Frequently asked questions
1) Have you tried using NHEJ mutants?
In the direct selection experiments I performed, using a lig-4 null, I did seem to yield better efficiency. However, it was always hard to tell if it was significant when screening hundreds of F1s and getting a single knock-in! One consideration that could be either beneficial or a hindrance is that the NHEJ mutations are all near pha-1, to varying extents: cku-70 is 0.49 cM from pha-1, cku-80 is 10.11 cM, and lig-4 is 3.84 cM. I had tried to knock-in a stop codon into cku-70 exon 1; such a strain could be maintained with a GFP marked hT2 balancer. Furthermore, due to tight linkage, one could easily outcross both pha-1 and the cku-70 mutation. Unfortunately, the sgRNAs I tested were inactive. I also didn't have any success with two lig-4 sgRNAs. One could cross in existing deletion alleles to the pha-1 strain; I didn't do this as these NHEJ mutant strains needed outcrossing, and then a recombination event was requited to get the pha-1 linked mutation. Unfortunately, I did not have time for these experiments with the revision timeline for the manuscript. Still, I think that it’s an outstanding question as to whether the NHEJ mutants could lead to further increases in efficiency, though potentially at a cost of background mutations.
2) What about other NHEJ RNAis?
Reviewing the literature, lig-4 and cku-80 RNAi both produced phenotypes; I did not find reports of robust phenotypes for cku-70. In a single repetition of the klp-12 MfeI site deletion assay and a single test of a GFP frameshift reporter (NHEJ restores GFP expression, similar to that described by Waaijers et al, 2013), cku-80 RNAi produced phenotypes, whereas lig-4 RNAi was inactive. As these were single experiments, I did not report them in the Genetics manuscript, but my feeling was that only cku-80 RNAi was worth pursuing.
3) Have you tried linear dsDNA templates, à la Paix et al. (2014)?
Regarding dsDNA repair templates generated by PCR, I don't see any reason why it shouldn't work with pha-1(ts) co-conversion, but I have not yet tried using these templates. I'm now trying to do some of those experiments, knocking GFP constructs into the same locus which ttTI5605 is inserted. This sgRNA has been validated (Dickinson et al. 2013), and I will post updates as the experiments progress. Again, this is an important question, as we still know very little about the mechanism of homologous recombination using ssDNA vs dsDNA templates.
In the direct selection experiments I performed, using a lig-4 null, I did seem to yield better efficiency. However, it was always hard to tell if it was significant when screening hundreds of F1s and getting a single knock-in! One consideration that could be either beneficial or a hindrance is that the NHEJ mutations are all near pha-1, to varying extents: cku-70 is 0.49 cM from pha-1, cku-80 is 10.11 cM, and lig-4 is 3.84 cM. I had tried to knock-in a stop codon into cku-70 exon 1; such a strain could be maintained with a GFP marked hT2 balancer. Furthermore, due to tight linkage, one could easily outcross both pha-1 and the cku-70 mutation. Unfortunately, the sgRNAs I tested were inactive. I also didn't have any success with two lig-4 sgRNAs. One could cross in existing deletion alleles to the pha-1 strain; I didn't do this as these NHEJ mutant strains needed outcrossing, and then a recombination event was requited to get the pha-1 linked mutation. Unfortunately, I did not have time for these experiments with the revision timeline for the manuscript. Still, I think that it’s an outstanding question as to whether the NHEJ mutants could lead to further increases in efficiency, though potentially at a cost of background mutations.
2) What about other NHEJ RNAis?
Reviewing the literature, lig-4 and cku-80 RNAi both produced phenotypes; I did not find reports of robust phenotypes for cku-70. In a single repetition of the klp-12 MfeI site deletion assay and a single test of a GFP frameshift reporter (NHEJ restores GFP expression, similar to that described by Waaijers et al, 2013), cku-80 RNAi produced phenotypes, whereas lig-4 RNAi was inactive. As these were single experiments, I did not report them in the Genetics manuscript, but my feeling was that only cku-80 RNAi was worth pursuing.
3) Have you tried linear dsDNA templates, à la Paix et al. (2014)?
Regarding dsDNA repair templates generated by PCR, I don't see any reason why it shouldn't work with pha-1(ts) co-conversion, but I have not yet tried using these templates. I'm now trying to do some of those experiments, knocking GFP constructs into the same locus which ttTI5605 is inserted. This sgRNA has been validated (Dickinson et al. 2013), and I will post updates as the experiments progress. Again, this is an important question, as we still know very little about the mechanism of homologous recombination using ssDNA vs dsDNA templates.