Design and analysis of CRISPR-based underdominance toxin-antidote gene drives

Abstract

CRISPR gene drive systems offer a mechanism for transmitting a desirable transgene throughout a population for purposes ranging from vector-borne disease control to invasive species suppression. In this simulation study, we assess the performance of several CRISPR-based underdominance gene drive constructs employing toxin-antidote principles. These drives disrupt the wild-type version of an essential gene using a CRISPR nuclease (the toxin) while simultaneously carrying a recoded version of the gene (the antidote). Drives of this nature allow for releases that could be potentially confined to a desired geographic location. This is because such drives have a nonzero invasion threshold frequency, referring to the critical frequency required for the drive to spread through the population. We model drives which target essential genes that are either haplosufficient or haplolethal, using nuclease promoters with expression restricted to the germline, promoters that additionally result in cleavage activity in the early embryo from maternal deposition, and promoters that have ubiquitous somatic expression. We also study several possible drive architectures, considering both "samesite" and "distant-site" systems, as well as several reciprocally targeting drives. Together, these drive variants provide a wide range of invasion threshold frequencies and options for both population modification and suppression. Our results suggest that CRISPR toxin-antidote underdominance drive systems could allow for the design of highly flexible and potentially confinable gene drive strategies.

Performance analysis of novel toxin-antidote CRISPR gene drive systems

“…Alternative configurations are also possible that would change the dynamics of TA drives [53]. For example, to achieve a greater degree of local confinement (at the costs of greater required release sizes, as is usually the case with such systems), a 2-locus 2-toxin-antidote system [45,46,52] could be engineered by using two TARE drives, each providing rescue for the target gene of the other system [53]. Such a system could presumably be engineered quite Fig.…”

“…Alternatives would be to use TADE drives with germline-only Cas9 expression, different nucleases in the TARE components, or expression of the tethered gRNAs with a germline promoter while retaining a Cas9 promoter that allows for cleavage in both the germline and early embryo. Highly localized suppression could also be obtained with a 2-locus TADE system, with one of the TADE alleles disrupting a sex-specific fertility gene, as in TADE suppression [53]. Indeed, a standard TADE suppression system with a promoter that has high embryo activity could itself be a feasible method for local population suppression, with the level of embryo activity allowing a variable introduction threshold, even without fitness costs [53].…”

“…Highly localized suppression could also be obtained with a 2-locus TADE system, with one of the TADE alleles disrupting a sex-specific fertility gene, as in TADE suppression [53]. Indeed, a standard TADE suppression system with a promoter that has high embryo activity could itself be a feasible method for local population suppression, with the level of embryo activity allowing a variable introduction threshold, even without fitness costs [53]. Furthermore, either of these TADE-based methods could be used for population modification if not located in a fertility gene, and a TARE/ClvR drive with a target that is not fully haplosufficient will also have a nonzero introduction threshold even without a fitness cost [40,53].…”

“…Indeed, a standard TADE suppression system with a promoter that has high embryo activity could itself be a feasible method for local population suppression, with the level of embryo activity allowing a variable introduction threshold, even without fitness costs [53]. Furthermore, either of these TADE-based methods could be used for population modification if not located in a fertility gene, and a TARE/ClvR drive with a target that is not fully haplosufficient will also have a nonzero introduction threshold even without a fitness cost [40,53]. Each of these systems should be possible to engineer with current techniques and target genes that are already characterized.…”

“…However, other genes in which disrupted alleles are recessive lethal or sterile (including sex-specific) could also be used. TADE targets should be haplolethal, but some level of haploinsufficiency will usually also be tolerable [53]. On the other hand, TADS targets are highly specific, and this will likely be the limiting factor in the engineering of these systems.…”