Population genomics of Culiseta melanura, the principal vector of Eastern equine encephalitis virus in the United States

Abstract

Background: Eastern Equine Encephalitis (EEE) (Togaviridae, Alphavirus) is a highly pathogenic mosquito-borne arbovirus that circulates in an enzootic cycle involving Culiseta melanura mosquitoes and wild Passeriformes birds in freshwater swamp habitats. Recently, the northeastern United States has experienced an intensification of virus activity with increased human involvement and northward expansion into new regions. In addition to its principal role in enzootic transmission of EEE virus among avian hosts, recent studies on the blood-feeding behavior of Cs. melanura throughout its geographic range suggest that this mosquito may also be involved in epizootic / epidemic transmission to equines and humans in certain locales. Variations in blood feeding behavior may be a function of host availability, environmental factors, and/or underlying genetic differences among regional populations. Despite the importance of Cs. melanura in transmission and maintenance of EEE virus, the genetics of this species remains largely unexplored.

Methodology and principle findings: To investigate the occurrence of genetic variation in Cs. melanura, the genome of this mosquito vector was sequenced resulting in a draft genome assembly of 1.28 gigabases with a contig N50 of 93.36 kilobases. Populations of Cs. melanura from 10 EEE virus foci in the eastern North America were genotyped with double-digest RAD-seq. Following alignment of reads to the reference genome, variant calling, and filtering, 40,384 SNPs were retained for downstream analyses. Subsequent analyses revealed genetic differentiation between northern and southern populations of this mosquito species. Moreover, limited fine-scale population structure was detected throughout northeastern North America, suggesting local differentiation of populations but also a history of ancestral polymorphism or contemporary gene flow. Additionally, a genetically distinct cluster was identified predominantly at two northern sites.

Conclusion and significance: This study elucidates the first evidence of fine-scale population structure in Cs. melanura throughout its eastern range and detects evidence of gene flow between populations in northeastern North America. This investigation provides the groundwork for examining the consequences of genetic variations in the populations of this mosquito species that could influence vector-host interactions and the risk of human and equine infection with EEE virus.

Fig 1. Culiseta melanura and two trapping methods used in this study.

A) An engorged Cs. melanura. B) Collection of adult mosquitoes from a resting box. C) A CO₂ baited CDC light trap used for collecting adult mosquitoes.

Fig 2. Culiseta melanura sampling sites from EEE virus foci across eastern North America.

Additional information for each sampling site is given in Table 1. Map created using the package mapplots.

Fig 3. The BUSCO results for five publicly available Dipteran vector genomes, and the Culiseta melanura draft genome polished with Arrow.

The draft genome generated in this study for Culiseta melanura has similar levels of completeness to several other vector genomes, with at least partial evidence of 88.7% single copy orthologs.

Fig 4. Results of SNMF genetic clustering analysis on 119 individuals from 10 populations of Culiseta melanura.

(A) By individual, considering K = 2, K = 3, and K = 4. K = 2 had the lowest cross entropy, but both K = 3 and K = 4 were similar to K = 2 in cross entropy. (B) By population, with mean proportion of ancestry to each cluster for K = 4 represented as pie slices.

Fig 5. The fineRADstructure coancestry matrix and MAP tree.

This analysis resolves the same genetic clusters as in other analyses and identifies fine-scale coancestry among populations and individuals. Cluster A was removed for this analysis to visualize finer scale structure between remaining clusters and populations (see Fig G in S1 File for the coanestry matrix containing Cluster A). A) In the coancestry matrix, individual coancestry is shown above the diagonal, while mean terminal clade coancestry is shown below the diagonal. Colored boxes over the MAP tree corresponds to the clusters in Fig 4 for K = 4. Posterior probability shown on branches. B) Results of the SNMF analysis for K = 4, ordered by coancestry.

Fig 6. A genotype composition plot showing the posterior membership probability of mosquitoes following DAPC analysis.

Individuals are grouped by the population. The majority of individuals are appropriately assigned to their respective populations.