Dengue and chikungunya virus loads in the mosquito Aedes aegypti are determined by distinct genetic architectures

Aedes aegypti is the primary vector of the arboviruses dengue (DENV) and chikungunya (CHIKV). These viruses exhibit key differences in their vector interactions, the latter moving more quicky through the mosquito and triggering fewer standard antiviral pathways. As the global footprint of CHIKV continues to expand, we seek to better understand the mosquito's natural response to CHIKV-both to compare it to DENV:vector coevolutionary history and to identify potential targets in the mosquito for genetic modification. We used a modified full-sibling design to estimate the contribution of mosquito genetic variation to viral loads of both DENV and CHIKV. Heritabilities were significant, but higher for DENV (40%) than CHIKV (18%). Interestingly, there was no genetic correlation between DENV and CHIKV loads between siblings. These data suggest Ae. aegypti mosquitoes respond to the two viruses using distinct genetic mechanisms. We also examined genome-wide patterns of gene expression between High and Low CHIKV families representing the phenotypic extremes of viral load. Using RNAseq, we identified only two loci that consistently differentiated High and Low families: a long non-coding RNA that has been identified in mosquito screens post-infection and a distant member of a family of Salivary Gland Specific (SGS) genes. Interestingly, the latter gene is also associated with horizontal gene transfer between mosquitoes and the endosymbiotic bacterium Wolbachia. This work is the first to link the SGS gene to a mosquito phenotype. Understanding the molecular details of how this gene contributes to viral control in mosquitoes may, therefore, also shed light on its role in Wolbachia. Author summaryThe virus chikungunya (CHIKV) that causes long term arthritis symptoms in humans is transmitted to through the bite of the Aedes aegypti mosquito. CHIKV, for which there is no vaccine, is becoming increasingly common across the globe. We therefore need to understand the mosquito's own ability to control CHIKV, as we may use that knowledge to create resistant mosquitoes through genetic modification. We show that the mosquito has very little ability to respond genetically to CHIKV, indicating low potential for the mosquito to evolve resistance. We also found that the genetic basis of CHIKV viral loads appears distinct from dengue, another common virus. As such, any strategy for engineering virus-resistant mosquitoes may need to be virus-specific or focus on the few overlapping genes in the mosquito response. Last, when we examined the mosquito genes whose expression differed between high and low-load virus lineages, we discovered a gene that was highly expressed in low-load families and therefore, potentially acting as a virus controller. Interestingly, a homolog of this gene has been discovered in the genome of the Wolbachia endosymbiont, itself known to limit virus replication inside its insect hosts. The functional importance of this homolog in virus control should therefore be explored in both mosquitoes and Wolbachia.