Arboviral infections are becoming increasingly aggressive on a global scale due to climate change, global travel and the development of insecticide resistance in vectors. They are vectored by mosquitoes, ticks, sandflies or biting midges. Among these, mosquito-borne viruses contribute heavily to the disease burden in human populations especially in developing countries (Franklinos et al., 2019). Flaviviruses such as dengue, Zika and West Nile and alphaviruses such as chikungunya are amongst the most common mosquito-borne viruses causing human disease. Dengue viruses (DENVs) are responsible for ~400 million infections each year and more than a quarter of the global population lives in endemic areas (Bhatt et al., 2013). Zika virus (ZIKV) has caused severe epidemics and according to the World Health Organization (WHO), a total of 86 countries and territories have reported ZIKV cases to date (World Health Organization, 2022). In addition to being transmitted by the mosquito vector, ZIKV can be transmitted from mother to fetus during pregnancy, through sexual contact, blood transfusions and organ transplants thus widening its transmission capacity. It is estimated that during the last epidemic in 2015-2016, approximately 1.5 million people were infected by ZIKV in Brazil with over 3,500 microcephaly cases reported (European Center for Disease Prevention and Control, 2015). West Nile virus (WNV) is another important agent causing disease in both humans and horses and is the most common etiological agent of viral encephalitis (Chancey et al., 2015). There is currently no vaccine available for WNV. Besides these flaviviruses, chikungunya virus (CHIKV) is an alphavirus that has caused severe outbreaks in Asia, Africa, Americas, and Europe making it a public health concern globally (World Health Organization, 2020). As of October 2022, nearly 3,400,000 chikungunya cases and 70 deaths have been reported globally with Brazil having the most cases (European Center for Disease Prevention and Control, 2022). During the massive outbreak in 2005-2006 in La Reunion Island, the virus acquired the ability to transmit via its secondary vector, Aedes albopictus due to an amino acid change in the E1 glycoproteinintheEast-

Central-South African genotype of the virus. In addition, other mutations in the E1 and E2 glycoproteins have further increased mosquito infectivity of the virus (Tsetsarkin et al., 2007).

Ae. aegypti (Diptera: Culicidae) is the major mosquito vector that transmits the viruses discussed above. These mosquitoes inhabit various regions of the world including both tropical and subtropical areas across several continents including Asia, Africa, North and South America, Europe and Australia (Kraemer et al., 2015). Since they are anthropophilic, these mosquitoes are well adapted to rapid urbanization and prefer artificial water containers for egg laying (Scott and Takken, 2012; Kraemer et al., 2015). Interestingly, the geographical distribution of the mosquito vector is temperature dependent (Brady et al., 2012; Brady et al., 2014; Kraemer et al., 2015) and it has been predicted that the mosquito habitats will be expanded to currently more temperate regions due to climate change (Khormi and Kumar, 2014; Mweya et al., 2016). Therefore, it is anticipated that even larger human populations will be exposed to these disease carrying vectors in the future.

Living organisms are vastly diverse. Every organism has signature characteristics in morphology, anatomy and physiology. Further, there is significant diversity in factors such as behavior and ecology. On the contrary, organisms are also remarkably analogous to each other based on fundamental traits at the molecular level. These similarities are basically mirrored in metabolism and biochemical mechanisms of inheritance (Robert Burger et al., 2021). Metabolites are universal molecules that do not vary across species or ecological barriers. These molecules are reflective of the output of genetic expression (DNA/RNA/protein interactions). Metabolic adaptations occur within an organism when in need of maintaining energy homeostasis and development under different environmental stimuli (Koyama et al., 2020). Therefore, metabolism is intimately associated with the biology of an organism. Investigating mosquito metabolism can provide the blueprint of a mosquito’s response to stimuli such as a viral infection, the changing microbiome, insecticide resistance and environmental changes. Importantly, metabolites can be traced back to identify the genotype of a particular phenotype helping to understand the molecular basis of biochemical responses.

Lipid metabolism in mosquitoes has been studied since the mid-1900s. Due to the lack of advanced technologies, these studies focused on the response of mosquito lipids to different diets, the conversion of food to fat, storage of fat in the fat body of the insects, the utilization of fat for energy during flight, metamorphosis, starvation, and the deposition of fat for oogenesis. The recent availability of the genome sequences of certain mosquitoes, advanced molecular biology techniques and the advent of systems biology approaches especially techniques in metabolomics have helped us understand lipid metabolism at a molecular level in mosquitoes as well as other insects. The first half of this review will focus on the findings from the late 1900s to the early 2000s, on the utilization of lipids in several physiological processes of mosquitoes. The second half of this review will focus on the discovery (or re-discovery) of mosquito lipids from the molecular biology/omics era.

Eggs are laid in water (oviposition). Being anautogenous insects (ex: Ae. aegypti), female mosquitoes require a vertebrate blood meal in order to produce eggs (Clements, 1992). Studies on eggs of other insects like Manduca sexta (tobacco hornworm) have shown that approximately 40% the dry mass of eggs represents lipids. Most of these lipids are acquired through maternal depositions and only 1% is generated in the egg (Canavoso et al., 2001). Provided that lipids have various functions, distribution of lipids in suitable tissues in the developing embryo is important for the emerging neonates (Atella and Shahabuddin, 2002).

After consumption of an adequate blood meal to facilitate ovarian development, female mosquitoes engage in the quest of finding suitable oviposition sites. Different mosquito species have varying preferences for oviposition sites. Ae. aegypti mosquitoes prefer freshwater habitats for egg laying while some species of Culex lay eggs in a wide range of sites from salt marshes to artificial containers. As reviewed by Bentley and Day, mosquitoes select their oviposition sites based on chemical and physiological cues at the site (Figure 1B) (Bentley and Day, 1989). Experiments on gravid colony Ae. aegypti mosquitoes have reported olfactory responses to fatty acid esters (Perry and Fay, 1967). A different study (Davis, 1976) reported one of the fatty acid esters, methyl propionate, as an active chemoattractant in oviposition. Additional studies have reported the influence of metabolites such as 7,11-dimethyloctadecane produced by the bacterium Pseudomonas aeruginosa as an oviposition attractant (Ikeshoji et al., 1975; Ikeshoji et al., 1979). According to a review by Bentley et al., there are multiple cues that are either of larval, pupal or adult origin that influence oviposition site selection in mosquitoes. These could be metabolites released from the previous life stages signaling the newly gravid mosquitoes the suitability of the site for safe oviposition (Bentley and Day, 1989).

Eggs hatch to produce larvae that undergo four instar stages before developing into a non-feeding pupa. Neonate larvae acquire most of the lipids from the mother through the maternal deposition of lipids in eggs (Ziegler, 1997; Ziegler and Ibrahim, 2001; Atella and Shahabuddin, 2002). As discussed previously, there is a considerable functional variability between lipids. Sequestering appropriate lipids in suitable sites in the embryo is critical for the health of neonates (Atella and Shahabuddin, 2002). Using fluorescently labeled fatty acids and phospholipids, Atella and Shahabuddin were able to track the distribution of maternal lipids in developing mosquito eggs and larvae. They found that fatty acids were distributed along the sides of the larval body especially where the muscles are located, while phospholipids aggregated along the intestinal gastric caeca (Figure 1C) (Atella and Shahabuddin, 2002). The authors justify this distribution mentioning the different functions owned by the lipids. Fatty acids deposited alongside the larvae body, especially in association with muscles are assumed to provide energy to support locomotion and rapid movements of newly emerged larvae to find food. Maternal phospholipids that are accumulated in the motile gastric caeca secrete lubricants into the lumen of the gut. These are possibly aiding the neonate larvae in assimilating ingested food. Larvae further acquire lipids, especially the essential polyunsaturated fatty acids, from aquatic food sources such as diatoms and algae (Sushchik et al., 2013). These lipids are required for the proper functioning of innate immunity, developmental processes, and the ability to fly in their adult stage (Dadd and Kleinjan, 1979; Dadd et al., 1987; Stanley and Miller, 2006). Fatty acids that are acquired during larval stages are transferred to the adult stages. However, during metamorphosis, fatty acid conversions occur where eicasopentanoic acid (EPA) and Arachidonic acid (AA) are transferred to the adult mosquito from the triacylglycerol (TAG) stores of the larvae to generate more polar lipids (Sushchik et al., 2013). Further, studies have shown how larval diet can alter vector competence in the adult Ae. aegypti mosquito (Nasci and Mitchell, 1994; Muturi et al., 2011). Female Ae. aegypti mosquitoes developing from larvae that are fed with a high nutrient diet have presented larger body size which is related with a greater metabolic reserve (Briegel, 1990). The feeding success of these female mosquitoes on vertebrate hosts is also significantly greater. This suggests that better nutrition at larval stages can impact the adult vectorial capacity. Moreover, studies by Silva et al., 2021 has shown how higher larval rearing densities can elevate stored TAG levels within adult mosquitoes and also influence the size and fecundity of the mosquito (Silva et al., 2021).

The pupae stage is a non-feeding period solely relying on energy stored at the larval stage. These energy reserves thus determine the ability of a newly emerged adult to survive, reproduce and transmit disease. If metabolic rate is determined by the rate of oxygen consumption of the organism, pupae show the least rate in comparison to larvae and the adult stages (Chapman, 2013). However, this metabolic rate fluctuates with time where it is initially high and then drops before it rises again by the time of eclosion (Chapman, 2013). The trend of metabolic variation is known to have a characteristic ‘U’ curve indicating the fall of metabolic rate initially at histolysis followed by an increase at histogenesis and differentiation (Figure 1D). Interestingly, the main source of energy is gained via fats (supplemented with a smaller portion of carbohydrates) during the pupal period (Chapman, 2013).

Adult mosquitoes are able to synthesize lipids from carbohydrate (sugar) meals (Ziegler and Ibrahim, 2001). Both males and females possess de novo fatty acid biosynthesis machinery, such as fatty acid synthase (FAS) and Δ-9 fatty acid desaturase enzymes, but males have higher FAS gene expression than females to fulfil lipid requirements (Figure 1E) (Jenkin et al., 1975; Sushchik et al., 2013; Chotiwan et al., 2022). By feeding on sugar meals alone, the females are capable of increasing their lipid content up to 300µg within 5 days (Ziegler and Ibrahim, 2001).

A blood meal taken by a female mosquito can induce a pronounced metabolic change in its physiological status (Figure 2A). This phenomena is known as a ‘ metabolic switch’ (Das De et al., 2022). Das De et al. compared the transcriptome of sugar-fed and blood-fed mosquitos using RNA-seq and showed that feeding blood (a high protein diet), induced expression of transcripts in the brain that are related to mitochondrial function and energy metabolism (Figure 2B) (Das De et al., 2022). Blood meals can also serve as an indirect source for lipids. In blood, lipids compose only about 4% of the nutrients (Lehane, 1991). Although no direct evidence has shown that the mosquito midgut epithelium can directly absorb lipids from the blood meal, increases in the expression of the genes that encode the proteins that absorb lipids from food, fatty acid binding protein and long chain fatty acid transport protein, have been reported (Sanders et al., 2003). The other 95% of nutrients in the blood meal are protein (Lehane, 1991). Using [¹⁴C]-labeled protein meals, Zhou et. al., have shown that approximately 30% of blood meal amino acids were oxidized to CO₂ or excreted as waste (Figure 1F) (Zhou et al., 2004a). To determine how female Ae. aegypti mosquitoes detoxify ammonia that is generated during the oxidation of amino acids in a blood meal, mosquitoes were fed with labeled ¹⁵NH₄Cl. The labeled ¹⁵N was traced in the whole mosquito body using electrospray ionization (ESI)-mass spectrometry and stable label isotope tracing (Horvath et al., 2021). The study showed that ¹⁵N was rapidly incorporated into glutamine (Gln) via glutamine synthase (GS) and with the aid of other enzymes, additional N-containing metabolites were generated in the mosquito (Horvath et al., 2021). However, 16% of the meal was converted to TAG, the storage lipid (Zhou et al., 2004a). The expression of several genes involved in lipid synthesis also increased after the blood meal was taken (Sanders et al., 2003). This finding provides more evidence that the blood meal can serve as a source of lipid reserves in mosquitoes. The reserve lipids are required for several physiological processes, such as oogenesis, metamorphosis, diapause, and prolonged flight (Zhou and Miesfeld, 2009; Arrese and Soulages, 2010; Sushchik et al., 2013). They also serve as a source for fatty acids which are precursors for synthesizing eicosanoids, pheromones, glycerophospholipids (GPs) and wax (Arrese and Soulages, 2010).

In the adult mosquito, the fat body is the central location for lipid synthesis, storage and degradation for energy production (Figure 2C) (Arrese and Soulages, 2010). It is an organ composed of loose tissues distributed throughout the insect body, lining the underneath of the cuticle and surrounding the gut and reproductive tissues (Dettloff et al., 2001). The majority of the cells in the fat body are adipocytes. These cells contain numerous lipid droplets which serve as the center of cellular lipid storage and energy metabolism (Olofsson et al., 2009). More than 50% of the dry weight of the fat body are lipids (Ziegler, 1991). The fat body also stores carbohydrates in the form of glycogen which constitutes about 25% of the dry weight. The rest of the carbohydrates (> 50% of intake glucose in Ae. aegypti) are oxidized or converted to lipids (Ziegler, 1991; Zhou et al., 2004b). This organ also serves as a source for synthesizing most of the hemolymph proteins. These proteins include lipophorin, the protein that is responsible for transporting lipids between cells or tissues, and vitellogenin, the protein that is required for egg maturation during oogenesis (Ziegler and Vanantwerpen, 2006).

Nutrients that are absorbed in the gut are transported to the fat body and converted to glycogen and lipids (Arrese and Soulages, 2010). Muscle cells only contain a small amount of energy reserves. As a result, the energy required for prolonged flight is provided by the fat body (Kaufmann and Brown, 2008). Less than 1% of lipids in eggs are locally synthesized. More than 80% of lipids in eggs are transferred from the fat body (Figures 2C, D) (Ziegler, 1997; Ziegler and Ibrahim, 2001).

Lipophorin is the main hemolymph lipoprotein. It plays a role as a reusable shuttle transporting lipids between tissues (Ziegler and Vanantwerpen, 2006). Similar to human high-density lipoprotein (HDL) and low-density lipoprotein (LDL), there are high- and low-density lipophorins (HDLp and LDLp) in insects. LDLp contains up to 63% of the lipids, while HDLp contains 30-50% of the lipids (Beenakkers et al., 1985). Apolipophorin I and II are integral components of the lipophorin particles whereas Apolipophorin III is transiently associated with the lipophorin particle (Van der Horst and Ryan, 2017). Lipophorins in most insects are enriched in diacylglycerol (DAG). However, lipophorins in mosquitoes and some other dipterans, but not Drosophila melanogaster (D. melanogaster) are enriched in triacylglycerol (Pennington et al., 1996; Pennington and Wells, 2002). The mechanism/s of lipid uptake from lipophorins into the oocytes are still unclear. Both receptor-mediated endocytosis of the intact lipoprotein particles and extracellular hydrolysis of lipids from the lipoprotein core have been observed (Ziegler and Vanantwerpen, 2006).

Female mosquitoes require a considerable amount of energy and intense metabolic support during reproduction. The gonadotrophic cycle (egg production cycle) of an Ae. aegypti female is regulated by altering titers of two major hormones, juvenile hormone (JH) and a steroid hormone called 20-hydroxyecdysone (20E) (Attardo et al., 2005; Roy et al., 2015). The gonadotropic cycle has two phases. In the first phase, the posteclosion (PE) phase, JH regulates the development of the mosquito which drives physiological functions related to egg maturation and blood digestion. The female mosquito is physiologically prepared for blood meal digestion and egg maturation by this hormone (Wang et al., 2017; Ling and Raikhel, 2021). The fat body and ovaries need to be exposed to JH in order for the synthesis and accumulation of yolk protein precursor vitellogenin (Vg) (Gwadz and Spielman, 1973) following PE period (previtellogenic maturation), the vitellogenic phase starts with the mosquito taking a blood meal. During this post blood meal phase (PBM), the titers of JH are reduced while 20E titers are increased (Arrese and Soulages, 2010). Cholesterol ingested in a blood meal acts as a precursor of 20E synthesis (Clayton, 1964; Ekoka et al., 2021). Cholesterol stored in the prothoracic glands of the mosquito larvae and pupae can also be used to synthesize 20E (Jenkins et al., 1992). 20E regulates and supports blood meal digestion and egg development in female mosquitoes (Gabrieli et al., 2014; Wang et al., 2017). Interestingly, pathways related to carbohydrate metabolism were shown to be upregulated during the peak of 20E synthesis in females (18-24hPBM) (Hou et al., 2015). Later, Dong et al., have showed that 20E regulated the carbohydrate metabolism through a nuclear transcription factor HR38 (Dong et al., 2018). Additionally, studies by Hou et al. have demonstrated how major carbohydrate metabolic pathways (glycolysis, glycogen and sugar metabolism and the citrate cycle) were considerably downregulated in the mosquito fat body at PE (Hou et al., 2015). However, these pathways were upregulated at the PBM stage. In addition, TAG levels were also decreased at PE but elevated PBM.

Lipids that are synthesized in the fat body are transported and deposited in eggs. These lipids contribute to about 35% of the weight of Ae. aegypti oocytes (Troy et al., 1975). It should be noted that lipids that are synthesized from carbohydrate meals are not sufficient to trigger the maturation of oocytes. Blood meals, or to be specific, amino acids in the meal, are needed to trigger the release of vitellogenin stimulating hormone in the ovaries to initiate the maturation process of the oocytes (Hagedorn et al., 1979). Accumulation of lipids in the oocytes starts only after a blood meal is taken (Troy et al., 1975). Although ovaries are capable of synthesizing complex lipids, especially GPs, less than 1% of locally synthesized lipids were found in the egg (Ziegler and Vanantwerpen, 2006). Using radioactively labeled lipids, Ziegler et. al., found that the majority of the lipids in eggs were TAG that was transferred from the fat body (Ziegler, 1997; Ziegler and Ibrahim, 2001; Sanders et al., 2003).

Ae. aegypti possess a mechanism to maintain metabolic homeostasis during the gonotrophic cycle. Zhou et al., did not observe differences in lipid and protein content and the number of eggs laid from females that underwent starvation before a blood meal (Zhou et al., 2004b). However, they observed significantly lower lipid and glycogen content in the mother after the eggs were laid. This indicates a trade-off between fecundity of the mother and the quality of the eggs. Although a significant portion of lipids accumulating in the oocytes from the first gonotrophic cycle comes from larval food and pre-existing maternal stores (Zhou et al., 2004b), the ability to de novo synthesize fatty acids is still important to produce viable eggs. Transient knockdown (KD) of two key enzymes in the de novo fatty acid biosynthesis pathway, acetyl-CoA carboxylase (ACC) and fatty acid synthase (FAS), caused a significantly lower number of eggs in the first gonotrophic cycle (Alabaster et al., 2011). Eggs that were produced from ACC-deficient mosquitoes also lacked eggshells and were nonviable.

As discussed in this review, fatty acids are a group of vital lipids in mosquitoes serving as structural components in cellular membranes, energy homeostasis, signaling, innate immunity and reproduction. Fatty acid synthesis is conducted by a multifunctional enzyme complex called the fatty acid synthase complex (FAS) (Maier et al., 2008). In Ae. aegypti several paralogues of FAS were found (Chotiwan et al., 2022). The study investigated the dynamic expression of FAS genes in relation to developmental stages. Larval and pupal stages showed negligible FAS expression in comparison to adult stages (Chotiwan et al., 2022). This is consistent with the fact that larvae and pupae utilize maternal lipids deposited in the eggs and do not need to synthesize fatty acids in early life stages. All FAS genes except one isoform were highly expressed in adult male mosquitoes in comparison to other life stages. Since male mosquitoes do not feed on blood, they solely depend on nutrients taken up in a plant meal (nectar) and need FAS function to fulfil their lipid requirements (Chotiwan et al., 2022).

Sugars obtained in the diet of mosquitoes are partially hydrolyzed by enzymes in the saliva when stored in the crop. During this temporal storage, salivary enzymes partially hydrolyze the ingested sugars to produce hexoses. Clements describes how these end products are utilized in the synthesis of glycogen (glycogenesis), fatty acids and triglycerides (lipogenesis) in proportions that are species and life stage specific (Clements, 1992). Several studies have also extensively investigated adult energy metabolism in multiple mosquito species and reported age specific trends of glycogenesis and lipogenesis (Briegel, 1990; Briegel and Timmermann, 2001; Ziegler and Ibrahim, 2001). Briegel et al., and Briegel and Timmermann investigated the accumulation of glycogen during the first week and lipids during the first two weeks in the adult life stages of Ae. aegypti and Ae. albopictus mosquitoes (Briegel, 1990; Briegel and Timmermann, 2001). In similar studies in Culex tarsalis, carbohydrates and lipids followed the same trend as in other mosquito species. However, lipid synthesis was more rapid than carbohydrate synthesis in C. tarsalis (Gray and Bradley, 2003). Further, C. tarsalis mosquitoes exhibited higher lipid storage trends in young adults in comparison to Ae. aegypti mosquitoes. These trends could be accounting for the autogenous potential of the Culex species which imply that the adults do not need a blood meal to lay eggs but can utilize lipid reserves (Gray and Bradley, 2003).

Glutathione (GSH) is an important molecule for insects. This enzyme plays a critical role in a number of biosynthetic and detoxification reactions (Forman et al., 2009). Glutathione transferases (GST) are also important enzymes that play a role in detoxification of substances that can be both endogenous or xenobiotic. In insects, these enzymes are known to play a role in insecticide resistance (Enayati et al., 2005). Besides these functions, studies also discuss the alteration of GSH during aging of Ae. aegypti mosquitoes (Hazelton and Lang, 1983). The GSH biosynthesis rate was observed to be distinctly reduced in aging adult mosquitoes. Further, a considerable decrease in biosynthetic rates were observed during senescence of the mosquitoes (Hazelton and Lang, 1983). Impaired GSH biosynthesis leading to low levels of GSH implies that regular cellular functions of GSH such as detoxification of peroxides and xenobiotics will be impaired ultimately leading to tissue damage and death.

Vector control programs take the average age of a mosquito population as a crucial determinant of vectorial capacity and potential of disease transmission (Johnson et al., 2020). This makes mosquito age grading important in vector control. The technique in current use for age grading is the Detinova parity method that assesses the age of a female mosquito by taking the changes in ovary appearance into account (Gray et al., 2022). In addition, novel tools like surface – enhanced Raman spectroscopy (SERS) is currently being studied as a potential mosquito age grading technique (Wang D. et al., 2022). In a study by Wang et al., where age of a mosquito is determined by both SERS and Infrared spectroscopy states that key biological molecules are altered in a mosquito with age which subsequently alter the spectra obtained by either of the mentioned methods (Wang D. et al., 2022). However, in the field, there is a lack of capacity to accurately determine the age of a mosquito caught in the wild. Often the techniques used are impractical or unreliable (Johnson et al., 2020). Therefore, understanding metabolic changes occurring with aging in mosquitoes can be used to develop biomarker point-of-use tests for age grading as well as vector control (Iovinella et al., 2015).

When a mosquito takes a viremic blood meal, the virus particles have to pass through several physical barriers in the mosquito in order to establish a successful infection, disseminate through the mosquito and be transmitted to a human host (Bosio et al., 2000). Infection of the midgut epithelium is the first barrier to infection. Presence of DENV, serotype 2 (DENV2) in the midgut tissue can be detected with the 3H5 monoclonal antibody as early as 2 days after the infectious blood meal was taken (Salazar et al., 2007). Staining at this early stage shows infected foci, indicating that the infection spreads laterally from the initial infected epithelial cells to the neighboring cells and eventually throughout the midgut (Salazar et al., 2007).

Upon successful infection of midgut epithelial cells, the virus must pass through the midgut escape barrier and continue to replicate in other tissues. Studies using electron-microscopy have shown that flaviviruses such as WNV and St. Louis encephalitis virus, escape from the midgut to the secondary tissues by passing through the basal lamina, the layer of extracellular matrix surrounding the midgut (Whitfield et al., 1973; Girard et al., 2005). Interestingly, a second non-infectious blood meal ingested by the mosquito enhances viral escape due to micro-perforations in the mid-gut (Armstrong et al., 2020) A study on DENV2 tropism in Ae. aegypti detected viral antigen in the trachea from the abdominal areas, suggesting that the trachea may also serve as an escape route for the virus from the midgut (Salazar et al., 2007). Following escape of the midgut barrier, the virus must replicate and amplify the infection in secondary tissues. Each of these tissues presents infection and escape barriers. A study has shown that DENV2 replicates in the fat body, hemocytes, nerve tissues, ommatidia of the compound eyes, esophagus, hindgut, cardia, trachea and Malpighian tubules (Salazar et al., 2007). Unlike WNV, DENV2 was not found to infect muscles (Girard et al., 2005; Salazar et al., 2007). Efficacious pass-through these barriers allows the virus to infect the salivary glands where the virus can be shed in the saliva when the next blood meal is taken to transmit to another host (Bosio et al., 2000).

In order for arboviruses to better survive in nature, they must evolve to survive in the invertebrate vector as well as the vertebrate host (Rückert and Ebel, 2018). Flaviviruses infect and rearrange the membrane architecture in their arthropod host cells like that observed in infected human cells (Girard et al., 2005; Welsch et al., 2009; Gillespie et al., 2010; Offerdahl et al., 2012; Junjhon et al., 2014; da Encarnação Sá-Guimarães et al., 2021; Mazeaud et al., 2021). Interestingly, these virus-induced membrane structures are morphologically and functionally conserved between these evolutionary distant hosts. These structures are summarized in Table 1. They are mostly endoplasmic reticulum (ER)-derived and include i) vesicles (Ve), the circular vesicular structures that house the viral replication complex, ii) vesicle packet (Vp), the larger vesicles that surround Ve, iii) convoluted membranes (CM), the site of viral protein translation and processing, and iv) tubular structures (T) with unknown function. In C6/36 cells infected with DENV2, Junjhon et al. observed Vp, Ve and T, with the number of Ve increasing with viral RNA copy number indicating a linear correlation between membrane structures and viral RNA replication (Junjhon et al., 2014). However, in contrast to DENV2 infected human Huh7 cells, CM were not found in infected C6/36 cells (Junjhon et al., 2014).

Similar vesicular structures were observed in mosquito cell lines infected with other flaviviruses. Electron micrographs revealed the induction of similar vesicular structures in Ae. albopictus cells infected with Kunjin virus and Ae. aegypti cells infected with yellow fever virus (Ishak et al., 1988). These membrane-rearrangements were also observed in WNV infected Culex quinquefasciatus mosquito tissues (Girard et al., 2005). The above-mentioned vesicle structures were observed in the midgut epithelium, midgut muscle and salivary gland tissues indicating that this specific membrane architecture was universally induced in both human and mosquito hosts in response to infection with most flaviviruses. In addition to these discussed structures, DENV infection of C6/36 cells was reported to produce exosomes that could infect naive C6/36 cells. Based on these observations, the authors suggested that virus induced exosomes have infectious potential and supports viral dissemination in C6/36 cells (Reyes-Ruiz et al., 2019). These studies on membrane architecture, together with the studies on lipid composition of infected cells, suggest that in addition to rearranging cellular membrane architecture during infection, existing membrane lipids are reorganized (Vial et al., 2020) as well as additional lipids (such as GPs and sphingolipids, SPs) are synthesized and incorporated into these membranes to promote expansion of membrane mass. Lipids with unsaturated fatty acyl chains and cone-shaped lipids such as PE, lysoGPs and ceramides (Cer) were increased likely to provide curvature and membrane-bending capabilities to facilitate the specific architecture required. Essentially, there is a concerted effort (by viral gene products) to alter both lipid metabolism and cellular membrane architecture to acquire an intracellular environment conducive to viral replication (Roosendaal et al., 2006; Miller et al., 2007; Perera and Kuhn, 2008; Perera et al., 2012).

Several studies to date have highlighted the impact of viral infection on the mosquito metabolic landscape (Table 2) (Perera et al., 2012; Melo et al., 2016). The initial study to utilize metabolomics to profile the metabolic landscape was carried out by Perera et al, on C6/36 Ae. Albopictus cells infected with DENV2. The study highlighted that ~15% of the metabolome was altered by infections of these cells. Lipid changes included those that had the capacity to alter membrane curvature and destabilize architecture, fluidity, and permeability. Specifically, GPs with smaller head groups such as phosphatidylethanolamine (PE) and cone-shaped (ie: ceramide) and inverse-cone shaped lipids (ie: lysophospholipids) were elevated in infected cells, specifically in membranes enriched in the replication complex. They also observed changes in SPs and glycerolipid intermediates such as monoacylglycerols (MAG) and DAG which are bioactive signaling molecules that participate in membrane fusion, fission, and trafficking and capable of enhancing a conducive environment for viral replication (Perera et al., 2012). These studies also demonstrated that like observations in human cells, de novo fatty acid biosynthesis via the enzyme, FAS, was important for viral replication in mosquito cells. Subsequently, a metabolomics study on Zika virus infected C6/36 cells identified 13 similar lipid species as specific biomarkers of infection (Melo et al., 2016). These included several species of SPs, GPs such as phosphatidylcholine (PC), phosphatodylserine (PS) and PE, as well as the bioactive intermediates such as DAG.

Interestingly, in an elegant study using isotopically labelled precursors, Vial et al, determined that at early time points, de novo biosynthesis of aminophospholipids such as PC and PE was actively blocked by DENV infection of Ae. Aegypti (Aag2) cells and instead that existing amino PLs were reorganized into replication complexes (Vial et al., 2020). These studies also demonstrated that in Ae. Aegypti mosquitoes, the rate-limiting enzyme that catalyzes aminoPL biosynthesis, acylglycerolphosphate acyltransferase (AGPAT), was decreased by DENV infection further supporting the hypothesis that aminoPL reorganization rather than de novo biosynthesis was activated during early infection (Vial et al., 2019)

In the adult mosquito, Chotiwan et al. demonstrated how lipid metabolism was temporally altered in infected Ae. aegypti mosquito midguts (the first site of viral replication) during infection with DENV2. In this study, GPs, SPs and fatty acids were significantly elevated and correlated temporally with the development of viral replication in the midgut (Chotiwan et al., 2018). GPs in insects play a critical role in tolerance to environmental changes (Guan et al., 2013). Elevation of glycerolipid intermediates suggested that resources were diverted from energy storage to biosynthesis during infection. Increased acyl-carnitines signaled functional disruptions in mitochondrial activities and energy production. Therefore, this study highlighted that significant metabolic perturbations occurred at early stages of viral replication in the mosquito.

Vector control has been critical in the prevention and control of several vector-borne diseases. The use of insecticides to kill or deter vectors has been the mainstay of vector control globally (van den Berg et al., 2021). Intense insecticide use has selected mechanisms of resistance that are now prevalent in malaria vectors such as Anopheles gambiae, Anopheles sinensis, and Anopheles funestus (Hancock et al., 2018), as well as arbovirus-vectors such as Ae. aegypti and Ae. albopictus (Moyes et al., 2017). Insecticide resistance mechanisms include decreased cuticular penetration of insecticides, increased enzyme metabolism, and decreased sensitivity of insecticide target sites (Oppenoorth, 1984; Scott, 1990). The mechanisms of metabolic resistance will be the focus of this section (Figure 3).

Metabolic resistance protects against all insecticides used in public health, including pyrethroids, organophosphates, carbamates, and organochlorines (Smith et al., 2016) Insecticide metabolism is conducted by three enzyme families, cytochrome P450 monooxygenases (P450), glutathione transferases (GST), and carboxy/cholinesterases (CCE) (Strode et al., 2008; Feyereisen, 2012). Enhanced metabolism in resistant insects can be caused by gene over-expression (cis/trans regulation and gene amplification) or allelic variation of members of these enzyme families.

Because of the complexity of the enzyme family systems and the difficulty in purifying these enzymes (e.g., substrate overlap, instability, yields), understanding the mechanisms of resistance have proved difficult. The development of molecular and bioinformatics tools enabled the identification of genes and associated regulatory processes in resistant insects resulting in significant progress over the last decade. Several studies have investigated the differential gene expression of CYPs, GSTs, and CCEs in resistant Anopheles, Culex, and Aedes mosquitos (Smith et al., 2016; Moyes et al., 2017; Vontas et al., 2020). However, relatively few detoxification enzymes have been examined in vitro to validate their ability to metabolize insecticides. For example, CYP6P3, CYP6M2, and GSTE2 in An. gambiae have shown to metabolize pyrethroids, DDT, and bendiocarb (Müller et al., 2008; Stevenson et al., 2011; Mitchell et al., 2012; Yunta et al., 2019). Pyrethroids are metabolized by CYP6P9a and CYP6P9b in An. funestus (Riveron et al., 2014) and by CYP9M6, CYP6BB2, CYP9J24, CYP9J26, CYP6J28 and CYP9J32 in Ae. aegypti (Mitchell et al., 2012; Kasai et al., 2014). The role of these enzymes in pyrethroid resistance has been validated in vivo utilizing heterologous expression in Drosophila (Pavlidi et al., 2012; Edi et al., 2014; Reid et al., 2014; Riveron et al., 2014; Ibrahim et al., 2015) or RNA interference (RNAi) technologies in Ae. albopictus (Xu et al., 2018). A GAL4/UAS expression system was recently developed in An. gambiae to confirm in vivo that overexpression of GSTE2 conferred organophosphate and organochlorine resistance, CYP6P3 conferred pyrethroid and carbamate resistance, and CYP6M2 conferred pyrethroid resistance when overexpressed in the same tissues (Adolfi et al., 2019).

Metabolic technologies have recently been utilized to investigate the mosquito’s reaction to insecticides and to find metabolic routes of insecticide detoxification. For example, Prud’homme et al used targeted gas chromatography mass spectrometry (GC/MS) in conjunction with transcriptomics to assess the effect of ibuprofen on Ae. aegypti (Prud’homme et al., 2018). Direct ibuprofen exposure in larvae and adults had no effect on the 53 quantified polar metabolites, but F1 eggs from ibuprofen-exposed parents had lower levels of amino acids, carbohydrates, polyols, phosphoric acid, and ornithine, implying that ibuprofen exposure affected metabolic resource internalization in eggs (Prud’homme et al., 2018).

The profile of 12 amino acids and 31 acylcarnitines in Culex quinquefasciatus larvae treated with chlorpyrifos, temephos, and permethrin was determined using liquid chromatography tandem mass spectrometry (LC-MS/MS) (Martin-Park et al., 2017). Two acylcarnitines (C0 and C2) and the amino acid arginine were shown to be differentially associated with insecticide exposure. C0 concentrations were considerably higher in permethrin-exposed larvae, whereas C2 concentrations increased in permethrin-exposed but dropped in temephos-exposed larvae. Permethrin and temephos exposure enhanced arginine levels in larvae (Martin-Park et al., 2017). Studies have also shown that permethrin resistant mosquitoes have altered gut microbiota in comparison to the wild type and therefore have altered metabolic processes (Muturi et al., 2021).

In a separate study, Culex pipiens larvae subjected to varying concentrations of the neonicotinoid clothianidin showed differences in three groups of metabolites important in energy metabolism, including acylcarnitines, glycerophospholipids (GPs), and biogenic amine abundance (Russo et al., 2018) The unipolar and polar metabolites were quantified using flow injection tandem mass spectrometry (FIA-MS/MS) and LC-MS/MS, respectively. The highest dosage of clothianidin reduced acylcarnitines, GPs, and biogenic amines after 24 hours of exposure. Low and medium amounts reduced GPs and biogenic amines while increasing acylcarnitines. GPs and acylcarnitines were reduced at low and medium concentrations after 48 hours of exposure. These findings imply that low pesticide doses raised the energy requirements of exposed species (Russo et al., 2018).

Nuclear magnetic resonance (NMR) spectroscopy was utilized in Ae. aegypti to discover changes in metabolites caused by temperature and/or exposure to DDT, malathion, and deltamethrin. Metabolites involved in the tricarboxylic acid cycle, branched amino acid degradation, glycolysis/gluconeogenesis, amino acid, lipid and carbohydrate, nucleotide PRPP pathway, and phospholipid metabolism were the most affected (Singh et al., 2022). Seven of the nine discovered compounds were shown to be significantly altered by individual temperature and insecticide exposure. These included pyruvates, maltose, citrate, nicotinate, and -hydroxybutyrate, all of which are components of the glycolysis/gluconeogenesis, tri-carboxylic acid cycle energy producing pathways.

A recent study used LC-MS/MS to compare the metabolites of deltamethrin resistant and susceptible An. sinensis in larva and adult stages. Deltamethrin exposure resulted in the identification of 127 and 168 distinct metabolites in larvae and adults, respectively. Organooxygen compounds, carboxylic acids, GPs, and purine nucleic acids were shown to be different between resistant and susceptible mosquitos. The GPs route was shared by resistant larva and adults, and it may play an essential role in the metabolic deltamethrin detoxification (Li et al., 2022).

Pyrethroid resistance is typically associated with biological fitness costs, including size, longevity, fecundity, and mating behavior (Brito et al., 2013; Smith et al., 2016; Vera-Maloof et al., 2020). Moreover, pyrethroid resistance may influence the mosquito’s inherent permissiveness to viral infection, replication, and transmission. Few studies have reported the association between pyrethroid resistance and Aedes vectorial competence (VC), and the outcomes have been widely inconsistent among virus, mosquito strains and stage of viral infection. Few studies have evaluated the relationship between vector competence (VC) and pyrethroid resistance in laboratory Ae. aegypti (Zhao et al., 2018; Chen et al., 2020; Parker-Crockett et al., 2021; Stephenson et al., 2021; Wang L. et al., 2022) and Ae. albopictus strains (Richards et al., 2017; Chen et al., 2021). The direction of VC response differed among virus, viral replication stage, and mosquito strains. For example, VC for ZIKV (Zhao et al., 2018; Parker-Crockett et al., 2021) and DENV (Chen et al., 2021) was higher in the pyrethroid resistant strain. In contrast, three investigations have shown that the pyrethroid-resistant strains had significantly lower viral infection rates than the susceptible strains (Deng et al., 2021; Stephenson et al., 2021; Wang L. et al., 2022). Interestingly, only one study found that high kdr allele frequencies were associated with lower DENV-infection rates in field populations in Florida (Stephenson et al., 2021). As more studies elucidate the specific interactions between arboviruses and field resistant mosquitoes, understanding the mechanisms of interaction in specific mosquito tissues remains to be explored.

Recent studies on the mosquito microbiome have highlighted its critical impact on various stages of the mosquito life cycle including development and reproduction as well as ecological adaptation, pathogen infection, immunity and vector competence (Figure 2A). Increasing evidence suggests a critical metabolic inter-dependency between the microbiome and the mosquito vector that drive the resulting outcomes with detrimental effects if the metabolic relationship is perturbed.

The metabolic impact of the microbiome is evident throughout the life cycle of the mosquito. A comparison of Ae. Aegypti (requires a blood meal for egg production) versus Ae. Atropalpus (does not require a blood meal for egg production) identified that the microbiome played a crucial role in providing the nutritional resources required for Ae. Atropalpus egg production in the absence of a blood meal (Coon et al., 2016). Studies have also shown that bacterial microbiota in aquatic habitats significantly impact larval nutrient acquisition and/or assimilation influencing growth, development, and survival into pupation (Coon et al., 2017; Vogel et al., 2017; Correa et al., 2018; Valzania et al., 2018). These studies compared transcriptional responses in axenic (bacteria free) larvae and larvae containing either native or gnotobiotic (single species) microbiomes and discovered that processes such as amino acid transport, hormonal signaling, and metabolism were differentially expressed, suggesting a key role in larval development and survival. Axenic larvae exhibited delayed development time and stunted growth in comparison to their bacterially colonized counterparts.

A study by Feng et al., demonstrated the influence of gut microbiota of Anopheles stephensi (a potent malaria vector) on tryptophan metabolism. The study reported how elimination of gut microbiota via antibiotics increased accumulation of tryptophan and its metabolites (kynurenine and 3-hydroxykynurenine, 3H-K) in the mosquito with high levels of 3H-K leading to structural impairments in the peritrophic matrix which in turn facilitated Plasmodium berghei infection (Feng et al., 2021). Similar observations were also made by Das de et al (Das De et al., 2022),. Blood meal digestion also increased these specific metabolites in the midgut, and the gut microbiome was critical for the catabolism that maintains normal levels of these metabolites. In a second study, Bottino-Rojas et al., showed that the product of the kynurenine pathway, xanthurenic acid, was critical for controlling levels of microbiota as well as reproduction and survival of the mosquito (Bottino-Rojas et al., 2022). Mutations in orthologs of kynurenine hydroxylase impaired reproduction and survival in Ae. aegypti, An. stephensi and Culex quinquefasciatus as well as disrupted the midgut permeability barrier in An. stephensi. Kynurenine is a tryptophan metabolite. Together these studies suggest that tryptophan catabolism is critical for maintaining the integrity of the peritrophic matrix, preventing infection, maintaining normal levels of the gut microbiome, and ensuring survival and reproduction of the mosquito vector (Okech et al., 2006; Lima et al., 2012; Bottino-Rojas et al., 2022). Interestingly, tryptophan metabolism is a key pathway altered in the human metabolome during severe disease caused by DENV infection. Serotonin (another product of tryptophan metabolism) and kynurenine were identified to show differential abundance in DENV-infected patient serum. Serotonin levels were significantly lower in patients with dengue hemorrhagic fever (DHF) than those in the febrile phase (Cui et al., 2016).

On the other hand, there is evidence to show that metabolism of the mosquito influences the modulation of microbial density in different Ae. aegypti strains, in turn altering vector competence. A study by Short et al., demonstrated how specific metabolic pathways such as the branched chain amino acid (BCAA) degradation pathway can influence mosquito midgut microbiota. When the BCAA degradation pathway was silenced in two mosquito strains with broadly differing gut microbiota, a significant alteration in microbiota composition was observed in both strains. This resulted in levelling the variation between the two microbiomes. The authors hypothesize that variations in amino acid metabolism can be a crucial determinant of microbiota in a vector (Short et al., 2017). In relation to these results, there are previous studies that have shown how alterations of the mosquito midgut microbiota can alter vector competence in Ae. aegypti mosquitoes. An increase in DENV titers was observed in Ae. aegypti mosquitoes after reduction of bacterial loads in the midgut through antibiotic treatment (Xi et al., 2008). Moreover, introduction of multiple bacterial species into a native mosquito gut microbiome has been shown to decrease DENV titers (Ramirez et al., 2012; Ramirez et al., 2014). However, there is also evidence to show how specific midgut bacterial populations can enhance susceptibility of a vector to DENV and CHIKV (Apte-Deshpande et al., 2012; Apte-Deshpande et al., 2014). These studies highlight gaps that remain in our understanding of the mechanisms underlying the microbiome-mosquito-pathogen multipartite interactions that result in the outcomes observed, stimulating further studies.

As discussed earlier, vector control is an efficient way of preventing the spread of arboviruses. However, during the past two-three decades, the major vectors of these arboviruses have expanded their geographic range and population size (Kraemer et al., 2015). Rapid urbanization, climate changes and development of resistance to insecticides are some of the major factors governing the emergence of many medically important arboviruses. Efficient, vector control strategies are considered necessary to overcome these problems. In the last decade, interventions with biocontrol agents like Wolbachia have revolutionized the field of vector control. Wolbachia is a maternally inherited, endosymbiotic bacterium found naturally in ~ 60% of the insect population (Hilgenboecker et al., 2008; Zug and Hammerstein, 2012). Due to the unique capability of Wolbachia to block many arthropod-borne viruses, it has been widely used for biological control of mosquito borne viruses including dengue, West Nile and Zika (Moreira et al., 2009; Walker et al., 2011; Aliota et al., 2016). Wolbachia is a safe vector control strategy as there is no evidence of impact on the environment, animals or humans by the bacterium (Centers for Disease Control and Prevention, 2022). The mechanisms through which pathogen blocking is occurring are yet unknown. Understanding these mechanisms is critical to preclude any resistance development in the mosquitoes against the bacterium. Interestingly, there is evidence to show that Wolbachia alters metabolic homeostasis of the mosquito vector during infection (Figure 2D) (Caragata et al., 2014; Molloy et al., 2016; Koh et al., 2020; Manokaran et al., 2020; Nascimento da Silva et al., 2022). Given the reliance on metabolism for viral replication in the mosquito (discussed above), several studies have hypothesized that metabolic competition versus commensalism may be a mechanism for Wolbachia-mediated pathogen blocking.

Manokaran et al., showed that Aag2 cells infected with the Wolbachia (wMel strain) had reduced abundances of acyl-carnitines which seemed detrimental for both DENV and ZIKV replication. Supplementation with acyl-carnitines restored DENV and ZIKV replication in Wolbachia infected cells (Manokaran et al., 2020). In the metabolomic study by Chotiwan, et al., acyl-carnitines were increased significantly in DENV2-infected mosquito midguts reinforcing the hypothesis that acyl-carnitines may play a proviral role in DENV replication and their regulation could be a possible mechanism of pathogen blocking by Wolbachia. Two reports exist for the intersection of SPs in insects and Wolbachia. Initially Rong et al. showed that Wolbachia infection induced the expression of miRNAs known to regulate genes with functions in sphingolipid metabolism (Rong et al., 2014). This was confirmed by Molloy et al, that recently showed a complete depletion of sphingolipids during Wolbachia infection of Ae. albopictus cells (Molloy et al., 2016).

Infection with wMel has also been shown to influence oviposition, expression of egg yolk precursor genes as well as altered excretion of the blood meal in Ae. aegypti mosquitoes (Pimenta de Oliveira et al., 2017). The results of this study indicated delayed expression of genes required for development of eggs which in turn might have delayed yolk deposition creating a lag in oviposition. To further support their results, they call attention to studies by Caragata et al., that demonstrates how cholesterol levels were reduced by 15-25% in Wolbachia (wMelPop strain) infected Ae. aegypti mosquitoes (Caragata et al., 2014). Cholesterol is considered a primary precursor of the 20-hydroxyecdysone hormone that is an important regulator of oogenesis (Raikhel et al., 2002).

Competition for host resources such as cholesterol is also a proposed mechanism of pathogen blocking by Wolbachia (Caragata et al., 2013). The study conducted in Wolbachia (wMelPop and wMelICS) infected Drosophila flies fed with a cholesterol enriched diet has shown weakened pathogen blocking capabilities in comparison to flies fed with a standard diet. Both insect and virus are unable to synthesize their own cholesterol and depend on cholesterol taken up in the diet. Further, many of the arboviruses require host cholesterol to enter host cells and replicate (Mackenzie et al., 2007; Lee et al., 2008; Puerta-Guardo et al., 2010; Soto-Acosta et al., 2013; García Cordero et al., 2014). Thus, it is not surprising that manipulation of host cholesterol by Wolbachia could lead to perturbed viral entry and replication (Caragata et al., 2013). These observations are further supported by quantitative proteomics studies on Aag2 cells and Ae. aegypti midguts that revealed differential expression of proteins related to vesicular trafficking, lipid metabolism and unfolded protein response in response to Wolbachia infection (Geoghegan et al., 2017). Elevated levels of esterified cholesterol were detected in Wolbachia infected cells possibly resulting from perturbed intracellular cholesterol trafficking (Geoghegan et al., 2017). Reversal of this accumulation by treating with cyclodextrins restored dengue replication in these cells, suggesting free cholesterol was required for viral replication and sequestration of cholesterol through esterification seemed to enhance pathogen blocking. However, in this same study, Neiman Pick Type C1 and C2 (NPC1 and 2) proteins, responsible for retro recycling of cholesterol were elevated in expression compared to controls. Several studies have shown that elevated levels of NPC1 and 2 are beneficial to DENV replication. Therefore, the exact nature of the block in cholesterol trafficking and its relationship to NPC1 activity and viral replication remains to be explored further.