Cas9 induced mutations at yellow

As a proof of principle for the functionality of the Cas9/gRNA system in Drosophila, we chose the yellow locus to conduct our first target mutagenesis. We designed gRNAs to target two different sites in the yellow coding region (y1-gRNA and y2-gRNA; Supporting Information, Figure S1). The gRNA along with in vitro transcribed Cas9 mRNA were injected into y⁺ w¹¹¹⁸ embryos. The results from targeting yellow were summarized in Table 1.

For y1-gRNA, 80% of the injected F₀ males showed somatic yellow mosaicism suggesting that Cas9/y1-gRNA induced yellow mutations in somatic cells (Figure S1C). We crossed F₀ adults to y w flies to recover Cas9 induced yellow mutations in the germline. For y1-gRNA, between 30 and 67% of the F₀ yield at least one yellow offspring; 7–13% of the fertile flies inherited the Cas9-cut chromosome appearing as yellow mutants. For y2-gRNA, 36% of the F₀ were mosaic for yellow; 59–68% of them yield yellow offspring and 7–9% of the total progeny were yellow mutants. Sequencing identified small insertions and/or deletions (indels) at y1 or y2 in these individuals that bear signatures of nonhomologous end joining (NHEJ) repair (Figure S1D). Therefore, Cas9 can efficiently induce indel mutations at yellow in both the male and female germline.

Cas9 induced mutations at other euchromatic loci

To further demonstrate the functionality of the Cas9 system in Drosophila, we selected three additional euchromatic loci. Mutations of the ms(3)k81 locus cause male sterility (Fuyama 1986). No mutation has been characterized for either CG3708 or CG9652. Results for this series of targeted mutagenesis are reported in Table 2.

For ms(3)k81, we designed a gRNA that would result in Cas9 cutting at an RsaI restriction site. We reasoned that indels at this site would destroy the cut site and render a PCR fragment across this site resistant to RsaI digestion. Interestingly, all 17 F₀ males were sterile. Some of the sterility must be an artifact from the micro-injection as 12 of the 20 females were also sterile. However, another possibility is that some of the male sterility could be due to simultaneous disruption of both ms(3)k81 copies on homologous chromosomes in the germinal nuclei, which are determined to develop into germ cells. We subjected these sterile males along with several randomly selected F₀ female to PCR amplification followed by RsaI digestion (Figure S2A). While the PCR fragment from uninjected controls could be fully digested with RsaI, the PCR from F₀ flies showed no sign of digestion, suggesting that in most, if not all, of the PCR fragments, the RsaI site had been mutated due to Cas9 cutting. Our proposition was supported when we assayed the mutation frequency in the female germline. Of all the 88 progeny randomly selected from 8 fertile F₀ females, 87 were mutant for ms(3)k81 as assayed by RsaI digestion, DNA sequencing, and complementation with an existing ms(3)k81 mutation (Figure S3).

For CG3708, we designed a gRNA to direct Cas9 cutting at a BpmI site (Figure S2B). About 35% of the randomly selected progeny carried an indel at the BpmI site as assayed by both BpmI digestion of PCR fragments and DNA sequencing (Figure S4). For CG9652, we recovered two germline events with an indel out of 94 events sequenced (Figure S5). This frequency seemed rather low when compared to those from targeting the other loci. The underlying cause requires further investigations. In summary, we successfully targeted three different euchromatic loci using Cas9/gRNA. The germline frequency ranged from about 2% to essentially 100%.

Cas9 induced mutations at heterochromatic loci

The heterochromatic portion of the genome is generally considered inert. Yet it contains functional genes, some of which are essential for viability or fertility. We selected three different heterochromatic loci for targeted mutagenesis: the kl-3 fertility factor on the Y chromosome, the light, and RpL15 loci on autosomes.

For kl-3, all of the 29 F₀ males were sterile, yet only 4 of 37 F₀ females were sterile (Table 2), suggesting that in all of the F₀ males, most if not all of the germ cells carried a nonfunctional kl-3 locus, possibly due to Cas9-induced indel mutations. To provide support for this proposition, we PCR amplified the Cas9 targeted region in kl-3 from individual F₀ males. DNA sequencing revealed multiple peaks around the Cas9 cut site, consistent with the PCR products being a mixture carrying different indels of kl-3 (Figure S2D and Figure S7).

For the RpL15 locus, we observed something similar to that observed when targeting ms(3)k81, which might have been frequent alteration of both gene copies in addition to a possible off-target effects. Of the 800 embryos injected, none developed beyond the first-instar stage. This lethal phase matches that of existing RpL15 mutations (Schulze et al. 2001), suggesting that both RpL15 copies had been disrupted in the injected individuals. Consistently, PCR products of the Cas9-targeted region amplified from dying F₀ larvae could not be digested with the NspI enzyme, although in wild-type DNA they could (Figure S2C). DNA sequencing of PCR products from F₀ animals revealed the presence of indels (Figure S8).

For the light gene, DNA sequencing of 50 randomly selected F₁ adults identified 9 indel events representing a frequency of 18% (Figure S6)