The urban-adapted underground mosquito, Culex molestus, maintains exogenously influenced circadian rhythms despite an absence of photoperiodically induced dormancy

Abstract

In temperate climates, the mosquito Culex molestus lives almost exclusively in urban underground locations such as flooded basements, sewer systems and subway tunnels. Unlike most other mosquito taxa found at higher latitudes, Cx. molestus remains active and continues to breed throughout the winter. This is attributable to year-round above freezing temperatures in its preferred underground habitats combined with an inability to enter a state of arrested development ( diapause) in response to shortening photoperiods in autumn. Prior studies have shown that the genes associated with circadian rhythms (i.e. clock genes) also influence the photoperiodic induction of diapause in the closely related mosquito, Cx. pipiens. These results suggest that molecular changes in one or more clock genes could contribute to the absence of photoperiodically induced diapause in Cx. molestus. As Cx. molestus predominantly inhabits underground locations generally devoid of a predictable photoperiod, such mutations may not be removed by purifying selection as they would have minimal fitness consequences. To examine the possibility that Cx. molestus- specific genetic changes in one or more clock genes correlate with its inability to enter a photoperiodically induced dormant state, we first used genomic data to search for inactivating mutations or other structural variants in genes known to influence circadian rhythms in Diptera ( flies). We further investigated non-synonymous, derived genetic divergence in this same class of genes. Next, we generated transcriptome data from multiple life-stages of Cx. molestus to survey binary expression of annotated clock genes throughout this mosquitos lifecycle. Finally, we carried out experimental studies to assess the extent to which Cx. molestus retains exogenously influenced circadian rhythms, and whether it harbors any tendencies towards dormancy when exposed to a shortened photoperiod. Our results indicate that the gene Helicase domino (dom) has a nine-nucleotide, in-frame deletion specific to Cx. molestus. Previous work has shown that splice variants in this gene influence circadian behavior in Drosophila melanogaster. We also find derived, non-synonymous single nucleotide polymorphisms (SNPs) in eight genes that may also affect circadian rhythms and/or diapause induction in Cx. molestus. Four other circadian genes were found to have no quantifiable expression during any examined life stage, suggesting potential regulatory variation. Our experimental results confirm that Cx. molestus retains exogenously-influenced circadian rhythms but is not induced to enter a dormant state by a shortened photoperiod. Collectively, these findings show that the distinct, but potentially molecularly interconnected life-history traits of diapause induction and circadian rhythms are decoupled in Cx. molestus and suggest that this taxon may be a valuable tool for exploring exogenously influenced phenotypes in mosquitoes more broadly.