Targeting mosquito X-chromosomes reveals complex transmission dynamics of sex ratio distorting gene drives.

Engineered sex ratio distorters (SRDs) have been proposed as a powerful component of genetic control strategies designed to suppress harmful insect pests. Two types of CRISPR-based SRD mechanisms have been proposed: X-shredding, which eliminates X-bearing sperm, and X-poisoning, which eliminates females inheriting disrupted X-chromosomes. These differences can have a profound impact on the population dynamics of SRDs when linked to the Y-chromosome: an X-shredder is invasive, constituting a classical meiotic Y-drive, whereas X-poisoning is self-limiting, unable to invade but also insulated from selection. Here, we establish X-poisoning strains in the malaria vector Anopheles gambiae targeting three X-linked genes during spermatogenesis, resulting in male bias. We find that sex distortion is primarily driven by a loss of X-bearing sperm, with limited evidence for postzygotic lethality of female progeny. By leveraging a Drosophila melanogaster model, we show unambiguously that engineered SRD traits can operate differently in these two insects. Unlike X-shredding, X-poisoning could theoretically operate at early stages of spermatogenesis. We therefore explore premeiotic Cas9 expression to target the mosquito X-chromosome. We find that, by pre-empting the onset of meiotic sex chromosome inactivation, this approach may enable the development of Y-linked SRDs if mutagenesis of spermatogenesis-essential genes is functionally balanced.

Neofunctionalization driven by positive selection led to the retention of the loqs2 gene encoding an Aedes specific dsRNA binding protein.

BACKGROUND: Mosquito borne viruses, such as dengue, Zika, yellow fever and Chikungunya, cause millions of infections every year. These viruses are mostly transmitted by two urban-adapted mosquito species, Aedes aegypti and Aedes albopictus. Although mechanistic understanding remains largely unknown, Aedes mosquitoes may have unique adaptations that lower the impact of viral infection. Recently, we reported the identification of an Aedes specific double-stranded RNA binding protein (dsRBP), named Loqs2, that is involved in the control of infection by dengue and Zika viruses in mosquitoes. Preliminary analyses suggested that the loqs2 gene is a paralog of loquacious (loqs) and r2d2, two co-factors of the RNA interference (RNAi) pathway, a major antiviral mechanism in insects. RESULTS: Here we analyzed the origin and evolution of loqs2. Our data suggest that loqs2 originated from two independent duplications of the first double-stranded RNA binding domain of loqs that occurred before the origin of the Aedes Stegomyia subgenus, around 31 million years ago. We show that the loqs2 gene is evolving under relaxed purifying selection at a faster pace than loqs, with evidence of neofunctionalization driven by positive selection. Accordingly, we observed that Loqs2 is localized mainly in the nucleus, different from R2D2 and both isoforms of Loqs that are cytoplasmic. In contrast to r2d2 and loqs, loqs2 expression is stage- and tissue-specific, restricted mostly to reproductive tissues in adult Ae. aegypti and Ae. albopictus. Transgenic mosquitoes engineered to express loqs2 ubiquitously undergo developmental arrest at larval stages that correlates with massive dysregulation of gene expression without major effects on microRNAs or other endogenous small RNAs, classically associated with RNA interference. CONCLUSIONS: Our results uncover the peculiar origin and neofunctionalization of loqs2 driven by positive selection. This study shows an example of unique adaptations in Aedes mosquitoes that could ultimately help explain their effectiveness as virus vectors.