Construct and oligo design
- Identify the desired insertion or knock-in site in the genomic DNA.
- Take 50 bp of flanking sequence 5’ and 3’ to the insertion site and identify possible CRISPR cut sites (PAMs). I typically use Feng Zhang’s MIT site, though there are others out there. http://crispr.mit.ed
- Pick high scoring PAMs with no off-target sites in genes, as close as possible to the insertion site. My data suggests that DSBs closer to an insertion site may be more effective, though differences in PAMs may play a role as well. I typically choose two PAMs per desired modification. Save the output as a webarchive so that you have a list of the sgRNAs, their scores, and potential offtarget sites.
- If using PCR-derived sgRNAs, order oligos with the sgRNA target sequence in place of “N20” in 5’‑cctcctattgcgagatgtcttg(N20)gtttaagagctatgctgg‑3’. This keeps the size at 60 bp, which is the cheapest synthesis scale at IDT, the oligo supplier that I use. Note: do not include the “NGG” PAM sequence in the oligo.
- For oligo-templated knock-ins, use 35-80 basepairs of flanking sequence. For epitope tags, it helps to include a flexible spacer sequence encoding a restriction site (ie. a BamHI site encodes glycine-serine). This inclusion facilitates knock-in screening by restriction digestion. One can have the oligos PAGE purified, which leads to higher knock-in efficiency, or avoid this cost and inject more animals and screen more F1s. Inactivate PAMs in repair template. If possible, silently mutate one of the guanines in the “NGG” sequence. If this is not possible (ie. some codons do not allow the possibility of silent mutation of these guanines), make 5-6 silent mutations in the sgRNA target sequence.
- Design oligos for amplifying the insertion site (typically aiming for a 600 bp product); I use Primer3Plus (http://primer3plus.com/cgi-bin/dev/primer3plus.cgi) to design these oligos.
- Also design knock-in specific screening oligos. The oligo sits at an insertion junction with the two most 3’ bases binding to the insert and the remaining sequence binding to genomic sequence.
- In theory, it should be possible to enrich for similar modifications as described by Paix et al. (2014)(ie. oligo-templated insertion of stop codons or deletions, knock-in of PCR-derived GFP cassettes), but I have not formally tested these templates yet.