Introduction

The living world emerges from the interaction of different organisms that maintains between them complex and diverse interactions (Chow et al. 2010). Among them, symbiotic relationship was defined as the sustainable coexistence of two or more organisms (Dheilly 2014) that concerns a part or the whole life cycle of such organisms, regardless of the nature of trade arising between them (Brossard et al. 1991). Microorganisms have colonized and predominated throughout the world and demonstrated an extraordinary variety. In terms of adaptation, few microorganisms could even survive in extreme conditions, and settle and grow in hostile environments (Minard et al. 2013a). Then, this ability to live everywhere allowed them to establish special relationships with living hosts. Especially, the symbiotic relationships that different types of microorganisms have established with mosquitoes have probably played a key role in the evolutionary success of these insects. Symbiotic microorganisms influenced various biological functions and integrated those under the control of the host genome, providing to mosquitoes an improved adaptability to their environment (Ricci et al. 2012a). Currently, more than 3,500 species of Culicidae were described in the World, classified in two subfamilies and 112 genera. The subfamily Anophelinae has three genera while the Culicinae has 109 genera segregated into 11 tribes (Harbach 2013). Studying mosquito biology, ecology, and control is important because females of many species are potential vectors of parasites leading to infectious diseases such as malaria, dengue fever, Chikungunya disease, West Nile fever, and dirofilariosis affecting animals including humans (Bawin et al. 2015; Ricci et al. 2012b). Mosquitoes are the indirect cause of more morbidity and mortality among humans than any other group of organisms. The extension of human activity, trade, and tourism contributed to the emergence of new strains of diseases in areas previously free of these diseases (Medlock et al. 2015). Indeed, the development of transport networks, the rapidity of travel, and the facilitation of human movement increased the risk of vector invasion (Manguin and Boëte 2011). The emergence or re-emergence of a pathogen could have a significant impact on human and animal health in naive environments and countries. Moreover, the role of climate change had a significant impact in the spread of invasive mosquito species (Slenning 2010). The loss of the biodiversity also increases the risk of human exposure to vector-borne diseases (Ostfeld 2009). The absence of a vaccine against most of arbovirosis implies efficient vector control strategies, which are up to date mainly based on the use of chemical insecticides. However, most vectors develop resistance against currently used products (Corbel and N’Guessan 2013) whereas new chemical compounds was limited firstly by financial costs in research and development, and secondly by the toxicity of derivatives to human population, biodiversity, and environment. The exploitation of new methods, such as the use of microbial communities as a bio-pesticide for vector control, will be therefore a promoted strategy at the global level (Bawin et al. 2015).

In Belgium, the introduction and spread of mosquito-borne viruses could be prevented yet. Although there was no case of autochthonous transmission of arbovirosis by mosquitoes in the last decades, the risk for the introduction of new pathogens is increasing. Belgium has multiple relationships with neighboring countries which have competent vectors and where autochthonous cases of arbovirosis were recently recorded (e.g., Chikungunya and dengue in France; Vega-Rua et al. 2013). Unfortunately, the existence of competent vectors in Belgium increases the risk of contingency. In parallel, the country is not protected with the increase of vector resistance due to the widespread use of insecticides. The aims of this study were (1) to check the presence of the six symbiotic bacteria in field mosquito species in Belgium and (2) to suggest new candidates for future paratransgenesis strategies.

Material and method

Sampling areas and mosquito collection

The mosquito individuals used for this study was originated from recent inventories in Belgium of (Boukraa et al. 2013), 2015b (submitted), Raharimalala et al. (2015; submitted), and also some specimens from the present study. In total, we examined mosquitoes from 13 sites in Belgium (Fig. 1) belonging to various environments including equestrian farm, cattle farm, company of import/recycling used tires, and quarry of recycling of materials from construction. The collection of mosquitoes was performed in 2011–2013 where both larval sampling (dipping) and adult trapping (CO2 baited trap Mosquito Magnet Liberty Plus) were used. To avoid breeding site influence on the screening, larvae were collected in various environments differing by their type (artificial/natural), kinds (buckets/water troughs/puddle/pond/used tires), state water (permanent/temporary and standing/running), and vegetation situation (presence/absence of floating and/or emergent vegetation). All specimens were stored either in 70–95 % ethanol (larval stages) or in a freezer at −20 °C (adult stages) until being used for polymerase chain reaction (PCR) analyses. Mosquito species identification was based on morphotaxonomy (larvae, adults, and genitalia analyses), by using identification keys (Becker et al. 2010; Schaffner 2001).

Fig. 1
figure 1

Localization map of study sites used for mosquito collections: Br Brecht, Ch Chénée, Gx Gembloux, Ma Maasmechelen, Mn Malonne, Na Nattoye, Ne Neufchateau, Lo Lochristi, Rd La Reide, SL Somme-Louze, Sp Sprimont, Vr Vrasene, Wg Warsage

Full size image

Nucleic acid extraction

DNeasy Blood & Tissue Kits (Qiagen) were used to extract genomic DNA following the instructions of the manufacturer. Briefly, all 173 samples were washed two times in sterile distilled water and one time in 70 % alcohol to avoid contamination. The samples were then individually crushed with plastic pestles in 1.5 ml tubes with 180 μl of buffer ATL, and 20 μl of proteinase K were added. Mixtures were vortexed thoroughly and incubated 3 h in a thermomixer (56 °C, 450 rpm). After processing and washing with adequate product in the kit, total DNA was eluted with 70 μl of buffer AE. The purity and quantity of DNA contained in each sample were assessed using a spectrophotometer (Nanodrop ®), and DNA concentration was adjusted to 30 ng μl−1 before being stored at −20 °C until used.

PCR amplification and sequencing data

A conventional PCR was used to screen the presence of six genera of (endo-) symbiotic bacteria (Wolbachia sp., Comamonas sp., Delftia sp., Pseudomonas sp., Acinetobacter sp., and Asaia sp.) among the 173 specimens belonging to 11 mosquito species from 13 sites in Belgium (Table 1). In a first time, universal eubacteria primers pA: 5′-AGAGTTTGATCCTGGCTCAG-3′ and pH: 5′-AAGGAGGTGATCCAGCCGCA-3′ that target for 16S rRNA gene sequences were used (Bruce et al. 1992) to check the presence of bacterial DNA in the samples. PCR conditions were set as an initial denaturation at 95 °C for 3 min, followed by 30 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 40 s, extension at 72 °C for 1 min 15 s, and a final extension at 72 °C for 10 min. Then, each positive sample was checked with six specific primers that target (endo-) symbiotic bacteria (Table 2). For all process, PCR was performed using 60 ng of genomic DNA per microliter in a final volume mix of 50 μl, with 5 μl buffer ×1, 1 μl dNTP 2 mM, 0.7 μl of each primer (forward/reverse) to 10 μM of final concentration, and 0.5 μl of 0.4 U Taq DNA polymerase (Eurogentec). Results were checked by electrophoresis on 2 % agarose gels (70 V, 45 min), stained with SYBR® Safe DNA gel stain (Invitrogen) and photographed with Gel Doc 2000 system (BioRad). A part of positive PCR products to the specific primers were sequenced (GATC Biotech AG, Germany), and sequences data were edited and aligned with BioEdit and Multialn softwares. All sequences were blasted in NCBI (http://blast.ncbi.nlm.nih.gov/Blast.cgi). Results were compared with data available in GenBank to confirm the identity of bacteria and deposited under accession numbers [GenBank: KJ512993–KJ512997].

Table 1 Mosquito species screened and sampling sites
Full size table
Table 2 Primers used in this study to target potential symbiotic bacteria in sampled field mosquitoes
Full size table

Statistical analysis

Chi-square tests of independence were applied to compare the presence of symbiotic bacteria between mosquito species in larvae and adults. All tests were performed using Minitab® v.16 software (http://www.minitab.com/fr-FR/default.aspx).

Results

Eleven mosquito species (144 larvae and 29 adults) belonging to five genera were used for the screening (Table 1): Culex pipiens s.l. L, Culex torrentium (Martini, 1925), Culex hortensis (Ficalbi, 1889), Anopheles claviger (Meigen 1804), Anopheles maculipennis s.l. (Meigen, 1818), Anopheles plumbeus (Stephens 1828), Culiseta annulata (Schrank, 1776), Ochlerotatus geniculatus (Olivier, 1791), Ochlerotatus dorsalis (Meigen, 1830), Aedes albopictus (Skuse, 1894), and Coquillettidia richiardii (Dyar, 1905). Mosquito species were selected for the study because of their susceptibility as vectors.

The presence of symbiotic bacteria statistically differed between mosquito species in both larvae (Table 3) and adults (Table 4). Considering larvae, Wolbachia sp. was only found in Cx. pipiens s.l. (54.2 %) and Cx. hortensis (16.7 %; χ 2 (5) = 60.3; P < 0.001; Table 3). Pseudomonas sp. was found in all mosquito species (Fig. 2), and dominated in Cx. pipiens s.l. (93.8 %) and Cx. torrentium (83.3 %), followed by Cs. annulata (75.0 %), An. maculipennis s.l. (72.2 %), An. claviger (50.0 %), and Cx. hortensis (33.3 %; χ 2 (5) = 19.9; P = 0.001; Table 3). Acinetobacter sp. was also found in all tested species and dominated in Cx. hortensis (100.0 %), followed by Cs. annulata (54.2 %), Cx. pipiens s.l. (52.1 %), An. maculipennis s.l. (50.0 %), Cx. torrentium (36.1 %), and An. claviger (33.3 %; χ 2 (5) = 12.6; P = 0.027; Table 3). Asaia sp. was found in An. maculipennis s.l. (33.3 %), Cs. annulata (29.2 %), Cx. torrentium (19.4 %), and An. claviger (16.7 %) but not in Cx. pipiens s.l. (0.0 %) and Cx. hortensis (0.0 %; χ 2 (5) = 18.6; P = 0.002; Table 3). No Comamonas sp. and Delftia sp. were found in larvae.

Table 3 PCR results for the screening of the six symbiotic bacteria among six mosquito species (larvae) from Belgium
Full size table
Table 4 PCR results for the screening of the six symbiotic bacteria among six mosquito species (adults) from Belgium
Full size table
Fig. 2
figure 2

Distribution of symbiotic bacteria in mosquito species: a–d Wolbachia, Pseudomonas, Acinetobacter, and Asaia distribution, respectively. Acl Anophles claviger, Aea Aedes albopictus, Am s.l. Anopheles maculipennis s.l., Aplu Anopheles plumbeus, Ch Culex hortensis, Cp Culex pipiens s.l., Cqr Coquillettidia richiardii, Csan Culiseta annulata, Ct Culex torrentium, Ocd Ochlerotatus dorsalis, Ocg Ochlerotatus geniculatus, Mossp mosquito species

Full size image

Considering adults, Wolbachia sp. was found in all Cx. pipiens, Oc. dorsalis, Ae. albopictus, and Cq. richiardii individuals but not in An. plumbeus and Oc. geniculatus (χ 2 (5) = 29.0; P < 0.001; Table 4). Pseudomonas sp. was found in Cq. richiardii (100.0 %), Ae. albopictus (100.0 %), Oc. dorsalis (100.0 %), Oc. geniculatus (88.9 %), An. plumbeus (75.0 %), and Cx. pipiens s.l. (66.7 %) but no statistical differences between species was found (χ 2 (5) = 2.3; P = 0.814; Table 4). Acinetobacter sp. was only found in Ae. albopictus (100.0 %) and Cx. pipiens (11.1 %; χ 2 (5) = 15.2; P = 0.010; Table 4). Asaia sp. was only found in Ae. albopictus (100.0 %) and Cx. pipiens (11.1 %; χ 2 (5) = 15.2; P = 0.010; Table 4). No Comamonas sp. and Delftia sp. were found in adults.

Concerning the distribution of bacteria in mosquito species, Pseudomonas was largely predominated in each species, in larvae and adults, and in different biotopes (Fig. 3a–d).

Fig. 3
figure 3

Pseudomonas repartition in different biotopes: a–d Breeding sites, mosquito species, stade of mosquito, and type of sites, respectively. Acl Anophles claviger, Aea Aedes albopictus, Am s.l. Anopheles maculipennis s.l., Aplu Anopheles plumbeus, Ch Culex hortensis, Cp Culex pipiens s.l., Cqr Coquillettidia richiardii, Csan Culiseta annulata, Ct Culex torrentium, Ocd Ochlerotatus dorsalis, Ocg Ochlerotatus geniculatus, Mossp mosquito species, Br bovine farm, Bw backwater, Hf horse feetprint, W.trough water trough

Full size image

Discussion

In this study, we tested six symbiotic bacteria genera in eleven potential vector mosquito species in Belgium. The objective of this study was to find potential candidates for a symbiotic bacteria paratransgenic control. The study choice of these six symbiotic bacteria genera was due to the known of their ability to manipulate part or entire biology of their host. Our results have shown that four out of the six bacteria studied (Wolbachia, Pseudomonas, Acinetobacter, and Asaia) were found while Comamonas and Delftia were not.

Decision to study Comamonas genera was due to its possible involvement in decreasing pathogen transmission, mainly in Anopheles species (Minard et al. 2013a) while Delftia genera could principally act in delaying larval development, also mainly in Anopheles species (Chavshin et al. 2014). These two genera were often detected in some Anopheles species, Cx. pipiens, Aedes (Lindh 2007), and recently, Zouache et al. detected them in Aedes albopictus of Madagascar (Zouache et al. 2011). But surprisingly, we did not find them including the three Anopheles species and in Cx. pipiens s.l. studied here. Comparing into the previous study, these bacteria were found in every tropical countries (Chavshin et al. 2014; Lindh 2007; Minard et al. 2013a) but never detected in temperate country. We supposed that the main cause could be that the environment is too hostile for these bacteria and limited their growth in Belgium. For that, we could not suggest their uses as paratrangenesis tool in Belgium.

Wolbachia genera constituted one of the symbiotic bacteria that we have used in this study. These endosymbiotic bacteria are well studied and reduce the life span of the mosquito species (Almeida et al. 2011; Atyame et al. 2014) or interfere with pathogen replication and dissemination in mosquito species (Bian et al. 2013; Caragata et al. 2014). In this research, Wolbachia is in low proportion in the studied mosquitoes. We supposed that this lowest proportion has the same reason of the case of Delftia and Comamonas, as the environment is not favorable for them. Despite that these bacteria were already used elsewhere in the fight against Culex and Aedes (van den Hurk et al. 2012; Zélé et al. 2014), we do not advocate its use in Belgium.

Pseudomonas was largely predominated in all stages and mosquito species studied. This high prevalence could suggest that Pseudomonas would be a best candidate to paratransgenesis because his grow in cell-free and ordinary culture media is facilitate and suitability for genetic transformation (Chavshin et al. 2014). In the literature, three key components were required for used bacteria as paratrangenesis: an existing of effector molecule that could achieves the desired effect, a mechanism to display or excrete the effector molecule on the surface of the bacteria, and bacteria that can survive in the mosquito long enough to produce the expected amount of effector molecules and thereby achieve the desired effect in the mosquito (Wilke and Marrelli 2015). These three parameters seem to be complete in Pseudomonas. Anterior study attested the use of chitinolytic activity of extra cellular mosquitocidal proteins of P. fluorescens (Brammacharry and Paily 2012). Moreover, the high prevalence of Pseudomonas may indicate existence of intimate symbiotic relationship between a given bacterium and its host (Minard et al. 2013b). This relationship would be exploited to use these bacteria as symbiotic control by exploiting its influence in mosquito fitness with introduction of a modified gene in the bacteria. Secondly, the fact that Pseudomonas could be found in both larvae and adults may suggest a transstadiale transmission. Or, another way to perform paratrangenesis approach was to target bacteria that are well established in mosquitoes and transmitted to the next generation (Wilke and Marrelli 2015). The studies conducted by Chavshin et al. (2012, 2014, 2015) confirmed this transstadial passage based of phenotypic marker as used for Asaia bacterium, such as green fluorescent protein (GFP; Favia et al. 2007). Normally, most bacteria are lost during the metamorphosis from larval to adult stage and only a few bacteria are transmitted from the larva to the adult (Moll et al. 2001). Indeed, during emergence and transformation of adult mosquitoes, a sterilization mechanism is set up to sequester bacteria in the larval gut. Just a few bacteria could cross the barrier of the meconial peritrophic membranes and be transmitted to adult stages (Chavshin et al. 2012). Most of transstadial bacteria (as the case of Pseudomonas) are Gram negative. These bacteria are generally transformable for useful in paratransgenic method (Wirth et al. 1989) and could affect the parasite load in the infected mosquitoes. They could dramatically reduce either Plasmodium falciparum Welch, 1897 or Plasmodium vivax Grassi and Feletti, 1890, parasite prevalence and density within vectors (Chavshin et al. 2012).

In the other part of our results, Acinetobacter sp. represented a non-negligible value in all larval species studied, likewise the case of Asaia sp. with the lowest prevalence. The same results were observed by Minard et al. (2013b) and Zouache et al. (2011) during a prevalence study of symbiotic bacterial in Ae. albopictus and Ae. aegypti L. of Madagascar. Considering their high prevalence, candidature of these two symbiotic bacteria to paratransgenetic bacteria could also pay attention. Firstly, high prevalence of Acinetobacter could indicate and suggest that these bacteria could play a role in the biology of the mosquito in the field (Minard et al. 2013b); even this role in insect biology has been only confirmed in Stomoxys calcitrans L. where it ensured complete development of the fly larvae (Lysyk et al. 1999). Anterior study conducted by Mweresa et al. (2015) demonstrated that Acinetobacter baumannii played a role in mosquito attractiveness during host biting. Its presence is also detected to have increased considerably after feeding on blood meals than feeding on sugar of the adult female mosquitoes (Zayed and Bream 2004). But most of the species of Acinetobacter are associated with bacteraemia, pulmonary infections, meningitis, diarrhea, and notorious nosocomial infections and constituted a veritable danger for human and are responsible of rapid development of resistance to a wide range of antimicrobials (Doughari et al. 2011). Secondly, candidature of Asaia for use in mosquito paratransgenesis may pay attention also. These bacteria have a peculiar relationship with mosquitoes because of its localization in the gut, in the salivary glands, and in the reproductive organs of both sexes (Favia et al. 2007). Anterior study (Ricci et al. 2012b) found Asaia in all the developmental stages of all tested malaria vectors and in Aedes aegypti with high prevalence. It was demonstrated that like Pseudomonas, this bacteria is easily cultivable outside in cell-free media that facilitates his transformation.

Moreover, this bacteria could use multiple ways of transmission within and between population hosts like vertically transmission (by progeny by maternal or paternal), horizontal routes by mating and co-feeding (Ricci et al. 2012b). However, even the three symbiotic bacteria were promising as a good candidate; some characteristics such as transstadial transmission, non-pathogenicity, immunological and physiological adaptation to mosquito midgut conditions, colonization in the mosquito midgut including effective competition with resident bacteria, and persistency in the gut for a reasonable time (Chavshin et al. 2014) should be well examined. In fact, anterior studies reported that members of Acinetobacter and Asaia may cause opportunistic infections in humans and their presence could provoke cross-infection when mosquitoes bite hosts (Alauzet et al. 2010; Doughari et al. 2011; Minard et al. 2013b). These reason motivated us to suggest Pseudomonas as the best candidate latest.

Conclusion

This study permitted to demonstrate the prevalence of four symbiotic bacteria, among 6, in 11 mosquito species in Belgium. It also allowed the highlighting of the possible role of three bacteria genera for paratransgenesis in Belgium. Due to its omnipresence, Pseudomonas seemed to be the most interesting candidate for paratransgenic bacteria in order to reduce and/or inhibit the transmission of pathogens.