Efficient Screening of CRISPR/Cas9-Induced Events in Drosophila Using a Co-CRISPR Strategy

Abstract

Genome editing using the Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) and associated nuclease (Cas9) enables specific genetic modifications, including deletions, insertions, and substitutions in numerous organisms, such as the fruit fly Drosophila melanogaster One challenge of the CRISPR/Cas9 system can be the laborious and time-consuming screening required to find CRISPR-induced modifications due to a lack of an obvious phenotype and low frequency after editing. Here we apply the successful co-CRISPR technique in Drosophila to simultaneously target a gene of interest and a marker gene, ebony, which is a recessive gene that produces dark body color and has the further advantage of not being a commonly used transgenic marker. We found that Drosophila broods containing higher numbers of CRISPR-induced ebony mutations ("jackpot" lines) are significantly enriched for indel events in a separate gene of interest, while broods with few or no ebony offspring showed few mutations in the gene of interest. Using two different PAM sites in our gene of interest, we report that ∼61% (52-70%) of flies from the ebony-enriched broods had an indel in DNA near either PAM site. Furthermore, this marker mutation system may be useful in detecting the less frequent homology-directed repair events, all of which occurred in the ebony-enriched broods. By focusing on the broods with a significant number of ebony flies, successful identification of CRISPR-induced events is much faster and more efficient. The co-CRISPR technique we present significantly improves the screening efficiency in identification of genome-editing events in Drosophila.

Keywords: CRISPR; Cas9; Drosophila; co-CRISPR; co-conversion.

Figure 1

Schematic representation of the lbk gene, positions of gRNAs, and knock-in cassette. We selected two guides: one at the N-terminus between two putative start sites on the antisense strand (bottom), gRNA-lbk1, and the other overlapping the termination codon on the sense strand (top), gRNA-lbk2. The gRNA sequences are shown in black with PAM sites in red, while gray indicates the complementary sequence. The underline in gRNA-lbk2 indicates the termination codon. The knock-in cassette consisted of a 225 bp 3xFLAG-3XHA flanked by 387 and 384 bp lbk homology arms. The length of genomic lbk from the first putative ATG start codon to the TGA stop codon is 4629 bp.

Figure 2

Genetic scheme for screening with co-CRISPR. Transgenic nos-Cas9 embryos were injected with a mix of either gRNA-e, gRNA-lbk1, and gRNA-lbk2 expression plasmids or with gRNA-e, gRNA-lbk2, and a knock-in repair template containing a FLAG-HA tag. Injected individuals were crossed to the ebony double balancer (P cross). F1 offspring were scored for e and e+. Individual F1 curly, ebony flies were crossed to a second chromosome balancer (F1 cross) to generate second chromosome lbk balanced lines. The F1 cross shown is with a single curly, ebony male. After the e broods are identified, following e is unnecessary. Homozygotes were then analyzed for mutations in lbk. Fly images were created on Genotype Builder Photoshop file S5 (Roote and Prokop 2013).

Figure 3

Broods with a high percentage of co-CRISPR marker e are highly enriched for mutations in lbk. We plot broods from the F1 generation in three groups (0% ebony, 1–50% ebony and >50% ebony) against the percentage of F1-derived balanced with lbk-1 (A) and lbk-2 (B) mutations. We observed a strong correlation between the percentage of e in F1 broods and mutations in lbk target sites (P < 0.05 for lbk-1 and P < 0.01 for lbk-2; Pearson correlation: r² = 0.21 for lbk-1 mutants, r² = 0.32 for lbk-2 mutants). Total numbers of lines sequenced for lbk-1 were 24 (0%), 54 (1–50%), and 42 (>50%), and 24 (0%), 51 (1–50%) and 40 (>50%) for lbk-2.

Figure 4

CRISPR-induced mutations in lbk. Sequence alignments of wild-type lbk in transgenic nos-Cas9 flies and some representative lines of the ebony-enriched jackpot lines containing indels in genomic regions lbk-1 (A) and lbk-2 (B). The PAM sequence is highlighted, with genomic targets underlined. The number of base pairs deleted [−#], inserted [+#] and/or substituted [#] are indicated in brackets. A total of 120 balanced lines were sequenced (235 sequences for the two genomic regions); additional sequence alignments are in Figure S2. The stop codon in lbk-2 in this and subsequent figures is shown in red text.