Reliable CRISPR/Cas9 Genome Engineering in Caenorhabditis elegans Using a Single Efficient sgRNA and an Easily Recognizable Phenotype

Abstract

CRISPR/Cas9 genome engineering strategies allow the directed modification of the Caenorhabditis elegans genome to introduce point mutations, generate knock-out mutants, and insert coding sequences for epitope or fluorescent tags. Three practical aspects, however, complicate such experiments. First, the efficiency and specificity of single-guide RNAs (sgRNA) cannot be reliably predicted. Second, the detection of animals carrying genome edits can be challenging in the absence of clearly visible or selectable phenotypes. Third, the sgRNA target site must be inactivated after editing to avoid further double-strand break events. We describe here a strategy that addresses these complications by transplanting the protospacer of a highly efficient sgRNA into a gene of interest to render it amenable to genome engineering. This sgRNA targeting the dpy-10 gene generates genome edits at comparatively high frequency. We demonstrate that the transplanted protospacer is cleaved at the same time as the dpy-10 gene. Our strategy generates scarless genome edits because it no longer requires the introduction of mutations in endogenous sgRNA target sites. Modified progeny can be easily identified in the F1 generation, which drastically reduces the number of animals to be tested by PCR or phenotypic analysis. Using this strategy, we reliably generated precise deletion mutants, transcriptional reporters, and translational fusions with epitope tags and fluorescent reporter genes. In particular, we report here the first use of the new red fluorescent protein mScarlet in a multicellular organism. wrmScarlet, a C. elegans-optimized version, dramatically surpassed TagRFP-T by showing an eightfold increase in fluorescence in a direct comparison.

Keywords: CRISPR/Cas9 genome engineering; Caenorhabditis elegans; coconversion; dpy-10; mScarlet.

Figure 1

Generation of d10-entry strains. (A) Insertion of the d10 sequence into sup-9, egl-23b, egl-23 (C-terminus), and twk-18 using a single-strand oligonucleotide repair template compatible with multiple sgRNAs. Genes and their intron/exon structure are displayed in the 5′–3′ orientation. The ssON repair templates are represented by black arrows (containing the d10 sequence in green) above the coding strand and translation of the target gene. Correspondence of homology regions between the ssON repair template and genomic locus is indicated in gray. sgRNA binding sites are indicated by blue open arrows. (B) unc-58 coconversion is used to detect the insertion of d10 sequences into a gene of interest. unc-58(e665) mutants are easily identified in the F1 progeny of injected P0 animals based on their straight body posture, lack of mobility, and characteristic rotation around the antero-posterior body axis. RT, repair template. (C) BanI, BssSI, and BsrBI restriction sites are present in the d10 protospacer sequence and are used for RFLP analysis. The Cas9 double-strand break site is indicated by an arrowhead. (D) R12E2.15 contains the only predicted off-target site of the d10 sgRNA. Four base changes (in pink) distinguish both sites. A BsrBI site follows the Cas9 double-strand break site (indicated by an arrowhead), between the −3 and −4 bases relative to the protospacer adjacent motif (PAM).

Figure 2

Generation of multiple knock-in lines using a single d10-entry strain. (A) A single d10-entry strain is used to engineer N-terminal TagBFP, TagRFP-T, and wrmScarlet fusions in the twk-18 locus. Correspondence of homology regions between the plasmid repair template and twk-18 genomic locus is indicated in gray. RT, repair template. (B) Two to 3 days following injection of a d10-entry strain with a CRISPR/Cas9 mix, F1 progeny with Dpy-10 phenotypes (Rol or Dpy) can be easily recovered, and further screened in the F2 generation to identify the desired genome edits by PCR or phenotype.

Figure 3

Comparison of TagBFP, TagRFP-T, and wrmScarlet using reliable editing of the twk-18 locus. (A) wrmScarlet::TWK-18 is visibly brighter than TagRFP-T::TWK-18. Side-by-side comparison of two young adult hermaphrodites. wrmScarlet-associated fluorescence is visible by eye in freely moving worms on NGM plates, while TagRFP-T is not detectable by eye in this context. (B) The two-pore domain potassium channel TWK-18 decorates the plasma membrane of body wall muscle cells. Representative images of head muscle cells labeled with N-terminal fusions of TWK-18 to TagBFP, TagRFP-T, and wrmScarlet. Head is left. Bar, 10 µm. (C) Quantification of fluorescence intensity shows an eightfold increase in fluorescence between TagRFP-T and wrmScarlet. Mean ± SD. Student’s t-test, *P < 0.0001.