Optimized CRISPR tools and site-directed transgenesis towards gene drive development in Culex quinquefasciatus mosquitoes

Culex mosquitoes are a global vector for multiple human and animal diseases, including West Nile virus, lymphatic filariasis, and avian malaria, posing a constant threat to public health, livestock, companion animals, and endangered birds. While rising insecticide resistance has threatened the control of Culex mosquitoes, advances in CRISPR genome-editing tools have fostered the development of alternative genetic strategies such as gene drive systems to fight disease vectors. However, though gene-drive technology has quickly progressed in other mosquitoes, advances have been lacking in Culex. Here, we develop a Culex-specific Cas9/gRNA expression toolkit and use site-directed homology-based transgenesis to generate and validate a Culex quinquefasciatus Cas9-expressing line. We show that gRNA scaffold variants improve transgenesis efficiency in both Culex quinquefasciatus and Drosophila melanogaster and boost gene-drive performance in the fruit fly. These findings support future technology development to control Culex mosquitoes and provide valuable insight for improving these tools in other species.

modified by adding the gRNA target sequences or the scaffold variants, such further edits are not reported in this table. Next to the construct structure are reported the genes from which genomic sequences were amplified and the laboratory line from which genomic DNA and PCR amplification was run. (b) Variation of the Cas9 constructs used to test activity in cells and in embryos. These constructs were tested instead of the simpler ones as we originally intended to target the white gene for transgenesis. Additionally we tested two exogenous promoters, Actin5C from Drosophila melanogaster (D. mel.) and PUb from Aedes aegypti. (c) HR template plasmids built for transgenesis of the vasa-Cas9 cassette in Culex quinquefasciatus. (d) Constructs used to evaluate transgenesis and gene drive performance in Drosophila melanogaster. Accession numbers for the deposited plasmids are also provided below and in the Data Availability Statement. Editing efficiency (%) at kh3 target locus in Hsu cells when transfected with a mixture of Cq-Actin5C>Cas9 paired either to U6:2 or U6:6 pol-III promoters expressing the kh3-gRNA with different gRNA scaffold variants. Histogram bars represent the mean, error bars and dots represent SD and distribution of 3 biological replicates. (a-b) Statistical comparisons were generated with a two way ANOVA multiple comparison of the mean, with Tukey's corrections. **** = P tukey < 0.0001, ns = P tukey > 0.05. P-values: (a) 12d-18d<0.0001; (b) U6-2_O-L=0.9997; U6-2_O-L+M>0.9999; U6-6_O-L<0.0001; U6-6_O-L+M<0.0001. (c) Modified version of our protocol in Fig. 1d that we used to evaluate in vivo the allele editing rates of the gRNA variants. The Cq-Actin5C>Cas9 plasmid was co-injected with U6:1>cd1-gRNA with gRNA scaffold variants, then we collected DNA from all developing embryos at~36h post injection for the subsequent deep sequencing analysis. (d) Edited allele percentages observed at the cd1 site, for the scaffold variants evaluated. The control was generated from a pool of uninjected eggs. Pupae about to hatch were collected and individually deposited in different Eppendorf tubes with about 1mL of water. Each tube was then coupled with a capped plastic Drosophila vial with the bottom sawed off, using a 3D-printed adaptor. (b) After~24 hours the hatched mosquitoes were maintained in the respective tubes, while (c) the pupal case was removed and used for PCR amplification of the cardinal locus. (d) Resulting amplicons were used for Sanger sequence analysis of the cardinal locus and evaluate the individuals' genotype (A/A, A/B or B/B; we did not recover any B/B animals suggesting lethality associated with this allele). (e) All A/A male and female mosquitoes were pooled in a cage to generate a line homozygous for the cardinal-A allele, about~24h after hatching. Note that the mosquitoes were able to survive overnight within the collection vials without needing a source of food or water as they were kept in a high-humidity environment.  * The G0 germline cutting and transgenesis efficiencies were calculated as numbers of independent pools that produce either cd-/cd-mutant (cutting) or DsRed+ (transgenesis) animals, divided by the total number of crossed G0s. While each pool contains several G0 individuals, in our calculations we assume only one editing event happens in each positive pool, therefore our calculations might underestimate the cutting and transgenesis rates. # The overall cutting and transgenesis efficiencies were calculated as the number of G1 individuals with either cd-/cd-(cutting) or DsRed+ (transgenesis) phenotypes divided by the total number of G1s.^In this injection we used an older batch of Cas9 protein which later showed low activity in subsequent studies. V1157 ACACTCTTTCCCTACACGACGCTCTTCCGATCTTGGTCATGATCAAGTGCAAACCG

Primer used for NGS analysis in Hsu cells (Kh-3 target site)
NGS218 ATGCGGCACACGCCATGGTT

Supplementary Data 1 -Genome editing quantification data.
Summary of the CRISPResso2 batch analysis on allele editing frequencies performed on deep sequencing data derived from in vivo and in vitro experiments. For each sample listed in tab 1 and tab 2 are specified: experimental details including names and ratio (%) of plasmid DNA transfected/injected and the experimental group; The raw CRISPResso allele frequency analysis output; The data extrapolated from the batch analysis to calculate allele editing and plotted in Fig1 and Supplementary  Figure 2. The data as described is subdivided in the following file tabs: 1. Cell line editing quantification 2. Embryo editing quantification 3. CRISPResso2 parameters: Includes the running parameters used for all analysis on CRISPResso2 with one example representative of each batch experiment run.

Supplementary Data 2 -Culex quinquefasciatus transgenesis counting data.
Different injection conditions with G0 survival efficiencies were listed. Raw counting data of the G1 progeny phenotype indicating the cutting/ transgenesis in different conditions. cd-/cd-phenotype indicates the cutting and DsRed+ indicates the integration. All G0s were divided as male and female pools and crossed with cd-/cd-. Egg rafts from each male pool were scored together, while rafts of each female pool were divided and hatched singly in different batches, which gives more precise evaluation of germline rates. The data is subdivided in the following file tabs: 1. Figure 2 - Table 1 Data: The cutting/transgenesis efficiencies in conditions of injecting different HDR template variants (HR+Original gRNA; HR+"Loop"; HR+"Loop+Mutation") were recorded. The germline and overall efficiencies were calculated. Table 1 Data: The cutting/transgenesis efficiencies in other conditions by injecting HDR templates supplemented with Cas9 resources (Cas9 protein; Cas9 plasmid mixture; and both) were recorded. The germline and overall efficiencies were calculated. 3. Statistical Analysis. A Randomization Test for a Difference in Proportions was performed to evaluate differences between: 1) "Original" injection vs. "Loop" injection (per G0 germline); 2) "Original" injection vs. "Loop+mutation" injection (per G0 germline); 3) "Original" injection vs. "Loop" injection in overall scored G1s; and 4) "Original" injection vs. "Loop+mutation" injection in overall scored G1s.

Supplementary Data 3 -Cas9 line validation counting data.
Two experimental conditions of injecting in-vitro-transcribed Kh3-gRNA or Kh3-gRNA plasmid were listed. The embryos of the heterozygous Cas9 line were used for injection. The injection conditions and G0 situations were recorded. The Cas9 positive G0s were crossed with kh-/kh-mutants to evaluate gRNA activity in G1. The cutting efficiencies were calculated based on numbers of individuals giving khmutant phenotype in G1s. The data is subdivided in the following file tabs: 1. Figure 3e Data: The editing rate observed for either IVT-gRNA or plasmid is calculated.

Supplementary Data 4 -Effect of the gRNA scaffold with the loop modification on transgenesis efficiency.
Raw counting data of the G1 progeny phenotypic scoring indicating total females and males screened. All G0 crosses were performed in single-pairs between one injected individual and one wildtype (non-injected) animal. Number of recovered GFP positive (GFP+) and non-GFP (GFP-) individuals from the G1 are indicated. This allows us to precisely evaluate single-germline transgenesis rates in our experimental conditions. The data is subdivided in the following file tabs: 1. Fig. 4 (w5 original): Trangenesis rates after injecting the w5-gRNA element carrying the "Original" gRNA scaffold version. Independent germlines that gave rise to transformants (GFP+) and total germlines analyzed as well as transformation rates are indicated. 2. Fig. 4 (w5 loop): Trangenesis rates after injecting the w5-gRNA element carrying the "loop" gRNA scaffold modification. Independent germlines that gave rise to transformants (GFP+) and total germlines analyzed as well as transformation rates are indicated. 3. Statistical Analysis. Test for the Difference in G0 germline Proportions between w5 "Original" injection vs. w5 "Loop" injection, and Test for the Difference in G1 scored individuals Proportions between w5 "Original" injection vs. w5 "Loop" injection.

Supplementary Data 5 -Gene drive experiments with different gRNA scaffold variants in Drosophila melanogaster.
Raw counting data of the F2 progeny phenotypic scoring indicating females and males recovered. Red marker (DsRed+), green marker (GFP+), both fluorophores (both), no fluorescence (none), wild-type eye (w+), white-eye (w-), and mosaic eyes (mosaic) were scored in order to track Cas9 (red marker) and the gRNA (green marker) transgenes, as well as other outcomes of the cross. Transgene inheritance rates in the F2 progeny for each specific tube (marked as "F1 Cross" in the table) were calculated by combining data from males and females. Average inheritance for both markers and the standard deviation are calculated as well. The data is subdivided in the following file tabs: 1. Fig. 5 (Original): Inheritance of the w5-CopyCat targeting the white gene carrying the "Original" gRNA scaffold variant, driven by a transgene expressing Cas9 under the vasa promoter. 2. Fig. 5 (Loop): Inheritance of the w5-CopyCat carrying the "Loop" gRNA scaffold variant. 3. Fig. 5 (Mutation): Inheritance of the w5-CopyCat carrying the T>C "Mutation" gRNA scaffold variant. 4. Fig. 5 (Loop+Mutation): Inheritance of the w5-CopyCat carrying the gRNA scaffold variant with both the additional "Loop" as well as the T>C "Mutation". 5. Summary: This tab summarizes the total of flies counted in the four experimental conditions, and reports the average values for inheritance, cutting and conversion rates observed. 6. St: Inheritance rates statistics: Statistical analysis comparing different inheritance rates observed. The "Loop", "Mutation" and the "Loop+Mutation" conditions were compared to the "Original" condition. 7. St: Cutting rates statistics: Statistical analysis comparing different cutting rates comparison.
The "Loop", "Mutation" and the "Loop+Mutation" conditions were compared to the "Original" condition.
8. St: Conversion efficiency statistics: Statistical analysis comparing different conversion efficiencies comparison. The "Loop", "Mutation" and the "Loop+Mutation" conditions were compared to the "Original" condition.
The cd1-sgRNA and Kh3-sgRNA were synthesized and later replaced the BbsI restriction site to build different U6-gRNA plasmids for experiments.

HDR templates for Culex quinquefasciatus transgenesis 1) pVMG252 (Cq-vasa>Cas9_cdHAs_O)
The vasa>Cas9 element was amplified from pVMG213 then ligated with Opie2>DsRed marker, and this transgene later inserted between two~1.5 kb homology arms (HAs) matching the genomic sequences of the cardinal locus (CPIJ005949) abutting the cd1-gRNA target site. The U6:1>cd1-gRNA component was placed outside of the HAs.

2) pVMG280 (Cq-vasa>Cas9_cdHAs_L)
The HAs and transgene are the same as pVMG252, the difference is the gRNA scaffold with a modified "Loop" structure.

3) pVMG277 (Cq-vasa>Cas9_cdHAs_L+M)
The HAs and transgene are the same as pVMG252, the difference is the gRNA scaffold with a modified "Loop+Mutation" structure.