplicons — and by the time the overall process requires.

Streamlining Protocols

The Genome Center at Washington University — one of RainDance's early-access customers, along with the Broad Institute and the J. Craig Venter Institute — has been working on protocol changes to streamline template preparation, decrease the DNA input requirements, and increase the specificity.

Vince Magrini, a research assistant professor at the center whose team has been working with a beta-version of the RainDance system, said that he and his colleagues have explored fragmenting genomic DNA by HydroShear instead of incorporating biotin-dATP during nick translation, thus reducing the amount of genomic DNA needed to generate the 2- to 4-kilobase DNA fragments.

They have also added two Solid Phase Reversible Immobilization cleanup steps after the amplification to quickly remove genomic DNA and small non-specific DNA, so less non-specific DNA enters the sequencer.

So far, he and his colleagues have worked with the 384-amplicon primer library from RainDance and 3 micrograms of genomic DNA, and have been able to do up to eight "merges" of primer libraries with PCR reagents and DNA per day.

After sequencing the amplification products on the Illumina GAII platform, they found a number of the amplicons to be underrepresented, which correlated in many cases with their length, Magrini said.

The WashU researchers are now planning to scale up to 4,000 amplicons, probably a set of targets the center has already sequenced by conventional PCR in the past in its medical resequencing pipeline. This will help them determine the best coverage and read length in order to analyze the targets for sequence variants, he said.

He and his colleagues will also explore using whole-genome amplified DNA as input material, which would be important for projects where the amount of starting DNA is limited, he added.

Overall, Magrini said, the RainDance technology has been easy to use — as the company provides the primer library, chips, and fluidics — although there have been occasional technical problems, some of which resulted from user error. "Everything comes shipped, and you just prep the DNA template, plug in the reagents and press 'go' and start doing the merges," he said.