Increasing the efficiency of precise genome editing with CRISPR-Cas9 by inhibition of nonhomologous end joining.
They confirm what Singh et al previously reported in a small but exciting data morsel last fall, which is that substantially higher rates of HDR-mediated insertion can be achieved in mouse zygotes by treating them with the NHEJ inhibitor, SCR7, during the injection process. They actually mixed SCR7 (final conc. 1mM) directly into the injection cocktail of gRNA + Cas9mRNA + donor ssDNA oligo.
After some preliminary tests in cell lines, they moved to zygotes. Using a donor oligo to insert a short peptide tag and validated CRISPR targets/gRNAs, they did tests with and without SCR7. Bottom line: HDR-mediated insertion rates increased by several fold for the two genes they tested. Although that may not sound like a breakthrough to some of you, many of you will know that in the world of mouse engineering it's key, because it will probably often mean the difference between getting zero versus a few correctly engineered pups out of an injection series.
Some other highlights are:
1. The embryos seem to tolerate SCR7 application under these conditions with no problem; no toxicity or increased death was noted. Various other studies seem to support that transient inhibition of NHEJ is well tolerated. Note that the SCR7 target, ligase IV, is critical for embryonic development so it can't be globally knocked out.
2. No increase in off-target effects. Cool.
1. Yesterday's google searching quickly turned up 3 companies selling SCR7. Yay.
2. SCR7 must be dissolved in DMSO. I think making a stock solution of 100 mM SCR7 in DMSO is reasonable. So the final injection mix, with 100-fold SCR7 dilution from the stock, will have 1 mM DMSO and also 1% DMSO. I couldn't dig out the SCR7 stock solution details from the paper but it's probably close to these parameters.
3. SCR7 very strongly inhibits the recovery of NHEJ-style mutations from the CRISPR targets.
4. The zygote injections were all done cytoplasmic, not pronuclear, although they were done at the pronculear stage. Thus it is clear that HDR edits with ssDNA oligos can be efficiently done by cytoplasmic injections. This is great because it results in higher rates of pup survival than pronuclear injection.
Still lingering questions for me:
1. Although the authors showed they could increase the insertion rate of a "large" cassette - a GFP-style reporter ORF - in cell culture, they did not repeat this experiment in embryos. Or at least they didn't show the data. Was there negative data to report? Or just not enough live pups yet for them to feel comfortable with publishing a negative result? Or have they not tried it yet? The routine insertion of kilobase-sized cassettes in embryos is now my next CRISPR mountain to climb!
2. I would kinda like to know if there may be an increased rate, or change, in the genome-wide mutation rate by SCR7 treatment. After all we are mucking around with the DNA repair pathway here. Since each mammal embryo probably has on the order of 50-100 new mutations anyway, it would have to be a pretty substantial change in mutation rate to scare me off. I'll bet there is no detectable effect. Besides, NHEJ usually results in new mutations anyway - so I would imagine that we'd observe cell or embryo death following SCR7 treatment, long before we could observe a change in mutation rates or spectrum in surviving embryos.