Search

The variation in viral gene frequencies remains a largely unexplored aspect of within-host genetics. This phenomenon is often considered to be specific to multipartite viruses. Multipartite viruses have segmented genomes, but in contrast to segmented viruses, their segments are each encapsidated alone in a virion. A main hypothesis explaining the evolution of multipartism is that, compared to segmented viruses, it facilitates the regulation of segment abundancy, and the genes the segments carry, within a host. These differences in gene frequencies could allow for expression regulation. Here, we show that wild populations of a segmented virus, bluetongue virus (BTV), also present unequal segment frequencies. BTV cycles between ruminants and Culicoides biting midges. As expected from a role in expression regulation, segment frequencies tended to show specific values that differed between ruminants and midges. Our results expand previous knowledge on gene frequency variation and call for studies on its role and conservation beyond multipartite viruses., Genome segmentation is mainly thought to facilitate reassortment. Here, we show that segmentation can also allow differences in segment abundance in populations of bluetongue virus (BTV). BTV has a genome consisting in 10 segments, and its cycle primarily involves periodic alternation between ruminants and Culicoides biting midges. We have developed a reverse transcription-quantitative PCR (RT-qPCR) approach to quantify each segment in wild BTV populations sampled in both ruminants and midges during an epizootic. Segment frequencies deviated from equimolarity in all hosts. Interestingly, segment frequencies were reproducible and distinct between ruminants and biting midges. Beyond a putative regulatory role in virus expression, this phenomenon could lead to different evolution rates between segments. IMPORTANCE The variation in viral gene frequencies remains a largely unexplored aspect of within-host genetics. This phenomenon is often considered to be specific to multipartite viruses. Multipartite viruses have segmented genomes, but in contrast to segmented viruses, their segments are each encapsidated alone in a virion. A main hypothesis explaining the evolution of multipartism is that, compared to segmented viruses, it facilitates the regulation of segment abundancy, and the genes the segments carry, within a host. These differences in gene frequencies could allow for expression regulation. Here, we show that wild populations of a segmented virus, bluetongue virus (BTV), also present unequal segment frequencies. BTV cycles between ruminants and Culicoides biting midges. As expected from a role in expression regulation, segment frequencies tended to show specific values that differed between ruminants and midges. Our results expand previous knowledge on gene frequency variation and call for studies on its role and conservation beyond multipartite viruses.

Depuis la réémergence du sérotype 8 de la fièvre catarrhale ovine (FCO) en France continentale en 2015, une surveillance entomologique des populations de Culicoides a été redéployée à l'échelle de la métropole entre novembre 2015 et avril 2018, complétant le réseau de piégeages déjà en place en Corse depuis 2002. Pour la saison hivernale 2017-2018, cette surveillance aura permis de déclarer 16 départements en zone saisonnièrement indemne sur des durées variables. Suite à l'arrêt de la surveillance en France continentale en avril 2018, le réseau a continué de fonctionner en Corse et au port de Sète. De plus, des missions ponctuelles de surveillance de l'espèce d'origine afrotropicale Culicoides imicola, vectrice majeure du virus de la FCO, ont eu lieu en alternance une année sur deux, dans les Pyrénées-Orientales, où l'espèce n'est plus capturée depuis 2012, et dans le Var et les Alpes-Maritimes où l'on observe une stabilité de la distribution géographique et des abondances des populations depuis l'installation de l'espèce détectée en 2003. Les données d'abondance et de distribution relevée sur le long terme dans cette zone suggèrent que cette espèce exotique pourrait avoir atteint sa limite géographique nord dans le bassin méditerranéen.

Additional file 1: Figure S1. Yearly variation of the mean abundance for each country. The abundance (y axis) was calculated as the mean all the observations (log transformed) from each country. Figure S2. QQ-plots of the residuals per month for the Obsoletus ensemble. Figure S3. QQ-plots of the residuals per month for the Pulicaris ensemble. Figure S4. Comparison of the abundance maps for each month using Random Forest (RF) and Interpolations for the Obsoletus ensemble. a Maps from January to June. b maps from July to December. Figure S5. Comparison of the abundance maps for each month using Random Forest (RF) and Interpolations for the Pulicaris ensemble. a Maps from January to June. b Maps from July to December. Figure S6. Comparison of the abundance maps for each month using Random Forest (RF) and Interpolations for Culicoides imicola. a Maps from January to June. b Maps from July to December. Figure S7. At a local scale, interpolation maps produce a smother surface between the farms compared to environmental driven RF, for which the predictions differ between adjacent pixels. The example shown in the figure corresponds to the August maps for the Obsoletus ensemble. Green dots: farms used for training, purple dots: farms within the test set.

Our understanding on the structure and role of viral communities in animal hosts has not yet reached the knowledge level attained on the bacteriome. This situation is, among others, probably due to technical constraints in adapting metagenomics, a main tool in bacteriome studies, to the characterisation of viral communities. Technical developments have been achieved for most steps of viral metagenomics. However, there is yet a key step that has received little attention, that of library preparation. This situation is surprising because developments in library preparation have largely facilitated bacteriome studies. Here, we present a library preparation optimized for metagenomics of RNA viruses from animals. The library has been designed to provide advantages found in metabarcoding libraries used in bacteriome studies, while allowing shotgun sequencing as required in viral metagenomics. We have validated a full metagenomics approach incorporating our library preparation and using mock viral communities. We have further tested our approach in two pilot studies with field-caught insects, all vectors of viral diseases. Our approach provided a fold increase in virus-like sequences compared to other studies, and nearly-full genomes from new virus species. Moreover, our results suggested conserved trends in virome composition in a mosquito species, and thus the existence of deterministic forces shaping its virome. Finally, the sensitivity of our approach was relatively good when compared to a commercial diagnostic PCR for the detection of an arbovirus in field-caught vectors. Our approach could facilitate studies on viral communities from animals and the democratisation of metagenomics in virus ecology.

Additional file 2: Table S1. Characteristics of trapping sites. Table S2. Temporal variables considered on mixed-effect negative binomial hurdle models. Table S3. Non-dynamic variables considered on mixed-effect negative binomial hurdle models.

The original version of this Article contained an error in Fig. 1a, in which the sequences of the reverse and forward primers were swapped

International audience

International audience

Vector abundance maps are a necessary step in mapping the risks associated with vector-borne diseases to help target surveillance and control actions. Risk mapping requires sampling vector species at different scales, which is made difficult by heterogeneity in vector spatial distribution and dynamics. In Morocco, Culicoides were responsible for African horse sickness (AHS) outbreaks in the 1960s and late 1980s. Since 2004, Morocco is facing an endemic situation of bluetongue virus (BTV) transmission. Our study aimed to produce first abundance maps for the most abundant Culicoides species, and risk maps for BTV and AHSV, across the country. We conducted a national cross-sectional survey using a stratified sampling strategy based on eco-agronomic zoning, for which we hypothesized the homogeneity of Culicoides diversity and dynamics. Historical collection time series were used to determine maximum abundance periods based on climate zonation. This strategy allowed to plan a national sampling of Culicoides abundance during a single collection season. Collections were conducted twice a year with 144 farms trapped. Culicoides were sampled for consecutive 48h using a suction light trap. We obtained a total of 262 samples under identification process. The results so far have shown that C. imicola was the most common species (present in at least 70% of the sites), followed by C. circumscriptus (absent at mild altitudes) and C. kingi (mostly present in the south). Culicoides obsoletus/C. scoticus have been collected frequently in the country, including in the south. This highlighted the wide and unexpected spatial distribution of these species from Scandinavia to North Africa reinforcing the existence of sibling species. Other species have also been identified, such as C. puncticollis, C. newsteadi, C. catanei/C. gegjelensis (rather present in the north on the coast), C. punctatus (rare in the south), C. pulicaris/C. lupicaris and C. kurensis. Interestingly, although C. paolae is mentioned in the literature as being associated with prickly pears in the Mediterranean basin, this species was absent in many areas heavily vegetated by this fig tree. This original dataset will be used to determine the ecological factors of the abundance and distribution of Culicoides of veterinary interest in order to provide accurate mapping of vector-related risks across Morocco.

Following the re-emergence of the bluetongue virus serotype 8 (BTV8) in Septem-ber 2015 and in accordance with the European Commission's regulatory require-ments, a new entomological surveillance network has been set up throughout mainland France to monitor the activity of Culicoides, which are vectors of BTV. The new Culicoides surveillance network uses suction light traps that opera-te one night per week in one site per eco-climatic zone. Twenty-four eco-clima-tic zones were determined in mainland France by analyzing the data collected by the first surveillance network set up from 2009 to 2012 during the emergen-ce of BTV8. Each eco-climatic zone, regrouping one or more French departments, is homogeneous, taking into account the diversity of Culicoides species, the dura-tion of vector inactivity and the week of the beginning and the end of inactivity. This entomological surveillance active from November (Y0) to April (Y1) has made it possible to determine successfully periods without Culicoides vectors for the different eco-climatic zones defined on the continental territory. The informa-tion provided by this network, coupled with the surveillance of viral circulation, made it possible to report during winters 2015 to 2018 several French departmen-ts in BTV seasonally free zones, thus lifting restrictions on livestock movements. Research impact highlights: Following the re-emergence of BTV8 in September 2015, a new Culicoides surveillance network is operational throughout mainland France to determine the periods without vectors in each eco-climatic zone. The information pro-vided by this network, coupled with viral circulation monitoring, allows BTV's seasonal free zones to be declared each winter, thus lifting restrictions on livestock movements.

Reunion Island regularly faces outbreaks of epizootic haemorrhagic disease (EHD) and bluetongue (BT), two viral diseases transmitted by haematophagous midges of the genus Culicoides (Diptera: Ceratopogonidae) to animals of economic importance such as cattle, sheep and goats. To date, five species of Culicoides are recorded in Reunion Island: Culicoides bolitinos, C. enderleini, C. grahamii, C. imicola, and C. kibatiensis. Although epizootics and Culi-coides diversity are already well documented, abundance and seasonality of the five species are not. According to a recent viral screening of local Culicoides populations (unpublished data), at least four species are involved in the transmission of each virus. Therefore, cha-racterizing the risk period by modelling the temporal dynamics of the five Culicoides species is a key step to better understand BT and EHD epidemiology and improve their control. Between 2016 and 2018, 55 biweekly Culicoides catches using OVI traps were set up in 11 sites. A hurdle model (i.e. a presence/absence model combined with an abundance model) was developed for each species in order to determine climatic and environmental drivers of presence and abundance of Culicoides. Regarding abundance, average Culicoides catch per site ranges from 4 to 45,875 individuals. Also, diversity differ between sites with C. imicola being dominant at low altitude and C. kiba-tiensis at high altitude. A marked seasonality is observed for the 3 other species. Eleven me-teorological and environmental determinants were used to model presence and abundan-ce of each species: temperature, humidity, rain, wind, global radiation, vegetation index, eco-climatic area, land use, farm density, animal density and length of nearby watercourse. The association of these determinants to explain presence and/or abundance depends on the species, but each plays a role in at least one species. This is the first study to model Culicoides population dynamics in Reunion Island. In the absence of vaccination and vector control strategies, determining periods of high abundance of Culicoides is a crucial first step towards identifying periods at high risk of transmission for both viruses.

Landscape genetics is an approach for understanding how geographical and environ-mental features structure genetic variation. Whether landscape patterns, in particular those created by human land uses, influence gene flow and its intensity to understand the spread of vector-borne diseases. Culicoides dispersal capacities are described as 'dis-persive stratified' which results from the combination of processes occurring at a short distance combined with jumps over long distances. Although winds and local air move-ments have an undeniable role in the dispersal of Culicoides, anthropogenic factors also play a predominant role, particularly at the local level (host distribution, landscape patterns). Our overall objective is to characterize the dispersal abilities of two main vector species in Europe, C. obsoletus and C. chiopterus, showing different host-vector behaviors. However, Culicoides obsoletus is reported in sympatry with one morpholog-ically indistinguishable species, C. scoticus, and other morphologically related species, C. chiopterus, C. dewulfi and C. montanus. Moreover, recently, several authors have report-ed the existence of cryptic diversity within the species commonly called C. obsoletus. Se-quencing a portion of the Cox1 mitochondrial gene of 3,200 C. obsoletus/C. scoticus individ-uals from 17 European countries revealed two clades within C. scoticus and confirmed the presence of at least three undescribed phylogenetic clades (C. obsoletus clade O2, C. obso-letus clade Dark and one not yet named) close to C. obsoletus. These results are reinforced by rDNA16S mitochondrial gene sequences and a gene coding for ribosomal rDNA18S, over the entire haplotypic diversity resulting from Cox1 barcoding. Then, we investigated how dispersion shapes the spatial arrangement of C. obsoletus genetic diversity in Europe using 13 microsatellite markers by observing patterns of allelic frequency distributions. This work is the first step to a more comprehensive study on the landscape genetics of two main vector species in the Palearctic region, C. obsoletus and C. chiopterus.

The study of viral communities has recently been boosted by the use of the so-called high-throughput sequencing (HTS) technologies. Coupled with random amplification of nucleic acids, HTS allows the identification of viral species present in a given sample without a-priori knowledge on their identity but their resemblance to known viruses. The huge interest of this approach for virus ecology and diagnostics has triggered several technical improvements in most steps required in virus metagenomics, from nucleic-acid extraction to bioinformatics analysis of sequencing results. Surprisingly, there is yet a key step that has received little attention, that of library preparation. Here, we describe a library preparation for exploring viral diversity in a large number of samples with the use of high-throughput sequencing. This method uses custom adaptors that are PCR ligated, allowing low cost, high multiplexing and fully exploitable read lengths. After validation of our method with artificial viral communities, we have tested it with two samples set from wild-caught arthropod vectors showing that our approach allows to identify an arbovirus in highly-degraded samples with a relatively high sensitivity.

For vector-borne diseases, producing abundance maps is a necessary step for risk mapping, which is an important tool to help focusing surveillance actions. Such as mapping requires sampling vector insects at a national/regional scale, which is quite difficult due to important heterogeneity in spatial distribution and in temporal dynamics. In Morocco, Culicoides biting midges were responsible for African horse sickness (AHS) outbreaks in the 1960s and the most recently in the late 1980s. They are also involved in the enzootic transmission of bluetongue (BT) virus since 2004. We aimed to produce first abundance maps for most abundant Culicoides species, and then risk maps for BTV and AHSV, across Morocco. In 2016, we compared species diversity in ruminant farms and in horse holding to determine if abundance maps derived from sampling in ruminant farms could be used to generate risk maps for both AHS and BT viruses. In 2017, we carried out a cross-sectional survey at the national scale. We used a stratified sampling strategy using eco-agronomic zones, with which we hypothesized homogeneity of Culicoides diversity and dynamics. We used historical collection time series to determine periods of abundance peak according to climatic zonation. This strategy allowed us to plan a national sampling of Culicoides abundance. In 2016, we found the same species in cattle and horse holdings. In addition, Culicoides paolae – reported in the literature as associated with Barbary fig trees, and Culicoides scoticus were recorded for the first time in Morocco. These records were confirmed by molecular assay. In 2017, collection campaign was conducted twice a year (from late April to June and from September to October): 144 farms were trapped during this period. At each visit, Culicoides populations were sampled for 48 consecutive hours using a light trap. A total we obtained 262 samples. Samples were currently under identification process. The next step will be to determine ecological drivers of abundance and distribution to provide accurate Culicoides mapping across Morocco.

Recent epidemics caused by different mosquito‐borne viruses underline the viral diversity associated to mosquitoes. However, beyond human viral pathogens, we know little on other viruses of the mosquito virome. This unexplored diversity probably influences vector competence and other aspects of the mosquito biology, as shown for the mosquito bacteriome. We have analysed the virome of Culex pipiens, a mosquito vector of important arboviruses like Rift Valley fever virus or West Nile fever virus. To this end, we have coupled a metagenomics approach with a large sampling campaign involving different countries and habitats around the Mediterranean basin. Our results show for the first time conserved patterns in virus diversity among mosquito populations, as well as specificities probably linked to different environmental conditions. The discovery of a ubiquitous group of viruses strongly supports the existence of a core virome in Culex pipiens that is likely to influence mosquito physiology.

International audience

International audience

Culex pipiens is a mosquito vector of West Nile virus (WNV), the causing agent of West Nile fever. Epizooties of WNV have taken place in Europe, mainly in France, Italy and Spain, and in some instances, human cases have been diagnosed. However, beyond WNV, little is known on the viruses associated to Culex pipiens, above all on viruses non-pathogenic for humans or livestock. Like Wolbachia bacteria, such viruses could have a role in WNV epidemiology either limiting mosquito populations or modulating the chances of WNV infection of C. pipiens. Through metagenomics approach, we characterized the virome of C. pipiens in five countries around the Mediterranean basin Our results show for the first time conserved patterns in diversity and structure within the virome of C. pipiens with a large abundance of the Mesoniviridae, a relatively new family of mosquito-specific viruses in the Nidovirales order. To date, the few data available on Mesoniviridae show that they probably have a worldwide distribution and they can infect different species of mosquitoes. The sequences obtained allowed to assemble several full genomes of Mesoniviridae. Interestingly, phylogenetic analysis showed a highly similarity with strains of Nam-Dinh virus, a virus recently identified in Côte d'Ivoire (West Africa) and Vietnam but which had not yet been observed in Europe. In view of the importance of the Mesoniviridae dispersion throughout the world, we determined the relative abundance of this family in natural population of Culex pipiens in Camargue. (Texte integral)