Q: What is the minimum length of the homology arms for homologous recombination?
A: Geraldine Seydoux's lab has recently reported that as little as 30-60 bp of homology is enough to support recombination. We have verified that short homology arms can work sometimes; however, in our hands, using short homology arms results in much lower efficiency, along with a higher frequency of rearrangements and false positives. We routinely use homology arms that are 500-700 bp in length.
Q: How do you add LoxP sites to generate the floxed unc-119(+) cassette?
A: We do this by PCR, using any unc-119(+)-containing plasmid as template and including the LoxP sites in the primers. The LoxP sequence is 5’-ATAACTTCGTATAGCATACATTATACGAAGTTAT-3’. Including this sequence makes the primers a bit long, but this hasn't caused us any problems so far.
Q: Do you have a plasmid with unc-119(+) flanked by LoxP sites?
A: We have such a plasmid and can send it on request. However, we strongly recommend generating the floxed unc-119(+) cassette by PCR, using any unc-119(+)-containing plasmid as template and including the LoxP sites in the primers. Here's why: a LoxP site can't be used as the annealing part of a primer (that is, the 3' part for which the Tm is calculated), for two reasons: first, because there are two of them in any vector you'd want to amplify, and second, because LoxP sites are palindromes and fold into hairpins after denaturation. Therefore, there are two ways to generate a floxed unc-119(+) PCR product:
1) Use long primers containing the LoxP sites, as recommended above. The LoxP sites are still part of the primers, but they are in the "tails" that don't actually have to anneal. The long primers have not caused us any problems. This strategy does not required a floxed unc-119(+) template - any unc-119(+)-containing plasmid can be used.
2) Include spacer sequences on either side of the floxed unc-119(+) cassette, and design your primers to anneal to those. We didn't use this strategy and don't recommend it, because the goal of using Cre/Lox in the first place is to add as little extraneous sequence to the genome as possible.
Q: Can I insert the unc-119(+) marker into an intron?
A: Yes, we have done this and it works (see the "Broken GFP" strategy on our Construct Designs page). The marker will disrupt expression of the gene initially, but the LoxP site left behind after marker removal will not interfere with splicing. We suggest orienting unc-119(+) on the same strand as the gene you are disrupting, not in an antisense orientation, since transcriptional read-through between the gene of interest and an unc-119(+) on the opposite strand could produce double-stranded RNAs and cause unforeseen effects.
Q: Have you tried removing the selectable marker with FLP recombinase instead of Cre?
A: We haven't tried this, but it has been reported to work in a paper by Vázquez-Manrique et al. (2010).
Q: How large an insertion/deletion can one make using this technique?
A: We have not tried to rigorously define an upper limit on size. However, Christian Frøkjær-Jensen has reported making insertions of up to 16 kb and deletions of up to 25 kb using Mos1-based methods, which also rely on homologous recombination. Our lab has made deletions of up to 9 kb, and insertions of up to 12 kb. The efficiency of these modifications was similar to what we usually see for smaller changes, so it's likely we haven't yet approached the upper limit. If one wishes to make a very, very large insertion/deletion, in principle this could be done by repeatedly modifying the same locus in several homologous recombination steps.
Q: What is the correct nomenclature for knock-ins and targeted mutations?
A: Nomenclature guidelines for genome engineering applications were described in a 2013 Worm Breeder's Gazette article. Briefly, modified loci are given standard allele designations, with a bracketed notation that indicates the nature of the genome modification. For example, an nmy-2::gfp knock-in allele is nmy-2(cp7[nmy-2::gfp + LoxP unc-119(+) LoxP]) I. (For convenience we sometimes abbreviate this as nmy-2::gfp(cp7), but note that this is not approved nomenclature and should not be used in publications.) A phosphorylation-defective lin-31 mutant is lin-31(cp2[T145A T200A T218A T220A + LoxP unc-119(+) LoxP]). After removal of the unc-119(+) selectable marker with Cre recombinase, a new allele designation is assigned. For example, excision of unc-119(+) from the nmy-2::gfp(cp7) strain generated the new allele nmy-2(cp13[nmy-2::gfp + LoxP]).
Q: In your 2013 paper you used the unc-119 strain DP38, but your protocol recommends using HT1593 instead. What's the difference between the two strains?
A: HT1593 is an outcrossed derivative of DP38. Both strains carry the unc-119(ed3) allele. DP38 has a more severe phenotype than HT1593; DP38 worms are more Dpy and Egl, especially at 25°C, and are somewhat less active. The more severe phenotype of DP38 is undoubtedly due to additional mutations besides unc-119(ed3). We chose to use DP38 in our initial experiments because the more severe phenotype makes it easier to spot very rare rescued animals, and we did not know how efficient Cas9-triggered homologous recombination would be. However, the selection procedure turned out to be extremely efficient, such that using DP38 is not necessary. We have successfully used HT1593 in later experiments, and we recommend using it in future studies because it lacks secondary mutations that could complicate the interpretation of results.
Q: Will this approach work in other nematode species?
A: We have not tested Cas9-triggered homologous recombination in other species, but we expect it would work. The biggest challenge might be getting Cas9 and the sgRNA to express in the germline in other species. For closely-related species, our plasmid-based system using the eft-3 and U6 promoters might work, but for more distantly-related species an RNA injection approach might be a better choice (for example, see papers from Lo et al. and Chiu et al.).
Can't find what you're looking for here? Try the Worm CRISPR Forum.
A: Geraldine Seydoux's lab has recently reported that as little as 30-60 bp of homology is enough to support recombination. We have verified that short homology arms can work sometimes; however, in our hands, using short homology arms results in much lower efficiency, along with a higher frequency of rearrangements and false positives. We routinely use homology arms that are 500-700 bp in length.
Q: How do you add LoxP sites to generate the floxed unc-119(+) cassette?
A: We do this by PCR, using any unc-119(+)-containing plasmid as template and including the LoxP sites in the primers. The LoxP sequence is 5’-ATAACTTCGTATAGCATACATTATACGAAGTTAT-3’. Including this sequence makes the primers a bit long, but this hasn't caused us any problems so far.
Q: Do you have a plasmid with unc-119(+) flanked by LoxP sites?
A: We have such a plasmid and can send it on request. However, we strongly recommend generating the floxed unc-119(+) cassette by PCR, using any unc-119(+)-containing plasmid as template and including the LoxP sites in the primers. Here's why: a LoxP site can't be used as the annealing part of a primer (that is, the 3' part for which the Tm is calculated), for two reasons: first, because there are two of them in any vector you'd want to amplify, and second, because LoxP sites are palindromes and fold into hairpins after denaturation. Therefore, there are two ways to generate a floxed unc-119(+) PCR product:
1) Use long primers containing the LoxP sites, as recommended above. The LoxP sites are still part of the primers, but they are in the "tails" that don't actually have to anneal. The long primers have not caused us any problems. This strategy does not required a floxed unc-119(+) template - any unc-119(+)-containing plasmid can be used.
2) Include spacer sequences on either side of the floxed unc-119(+) cassette, and design your primers to anneal to those. We didn't use this strategy and don't recommend it, because the goal of using Cre/Lox in the first place is to add as little extraneous sequence to the genome as possible.
Q: Can I insert the unc-119(+) marker into an intron?
A: Yes, we have done this and it works (see the "Broken GFP" strategy on our Construct Designs page). The marker will disrupt expression of the gene initially, but the LoxP site left behind after marker removal will not interfere with splicing. We suggest orienting unc-119(+) on the same strand as the gene you are disrupting, not in an antisense orientation, since transcriptional read-through between the gene of interest and an unc-119(+) on the opposite strand could produce double-stranded RNAs and cause unforeseen effects.
Q: Have you tried removing the selectable marker with FLP recombinase instead of Cre?
A: We haven't tried this, but it has been reported to work in a paper by Vázquez-Manrique et al. (2010).
Q: How large an insertion/deletion can one make using this technique?
A: We have not tried to rigorously define an upper limit on size. However, Christian Frøkjær-Jensen has reported making insertions of up to 16 kb and deletions of up to 25 kb using Mos1-based methods, which also rely on homologous recombination. Our lab has made deletions of up to 9 kb, and insertions of up to 12 kb. The efficiency of these modifications was similar to what we usually see for smaller changes, so it's likely we haven't yet approached the upper limit. If one wishes to make a very, very large insertion/deletion, in principle this could be done by repeatedly modifying the same locus in several homologous recombination steps.
Q: What is the correct nomenclature for knock-ins and targeted mutations?
A: Nomenclature guidelines for genome engineering applications were described in a 2013 Worm Breeder's Gazette article. Briefly, modified loci are given standard allele designations, with a bracketed notation that indicates the nature of the genome modification. For example, an nmy-2::gfp knock-in allele is nmy-2(cp7[nmy-2::gfp + LoxP unc-119(+) LoxP]) I. (For convenience we sometimes abbreviate this as nmy-2::gfp(cp7), but note that this is not approved nomenclature and should not be used in publications.) A phosphorylation-defective lin-31 mutant is lin-31(cp2[T145A T200A T218A T220A + LoxP unc-119(+) LoxP]). After removal of the unc-119(+) selectable marker with Cre recombinase, a new allele designation is assigned. For example, excision of unc-119(+) from the nmy-2::gfp(cp7) strain generated the new allele nmy-2(cp13[nmy-2::gfp + LoxP]).
Q: In your 2013 paper you used the unc-119 strain DP38, but your protocol recommends using HT1593 instead. What's the difference between the two strains?
A: HT1593 is an outcrossed derivative of DP38. Both strains carry the unc-119(ed3) allele. DP38 has a more severe phenotype than HT1593; DP38 worms are more Dpy and Egl, especially at 25°C, and are somewhat less active. The more severe phenotype of DP38 is undoubtedly due to additional mutations besides unc-119(ed3). We chose to use DP38 in our initial experiments because the more severe phenotype makes it easier to spot very rare rescued animals, and we did not know how efficient Cas9-triggered homologous recombination would be. However, the selection procedure turned out to be extremely efficient, such that using DP38 is not necessary. We have successfully used HT1593 in later experiments, and we recommend using it in future studies because it lacks secondary mutations that could complicate the interpretation of results.
Q: Will this approach work in other nematode species?
A: We have not tested Cas9-triggered homologous recombination in other species, but we expect it would work. The biggest challenge might be getting Cas9 and the sgRNA to express in the germline in other species. For closely-related species, our plasmid-based system using the eft-3 and U6 promoters might work, but for more distantly-related species an RNA injection approach might be a better choice (for example, see papers from Lo et al. and Chiu et al.).
Can't find what you're looking for here? Try the Worm CRISPR Forum.