Introduction

West Nile virus (WNV; Flavivirus; Flaviviridae) is a significant global threat, causing disease in humans and animals worldwide. In nature, WNV is transmitted through mosquito vectors to birds, its main reservoirs, and occasionally to ‘dead-end’ hosts like horses and humans. WNV is underdiagnosed and associated with harmful health, social, and economic consequences [1]. During recent decades, WNV caused different outbreaks in Europe, with a large increase in autochthonous infections occurring in 2018 [2]. In Spain, the local circulation of WNV has been reported since 2003 [3, 4], and the first WNV-associated disease case in humans was reported in 2004. Despite endemic circulation supported by long-term studies using wild birds [5] and feral horses [3], human cases were sporadic in subsequent years (2 cases in 2010 and 3 in 2016). In 2020, a significant outbreak occurred with 77 human cases and 8 fatalities, most of them in Seville province [6], southern Spain. Human infections have also been documented during 2024 [7], supporting the necessity to investigate the factors affecting WNV epidemiology in the area.

Mosquitoes of the Culex (Cx.) genus are the primary WNV vectors, especially the species within the Culex pipiens complex [1, 8]. Yet, most evidence supports the role of other species in the natural circulation of the virus. This is the case of Culex perexiguus which is considered a major vector in southern Spain [9]. WNV has been repeatedly identify in Cx. perexiguus pools captured in the area [10, 11] and the abundance of Cx. perexiguus has been positively associated with the prevalence of WNV in wild bird populations [12]. During the 2020 outbreak, most WNV infections were detected in Cx. perexiguus (33 out of 419 mosquito pools tested using Real-Time RT–PCR; 7.88%), with a much lower incidence in Cx. pipiens (1 out of 152 mosquito pools tested using Real-Time RT–PCR; 0.66%) and a total absence in Cx. modestus (75 mosquito pools tested). Culex perexiguus was identified as the primary vector maintaining WNV enzootic circulation in natural and agricultural areas, while Cx. pipiens potentially aided in the WNV transmission to humans in urban areas [8, see also 9].

Growing evidence highlights mosquito microbiota’s significant role in affecting the different components of the vectorial capacity [13] and in disease transmission control [14, 15]. Mosquito microbiota of wild mosquitoes largely varies between species and areas [16, 17], and its composition may have profound effects on their vector competence and the dynamics of mosquito-borne pathogen transmission. Mosquito microbiota may directly affect the development of pathogens in mosquitoes by the competition with the pathogens for resources [18] and the secretion of anti-pathogen molecules [19], the hindrance of necessary interactions between the pathogen and vector epithelium [20] or the formation of the peritrophic matrix around the blood bolus after blood feeding, which is a barrier against pathogens [19], among other mechanisms. In addition, mosquito microbiota may modulate the immunological responses of mosquitoes finally determining pathogen development [14].

According to studies conducted under controlled conditions, WNV infection has been negatively correlated to Wolbachia relative abundance in Culex mosquitoes [21]. On the contrary, WNV infection positively correlated with the relative abundance of bacteria of the genera Enterobacter and Serratia and to bacterial diversity [21, 22]. However, despite the necessity to identify natural associations between mosquitoes and their microbiota and their effects on pathogen development, studies in the field using Culex mosquitoes are scarce and mainly focus on bacteria of the genus Wolbachia, showing no clear association patterns with WNV infections [21]. Furthermore, the microbiota of Cx. perexiguus and its relationship with pathogens such as WNV is virtually unknown.

Here, we characterize the bacterial community of the Cx. perexiguus microbiota and investigate the potential relationship between its bacterial diversity and composition and the WNV infection status. We used field-collected Cx. perexiguus mosquito females from a locality of the Seville province captured in August 2020, during the WNV outbreak in southern Spain, to identify the ecological interactions occurring under natural conditions.

Material and methods

Results

Overall, 1,000 Cx. perexiguus females grouped in 100 pools of 10 mosquitoes each were molecularly tested for the presence of WNV. Of them, 19 mosquito pools were positive (19%), representing an estimated prevalence of 0.02 (Confidence limits 2.5%: 0.01; 97.5%: 0.03; SE: 0.0047). These 19 WNV-positive mosquito pools together with 21 WNV-negative mosquito pools were used for the subsequent analyses.

Characterization of Cx. perexiguus microbiota

We analyzed the microbiota of 40 Cx. perexiguus pools obtaining a total of 6,453,096 reads, ranging from 46,513 to 419,497 (average = 161,327.4). The rarefaction curves reached saturation for all samples, indicating that our sampling effort captured most of the bacterial diversity (S1 Fig). We identified a relatively rich bacterial community in Cx. perexiguus pools with a total of 4,434 ASVs belonging to 18 phyla, 33 classes, 85 orders, 168 families, and 410 genera.

The microbiota of Cx. perexiguus mosquito pools was dominated by the phylum Proteobacteria. The Burkholderiaceae family was the most abundant (average relative abundance = 0.676, SD = 0.150) except in two samples. In these two samples, the Erwiniaceae family was the most abundant (average relative abundance = 0.085, SD = 0.137). Following these two families, the next most abundant families in mosquito pools were Rhodobacteraceae, Cellulomonadaceae, Micrococcaceae, Microbacteriaceae, and Enterobacteriaceae (Fig 1A; S2 Fig). Burkholderia (Burkholderiaceae) was the most abundant genus in mosquito pools (average relative abundance = 0.696, SD = 0.152), followed by Erwinia and Pantoea (Erwiniaceae) with a relative abundance of 0.050 (SD = 0.091) and 0.037 (SD = 0.093), respectively. The genera Wolbachia, Serratia, and Enterobacter were identified in 4, 4 and 33 mosquito pools, with a maximum relative abundance of <0.001, 0.002, and 0.004, respectively.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1.

A: Relative abundance of the 12 most abundant taxa at family level in negative and positive Cx. perexiguus mosquito pools for West Nile virus (WNV). B: Distribution of the Shannon diversity index (alpha-diversity) at the family level of Cx. perexiguus mosquito pools according to the infection status by WNV. The vertical lines go from the lower and upper quartiles to the minimum or maximum, respectively, and the horizontal line represents the median. C: Principal Co-ordinates Analysis (PCoA) for Bray-Curtis distance matrix (relative abundance) by WNV infection at family level. Percentages shown in MDS1 and MDS2 axes refer to the percentage of variation explained by each of the two selected main coordinate axes.

https://doi.org/10.1371/journal.pone.0314001.g001

Discussion

Mosquito microbiota has been highlighted as a major driver of vector competence by affecting the development of pathogens directly and indirectly, supporting the importance of studying their composition in natural mosquito populations. To our knowledge, here we characterized for the first time the microbiota composition of field-collected Cx. perexiguus mosquito females captured in an area of endemic circulation of vector-borne pathogens, including the zoonotic WNV.

Using a metabarcoding approach, we identified a relatively rich bacterial community in female Cx. perexiguus pools representing 410 different genera. Among them, bacteria of the genera Burkholderia (Burkholderiaceae), Erwinia and Pantoea (Erwiniaceae) dominated, in terms of relative abundance, the microbiota of Cx. perexiguus females. These genera are commonly found in Aedes spp. and Anopheles spp. mosquitoes, both in the wild and laboratory colonies [38]. However, they are rare in the microbiota of Culex mosquitoes [39]. In addition, Wolbachia showed a low prevalence in the studied mosquito population. While Cx. pipiens has a very high prevalence of Wolbachia in different populations [40], other related species of the Culex genus may show an opposite pattern [40, 41]. This is the case of species including Culex torrentum [41], Culex restuans [42] or Cx. perexiguus, as shown here. These results suggest that the microbiota composition of mosquitoes may largely differ between species, including those breeding in the same area [41, 42]. However, our results are limited by the fact that we only sampled mosquitoes a single population and time point, thus further studies are necessary to identify the natural variation of the microbiota composition of this species in a broader study area.

Mosquito microbiota has been highlighted as a major driver of vector competence by affecting the development of pathogens both directly and indirectly. Different field and laboratory studies on Culex spp. have associated WNV infection with changes in the microbiota community composition of mosquitoes, including changes in the abundance of the most prevalent microbiota genera [21]. For instance, Zink et al. [22] exposed Cx. pipiens mosquitoes to WNV and observed that WNV-exposed and infected mosquitoes showed lower relative abundance of Wolbachia but higher relative abundance of bacteria of the genera Enterobacter and Serratia and bacterial diversity. In this study, authors also observed an up-regulation of numerous genes related to the mosquito innate immune system, suggesting an indirect interaction between WNV and the mosquito microbiome through the immune response. On the other hand, Wolbachia is known to block replication of WNV and other viruses such as dengue, Chikungunya, and West Nile virus [18, 43, 44], while its absence may affect the susceptibility to pathogen infection by mosquitoes. For example, Wolbachia-free Cx. modestus mosquitoes are major vectors of WNV in endemic regions, relegating sympatric Cx. pipiens to a secondary role [41, 45]. Likewise, according to our results, Cx. perexiguus, the primary vector of WNV during the 2020 Spanish outbreak [9], exhibits a low prevalence of Wolbachia potentially explaining, at least in part, the role of this species in the transmission of WNV in the area. In addition, the low presence of this taxa in Cx. perexiguus could potentially explain the absence of differences in the microbiota composition of mosquitoes according to their WNV infection status. Further studies using Cx. perexiguus mosquitoes reared under controlled conditions could help to assess the link between mosquito microbiota and WNV-infection in wild mosquitoes in the future. These studies may consider using Wolbachia bacteria (strain wMel), which reduce the lifespan and partially block viral infections in other mosquito species such as Aedes aegypti [18, 46].

Conclusion

This study provides the first information about the microbiome composition of Cx. perexiguus, a major WNV vector in southern Spain, during the largest WNV outbreak ever recorded in the country. The low prevalence of Wolbachia in Cx. perexiguus mosquitoes could potentially explain, at least in part, the relevance of this mosquito species in the WNV transmission and makes necessary experimental tests to identify the impact of mosquito microbiota composition on Cx. perexiguus vector capacity. However, in order to conduct these studies, previous information about the composition of the microbiota of mosquitoes in the wild is essential.

Supporting information

Acknowledgments

We thank the contribution of Álvaro Solís during the fieldwork. Two anonymous reviewers provided valuable comments on a previous version of the manuscript