The Insect Microbiome Modulates Vector Competence for Arboviruses

Diseases caused by arthropod-borne viruses (arboviruses), such as Dengue, West Nile, and Chikungunya, constitute a major global health burden and are increasing in incidence and geographic range. The natural microbiota of insect vectors influences various aspects of host biology, such as nutrition, reproduction, metabolism, and immunity, and recent studies have highlighted the ability of insect-associated bacteria to reduce vector competence for arboviruses and other pathogens. This reduction can occur through mechanisms, such as immune response activation, resource competition, or the production of anti-viral molecules. Studying the interactions between insect vectors and their microbiota is an important step toward developing alternative strategies for arbovirus transmission control.


Wolbachia and Cytoplasmic Incompatibility (CI)
Bacteria of the genus Wolbachia are maternally inherited, obligate intracellular symbionts that have been estimated to infect 66% of insects [56]. Several arbovirus vectors such as Culex quinquefasciatus and Ae. albopictus are naturally infected with Wolbachia, but not Ae. aegypti [43,[57][58][59]. Recent research, however, has shown that stable transinfection of Wolbachia from Drosophila and Ae. albopictus into Ae. aegypti is possible [41,60,61] and, in fact, has great potential as an arboviral control strategy (described below in Section 3).
Wolbachia spreads quickly through populations because of its ability to alter insect reproduction through mechanisms such as feminization, parthenogenesis, and cytoplasmic incompatibility (CI), which increase the reproductive success of infected insects [62]. In mosquitoes, CI ensures that offspring will be infected by Wolbachia because uninfected eggs fertilized with sperm from infected males will not survive [63,64]. This phenomenon, which is maintained in stably trans-infected Ae. aegypti, is useful for the dissemination of Wolbachia in field mosquito populations [47].

The Insect Eukaryotic Microbiome
In addition to bacterial microbiota, studies have also isolated eukaryotic microorganisms such as fungi and yeast using culture-dependent methods. An early study identified 18 non-pathogenic yeast isolates in the genera Candida, Yarrowia, Rhodotorula, and Cryptococcus from larval and adult stages of Aedes, Culex, and Anopheles mosquitoes [65]. A later study isolated Candida and Pichia yeast from Ae. aegypti midguts [66]. Wickerhamomyces anomalus yeast has also been found in the midgut and reproductive organs of various mosquito species, suggesting a complex eukaryotic microbiome in various tissues [67,68]. These findings are not limited to mosquitoes. For example, 39 fungi were isolated from the cuticle and midgut of five sandfly species, suggesting that eukaryotic microbiota might be common among insects [69].
The eukaryotic microbiota has been much less well-studied than the bacterial microbiota, especially with metagenomic sequencing methods, and further work is required in order for us to fully understand its impact on insect biology and arbovirus transmission.
Paratransgenesis, which involves the genetic modification of insect microbiota to inhibit human pathogens, is considered a promising novel disease control approach [70,71]. The eukaryotic microbiome, especially yeast, have a high potential for paratransgenesis due to their safety, large scale production systems, and available genetic manipulation tools [72][73][74]. Yeasts can be genetically modified to inhibit arboviruses through secretion of antiviral anti-microbial peptides (AMP), such as a cecropin-like peptide possessing anti-DENV and anti-CHIKV activity [75]. Paratransgenesis can also be applied to entomopathogenic fungi, thus maximizing disease control potential through a combination of vector killing and reduction of vector competence. This approach has been employed with the fungus Metarhizium anisopliae, by engineering it to express the SM1 peptide, which inhibits Plasmodium development in Anopheles mosquitoes [76].

Microbiota-Driven Mechanisms Affecting Vector Competence
Insect microbiomes have long been co-evolving with their hosts. Early studies of insect symbionts suggested beneficial roles in nutrition; for example, the gut microbiota of termites greatly facilitate cellulose digestion [77,78]. Recent studies in medically important insect vectors also indicate the importance of microbiota for nutrient digestion, metabolism, egg production, development, and immune responses [23, [25][26][27][79][80][81][82]. In other cases, endosymbionts, such as Wolbachia and Spiroplasma, may require nutrients from the host for efficient replication [83,84]. These interactions have great potential to influence vector competence for pathogens, since arboviruses require host factors and cellular machinery for their replication and are also controlled by insect immune responses. In addition to these indirect effects, microbiota may also directly interact with arboviruses, since some bacteria species are known to secrete anti-viral compounds [85][86][87][88]. A novel Chromobacterium sp. (Csp_P) species isolated from field-caught Ae. aegypti can reduce mosquito susceptibility to DENV infection when introduced to the mosquito midgut tissue [89]. Certain microbiota can also increase vector competence for arbovirus infection [90,91].
The role of the microbiome on mosquito immunity and vector competence was first characterized in Anopheles mosquitoes and Plasmodium parasites [120]. Transcriptomic comparison between septic and aseptic An. gambiae using microarrays identified a number of immune-related genes up-regulated in the presence of midgut microbiota, which subsequently resulted in lower susceptibility to Plasmodium infection in septic mosquitoes. This study provided a fundamental basis for subsequent studies concerning the effect of the mosquito microbiome on arbovirus infection.
The effect of the microbiome on insect vector competence for arboviruses has been studied in DENV and Ae. aegypti [39,109]. Removal of mosquito gut microbiota by treatment with antibiotics results in higher midgut DENV titers [109], and gene expression analysis has revealed that aseptic Ae. aegypti have lower levels of AMP gene expression (attacin, cecropin, defensin, and gambicin), suggesting a lower level of immune activation [109]. DENV infection of Ae. aegypti salivary glands induces the Toll and IMD pathways and results in the expression of a putative cecropin-like peptide with antibacterial, anti-DENV, and anti-CHIKV activity [75]. Subsequently, the bacterium Proteus sp. (Prsp_P), derived from the gut of field mosquitoes, has been shown to up-regulate AMP gene expression and confer increased resistance to DENV infection of the mosquito gut [39]. These results emphasize the overlap between antibacterial and antiviral insect immune responses.
Wolbachia contributes to Drosophila's resistance to virus infection, and trans-infection of Wolbachia from Drosophila to Ae. aegypti also increases the mosquitoes' resistance to DENV, CHIKV, YFV, and Plasmodium infection [57,61,[121][122][123]. Introducing Wolbachia into a new insect host can elicit immune responses, as shown in the trans-infection of wMel and wMelPop from Drosophila to Ae. aegypti [124]. It has, however, also been shown that Wolbachia provides protection against DENV infection in Drosophila without activating Drosophila's immune response [124], suggesting that Wolbachia provides protection against arbovirus infection through both immunitydependent and -independent mechanisms, depending on the combination of Wolbachia strain and insect host. Wolbachia strain wAlbB from Ae. albopictus has also been trans-infected into Ae. aegypti and shown to contribute to DENV resistance [41,125,126]. Gene expression analysis of Ae. aegypti infected with wAlbB has revealed wAlbB-induced production of reactive oxygen species (ROS), which in turn induce the activation of the Toll pathway [126].
In addition to immune signaling, RNA interference (RNAi) is another major insect anti-viral mechanism. In the canonical exogenous small interfering (siRNA) pathway, viral genomes are recognized and degraded based on sequence complementarity, through the action of Dicer2 (Dcr2) and the RNA-induced silencing complex (RISC) [127][128][129]. The components of the exogenous siRNA pathway are constitutively expressed in the cytoplasm, and there is to date no evidence that this mechanism can be activated by microorganisms other than viruses. However, insects also rely on other small RNA pathways, such as the Piwi-interacting RNA (piRNA) [130] and microRNA (miRNA) pathways [131,132], to restrict arbovirus infection. Recent studies have shown that Wolbachia wMelPop-CLA can alter the mosquito's miRNA profile [133], and can also alter mosquito gene expression through the induction of host microRNAs (miRNAs) [134,135]. The induction of miRNA aae-miR-12 promotes the growth of Wolbachia through a down-regulation of DNA replication licensing factor (MCM6) and the monocarboxylate transporter (MCT1) genes; the increased Wolbachia growth then reduces vector competence for DENV in a density dependent manner [134]. Aae-miR-2940, another miRNA induced by Wolbachia, suppresses Ae. aegypti DNA methyltransferase (AaDnmt2) gene expression. The down-regulation of AaDnmt2 again promotes Wolbachia replication but reduces DENV titers in mosquito cells [135].
The impact of microbiota on immune activity and vector competence in insects has also been documented in insects other than arbovirus vectors. The tsetse fly symbiont, Wigglesworthia glossinidia, activates the IMD pathway and inhibits trypanosome parasite infection [136]. Studies of W. glossinidia in tsetse suggest the importance of the microbiota in the larval stages for immune maturation in adult insects. Wild-type flies that lack W. glossinidia during larval development appear to have compromised immune responses such as AMP expression, prophenol-oxidase activity, melanization, and increased hemocyte number [81,137]. These results emphasize the importance of certain microbiota in particular developmental stages for maturation of the insect immune system in adults; however, this phenomenon is yet to be studied in arbovirus vectors.
In some cases, the insect microbiota does not confer resistance to arbovirus infection but instead increases the insects' susceptibility to arbovirus infection. For example, trans-infection of wAlbB Wolbachia to Culex tarsalis increases the susceptibility of the mosquitoes to WNV infection [90]. Gene expression analysis has revealed down-regulation of Rel1, a transcription factor responsible for activating Toll pathway-dependent effectors, suggesting that Wolbachia can suppress insect immune responses [90]. The presence of Serratia odorifera in the Ae. aegypti midgut increases the mosquitoes' susceptibility to DENV infection, possibly through a suppression of immune responses via the binding of prohibitin [91]; however, this possibility has yet to be experimentally confirmed. Other than arboviruses, recent studies have shown that Wolbachia increases mosquito susceptibility to Plasmodium parasite infection [138,139]. One Wolbachia strain can result in either an increase or a decrease of Plasmodium infection, for example, wAlbB reduces P. falciparum infection but increases infection of An. gambiae with P. berghei [139,140]. Environmental factors such as temperature can also affect the outcome of Plasmodium infection when mosquitoes are infected with Wolbachia [141]. For example, Wolbachia wAlbB reduces P. yoelii infection at 28 °C, but increase parasite load at 20 °C. These observations emphasize that the relationship between insect vector, insect microbiota, and human pathogens is far more complex than anticipated, and environmental factor can influence these interaction.

Resource Competition
Certain combinations of Wolbachia strain-insect species do not result in the elicitation of insect immune responses, as reported in D. simulans and Ae. albopictus [142] or in infections of Drosophila with the wAu and wMel strains [143], suggesting that immune activation is not the only mechanism affecting vector competence. Both arboviruses and insect microbiota, especially endosymbionts, such as Wolbachia and Spiroplasma, require nutrients and host factors for efficient replication. The anti-viral effect of Wolbachia in an Ae. albopictus cell line has been shown to be density-dependent [144,145], suggesting that high densities of Wolbachia competing for limited resources can affect vector competence.
Lipids, for example, are required by both the microbiota and arboviruses. Several arboviruses such as DENV and WNV use receptor-mediated endocytosis for cell entry in both vertebrate and invertebrate hosts, a process that involves remodeling of lipid membranes [146][147][148][149][150]. After cell entry, viruses modify intracellular compartments of the host to facilitate protein processing and virus replication and assembly [151,152]. DENV influences expression of genes involved in lipid synthesis to alter the host's lipid composition, lipid homeostasis, and intracellular membrane trafficking [153][154][155]. Wolbachia also uses lipids from host cells for replication and therefore competes with and inhibits DENV and CHIKV replication [61,156].
In the Drosophila and honeybee models, Spiroplasma replication requires lipid and vitamins from its insect host [83,84]. Its requirements in mosquitoes have not been studied, but if similar to those of Drosophila and honeybees, it is plausible that the bacteria may also be able to influence vector competence through Wolbachia-like resource competition mechanisms.

Secondary Metabolite Production
Actinomycetes, bacteria commonly found in mosquito gut, have long been known to secrete secondary metabolites with anti-bacterial, anti-fungal, and anti-viral activity [88,157]. Another bacterium commonly found in soil and water, Chromobacterium violaceum, has also been studied for its anti-viral activity [86,158]. A recently characterized Chromobacterium sp. (Csp_P) isolated from field mosquito guts, has shown a promising potential as vector-borne disease control tool. Csp_P blocks infection of An. gambiae and Ae. aegypti with Plasmodium and dengue virus, respectively, and exerts entomopathogenic activity against larval and adult stages, likely though the production of secondary metabolites [89]. Bacteria isolated from the Ae. albopictus midgut, such as Pseudomonas rhodesiae, Enterobacter ludwigii, and Vagococcus salmoninarium, have been shown to directly inhibit La Crosse virus independently of the mosquito, suggesting that these bacteria may produce anti-viral molecules [159].
These discoveries suggest that certain species of disease vector's natural gut microbiome directly influences arbovirus infection through natural products. Isolation of these bacteria and anti-pathogen molecules may open an interesting avenue for the discovery and development of novel therapeutic drugs.

Field Applications of Insect Microbiota for Arbovirus Transmission Control
The concept of insect microbiota as an arbovirus control tool has great potential, but it also raises numerous practical and safety concerns. In addition to exploring and characterizing anti-viral mechanisms, studies to address the applicability of these microorganisms to the field should be pursued. For example, the microbial composition of field mosquito guts can be far more complex than in those of mosquitoes in laboratory settings, and this complexity may interfere with the proposed arbovirus transmission-blocking strategy. This complexity was addressed in a recent study of the effect of the mosquito microbiome on the ability of Wolbachia to establish itself in a new insect host [160]. Interactions between the microbiota and Wolbachia inhibited transmission of Wolbachia to the next generation, and also resulted in mosquito mortality.
To date, the most advanced field application of insect microbiota for controlling arbovirus transmission is the Eliminate Dengue program in Australia, which has released Wolbachia-infected Ae. aegypti to control DENV transmission [47,60,161]. In this case, Wolbachia successfully invaded the natural mosquito population, and a follow-up study has found that field-caught Wolbachia-infected Ae. aegypti still maintain their refractoriness to DENV [162]. This program has since been expanded to other countries, including China, Vietnam, Indonesia, Colombia, and Brazil [163].
For gut bacteria which, unlike Wolbachia, are unable to drive themselves into a population, achieving sustained delivery to mosquitoes in nature remains an important and understudied practical issue. Existing measures for mosquito population control such as oviposition traps, spraying of toxins or insect pathogens, and artificial nectar bait [164][165][166][167][168][169] can be adapted as dissemination strategies; however, continued release may be required to maintain these microbiota in the mosquito population.
Paratransgenic approaches, while not explored in the area of arbovirus control, have the potential to reduce arbovirus transmission in a number of ways. Microbiota can be engineered to (1) have enhanced entomopathogenic activity [170,171], (2) secrete anti-pathogen molecules (extensively studied for arthropod-borne parasites) [70,71]), or (3) secrete molecules that activate insect immune responses against the pathogen. Research to identify candidate genes and molecules that increase entomopathogenic activity, inhibit arboviruses, and activate insect immune responses is required to move the field forward.
Collectively, research that will pave the way for the use of insect-derived bacteria as an alternative arbovirus control strategy is still at an early stage. Extensive studies are required to ensure safety and effectiveness prior to a field release of insect microbiota as an arbovirus control strategy.

Conclusion
The global burden of arboviral diseases has been rapidly increasing in recent decades. Studies of insect-associated microbial species suggest that they can alter vector competence by modulating host immune responses, competing with arboviruses for resources, and secreting anti-viral factors. Understanding the tripartite relationships between the insect, its microbiome, and the arboviral pathogens it harbors will allow us to develop alternative strategies to reduce the burden of arboviral diseases.