Maternal lipid mobilization is essential for embryonic development in the malaria vector Anopheles gambiae

Reproduction is a highly energy-demanding process that requires tight metabolic control coordinated by hormones, metabolites, and immune factors [1,2]. In mammals, lipid stores in the form of triglycerides (TAGs) are deposited in the cells of adipose tissue, where they serve as a dynamic energy reservoir. As TAGs are unable to cross cell membranes, for transport to other tissues, they need to be either incorporated into vesicles or remobilized. Regulation of these processes is critically important during gestation as they coordinate the necessary energy supply to the growing fetus, and their perturbations can have negative effects on pregnancy outcomes [3,4]. In reproductive organs, lipid mobilization allows essential reproductive functions such as the development of ovarian follicles and the synthesis of steroid hormones, which prepare the endometrium for implantation [3].

In simplified terms (reviewed in [5]), TAG remobilization follows a hydrolytic process requiring their breakdown into free fatty acids (FFAs) and glycerol, regulated by hormonal stimuli that activate lipolytic enzymes like hormone-sensitive lipase (HSL) and adipose triglyceride lipase (ATGL) [6–8]. The canonical pathway for lipolysis activation in adipocytes is driven by catecholamines, adrenocorticotropic hormone (ACTH), and secretin, which upon binding to β-adrenergic G-protein-coupled receptors stimulate cyclic adenosine monophosphate (cAMP) synthesis, activating protein kinase A (PKA). PKA activation causes the phosphorylation of HSL, which induces its translocation onto the lipid droplet (LD) surface aided by PKA-mediated phosphorylation of perilipin-1 [9,10]. Perilipin-1 phosphorylation also results in the dissociation of the co-activator comparative gene identification-58 (CGI-58) [11,12], which can then bind to ATGL and activate lipolysis [13]. Upon completion of hydrolysis, FFAs will bind to carrier proteins that transport them to target tissues in the body for their use in energy production through ß-oxidation, hormone synthesis, or other cellular metabolic needs [14].

Although the last common ancestor of insects and mammals is estimated to have lived over 500 million years ago, insects possess a similar metabolic cascade to mobilize lipids. Lipolysis of TAGs is hormonally stimulated by a neuropeptide hormone, adipokinetic hormone (AKH), that has an analogous role to glucagon in mammals and is released from the corpora cardiaca [15,16]. Once released, AKH binds to its G protein-coupled receptor (AKHR) triggering an increase in intracellular calcium levels and cAMP synthesis, followed by PKA activation [17]. Activated PKA in turn phosphorylates lipid storage droplet protein 1 (LSD1), a homologue of perilipin 1-like protein, on the surface of LDs, ultimately inducing the activation of TAG lipases to catalyze the rate-limiting step in the hydrolysis of fatty acids from TAGs to generate FFA and diacylglycerides (DAGs) [18–22]. In hematophagous insects, after blood feeding, complex lipids such as TAGs and phospholipids are broken down into FFAs in the midgut lumen for transport across the luminal membranes [23], where they are reutilized to synthesize DAGs, TAGs, esters, and phospholipids [24]. These lipids are then loaded—in concert with lipid transfer proteins—onto circulating lipoproteins, principally lipophorin (Lp), on the basal side of the midgut epithelium for delivery to key tissues such as flight muscles and ovaries, or directed to the fat body for storage or further mobilization [19,20,25,26]. Dynamic changes occur in the lipid profile of the fat body during a reproductive cycle [27], with substantial proportions of stored lipids being mobilized for egg production [28,29]. Lp-transported lipids, as well as the yolk protein vitellogenin (Vg), are incorporated into developing oocytes [30,31] whereupon lipid is organized into LDs and Vg crystallizes into vitellin yolk granules [32] to be degraded after oviposition to nourish the developing embryo [33]. Coordinated lipolysis regulation is mediated by multiple conserved signaling pathways, including 20-hydroxyecdysone [34,35], Hedgehog [36], Decapentaplegic [37], NF-κB [38], as well as the neuropeptides FMRFamide [39] and AKH [40]. The components of the lipolytic machinery are shared by important insect vectors of human diseases including malaria, Chagas disease, and African Trypanosomiasis [23,41–44], and could potentially be targeted to reduce the size of vector populations and aid in disease control efforts. Orthologues of TAG lipase and AKHR are significantly increased after blood feeding in the fat body of Aedes aegypti mosquitoes [34,40], and TAG stores in the fat body, which are sharply reduced after blood feeding due to their mobilization and incorporation into developing eggs, are not mobilized when the AKH receptor AKHR is silenced by RNAi [40]. Similarly, disruption of lipid mobilization through targeting AKH signaling, ATGL or Perilipin/LSD1 in several insects decreases fecundity [43,45] and causes embryonic lethality [21,46]. However, information regarding lipid mobilization in Anopheles is relatively scarce.

In one of the most important vectors of human malaria, the Anopheles gambiae mosquito, lipid transport by Lp has been shown to impact the development of Plasmodium falciparum and Plasmodium berghei, respectively, a human and rodent malaria parasite. Indeed, silencing Lp before an infectious blood meal reduces oocyst numbers in both parasite species [47–49], while it impairs oocyst growth and sporozoite infectivity in P. berghei but not P. falciparum [48–50]. These studies also revealed that Lp, and to a lesser extent Vg, which transports approximately 5% of lipids within oocytes [51], are essential for egg development as their silencing drastically reduces the number of developing ovarian oocytes [31,47,48,50]. A role of upstream lipolytic pathways in the reproduction of this or other human malaria vectors has yet to be determined, and their function has only previously been studied in the context of pathogen infections. Silencing of the TAG lipase AgTL2 (a pancreatic lipase that is the functional homologue of ATGL and predominantly induced in the fat body after blood feeding [52]) by RNAi leads to accelerated P. falciparum oocyst growth, possibly due to the accumulation of TAGs, DAGs, and phosphatidylcholine observed in midguts and fat bodies [48], while depleting AKH1 and AKHR leads to reduced number of sporozoites in the salivary glands [53]. Whether impairing lipolysis impacts reproduction in this anopheline is unknown, and addressing this question could uncover key reproductive targets to reduce mosquito populations and curb malaria incidence in disease endemic countries.

Our data reveal an unappreciated role for maternal TAG remobilization in shaping the development of An. gambiae embryos. We show that impairing lipolysis via AgTL2 depletion in the fat body, the fat storage tissue, leads to a substantial decrease in TAGs in the ovaries of adult females, the destination of lipids shuttled via the function of lipoproteins. In turn, we demonstrate that inheriting vastly reduced TAG levels triggers a cascade of nutritional dysregulation that prevents embryos from hatching and leads to their death. The almost complete penetrance of this phenotype led us to test the possibility of targeting lipolysis in mosquitoes to reduce the size of field populations. We were pleasantly surprised by the efficacy in proof of principle experiments of the lipase inhibitor orlistat, which induced significant embryonic lethality when applied to females. Orlistat is a drug used in humans to counter obesity and therefore it would not be suitable for a mosquito-targeting strategy. However, in the future, it may be feasible to generate Anopheles-specific inhibitors that target the function of AgTL2, AgLSD1, or other factors critical to mosquito lipolysis, for incorporation in bed nets (or other vector control strategies like indoor residual sprays [64]). Developing these inhibitors for mosquito control may prove challenging given the close homology between mosquito and human lipases [21], although it is possible that allosteric inhibitors may show higher specificity against mosquito enzymes [65]. Potential impacts on nontarget organisms would need to be evaluated through toxicological testing in mammals and insect surveys during efficacy trials of the intervention. If necessary, effects could be mitigated through usage methods, dosages, or formulations that reduce exposure of other insects to these inhibitors, such as physical incorporation into bed nets, which generally only target night-biting species contacting the net. Importantly, as mentioned above, compounds that interfere with mosquito reproduction are already incorporated into bed nets utilized in malaria control strategies. Indeed, dual-ingredient nets including the insect growth regulator pyriproxyfen show promise in malaria endemic countries [59–61], as they act as secondary active agents in the event of mosquito resistance to the primary insecticide. This is particularly relevant at a time when insecticide-treated nets, after contributing to averting millions of malaria cases worldwide [66], are showing worrying signs of inefficacy [67], representing one of the principal causes behind the increase in malaria cases after years of decline [68,69]. Once developed, Anopheles-specific inhibitors of the lipolytic machinery could therefore contribute to feed the pipeline for the next generation of dual-ingredient bed net products.

Besides Lp-transported lipids, the yolk protein Vg conveys approximately 5% of total oocyte lipids [51], and so yolk deposition could potentially be affected in AgTL2 embryos and contribute to their abnormal development. Interestingly, embryonic lethality was reported when maternally deposited enzymes required for yolk usage in the embryo were silenced in Rhodnius prolixus embryos [70], suggesting that such phenotypes are due to an inability to mobilize yolk effectively. Our studies in An. gambiae have however shown that embryonic lethality occurs much earlier when Vg is silenced. Indeed, Vg depletion in mothers, which led to reduced yolk protein deposition into the oocyte, arrested embryonic development within the first few hours of oviposition [31], suggesting the late mortality observed in dsAgTL2 embryos is not due to defects in yolk deposition. The effects of AgTL2 silencing on Anopheles reproduction are in line with phenotypes observed in insects after manipulation of other components of the lipolytic machinery. In Drosophila, knockout of brummer (bmm), the orthologue of the mammalian adipose TAG lipase ATGL, in both parents leads to excessive fat accumulation and impaired lipid mobilization from fat bodies, and severely affects hatching of mutant embryos by unknown mechanisms [21], and similar results were obtained in recent studies in Bombyx mori [46] and R. prolixus [45]. Although these studies across insects did not determine the mechanisms triggered by inhibition of lipolysis, they highlight a degree of conservation in the importance of TAG mobilization in insect reproduction that could well be exploited to target populations of insect vectors of human and animal pathogens, as well as agricultural pests.

Our multi-omics analyses revealed that embryos from AgTL2-depleted mothers have considerable deficits in key metabolic processes. Carbohydrate metabolism and ß-oxidation are shut down in early developmental phases, followed by disruption of the TCA cycle. While these metabolic defects are not apparent in the histological samples and are not reflected by changes in transcriptomic profiles at the early RNAseq time points, it is likely that their accumulation leads eventually to the mis-regulation of several genes observed at the last time point analyzed (38 h post oviposition), a time when control embryos are close to hatching. Up-regulated genes are enriched in stress response and protein folding genes, revealing embryos that are under severe stress. This is paralleled with a strong down-regulation in key sensory signaling receptors and neuropeptides, eclosion hormone [54,55], and cuticular proteins that form the second cuticle needed before hatching [58]. LDs are known to be multifunctional storage compartments in the developing embryo [71], providing a source of not only TAGs and phospholipids but other lipids classes as well, such as the waxy saturated hydrocarbons required for cuticle synthesis [72]. Along with a significant decrease in expression of 26 annotated cuticle-related genes (S3 Data), a lack of LD breakdown in dsAgTL2 embryos may sequester these structural lipids away from cuticle synthesis enzymes, preventing the production of adequate cuticle for hatching [58]. Lipids stored within LDs may also serve as the source of several key signaling molecules, such as eicosanoids, hormones derived from polyunsaturated FFAs with roles in egg and eggshell formation [73]. Limited eicosanoid signaling may also explain embryonic lethality as eggshell defects and low hatching rates were recently reported in R. prolixus after bmm silencing [45], and eggshell defects have been observed after inhibition of eicosanoid signaling in Aedes mosquitoes [74].

We also observed a highly significant GO term enrichment for neural functions in down-regulated genes at 38 h post oviposition (top 4 terms), including neuropeptide and GPCR signaling pathways. CCHamide1-related proteins, induced in control but not dsAgTL2 embryos, are arthropod neuropeptides secreted by enteroendocrine cells in the gut upon sensing of different dietary conditions in the intestinal lumen. They regulate activity of dopaminergic neurons in the brain to control gut motility and other physiological behaviors [75–77], although interestingly they have not previously been implicated in embryogenesis and hatching. Lipids and their derivatives are required for neuronal signaling through their roles in synapse function and axonal wrapping [78,79]. Indeed, we observed increased levels of LPE and decreased ceramide phosphoethanolamine (CerPE) in dsAgTL2 embryos at 38 h post oviposition, changes that resemble Drosophila glial mutants unable to ensheathe axons [80], again suggestive of neuronal defects. In mammals, deficiencies in lipid metabolism and lipid signaling have been associated with developmental and cognitive problems as well as with neurodegenerative diseases. Imbalances in fatty acids during gestation causes altered brain development in the fetus, while severe fatty acid deprivation can impair cognitive functions (reviewed in [78]). Although we could not pinpoint the precise mechanisms behind embryonic lethality, our results show how reduced energy stored caused by impaired lipid remobilization and low TAG levels leads to the collapse of normal metabolic functions, eventually causing severe stress and cellular death.

Finally, our data confirm that P. falciparum parasites are affected by the mosquito lipolytic machinery. Depletion of AgTL2 (previous studies [48]) and AgLSD1 (this study) leads to accelerated oocyst growth rates and faster sporozoite invasion of salivary glands. Interestingly, however, such acceleration is not recapitulated by orlistat treatment, an observation that implicates fat body (rather than midgut) lipids in parasite growth. Indeed, while both AgTL2 depletion and orlistat treatment induce glyceride accumulation in the midgut (this study and [48]), differences in glyceride levels in the fat body are only detected after lipase silencing. This difference suggests that orlistat, while closely phenocopying the effects of AgTL2 depletion in terms of oogenesis and embryonic development, may inhibit different or additional aspects of the female lipolytic machinery. Regardless of the mechanism, when combined with the observed effects on fertility, these data support the idea that if Anopheles-specific lipase inhibitors were to be generated, they may represent promising tools for field interventions aimed at reducing the size of field populations of malaria vectors, aiding efforts to control this devastating disease.