Significance Aedes aegypti is the principal vector of multiple arboviruses that significantly affect human health, including dengue, chikungunya, and zika. Development of tools for efficient genome engineering in this mosquito will not only lay the foundation for the application of genetic control strategies, but will also accelerate basic research on key biological processes involved in disease transmission. Here, we report the development of a transgenic CRISPR approach for rapid gene disruption in this organism. Given their high editing efficiencies, the Cas9 strains we developed can be used to quickly generate genome modifications, allowing for high-throughput gene targeting, and can possibly facilitate the development of gene drives, thereby accelerating comprehensive functional annotation and development of innovative population control strategies.

Abstract The development of CRISPR/Cas9 technologies has dramatically increased the accessibility and efficiency of genome editing in many organisms. In general, in vivo germline expression of Cas9 results in substantially higher activity than embryonic injection. However, no transgenic lines expressing Cas9 have been developed for the major mosquito disease vector Aedes aegypti. Here, we describe the generation of multiple stable, transgenic Ae. aegypti strains expressing Cas9 in the germline, resulting in dramatic improvements in both the consistency and efficiency of genome modifications using CRISPR. Using these strains, we disrupted numerous genes important for normal morphological development, and even generated triple mutants from a single injection. We have also managed to increase the rates of homology-directed repair by more than an order of magnitude. Given the exceptional mutagenic efficiency and specificity of the Cas9 strains we engineered, they can be used for high-throughput reverse genetic screens to help functionally annotate the Ae. aegypti genome. Additionally, these strains represent a step toward the development of novel population control technologies targeting Ae. aegypti that rely on Cas9-based gene drives.

The yellow fever mosquito, Aedes aegypti, is the principal vector of many arboviruses, such as dengue, chikungunya, yellow fever, and zika. These pathogens are globally widespread and pose significant epidemiological burdens on infected populations, resulting in hundreds of millions infections and over 50,000 deaths per year (1⇓⇓–4). Due to the hazards they impose, many methods for controlling Ae. aegypti populations have been implemented, with the most common being chemical insecticides. However, chemical control has proven incapable of stopping the spread of Ae. aegypti, primarily due to the mosquito's ability to rapidly adapt to new climates, tendency to oviposit in minimal water sources, desiccation-tolerant eggs, and quick development of insecticide resistance (5, 6). Therefore, significant efforts are currently underway to discern the underlying molecular and genetic mechanisms important for arboviral vector competence, with the overall aim of developing insecticide-free ways to disrupt viral disease cycles (7). Importantly, uncovering these mechanisms hinges on the ability to stably insert and disrupt specific genes-of-interest through tailored genome engineering in a target-specific manner. Fortunately, several tools have been successfully employed in mosquitoes for targeted genome engineering that rely on either zinc finger nucleases (8⇓–10), transcription activator-like effector nucleases (TALENs) (11, 12), or even homing endonuclease genes (13). However, because they rely on context-sensitive modular protein–DNA-binding interactions, each of these designer nucleases are time-consuming and complicated to engineer and validate, making them onerous for routine use in most laboratories.

To overcome the significant limitations posed by previous genome-editing tools, the clustered regularly interspaced short palindromic repeats/CRISPR-associated sequence 9 (CRISPR/Cas9) system, originally discovered in bacteria and archaea (14⇓⇓⇓⇓⇓–20), has been adapted as a programmable (20, 21) precision genome-editing tool in a diversity of organisms (22⇓⇓⇓⇓⇓⇓–29), including mosquitoes (30⇓⇓⇓–34). Briefly, the CRISPR/Cas9 system is guided by a chimeric programmable synthetic short-guide RNA (sgRNA) (20) that binds Cas9, directing it to a user-specified genomic DNA target sequence, via Watson–Crick base pairing between the sgRNA and the target DNA sequence, thereby generating site-specific double-strand (ds) DNA breaks. This system can be easily reprogrammed to modify virtually any desired genomic sequence by recoding the specificity-determining sequence of the sgRNA. Recoding constraints are minimal and simply require that the target sequence is unique, compared with the rest of the genome, and is located just upstream of a protospacer adjacent motif (PAM sequence), typically consisting of the three nucleotide motif NGG (20, 21). Both the minimal constraints and ease of use make CRISPR/Cas9 a powerful tool for genome-engineering applications.

Importantly, CRISPR-mediated genome engineering of organisms has been achieved in a variety of different ways. For example, direct injection of in vitro-purified sgRNAs combined with either purified Cas9 RNA, recombinant Cas9 protein, or even with Cas9 expression plasmids, have all been successful for a variety of organisms (29, 32, 35⇓⇓⇓⇓–40), including mosquitoes (30⇓⇓⇓–34). However, the rate of mutagenesis and lethality have varied widely, both within and among these different studies, and such discrepancies are likely due to either the methods used to introduce the editing components, the variability in sgRNA functionality against target sites, or even unavoidable variability from manual injection of the components (41). To overcome these significant limitations, previous studies in other organisms have shown that stably expressing a transgenic provision of Cas9 in the germline can decrease toxicity to injected embryos, increase the rates of mutagenesis generated by both nonhomologous end joining (NHEJ) and homology-directed repair (HDR), and can also increase the rates of germline transmission of the disrupted allele to offspring (27, 41⇓⇓⇓⇓–46).

Germline expression of Cas9 is also essential for developing innovative technologies that rely on “gene drive,” which is a novel strategy proposed for the control of vector-borne diseases by rapidly spreading alleles in a population through super-Mendelian inheritance (47, 48). In mosquitoes, gene drives could potentially be used to rapidly disseminate a genetic payload that reduces pathogen transmission throughout a population, thereby suppressing vector competence and human disease transmission. Other possible applications include the suppression of the population by spreading alleles that impair fertility or viability (49, 50). A CRISPR homing-based gene-drive element consists of only a few components, such as an sgRNA and a germline-expressed Cas9 endonuclease that is positioned opposite its target site in the genome. The drive encodes the editing machinery (i.e., Cas9 and sgRNA), allowing it to cut the opposite allele and copy itself into this disrupted allele via HDR, thereby converting a heterozygote into a homozygote, enabling rapid invasion of the drive into a population. In fact, Cas9 has been used to develop highly promising homing-based gene drives in a number of organisms, including yeast (51), Drosophila (52), and Anopheles mosquitoes (52, 53); however, such a system has yet to be developed in Ae. aegypti.

We therefore aimed to develop a CRISPR/Cas9 transgenic system that would enable more robust and widespread Ae. aegypti genome-engineering applications, including disrupting genes important for vector competence, while also laying the foundation for the future development of gene drives. To do so, we utilized several previously described transcriptional regulatory elements (54, 55), many of which are active in the germline (56), to drive expression of Streptococcus pyogenes Cas9 in Ae. aegypti. In total, we developed six independent Cas9 expressing strains, and herein characterize and demonstrate their effectiveness for gene disruption and gene insertion via HDR. We demonstrate the efficiency of using these strains by disrupting six genes in Ae. aegypti, giving rise to severe phenotypes, such as having an extra eye (triple eyes), an extra maxillary palp (triple maxillary palps), a nonfunctional curved proboscis, malformed wings, eye pigmentation deficiencies, and pronounced whole-animal cuticle coloration defects. Furthermore, we also demonstrate the ease of generating double- and triple-mutant strains simultaneously from a single injection, a technique that will facilitate the ease of gene-function study in this nonmodel organism. Overall, gene-disruption efficiencies, survival rates, germline-transmission frequencies, and HDR rates were all significantly improved using the Cas9 strains we develop here. These strains should be highly valuable for facilitating the development of innovative control methods in this organism in the future.

Discussion Previous studies have demonstrated effective CRISPR/Cas9 genome editing in the mosquito Ae. aegypti (30⇓⇓⇓–34); however, these studies utilized a nontransgenic source of Cas9, limiting both survivorship and editing efficiencies. To overcome these previous limitations, and to reduce the complexity of injecting multiple components (i.e., Cas9 and sgRNA), herein we developed a simplified transgenic Cas9 expression system in Ae. aegypti, similarly to what is routinely used in other organisms such as D. melanogaster (27, 41, 42). Importantly, to achieve highly specific and consistent genome modifications, we demonstrate that embryos from these Cas9 strains need only be injected with easy-to-make sgRNAs. By using these Cas9 strains, we disrupted multiple genes that were either homozygous viable (kh, white, yellow, and ebony), or homozygous lethal (deformed, sine oculis, vestigial), resulting in dramatic phenotypes affecting viability, vision, flight, and blood feeding, and therefore may be useful for developing control strategies and genetic sexing techniques in the future. For example, in a yellow mutant background, the endogenous gene encoding yellow could be linked to the male-determining locus (32), using CRISPR-mediated HDR, to generate a robust genetic sexing system by which male embryos/larvae/adults would be dark and female embryos/larvae/adults would be yellow. An appealing advantage of our Cas9 transgenic system is the ability to efficiently disrupt multiple genes simultaneously. We have demonstrated that we can efficiently generate large deletions, or even double- (yellow-white; ebony-white; yellow-ebony), or triple- (yellow-ebony-white) mutants. Importantly, these multimutants can rapidly be generated in a single-step approach by injecting multiple sgRNAs into the embryos of the transgenic Cas9 strains, significantly reducing downstream efforts. This rapid multiplex gene knockout approach will be instrumental for dissecting gene networks in this nonmodel organism. While the Cas9 strains were generated in the Liverpool genetic background, these strains can be introgressed into other genetic backgrounds if desired. The germline Cas9 strains developed here may also bring us one step closer to engineering an effective CRISPR/Cas9-homing–based gene drive system (47, 48) in this organism. Homing-based drive systems rely on HDR to convert heterozygous alleles into homozygous alleles in the germline, and have recently been successfully engineered in two Anopheline mosquito species (51, 69). While these studies were fruitful at significantly biasing rates of Mendelian inheritance rates of the drive containing alleles, they were severely limited by the rapid evolution of resistance alleles generated by NHEJ, and are therefore not predicted to spread into diverse wild populations (70). It was hypothesized that these resistance alleles formed due to high levels of maternal deposition of Cas9 in the embryo, and by restricting Cas9 expression to the germline may subsequently increase rates of HDR and reduce rates of NHEJ. In addition to restricting expression to the germline, multiplexing of sgRNAs in the drive, and designing the drive to target a critical gene have also been proposed as innovative strategies to increase rates of HDR and reduce resistance caused by NHEJ (47, 48, 70); however, these hypotheses remain to be demonstrated. Notwithstanding, it will be interesting to determine if our Ae. aegypti Cas9 strains, each with varying expression in the germline, will be effective in a gene drive system designed for Ae. aegypti. This would be straightforward to test in a molecularly confined safe split-gene drive design where the Cas9 and the drive are positioned at different genomic loci. In this split-design, the Cas9 strains we developed can be directly tested without further modification by simply genetically crossing with a split gRNA-drive component and measuring rates of inheritance (49, 71⇓–73). While the CRISPR/Cas9 transgenic system developed here is quite effective, it would be useful to have the ability to supply the sgRNAs transgenically. In D. melanogaster, polymerase-3 promoters have been utilized to express sgRNAs, and through genetic crosses with Cas9 strains mutation efficiency could be increased up to 100% (42, 43). Furthermore, it should be noted, that a slight disadvantage of these Cas9 strains for some groups may result from the fact that these strains were generated in the Liverpool genetic background, which restricts genome modifications to only this background. However, this can easily be overcome by simply using genetics to introgress these promoter-Cas9 transgenes into other genetic backgrounds, or by generating new transgenic strains in the desired background using the plasmids generated here. Overall, our results demonstrate that our simplified transgenic Cas9 system has improved capacity to rapidly induce highly efficient and specific targeted genome modifications, including gene disruptions, deletions, and insertions. Given their high efficiencies, these Cas9 strains can be used to quickly generate genome modifications allowing for high-throughput gene targeting, thereby accelerating comprehensive functional annotation of the Ae. aegypti genome.

Materials and Methods Insect Rearing. Mosquitoes used in all experiments were derived from of the Ae. aegypti Liverpool strain, which was the source strain for the reference genome sequence (58). Generation of Ae. aegypti Cas9 Transgenic Lines. Transgenic Ae. aegypti Cas9 mosquitoes were created by injecting 0- to 1-h-old preblastoderm-stage embryos with a mixture of piggybac vector containing the Cas9 expressing plasmid designed above (200 ng/μL) and a source of piggyBac transposase [phsp-Pbac, (200 ng/μL)] (74⇓–76). Characterization of AAEL010097-Cas9 Insertion Site. To characterize the Cas9 insertion site for AAEL010097-Cas9, we utilized a previously described inverse PCR protocol (77). CRISPR Mediated Microinjections. Embryonic collection and CRISPR microinjections were performed following previously established procedures (30, 78).

Acknowledgments We thank Robert Harrell for providing the pBac-3xP3-dsRed and phsp-Pbac plasmids; Kate M. O’Connor-Giles for providing the p3xP3-EGFP/vasa-3xFLAG-NLS-Cas9-NLS plasmid and sequence; Alexander Knyshov for assisting with SEM imaging; and Anthony A. James for reading over the manuscript and providing insightful comments and edits. This work was supported by a private donation from https://www.maxmind.com/ (to O.S.A.); US NIH K22 Grant 5K22AI113060 (to O.S.A.); NIH R21 Grant 1R21AI123937 (to O.S.A.); NIH R21 Grant 1R21AI115271 (to B.J.W.); and Defense Advanced Research Project Agency Safe Genes Program Grant HR0011-17-2-0047 (to O.S.A.).

Footnotes Author contributions: M.L. and O.S.A. designed research; M.L., M.B., T.Y., and C.S.B. performed research; M.L. and O.S.A. analyzed data; and M.L., B.J.W., and O.S.A. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Data deposition: All Cas9 plasmids and sequence maps have been deposited and made available for order and download at www.addgene.org/ [Addgene IDs 100580 (AAEL006511-Cas9), 100581 (AAEL003877-Cas9), 100608 (AAEL005635-Cas9), 100705 (AAEL007097-Cas9), 100706 (AAEL007584-Cas9) and 100707 (AAEL010097-Cas9)].

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1711538114/-/DCSupplemental.