NEW YORK (GenomeWeb) – Bio-Rad Laboratories has found another application for its droplet digital PCR technology, hitching the instrument to one of the fastest moving trends in biology, CRISPR/Cas9 genome editing.
"Droplet digital PCR (ddPCR) is a tool that we've used for rare cancer mutation detection," Jen Berman, a scientist at Bio-Rad's Digital Biology Center, told GenomeWeb earlier this week at the Festival of Genomics in Boston. "This is just another flavor of rare mutation detection."
Berman has been making the conference rounds to show scientists how ddPCR can help them in their genome editing experiments. The efficiency of genome editing technologies is often low and is highly dependent on locus and cell type, she said. Oftentimes, researchers need to check their samples and see if any editing happened at all.
Berman makes the case that ddPCR can detect genome editing events at efficiencies less than 0.5 percent, and the quantification it allows can also reveal ratios between homology directed repair (HDR) and non-homologous end-joining (NHEJ) DNA repair mechanisms, as well as detect off-target effects.
Whether a researcher is trying to just create indel mutations for gene knockout or replace existing sequences in more precise editing, ddPCR can quantitatively detect the mutation "alleles," according to Berman.
Other methods to detect genome editing events include qPCR and gel-based T7 endonuclease surveyor assays. "You can observe on a gel when you get edits, but it's pretty finicky and it's semi-quantitative and not very sensitive below [editing efficiencies of] 10 percent," Berman said, which is often below the efficiency of CRIPSR/Cas9 as well as other methods such as zinc finger nucleases and transcription activator-like effector nucleases.
But ddPCR can also provide additional information on editing with a higher sensitivity, lower cost, and faster turnaround time that could sufficiently differentiate it from sequencing methods used to detect ratios between DNA repair mechanisms and off-target activity, she said.
After CRISPR/Cas9 makes its edits, it induces one of two DNA repair pathways. HDR can take a provided template and insert it at the cut site, while the more quick-and-dirty NHEJ simply rejoins the ends of the broken double strand.
"What people often want is HDR, and NHEJ represents collateral damage," when trying to introduce larger edits into the genome, Berman said. "There are a lot of papers trying out strategies to drive up HDR compared to NHEJ. Minimizing NHEJ will be particularly relevant to therapeutic editing where they want to introduce a precise change but need to be careful of knocking out other functions." Accidentally knocking out a tumor suppressor gene could be disastrous, for example.
Sequencing can help find the ratio of HDR to NHEJ repair, but it runs the risk of bias because it requires an amplification step. "Shorter things amplify better than longer things," Berman said, so in a case where NHEJ led to a deletion, instead of a larger insertion with HDR, the amplification could show more of the NHEJ-mediated repair.
With a little planning, ddPCR primers could be designed to detect the different possible kinds of mutation alleles that would result from either HDR or NHEJ.
Berman said that concerns with off-target effects could also be studied with ddPCR, especially as a way to triage edited cells for further study. The only other way to find off-target effects in, say, a cell, cell line, or organism is with whole-genome or targeted sequencing methods. "Let's say I'm editing my stem cells, and know I have danger of off-target events based on sequence homology. Let me look at these pools of cells, look at least at the frequency of off-targets to make sure their integrity is fine," she said.
"Some people won't move forward with a clone without sequencing for potential off-targets," she said, to guard against proceeding with extraneous mutations added in the editing process. Sequencing is more expensive and can take days. Digital PCR could screen for predictable off-target edits in just a few hours and cheaply, too, at a cost of about $3 per well, said Berman, adding that each well of the 96-well plate could be used to look for a different off-target edit.
So if researchers had a discrete list of off-targets they were interested in, they could run their samples through ddPCR with primers targeting those unwanted edits and know within a few hours which samples had experienced off-target edits.
"It's a first step to bin samples," she explained, so that if any cells displayed easily predictable off-target effects they could be discarded. Sequencing could be reserved for only the most clean-looking samples. "You might do WGS to make sure the only perturbation in the cell you're studying is the edit you made," as a last step before starting the study of that genomic edit.
Of course, the specificity of the programmable guide RNA (gRNA) is the root cause of off-target effects, and Berman says ddPCR assays could also help validate which gRNAs work best.
"A lot of computational work has been done to identify the best gRNA for a particular edit or editing strategy," Berman said, but she added that there's a great need for empirical validation of gRNA efficacy. "We have pretty good computational tools but connecting that to a quantitative readout from the lab will be important to advancing the field," she said. "A ddPCR assay could allow a readout for absolute quantification of gRNA efficiency."
Furthermore, Berman thinks that ddPCR could be used to infer the hidden rules that lead to one repair mechanism being favored over another. "Imagine a study looking at gRNA efficacy all over a region and looking at what the rules are for good editing," Berman said. "That type of work investigating mechanisms of repair could be of interest to the field."
This story has been updated to note that CRISPR/Cas9 can also have a rate of efficiency less than 10 percent.