The Impact of Wolbachia on Virus Infection in Mosquitoes

Mosquito-borne viruses such as dengue, West Nile and chikungunya viruses cause significant morbidity and mortality in human populations. Since current methods are not sufficient to control disease occurrence, novel methods to control transmission of arboviruses would be beneficial. Recent studies have shown that virus infection and transmission in insects can be impeded by co-infection with the bacterium Wolbachia pipientis. Wolbachia is a maternally inherited endosymbiont that is commonly found in insects, including a number of mosquito vector species. In Drosophila, Wolbachia mediates antiviral protection against a broad range of RNA viruses. This discovery pointed to a potential strategy to interfere with mosquito transmission of arboviruses by artificially infecting mosquitoes with Wolbachia. This review outlines research on the prevalence of Wolbachia in mosquito vector species and the impact of antiviral effects in both naturally and artificially Wolbachia-infected mosquitoes.


Introduction
Understanding the factors that contribute to the transmission of arboviruses may facilitate strategies to limit the spread of disease. Mosquito transmission of viruses is impacted by interactions between the virus, host and other microbes. Presence of the endosymbiotic bacterium Wolbachia pipientis can interfere with microbial and parasite infection in insects, including viruses in mosquitoes (reviewed in [1][2][3][4]). As a result of this characteristic, there is increased interest in exploiting Wolbachia as a means of biological control of arthropod transmitted infectious pathogens (reviewed in [1,5,6]). This review is focused on the impact of both natural and artificial Wolbachia infection on the outcome of virus infection in vector mosquitoes.

Wolbachia in Insects
Wolbachia is an alphaproteobacterium predicted to infect more than 40% of insect species [7,8]. Wolbachia infection can have a wide range of impacts on insects [9]. An obligate intracellular bacterium, Wolbachia lives in the cytoplasm of host cells and is dependent on host cell resources for replication. The primary transmission route of Wolbachia is vertical inheritance through the cytoplasm of the maternal line, although horizontal transmission between insect species also contributes to Wolbachia prevalence [10][11][12][13][14]. Invasion of invertebrate populations is generally achieved via Wolbachia induced modification of host reproductive systems. Invasion into insect populations can occur very rapidly; for example, Wolbachia swept through Californian Drosophila simulans (D. simulans) populations in three years [15]. The ability to invade populations together with the recent finding Viruses 2015, 7, 5705-5717; doi:10.3390/v7112903 www.mdpi.com/journal/viruses that Wolbachia can interfere with virus transmission has led to interest in utilising Wolbachia to control mosquito transmission of arboviruses [1,3,6]. Cytoplasmic incompatibility (CI) is a prevalent Wolbachia reproductive manipulation in insects [16], which increases the proportion of Wolbachia-infected individuals in the population. Wolbachia-infected females can successfully mate with an uninfected male or male infected with the same or a compatible Wolbachia type (see Figure 1). CI occurs when a Wolbachia-infected male mates with a female that is either not infected with Wolbachia (unidirectional CI) or infected with an incompatible type of Wolbachia (bidirectional CI) [17]. That is, "if the male is infected with an infection (type) that is not present in his mate, it is an incompatible cross" [18]. In mosquitoes, Wolbachia-induced CI skews the population toward Wolbachia-infected females. In contrast to the female gametes, Wolbachia is not present in the male sperm. The molecular events that lead to CI are not completely clear but involve changes in condensation of male chromatin in Wolbachia free zygotes and lack of mitotic synchrony between the parental chromosomes [19][20][21][22]. In diploid insects such as mosquitoes, viable progeny are not produced from these eggs. CI is rescued in Wolbachia-infected eggs, as there is a restoration of synchrony between the male and female chromosomes, therefore producing diploid Wolbachia-infected progeny [20,22]. For biological control approaches, CI can be harnessed to establish Wolbachia-infected populations in the field [23].
Viruses 2015, 7, page-page Cytoplasmic incompatibility (CI) is a prevalent Wolbachia reproductive manipulation in insects [16], which increases the proportion of Wolbachia-infected individuals in the population. Wolbachia-infected females can successfully mate with an uninfected male or male infected with the same or a compatible Wolbachia type (see Figure 1). CI occurs when a Wolbachia-infected male mates with a female that is either not infected with Wolbachia (unidirectional CI) or infected with an incompatible type of Wolbachia (bidirectional CI) [17]. That is, "if the male is infected with an infection (type) that is not present in his mate, it is an incompatible cross" [18]. In mosquitoes, Wolbachia-induced CI skews the population toward Wolbachia-infected females. In contrast to the female gametes, Wolbachia is not present in the male sperm. The molecular events that lead to CI are not completely clear but involve changes in condensation of male chromatin in Wolbachia free zygotes and lack of mitotic synchrony between the parental chromosomes [19][20][21][22]. In diploid insects such as mosquitoes, viable progeny are not produced from these eggs. CI is rescued in Wolbachia-infected eggs, as there is a restoration of synchrony between the male and female chromosomes, therefore producing diploid Wolbachia-infected progeny [20,22]. For biological control approaches, CI can be harnessed to establish Wolbachia-infected populations in the field [23].

Wolbachia-Mediated Antiviral Protection
Wolbachia-mediated antiviral protection was initially discovered in Drosophila melanogaster [24,25]. Flies infected with Wolbachia were protected from virus infection compared to paired groups of flies cured of their Wolbachia infection. In the presence of Wolbachia there is a significant delay in

Wolbachia-Mediated Antiviral Protection
Wolbachia-mediated antiviral protection was initially discovered in Drosophila melanogaster [24,25]. Flies infected with Wolbachia were protected from virus infection compared to paired groups of flies cured of their Wolbachia infection. In the presence of Wolbachia there is a significant delay in the mortality induced by the pathogenic RNA viruses Drosophila C virus (Dicistroviridae, DCV), Cricket paralysis virus (Dicistroviridae, CrPV) and Flock House virus (Nodaviridae, FHV) [24,25]. Accumulation of infectious DCV particles can be dramatically decreased early in infection, although a large change in FHV accumulation is not commonly observed. For the non-pathogenic Noravirus, a delay in virus accumulation was observed in the presence of Wolbachia without any impact on the survival of flies [24]. Thus presence of Wolbachia can have two different impacts on virus: firstly virus accumulation may be reduced/delayed, and secondly virus induced host mortality may be reduced/delayed. Both of these Wolbachia-induced effects are generally referred to as antiviral protection in Drosophila. However, the presence of some Wolbachia strains has no impact on virus-induced mortality or delay in accumulation [26]. In natural pairings, co-evolution of the host and Wolbachia mean that the contribution of each partner may be important for the outcome of antiviral protection.
In protective Wolbachia-host combinations, Wolbachia are found at high density. Different host-Wolbachia combinations have variable Wolbachia density in their insect hosts [27][28][29][30]. Protective Wolbachia are found at higher density than non-protective strains in D. simulans [26]; and if the Wolbachia density is experimentally decreased then protection is lost [27]. In Drosophila, Wolbachia density tends to be positively correlated with strength of antiviral protection and is sufficient to explain most of the variation in protection between Wolbachia strains [31,32]. In mosquitoes and mosquito cells Wolbachia density has also been implicated in antiviral effects [33][34][35][36]. The link between high Wolbachia density in the host and antiviral effects, leads to the hypothesis that all Wolbachia strains are capable of antiviral protection if a density threshold is reached, although this is yet to be tested experimentally.
The majority of studies assay Wolbachia density in whole insects; however, the importance of Wolbachia density to the outcome of infection may be at a cellular or tissue level. In addition to localisation in the germline tissues, Wolbachia has been identified in a number of somatic tissue types across insect species [27,28,30,[36][37][38][39][40][41]. Wolbachia distribution is influenced by both host and Wolbachia factors (reviewed in [42]) and density can vary across tissues within a host [27]. There is some evidence to suggest that presence and density of Wolbachia in cells/tissues that are the site of arbovirus replication is important in determining the outcome of infection [35,36].
The potential mechanisms of Wolbachia-mediated antiviral effects are not completely clear and have been discussed in recent reviews [2,3]. The focus of this review is the incidence of antiviral effects in mosquitoes infected with Wolbachia either naturally or artificially, but a summary of the research of mechanisms is warranted. The diverse variety of viruses affected by Wolbachia infection of insects, suggests that the mechanisms are not likely to target interactions specific to one type of virus. It is also likely that the mechanisms are largely independent of Wolbachia strain given the importance of density and lack of phylogenetic congruency in strains that protect Drosophila [31].
The role of Wolbachia density in antiviral effects suggests there may be competition for resources between the virus, host and Wolbachia or remodelling of the host cell environment and there is some evidence for this competition [3,26,35,36,55]. Another potential mechanism for antiviral effects is immune stimulation by the presence of Wolbachia, in particular this has been implicated in mosquitoes transinfected with Wolbachia [36,49,[56][57][58]. Interestingly, an increase in reactive oxygen species is observed in both mosquitoes and Drosophila infected by Wolbachia, and this is coincident with protection in Drosophila [58,59]. In addition, a role for microRNAs in controlling Wolbachia-mediated antiviral effects has been proposed [60,61]. Research on mechanisms of Wolbachia-mediated antiviral effects may be confounded by the difference between host species or mode of Wolbachia infection and further experimentation is required to delineate the important mechanisms.

Arboviruses in Mosquitoes
A subset of the over 3000 species of mosquitoes vector viruses that cause human disease ( Table 1). The major mosquito-borne human pathogenic viruses come from three RNA virus families: Flaviviridae genus Flavivirus, Togaviridae genus Alphavirus and Bunyaviridae genera Orthobunyavirus and Phlebovirus (Table 1) [62]. These viruses are often found in both mosquitoes and animals, with virus replication occurring in both host types. Human disease occurs following transmission of the virus via a bite from an infected mosquito. For many of these viruses, humans are a dead-end host with viral population being necessarily maintained in other animal hosts. Viruses including dengue virus (DENV) and chikungunya virus (CHIKV) have adapted to a human-mosquito transmission cycle and no longer require amplification in other animals. While there are a complex range of factors that determine vector competence for transmission of arboviruses in mosquitoes [63], the presence of Wolbachia could influence vector competence by altering mosquito susceptibility to virus infection. * For references see [62,64] and references there in.

The Distribution of Wolbachia in Vector Mosquitoes
Wolbachia was first discovered in the mosquito Culex pipiens [65] and is present in populations of various wild mosquito species. Surveys focused on disease transmitting mosquito genera identified Wolbachia in 7%-42% of the Culex pecies analysed and 0%-30% of the Aedes species analysed; until recently, no Wolbachia was detected in any of the tested Anopheles species [66][67][68]. It is interesting to note that Wolbachia is frequently detected in several of the common arbovirus vectors including the Culex pipiens complex and Aedes species including Ae. albopictus but not Ae. aegypti. The establishment of Wolbachia in some mosquito species can be impeded by the native microbiome and this may in part explain the absence of Wolbachia in some mosquito species in nature [69]. Improved methods of detection will likely lead to detection of Wolbachia in a wider variety of species. For example Wolbachia was recently detected for the first time in a limited number of Anopheles gambiae mosquitoes using high throughput sequencing of the 16s rRNA gene amplified from field caught mosquitoes [70]. The presence of Wolbachia in various arbovirus mosquito vectors raises the question of the impact of Wolbachia infection on arbovirus transmission in natural vector populations.
The ability to experimentally transfer Wolbachia into naïve hosts can create new vector-Wolbachia associations that do not occur in nature. This is attractive as a way of introducing Wolbachia into Wolbachia-free mosquitoes such as Ae. aegypti. In the laboratory Wolbachia can be transferred between insects by a process called transinfection (see [71]). Wolbachia is extracted from infected donor insects and injected into naïve insects. A stable transinfected insect line is established if the Wolbachia productively infects the female gonads and is passed from one generation to the next. Stable transinfection of Wolbachia into mosquito species is challenging, but has been successfully achieved for several species of mosquitoes including: Ae. aegypti, Ae. albopictus, Ae. polynesiensis, Cx. pipiens, and An. stephensi [41,50,[72][73][74][75][76][77][78][79][80][81][82][83]. In addition, mosquitoes can be transiently transinfected by injection of Wolbachia into adult mosquitoes. In both cases Wolbachia invades various tissues of the mosquito and can be recognised as foreign, therefore stimulating the host immune responses [54,56,84].
The creation of a new stable association between a host and Wolbachia strain and can lead to phenotypic and genetic changes. In naturally infected insects, maternal inheritance of Wolbachia across many generations maintains a close association between the host and symbiont, leading to co-evolution and stable interactions. In comparison, theory predicts that new host-Wolbachia associations are likely to be maladapted [85], and artificial transfer of Wolbachia to a new host is known to induce novel host phenotypes [86][87][88], and can also result in a burst of changes in the Wolbachia genome [89]. However, adaptation in the new host can occur relatively rapidly [86,88].

The Intrinsic Effects of Wolbachia on Virus Infection in Mosquitoes
The presence of Wolbachia in mosquitoes has varied impacts on arbovirus infection. This is a complex tripartite system with contributions from the host, Wolbachia and virus on the outcome of virus infection. In addition, Wolbachia infections in the mosquitoes analysed in the laboratory can either be naturally occurring or introduced by transinfection. Thus studies of antiviral effects in mosquitoes may be confounded by the Wolbachia infection mode [2].
Wolbachia-mediated antiviral effects have been well documented in stably transinfected mosquitoes ( Table 2). The mosquito species Ae. albopictus, Ae. polynesiensis, and Ae. Aegypti have been stably transfected with one or more of three Wolbachia strains (wMel, wMelPop and wAlbB). These studies have included viruses from the families Flaviviridae (WNV, DENV and YFV) and Togaviridae (CHIKV) ( Table 2). Arboviruses from the Bunyaviridae family are yet to be analysed in mosquitoes. A range of parameters can be examined for Wolbachia antiviral effects, these include: measuring the number of virus infected individuals, measuring virus load in whole or parts of mosquitoes, measuring dissemination and measuring virus in saliva as a proxy for transmission. In contrast to Wolbachia-mediated protection in Drosophila, since arboviruses have little impact on the survival of mosquitoes, protection against virus-induced mortality is not documented in mosquitoes. The impact of the presence of stably transinfected Wolbachia on the outcome of virus infection can range from a modest reduction in rate of infection amongst individuals or virus accumulation within infected individuals, to near complete interference with virus replication and transmission. Studies differ in the parameters tested and methods used, so it is difficult to make comprehensive comparisons across study systems. For example, in several cases antiviral effects were more prominent when virus was delivered orally rather than by injection [33,46]. However, in all cases where mosquito lines stably transinfected with Wolbachia have been analysed the presence of Wolbachia has decreased virus infection in at least one of the evaluated parameters [33,36,41,46,49,50,53,90].
Transient transinfection of Wolbachia into adult mosquitoes can lead to enhancement of virus infection. WNV infection rate was enhanced following transient transinfection of wAlbB into Cx. tarsalis mosquitoes [54]. Interestingly once the mosquitoes were infected with the virus, there was no impact of Wolbachia on virus accumulation, dissemination or transmission. This is the only enhancement of virus infection in mosquitoes that is linked to the presence of Wolbachia. It should be noted that transient transinfection is very different to natural stable infections where the host and Wolbachia have co-adapted to each other over many generations. It will be interesting to see if the enhancement in infection rate is limited to transiently transinfected mosquitoes. Cx. tarsalis mosquitoes are naturally Wolbachia-free [68] and stable transinfections of this mosquito have not been achieved so to date this comparison cannot be made. The possibility of arbovirus enhancement is an important consideration given there is one example of a natural Wolbachia infection stimulating increased susceptibility to a DNA virus in the African armyworm [43] and other examples of Wolbachia-induced enhancement of plasmodium infection in mosquitoes [84,[91][92][93].
In contrast to stably transinfected mosquitoes, those naturally infected with Wolbachia do not ubiquitously exhibit antiviral effects ( Table 2). Ae. albopictus mosquitoes naturally co-infected with wAlbA and wAlbB have similar total CHIKV or DENV loads to Wolbachia-free mosquitoes [34,51,52] and a small impact of Wolbachia on dissemination to the salivary glands was noted in one study for DENV [52]. Wolbachia-mediated antiviral effects for CHIKV were not stimulated by wAlbA and wAlbB introgressed into a new Ae. albopictus host background. The results of two studies on Wolbachia antiviral effects in naturally infected Culex mosquitoes were contrasting. While wPip was shown to mediate reduced WNV loads and transmission in Cx. quinquefasciatus mosquitoes [94], no effects of natural Wolbachia infection were detected for Cx. pipiens mosquitoes infected with WNV [95]. Interestingly, the two studies were performed by the same research group and they identified that the laboratory population of Cx. quinquefasciatus used in the original study had much higher somatic density of Wolbachia than the recently caught Cx. pipiens mosquitoes used in the second study [95]. Further analysis of somatic Wolbachia density in recently caught Cx. quinquefasciatus mosquitoes was even lower than that of the Cx. pipiens. This suggests that while presence of Wolbachia in Cx. quinquefasciatus can lead to reduced vector competence, the Wolbachia density in natural Culex populations may not be high enough to support these antiviral effects. As a consequence, Wolbachia may not impact vector competence in the field. Taken together, these studies suggest that Wolbachia may not have a major impact on competence of mosquitoes with a naturally occurring Wolbachia infection to transmit arboviruses; however, a limitation is that there are few studies on recently caught populations of mosquitoes and further research in this area is required before conclusions can be made.
Comparison of transinfected and naturally infected mosquitoes may give insight into factors important for Wolbachia-mediated antiviral effects. Robust antiviral effects induced by Wolbachia strain wMel transinfected into Ae. albopictus shows that this host can support Wolbachia antiviral effects [50,90]. However, natural wAlbA and wAlbB infection in this same host has either no or little impact on virus infection [34,51,52]. It is interesting that wAlbB transinfected into a different host, Ae. aegypti, is able to induce antiviral effects [49]. Thus, while both Ae. albopictus and wAlbB are individually competent partners for Wolbachia-mediated antiviral effects, no antiviral effects are demonstrated with this host-Wolbachia combination. This indicates that it is not a feature of either host or the Wolbachia strain per se which determines antiviral effects, but involves the interaction between the two. Transinfection of the Wolbachia into a new host in both these cases has stimulated antiviral effects. In mosquitoes there are two common effects of transinfection combinations that have antiviral effects: increased Wolbachia density and increased immune stimulation. Immune stimulation has been noted following stable transinfection of mosquitoes, and has been proposed as a potential mechanism for antiviral effects [36,49,[56][57][58]. Wolbachia infection also induces reactive oxygen species in transinfected Ae. aegypti [58] and naturally infected Drosophila [59]. An increase in reactive oxygen species corresponds with Toll pathway restriction of DENV infection leading to the suggestion that Wolbachia mediates anti-DENV effects through stimulation of the Toll pathway [58]. Contrasting this, broad immune stimulation is not observed in Wolbachia-mediated antiviral protection in Drosophila, either in naturally infected or heterologous transinfected flies [44,96,97].
In addition, no factors that influence Wolbachia-mediated protection have been identified by experimental disruption of Drosophila immune pathways including the Toll pathway [98][99][100]. Involvement of the Toll pathway in Wolbachia-mediated antiviral effects in transinfected mosquitoes could be directly tested through analysis of virus infection in Toll impaired mosquitoes.
Density is key to antiviral effects both in Drosophila and mosquitoes. The implication from several studies is that antiviral effects in transinfected mosquitoes are linked to an increase in Wolbachia density [39,90].
The principle of density being important for Wolbachia-antiviral effects is also well supported from both direct and indirect studies in Drosophila and cell culture systems [26,27,31,32,34,35]. It is not currently clear whether stable transinfection itself leads to an increase in density, or whether other factors such as a new host-Wolbachia association is involved.
Whether Wolbachia-mediated antiviral effects in mosquitoes will attenuate over time following transinfection remains to be determined [101]. Reduction in antiviral effects may eventuate if adaptation of Wolbachia to the transinfected host leads to a decrease in Wolbachia density or an attenuation of features that lead to the antiviral effects. Alternatively, the virus could evolve to "escape" the antiviral mechanisms mediated by Wolbachia. There is strong evolutionary pressure for the virus to overcome the Wolbachia antiviral effect and lack of evidence of strong effects in natural mosquito populations may suggest that evolutionary adaptation will lead to reduction of antiviral effects [101].

Conclusions
Currently, there is a contrast of antiviral effects reported in naturally versus artificially infected mosquitoes. The current literature suggests that natural infection of vector species with Wolbachia may not have widespread impact of arbovirus transmission. However, there are few studies on natural populations so it is not appropriate to draw strong conclusions at this point. In contrast, stable transinfection of Wolbachia into heterologous mosquito hosts clearly produces antiviral effects against arboviruses including DENV, WNV, YFV and CHIKV. These antiviral effects are likely related to increased Wolbachia density and possibly immune stimulation in the new host, although direct evidence for this is lacking. If antiviral effects are stimulated by the new Wolbachia-host association, it is possible that as adaption occurs these effects may decrease, which will be an important consideration for release of artificially infected mosquitoes as biocontrol for arbovirus transmission.