Analysis of CRISPR gene drive design in budding yeast

Control of biological populations remains a critical goal to address the challenges facing ecosystems and agriculture and those posed by human disease, including pests, parasites, pathogens and invasive species. A particular architecture of the CRISPR/Cas biotechnology – a gene drive – has the potential to modify or eliminate populations on a massive scale. Super-Mendelian inheritance has now been demonstrated in both fungi and metazoans, including disease vectors such as mosquitoes. Studies in yeast and fly model systems have developed a number of molecular safeguards to increase biosafety and control over drive systems in vivo, including titration of nuclease activity, anti-CRISPR-dependent inhibition and use of non-native DNA target sites. We have developed a CRISPR/Cas9 gene drive in Saccharomyces cerevisiae that allows for the safe and rapid examination of alternative drive designs and control mechanisms. In this study, we tested whether non-homologous end-joining (NHEJ) had occurred within diploid cells displaying a loss of the target allele following drive activation and did not detect any instances of NHEJ within multiple sampled populations. We also demonstrated successful multiplexing using two additional non-native target sequences. Furthermore, we extended our analysis of ‘resistant’ clones that still harboured both the drive and target selection markers following expression of Streptococcus pyogenes Cas9; de novo mutation or NHEJ-based repair could not explain the majority of these heterozygous clones. Finally, we developed a second-generation gene drive in yeast with a guide RNA cassette integrated within the drive locus with a near 100 % success rate; resistant clones in this system could also be reactivated during a second round of Cas9 induction.

This study a Strains GFY-150 and GFY-153 were generated from a previous study 3 . GFY-153 was the parental strain to generate GFY-163. Briefly, the CDC11 locus was replaced by the Kan R deletion cassette in BY4741 or BY4742 WT yeast harboring a URA3-based covering vector (pJT1520 4 ) that expresses a copy of CDC11 under its native promoter. These strains were used in mating tests (Fig.  S3). b Strain GFY-3733 is similar to GFY-3207. Artificial CRISPR sites 5 (u1') are positioned flanking the entire cassette with the sequence 5'-TTTTCCGGTGGACTTCGGCTACGTAGGGAGT-3'. The bold and underlined sequences include the PAM sites for Cas12a/Cpf1 (TTTV at the 5' end) and S. pyogenes Cas9 (NGG at the 3' end) with a common target site. Strain GFY-3733 was generated from integration of the cassette from the pGF-IVL1511 vector at the HIS3 locus in BY4742 yeast. The HIS5 gene is from fission yeast S. pombe and is the functional equivalent of S. cerevisiae HIS3. The 5' T for the upstream (u1') site has been artificially added. For the downstream (u1') site, the T already existed within the MX(t) sequence. For the 3' T within the upstream (u1') site, this base was artificially added. For the downstream (u1') site, this T already existed as part of the HIS3(t) sequence. c Strains GFY-4325 and GFY-4326 were two separate isolates created in an identical manner. First, BY4741 yeast were transformed with six overlapping PCR fragments that were assembled in vivo through selection on rich medium containing G418 (the initial integration contained the prMX-Kan R -MX(t) drug resistance cassette). Second, CRISPR-based editing was performed by activating Cas9 expression (galactose metabolism) and co-transformation of the pGF-V1642 plasmid expressing the sgRNA(Kan R ) cassette with a PCR fragment, prMX-CaURA3-MX(t) (amplified from plasmid JT2869), to serve as donor DNA. The URA3 gene is from C. albicans.

First Generation Yeast CRISPR Gene Drive
BY4741 haploid yeast are met15∆0 whereas BY4742 haploid yeast are MET15. We included the two haploid parental strains harboring the drive (GFY-2383) and target constructs (GFY-3733) at the HIS3 locus (labeled "C1" and "C2," respectively). We chose ten isolates (1-10) from the gene drive assay (Fig. 1) that were tested as diploids using the mating test (described in Fig. S3). One set of oligonucleotides (PCR A) tested for the presence of the MET15 coding sequence. The second set (PCR B) amplified the entire MET15 locus from within the promoter and terminator regions.
For the first PCR (left), the BY4742 (target) haploid and all 10 isolates displayed a fragment at the expected size of 583 bp. For the second PCR, two fragment sizes were expected: (i) met15∆0 would yield a product size of 900 bp whereas the MET15 locus would yield a band of 3,298 bp.
These amplified fragments were seen for haploid controls. However, for the GD1 isolates, two bands were observed at both sizes (red asterisk marks the larger band). Note, PCR reactions were optimized for generation of the 900 bp fragment. These data support that these gene drive strains included both the met15∆0 and MET15 alleles and were diploid. chromosomal DNA preparations using similar conditions. The drive allele was confirmed (PCR B) using oligonucleotides F2/R2 (PCRs include identical labels from Fig. 1D). The target allele was amplified (PCR D) using primers F4/R4. The LYS2 locus was amplified (PCRs E,G) using primers F5/R5 and F7/R7, respectively. Red asterisks, positions of included DNA ladders ("L").
Two haploid controls (labeled "B" and "C") are the original drive haploid strain (GFY-2383) and target haploid strain (GFY-3733), as in Fig. 1D. White lines indicate separate DNA gels. Note, the first and second gels were run together on a single agarose gel (two separate rows); the third and fourth gels were run in a similar fashion. One set of haploid controls was run per complete gel (for example, the first and third gels). No additional image processing or editing (aside from cropping for clarity) has been done. The isolate number is included on the far right of each image. These data demonstrate these 100 isolates were diploids, contained the drive allele, and had lost the target allele.