A co-CRISPR strategy for efficient genome editing in Caenorhabditis elegans

Abstract

Genome editing based on CRISPR (clustered regularly interspaced short palindromic repeats)-associated nuclease (Cas9) has been successfully applied in dozens of diverse plant and animal species, including the nematode Caenorhabditis elegans. The rapid life cycle and easy access to the ovary by micro-injection make C. elegans an ideal organism both for applying CRISPR-Cas9 genome editing technology and for optimizing genome-editing protocols. Here we report efficient and straightforward CRISPR-Cas9 genome-editing methods for C. elegans, including a Co-CRISPR strategy that facilitates detection of genome-editing events. We describe methods for detecting homologous recombination (HR) events, including direct screening methods as well as new selection/counterselection strategies. Our findings reveal a surprisingly high frequency of HR-mediated gene conversion, making it possible to rapidly and precisely edit the C. elegans genome both with and without the use of co-inserted marker genes.

Keywords: CRISPR-Cas9 system; CRISPR-Cas9-induced indels; CRISPR-Cas9-mediated HR; Co-CRISPR; blasticidin selection.

Figure 1

Efficient CRISPR-Cas9-mediated gene disruption in transgenic animals. (A) Schematic of screen for CRISPR-Cas9 genome editing events. The dominant transformation marker rol-6 was co-injected with Cas9, pie-1a sgRNA, and donor plasmids. F1 rollers were screened for NHEJ-mediated indels by DNA sequencing. Among 93 F1 rollers, 22 indels were obtained. (B) Sequences of the wild-type pie-1 target site (top) and CRISPR-Cas9-mediated indels among F1 animals: (i) pie-1 homozygotes carrying the same indel on both alleles; (ii) pie-1 homozygotes carrying a different indel on each allele; and (iii) pie-1 heterozygotes. Lowercase letters indicate insertions, and dashes indicate deletions. The PAM is marked in red, and target sequences are marked in blue. The number of deleted (−) or inserted (+) bases is indicated to the right of each indel. The numbers in parentheses in (iii) represent the number of animals with the indels shown.

Figure 2

“unc-22” Co-CRISPR as a marker to indicate actively expressed Cas9. (A) Schematic of Co-CRISPR strategy to identify functional sgRNAs targeting avr genes. sgRNAs targeting avr-14 and avr-15 were co-injected with a functional unc-22 sgRNA, the Cas9 expression vector, and the rol-6 transformation marker. F1 rollers or twitchers were transferred to individual plates. The plates were allowed to starve, and then they were copied to plates containing 2 ng/ml ivermectin to identify CRISPR-Cas9-induced avr-14; avr-15 double mutants. (B) Indel sequences in avr-14; unc-22; avr-15 triple mutants. avr-15 isolate 15 carried different indels on each allele. Sequences labeled with a question mark could not be precisely determined. (C) Comparison of twitcher-based indel frequency and roller-based indel frequency.

Figure 3

HR-mediated knock-in to generate fusion genes at endogenous loci. (A) Schematic of the Cas9/sgRNA target site and the donor plasmid for gfp::pie-1 knock-ins. The donor plasmid contains the gfp coding sequence inserted immediately after the start codon of pie-1, 1 kb of homology flanking the CRISPR-Cas9 cleavage site, and a silent mutation in the PAM site. (B) Strategy to screen for gfp knock-in lines. We placed three F1 rollers at a time on a 2% agar pad and screened for GFP expression using epifluorescence microscopy. GFP-expressing worms were individually recovered and allowed to make F2 progeny for 1 day before being lysed for PCR and DNA sequence analysis. We confirmed Mendelian inheritance of gfp knock-in alleles among F2 progeny. (C) GFP::PIE-1 expression in the germline of two- to four-cell embryos of gfp::pie-1 knock-in strains. (D) Immunoblot analysis showing PIE-1 expression levels in wild-type animals, MosSCI-mediated gfp::pie-1 knock-in animals, and CRISPR-Cas9-mediated gfp knock-in animals. A MosSCI strain of gfp::pie-1; pie-1(zu154) was obtained by crossing gfp::pie-1 (LGII) with the pie-1(zu154) (LGIII) null mutant. (E) mCherry expression in late embryos of the mCherry::vet-2 knock-in strain. (F) Schematic of Cas9/sgRNA target sequence, PAM site, and donor plasmid for pie-1::flag knock-in. The PAM is located in the last exon of pie-1. The donor plasmid includes flag coding sequence immediately before the pie-1 stop codon and ∼800-bp homology arms flanking the target site. (G) PCR and restriction analysis of an HR event. PCR products were generated using the primers indicated in F, and the products were digested with NheI. The pie-1::flag gene conversion introduces an NheI RFLP that is observed in F1 heterozygous and F2 homozygous pie-1::flag animals.

Figure 4

A blasticidin-resistance marker to select pie-1 knockout mutants. (A) Schematic of the Cas9/sgRNA target sequence and an HR donor plasmid in which a heterologous blasticidin-resistance (BSD) gene replaces a region of pie-1 and is flanked by 1-kb homology arms. The BSD gene is under the control of the rpl-28 promoter (568 bp) and 3′-UTR (568 bp). (B) Schematic of the blasticidin selection strategy to precisely delete the pie-1 gene. pie-1a sgRNA was co-injected with the Cas9 expression vector, the rol-6 transformation marker, the pie-1∆::BSD donor construct, and the pCCM416::Pmyo-2::avr-15(+) counterselection vector. The indicated number of F1 rollers was transferred to the plates containing 2 ng/ml ivermectin to select against the extrachromosomal array and 100 μg/ml blasticidin to identify BSD knock-in lines. We identified two plates with resistant, fertile adults among 14 plates, 3–4 days after transferring animals.