Mutual exclusion of Asaia and Wolbachia in the reproductive organs of mosquito vectors

Background Wolbachia is a group of intracellular maternally inherited bacteria infecting a high number of arthropod species. Their presence in different mosquito species has been largely described, but Aedes aegypti, the main vector of Dengue virus, has never been found naturally infected by Wolbachia. Similarly, malaria vectors and other anophelines are normally negative to Wolbachia, with the exception of an African population where these bacteria have recently been detected. Asaia is an acetic acid bacterium stably associated with several mosquito species, found as a dominant microorganism of the mosquito microbiota. Asaia has been described in gut, salivary glands and in reproductive organs of adult mosquitoes in Ae. aegypti and in anophelines. It has recently been shown that Asaia may impede vertical transmission of Wolbachia in Anopheles mosquitoes. Here we present an experimental study, aimed at determining whether there is a negative interference between Asaia and Wolbachia, for the gonad niche in mosquitoes. Methods Different methods (PCR and qPCR, monoclonal antibody staining and FISH) have been used to address the question of the co-localization and the relative presence/abundance of the two symbionts. PCR and qPCR were performed to qualitatively and quantitatively verify the distribution of Asaia and Wolbachia in different mosquito species/organs. Monoclonal antibody staining and FISH were performed to localize the symbionts in different mosquito species. Results Here we provide evidence that, in Anopheles and in other mosquitoes, there is a reciprocal negative interference between Asaia and Wolbachia symbionts, in terms of the colonization of the gonads. In particular, we have shown that in some mosquito species the presence of one of the symbionts prevented the establishment of the second, while in other systems the symbionts were co-localized, although at reduced densities. Conclusions A mutual exclusion or a competition between Asaia and Wolbachia may contribute to explain the inability of Wolbachia to colonize the female reproductive organs of anophelines, inhibiting its vertical transmission and explaining the absence of Wolbachia infection in Ae. aegypti and in the majority of natural populations of Anopheles mosquitoes. Electronic supplementary material The online version of this article (doi:10.1186/s13071-015-0888-0) contains supplementary material, which is available to authorized users.


Background
Most insects harbour mutualistic microorganisms that provide specific traits to the host, conferring evolutionary/adaptive advantages. Many beneficial symbionts are acquired by environmental uptakes in every generation, while few are located in the reproductive organs. As a consequence of this localization, they can easily be acquired by the offspring to ensure their transmission to the next insect generations by vertical transmission [1,2]. To date, only two bacteria have been shown to be located in the reproductive organs in several mosquito species: the alfa-proteobacteria Wolbachia and Asaia [3].
Wolbachia is a group of obligate intracellular bacteria that infect arthropods and nematodes [4]. Within arthropods, Wolbachia infects a wide spectrum of insects, reflecting the ability of these bacteria to manipulate host reproduction, favouring their own maternal transmission [5].
These intracellular bacteria were firstly described in Culex pipiens [6], but more recently they have been detected in mosquitoes from several genera, including Aedes, Culex, Coquillettidia, and Mansonia. Interestingly, until very recently, Wolbachia has not been recorded to naturally infect representatives of the genus Anopheles, that comprises more than 300 different species, about 60 of whose are malaria vectors [7]. Only very recently, the presence of Wolbachia has been detected in a few individuals of a natural small population of Anopheles gambiae. Interestingly, the positive mosquito individuals detected in this study clustered into only a few of the examined mating swarms, pinpointing that ecological and environmental factors might play a key role in the establishment of Wolbachia infections in the An. gambiae host [8].
Similarly, Wolbachia has never been detected in natural populations of Ae. aegypti, a main vector of Dengue and yellow fever [9].
In the last few years, several studies have revealed the potential of Wolbachia to control mosquito-borne diseases. Indeed, it has been shown that the "forced" introduction of some strains of Wolbachia in Ae. aegypti reduces its competence in transmitting Dengue virus. The mechanisms at the basis of this reduced vectorial capacity have not been elucidated, even though an upregulation of the mosquito immune response might play a role in this phenomenon [10,11]. More recently, it has also been proven that Wolbachia-infected Ae. aegypti, resistant to Dengue virus infection, are able to rapidly replace natural/susceptible populations, thus validating the Wolbachia-mediated population-replacement strategy proposed to control mosquito borne diseases [12].
For all these reasons, there were many attempts to infect Anopheles mosquitoes with Wolbachia, in order to export the novel symbiont-based control strategies to malaria vectors. As pointed out by Hughes & Rasgon [13], the failure of the first attempts, coupled with the apparent lack of natural infection in natural populations of anophelines, suggests that this genus is somehow refractory to Wolbachia infection. Nevertheless, in the last few years, trans-infection of cell lines and somatic infection of adults with Wolbachia have been achieved, proving that anophelines can also be forced to harbour these bacteria [14]. Recently, a strain of the main Asian malaria vector, namely Anopheles stephensi, was stably transinfected by embryonic microinjection of the wAlbB Wolbachia strain derived from Aedes albopictus [15].
Asaia is a bacterium stably associated with numerous mosquito species, including several anophelines, often being the dominant microorganism of the mosquito microbiota. This acetic acid symbiont localises in the gut, salivary glands and reproductive organs of adult mosquitoes. Asaia is horizontally transmitted through an oral route during feeding both in pre-adult and adult stages and through a venereal pattern during mating in adults [16,17]. Moreover, Asaia is vertically transmitted from mother to progeny indicating that it may rapidly spread in natural mosquito populations [17,18].
Both Wolbachia and Asaia have been proposed as promising microorganisms for the development of symbiont-based control methods to contrast vectorborne diseases [3,19].
Here we show that the abundances of Wolbachia and Asaia are negatively related in the reproductive organs of several mosquito species, with a pattern that approximate a mutual exclusion, particularly in anophelines. This provides a possible explanation for the absence of Wolbachia in natural populations of Anopheles. For simplicity, the phenomenon investigated in this work, i.e. the interference between the two symbionts for the colonization of the mosquito gonad niche, will be referred to as 'competition'.

Mosquito strains
Insectary reared strains: The An. stephensi colony (Liston strain) has been maintained for 4 years in the insectary at the University of Camerino (Unicam). Insects were reared at 29°C and 85-90 % relative humidity with photoperiods (12:12 Light-dark). Adult insects were maintained in a 5 % sucrose solution, and adult females were fed with mouse blood for egg laying. Larvae were maintained in spring water and fed daily with commercial fish food.
The An. gambiae colony was established from samples obtained from the Centre National de Recherche et de Formation sur le Paludisme (CNRFP) in Ouagadougou (Burkina Faso, West Africa) and descend from a wild colony.
The Ae. albopictus colony was established from mosquitoes that were field-collected from S. Benedetto del Tronto, Central Italy (42°57′00″N; 13°53′00″E) in 2008. They were maintained in the Unicam insectary in the same condition as described above.
Cx. quinquefasciatus species mosquitoes, collected in Hawaii (U.S.A.) in 2008 and provided by the insectary of the Center for Vector Biology, Rutgers University (New Brunswick, NJ, U.S.A.), were maintained in the Unicam insectary in the same condition described above.
The colony of An. stephensi stably trans-infected with wAlbB Wolbachia strain (and the relative control treated with antibiotics to remove Wolbachia infection) is the one described in [15], that has been bred for two generations at Unicam insectary before performing the analysis.
The colony of Ae. aegypti, stably trans-infected with wMelPop Wolbachia strain (and the relative control treated with antibiotics to remove Wolbachia infection), is the one described in [20], that has been bred for two generations at Unicam insectary before performing the analysis.

Mosquito samples
Total DNA was extracted from whole mosquitoes and/ or organs (dissected in a drop of sterile 1× PBS using sterile needles under a stereomicroscope) as previously described [21].
Quantitative PCR (qPCR) detection of native Asaia in different mosquito species PCR assays were designed to detect DNA of bacteria Asaia in organs from 13-day-old An. stephensi, Ae. albopictus and Cx. quinquefasciatus mosquitoes. genes of the different mosquito species were amplified as housekeeping genes (As-rps7, Ae-rps7, CX-rps3 respectively) to allow the normalization of Asaia amount.
Amplification consisted of 50 ng DNA, 1X SybrGreen Master Mix (Fermentas, Vilnius, Lithuania), 200 nM of primers. Primers used to amplify target sequences of Asaia 16S rRNA and rps7 of An. stephensi (As-rps7) were described in [22], whereas the primers for Cx-rps3 (Cx. quinquefasciatus) and Ae-rps7 (Ae. albopictus) genes are listed below: Reactions were run on a CFX thermocycler (Bio-Rad, Hercules, California, USA) using the following cycling conditions: 1 cycle of 95°C for 10 min, 40cycles of 95°C for 1 min, 60°C for 1 min, and 74°C for 30 sec. The PCRs were performed on six pools of organs, from ten individuals, for each mosquito species. Each pool was tested in duplicate. The relative quantity of Asaia in the mosquito organs was estimated as gene copy ratio calculating the copy number of Asaia 16S rRNA gene/respective housekeeping genes.

Statistical analysis
The statistical analysis of the amounts of Asaia, quantified by qPCR assay, was estimated using the Bio-Rad CFX Manager Software and the GraphPad software (http://www.graphpad.com). Data were obtained from the average of six pools per organ. The value of each pool resulted from the average of two technical replicates that have been compared by the Mann-Whitney test.

Asaia localization on eggs-surface
An. stephensi, An. gambiae and Ae. albopictus eggs were fixed with 4 % paraformaldehyde for 10 min at 4°C and washed twice with PBS 1X. The slides were incubated in 1 % Bovine Serum Albumine (BSA) in 1X PBS for 30 min at room temperature and successively for 1 h at 37°C with anti-Asaia monoclonal antibody (patent pending N. MI2012A001529) diluted 1:1000. After three washings in PBS 1X, they were incubated for 30 min at 37°C with anti-mouse IgG Alexa Fluor 594 conjugate (Invitrogen, Carlsbard, California, USA) diluted 1:100 in 1 % BSA in PBS and washed three times for 10 min with 1X PBS. Slides were mounted with glycerol and visualized by epifluorescent microscopy (Carl Zeiss Axio Observer.Z1, Milan, Italy).
After hybridization, the organs were washed in 200 μl of washing buffer (0.2X SSC, 60 % formamide) for 15 min at 40°C, and subsequently once in 500 μl of 0.1X SSC and twice in 200 μl of 1X SSC for 10 min at room temperature. After that they were washed twice in 1X PBS for 5 min at room temperature. Then 50 ng of DAPI (4′, 6′-diamidino-2-phenylindole) were added, and incubated for 5 min at room temperature. After a wash in 1X PBS for 5 min at room temperature, samples were mounted in anti-fading medium and then observed at a laser-scanning confocal microscope SP2-AOBS (Leica, Wetzlar, Germany). Control experiments involved either treatment of slides with RNase prior probe hybridization or in absence of probe.
The mosquitoes were fed with a strain of Asaia expressing the Green Fluorescent Protein (namely Asaia SF2.1 and AA5.5). These bacteria were grown 24 h at 30°C in GLY medium (glycerol 25 g/L; yeast extract 10 g/L; pH 5). Cells were harvested by centrifugation, washed three times in 0.9 % NaCl and adjusted to 10 8 cells per ml −1 in 50 ml of H 2 O/5 % (wt/vol) sucrose solution, supplemented with 100 μg·ml −1 of kanamycin to avoid the loss of plasmid from bacterial cells. After 4 days of feeding, the cotton pad containing strain SF2.1(Gfp) was removed and replaced with a new sterile 5 % (wt/vol) sucrose solution supplemented with kanamycin. Mosquitoes were sampled and dissected every 2-3 days up to 15 days after initial exposure to the bacterium. Guts and reproductive organs were fixed with 4 % paraformaldeyde for 10 min at 4°C. The slides were then mounted in glycerol and examined in fluorescent microscopy (Carl Zeiss Axio Observer.Z1, Milan, Italy).

Tissue localization of Asaia in different mosquito species
Firstly, by specific PCR assays we have investigated the tissue distribution of Asaia in laboratory-reared strains of Ae. albopictus and Cx. quinquefasciatus, two species of mosquitoes that are naturally infected with Wolbachia. Interestingly, while Asaia was found in the midguts of these mosquitoes, it was never detected in the reproductive organs (Table 1). This represents a remarkable difference if compared with Asaia distribution in tissues of mosquito species that are not infected with Wolbachia, such as An. gambiae, An. stephensi, Ae. aegypti, where Asaia has been constantly found in the reproductive organs and salivary glands, in addition to the midgut [25]. We have also evaluated the circulation of Asaia in different mosquito organs of An. stephensi, Ae. albopictus and Cx. quinquefasciatus by quantitative PCR. As shown in Fig. 1, occasionally, it was possible to detect the presence of Asaia in the reproductive organs of the two latter species, but in a very limited amount compared with that present in the midguts and in the reproductive organs of An. stephensi. Indeed, the differences in the Asaia load in the midguts and gonads of Ae. albopictus and Cx. quinquefasciatus, as revealed by qPCR, were evident and statistically significant. Of course, we cannot rule out the possibility that, when dissecting organs for the molecular analysis, Asaia bacterial cells located in the guts have contaminated the reproductive organs.
To further corroborate the null or very limited circulation of Asaia in the reproductive organs of Wolbachia infected mosquitoes, we have analysed Ae. albopictus and Cx. pipiens mosquitoes field collected in Italy. A high proportion of these mosquitoes did not host Asaia (63 % and 85 % respectively), as revealed by specific PCR assay on whole mosquitoes (Additional file 1: Table A).

Asaia localization on eggs-surface
The specific localization of Asaia in different tissues of mosquitoes Ae. albopictus and Cx. quinquefasciatus may explain the inability of these mosquitoes to vertically transmit Asaia, differently to what has been shown in An. gambiae [17]. To further investigate this aspect, we used a specific antibody to localize Asaia on the surface of eggs produced by different mosquito species, some naturally hosting Wolbachia and some not. Antibodybased staining revealed the presence of Asaia on the surface of eggs of An. stephensi and Ae. aegypti, while no signal was detected on the eggs of Ae. albopictus and Cx. quinquefasciatus (Fig. 2). The results of these experiments are congruent with those of tissue localization, reinforcing the hypothesis that the presence of Wolbachia prevents the establishment of Asaia within the reproductive organs of Ae. albopictus and Cx. quinquefasciatus (and vice-versa in anopheline mosquitoes and Ae. aegypti).

Asaia-GFP colonization of midguts and reproductive organs of different mosquito species
To further investigate the hypothesis of a microbial competition between Asaia and Wolbachia within the reproductive organs of mosquitoes, a series of experiments has been performed using a strain of Asaia expressing the green fluorescent protein (GFP), provided to the mosquitoes with sugar meal to obtain body colonization. We focused our attention on midguts and gonads. In Ae. albopictus and Cx. quinquefasciatus, the colonization with Asaia-GFP strain was limited to the midguts of the examined specimens, that were colonized in the 81 % and 80,3 % of the cases respectively, while no colonization was detected in the reproductive organs in any of the 60 specimens analyzed for each species (Fig. 3), (Additional file 2: Table B). These findings are just the opposite with those previously reported in An. gambiae, An. stephensi and Ae. aegypti where the colonization has been repeatedly observed in both midguts and reproductive organs [3] and confirmed by the same analysis in 60 specimens An. stephensi mosquitoes, that has been  Fig. 1 Quantitative detection of Asaia in organs of three different lab-reared mosquito species obtained by qPCR. The relative amount of the bacteria is expressed as a ratio of bacterial 16S rRNA and mosquito rps7 genes (An. stephensi and Ae. albopictus) or rps3 gene (Cx. quinquefasciatus) copies in a logarithmic scale. Abundance results from the mean±SEM of six pools (10 organs) for each species. Statistically significant differences are represented by asterisks (p<0.01) as determined by multiple comparisons using Mann Whitney test used as control, that were colonized in 86 % of the guts and in 30 % of the reproductive organs (Additional file 2: Table B).
Asaia-GFP colonization of reproductive organs of a strain of Ae. aegypti stably trans-infected with Wolbachia wMelPop The colonization experiments with Asaia-GFP were then extended to a strain of Ae. aegypti stably trans-infected with a Wolbachia from Drosophila melanogaster, named wMelPop [26]. In these Wolbachia-infected Ae. aegypti mosquitoes, the Asaia colonization rate was similar to that of the control, represented by the wild type Wolbachia-uninfected Ae. aegypti from the same strain (100 % vs 100 % in the guts, 36 % vs 39 % in the reproductive organs) (Additional file 3: Table C). Even though, these data do not seem to support a competition between the two microbes, the trans-infected strain and the control strain of Ae. aegypti have also been analysed by FISH assay to quantify the rate of infection. By comparing the intensity of the signal in wMelPop infected and uninfected mosquitoes, it was fairly evident that in the presence of Wolbachia, the amount of Asaia in the reproductive organs was lower than that in the absence of Wolbachia (Fig. 4).
Asaia-GFP colonization of reproductive organs of a strain of An. stephensi stably trans-infected with Wolbachia wAlbB The colonization experiments were extended also to a strain of An. stephensi that has been trans-infected with a Wolbachia strain from Ae. albopictus, named wAlbB. The control was represented by the wild type Wolbachiauninfected strain of An. stephensi. In both the Wolbachia infected and uninfected mosquitoes the guts were constantly colonised by Asaia (100 % in both strains), while in the reproductive organs no colonization was detected in Wolbachia-positive mosquitoes. On the other hand, colonization was observed in the Wolbachia uninfected mosquitoes, in 76 % of the ovaries and 20 % of the testis, out of 25 specimens examined per each organ (Table 2).

Discussion
Despite its common presence in arthropods, including several mosquito species, Wolbachia has not been detected in anophelines, with just one recent description in a few mating swarms from a small population of An. gambiae [8]. To verify possible causes of this phenomenon, we have tested the hypothesis that microbial competition may contribute to explain the absence of Wolbachia infection within members of the Anopheles genus. In particular, we have tested the competition that may occur at the reproductive organs, since Wolbachia is a well-known manipulator of host reproduction that exploits effects exerted at the level of the gonads to increase its own fitness [27,28]. Our hypothesis is that microorganisms that should in some way reduce the amount of Wolbachia within the gonads would also reduce the overall capacity of Wolbachia to induce reproductive manipulations, such as cytoplasmic We have focused our attention on Asaia, the only other bacterium known to localize in the mosquito's reproductive organs, and thus a potential competitor of Wolbachia for these anatomical niches.
Our overall results support the hypothesis that a competition between the two bacteria occurs at the reproductive organs, and this competition is particularly evident in the ovaries, that are essential not only for the maternal vertical transmission of Wolbachia, but also for determining phenomena like the rescuing of the sperm in CI, and thus for ensuring the differential fitness of Wolbachia infected VS uninfected females.
First, we have shown that in Cx. quinquefasciatus and Ae. albopictus (that are naturally infected by Wolbachia), Asaia infects the gut, but does not infect the reproductive organs. Our previous works, however, showed that mosquito-species naturally uninfected with Wolbachia (i.e. An. gambiae, An. stephensi, Ae. aegypti) host Asaia in the reproductive organs, as well as in other anatomical districts [17,29,30].
Secondly, the IFA-based comparative investigation, aimed at detecting the presence of Asaia on the eggs surface of An. gambiae, An. stephensi and Ae. albopictus, showed that only this latter species, the only one naturally infected with Wolbachia, did not carry Asaia on the egg surface. This is congruent with the absence of Asaia in the reproductive organs in Ae. albopictus, which is thus to be regarded as a dead-end host for these bacteria, in contrast to what was observed in An. stephensi (this study) and An. gambiae [17], where egg-smearing appears as the mechanism for the vertical transmission of these bacteria.
Thirdly, the colonization with an Asaia-GFP strain was detectable only in the guts, and not in the gonads, in Cx. quinquefasciatus and Ae. albopictus, while this strain of Asaia has already been shown to be very Even in the magnified area no signal of colonization is detected effective in the colonization of the reproductive organs in An. gambiae, An. stephensi, Ae. aegypti [3,25].
Finally, we showed that Asaia-GFP is able to colonize two Wolbachia artificially and stably infected strains of mosquito (Ae. aegypti and An. stephensi). In both cases we observed that an unnatural presence of Wolbachia in guts and reproductive organs does not alter the ability of Asaia-GFP to efficiently colonize the guts of recipient mosquitoes. Conversely, the presence of Wolbachia in the reproductive organs may strongly influence the ability of Asaia-GFP to efficiently colonize the reproductive organs of trans-infected mosquitoes in respect to the control strains. In Ae. aegypti we detected a similar percentage of colonized mosquitoes between Wolbachia infected and uninfected mosquitoes, although Asaia has been shown to be less abundant in the reproductive organs of Wolbachia-positive mosquitoes, compared to Wolbachia-negative ones.
In An. stephensi this difference between Wolbachiapositive and -negative mosquitoes was stronger, in that no colonization was observed in the reproductive organs from Wolbachia-positive mosquitoes, while colonization of reproductive organs in Wolbachia-negative insects was detected in about 75 % of the specimens tested. The above reported differences between Ae. aegypti and An. stephensi may be due to intrinsic species-specific factors and/or to the fact that the two species were infected with different Wolbachia strains: Ae. aegypti with one from Drosophila, the virulent strain wMelPop of Wolbachia, and An. stephensi with one from Ae. albopictus.
All these evidences taken together, underline a role of Wolbachia in preventing some mosquito species from a stable and successful Asaia infection in the gonads. On the other hand, our data seem to support a role of Asaia in the inability of Wolbachia to infect anopheline mosquitoes.
Asaia seems to exert its action specifically at the level of reproductive organs and particularly in ovaries suggesting that these two bacteria may compete for the same anatomical niche or infection routes in the host reproductive organs so that in the presence of Asaia, Wolbachia cannot colonize these organs, or that colonization is in some way limited. As already emphasized, a limited colonization of male and female gonads is likely to reduce the strength of reproductive manipulations such as CI, and, in the case of the female gonad, a limited (or absent) colonization will obviously reduce the efficacy of vertical transmission.
These conclusions are congruent with the results of a recent study described by Hughes and collaborators that have shown that Asaia may impede vertical transmission of Wolbachia in Anopheles. [31]. Conversely, it has recently been demonstrated also that the introduction of wAlbB in An. stephensi reduces female fecundity and causes a minor decrease in male mating competiveness [32]. Table 2 Colonization experiments with Asaia-GFP of a strain of Anopheles stephensi stably trans-infected with Wolbachia (W+) and a wild type Wolbachia-uninfected An. stephensi strain (W − ). Asaia-GFP was provided to mosquitoes with the sugar meal. Twenty five organs were analysed per each strain and relative positivity is reported  4 Asaia and Wolbachia detection in gonads of wMelPop infected (W+) and uninfected (W − ) Ae. aegypti mosquitoes. Detection is shown through interferential contrast microscopy (column 1) and after whole mount in situ hybridization with Asaia (in magenta, column 2) and Wolbachia (in yellow, column 3) specific probes. Bar corresponds to 120 um