Sex-specific gene expression in the developing A. aegypti pupal head
This investigation revealed global patterns of sex-specific gene expression in the A. aegypti pupal head transcriptome, which was assessed through custom A. aegypti microarray experiments (Additional file 1). These studies provided insight regarding the genes, pathways, and processes impacted by sex-specification in the 24 h pupal head transcriptome. The data collected may be of interest to those studying a number of tissues of vector importance, including the fat bodies, antennae, maxillary palp, and proboscis. However, the primary focus of the present study is the brain, and so the results of the microarray experiments were validated through in situ hybridization experiments performed on female and male pupal brain tissue (Figures 4, 5, 6, 7, 8, 9, 10). The 24 h pupal brain was selected, as our recent work has identified this time point to be a critical period for nervous system development. At this stage, which follows periods of extensive proliferative activity and pupal histolysis, the brain has an anatomical organization structure that is similar to that of the adult A. aegypti brain. The neuropils that are characteristic of the adult brain, including the antennal lobe, central complex, and three optic lobe neuropils, have begun to form. At 24 h APF, extensive neural process outgrowth, targeting of higher order brain neurons, synapse formation, and arborization occur, ultimately generating the increased neuropil density observed in the adult [6],[8],[23].
These studies detected sexually dimorphic expression of obp 10 and takeout through both head transcriptome microarray experiments and through analysis of transcript expression in the developing brain (Additional file 1, Figure 4A,D). Although expression of these genes had not previously been spatially assessed in the developing brain, dimorphic expression of both genes had been noted in A. aegypti by Bohbot and Vogt [31], who hypothesized that the genes might regulate differential olfactory and feeding behaviors in mosquitoes. It was subsequently demonstrated through knockdown experiments in Anopheles gambiae that Takeout regulates blood-feeding propensity [32]. Our own ability to detect dimorphic expression of these genes further validates the effectiveness of the methodological approach employed, while also raising the question of whether their dimorphic expression in the brain is functionally significant. This is a topic to be explored in future studies, which would be greatly enhanced through technical advancements that permit the manipulation of gene expression in a brain-specific manner in mosquitoes.
Components of critical developmental signaling cascades, including cubitus interruptus (ci), a member of the Hedgehog signaling pathway (reviewed by [33]), as well as arrow, a component of the Wnt/Wingless (Wg) signaling pathway [34], are upregulated in the developing female pupal brain (Figure 4B,E, Additional file 1). It will be interesting to functionally examine the roles of these genes during brain development, particularly given that Drosophila Ci has both transcriptional activation and transcriptional repressor forms (reviewed by [33]) that might differentially shape male and female development. Furthermore, although the function of Wg signaling in the regulation of sex-specific development has not yet been assessed in the insect nervous system, Wg signaling has been shown to regulate sexually dimorphic abdomen development in Drosophila. In flies, dimorphic Wg regulation in conjunction with monomorphic segment-specific cell death generates sex-specific abdomen shape [35].
More broadly, the results of this investigation suggest that underlying differences in proteolytic, metabolic, catabolic, and biosynthetic processes, as well as cell growth/proliferation and ion transport underlie differences in A. aegypti males vs. females (Tables 1, 2, 3, 4; Figures 2 and 3). Here, we discuss the potential relevance of these findings, focusing this discussion on development of the nervous system.
Proteases
Proteases, especially serine proteases, are significantly upregulated in female pupal heads (Figure 3, Table 2). These include 23 Clip-domain serine proteases that are significantly overexpressed in females. Clip-domain serine proteases have been well studied with regard to their multiple roles in insect immunity, including antimicrobial peptide synthesis, hemolymph coagulation, and melanization of pathogen surfaces [36]. In Anopheles gambiae, Clip-domain proteases are known for their function in the regulation of malaria parasite melanization [37]. In reference to insect development, two Clip-domain proteases function as components of the Toll pathway in the regulation of dorso-ventral axis specification (reviewed in [38]). Although Clip proteases have not been well studied in the context of the developing nervous system, the results of this investigation suggest that they are dimorphically expressed in female and male pupal heads. Other types of serine proteases have important CNS functions, including the regulation of cell migration, neurite outgrowth, and pathfinding, glial and neuronal cell survival, and synaptic remodeling [39]-[42]. Thus, it is possible that these Clip proteases might function as modifiers of the extracellular matrix in the context of neural migration.
Cell cycle/cell death
Cell cycle components (Figure 2A) were noted as a significant GO term among DETs upregulated in females (Table 1). For example, cdk4/6, a known regulator of cellular growth in D. melanogaster[43],[44], is upregulated in female pupal heads (Figures 4C and 11D). In flies, overexpression of Cyclin D-Cdk4/6 promotes cellular growth, and loss of cdk4 function results in flies with a smaller body size [43],[44]. It is possible that increased expression of cdk4 in A. aegypti females upregulates the growth of A. aegypti female pupae/heads with respect to males, which have a significantly smaller body and head (but not brain) size [18]. In relation to the developing nervous system, it has been shown that cdk4 expression is positively regulated in response to axon guidance gene signaling [45]. Thus, upregulation of cdk4 in the developing A. aegypti female brain might contribute to differential neural outgrowth. Furthermore, the microtubule-based movement GO term was significantly associated with DETs upregulated in females (Figure 2A; Table 1), suggesting that spindle fiber movement may be regulated in a sex-specific manner.
p53, a regulator of G1-S transition (reviewed by [46]), is also differentially expressed in pupal heads (Additional file 1), in which increased p53 expression can be detected in the male 24 h pupal brain (Figures 4I and 11A). In flies, p53 regulates apoptosis during development (reviewed by [46]). The regulation of cell death is a critical aspect of nervous system development which has also been linked to the development of sexual dimorphism in fruit flies (reviewed by [9]). For example, Dsx regulates death of the female TN1 neurons during Drosophila metamorphosis. In males, these neurons survive and are believed to function in the neural circuitry that is involved in production of the male courtship song (reviewed by [9]). These findings, in conjunction with the results of our investigation, suggest that regulation of p53 expression by Dsx may generate differences in programmed cell death that might shape the development of sex-specific traits in mosquitoes. In further support of this notion, caspase 7, an ortholog of Drosophila death caspase 1 and Ice, was also found to be expressed in a sex-specific manner (Figure 4F; Additional file 1). In addition to regulation of apoptosis, p53 has also been shown to regulate neurite outgrowth and axonal regeneration [47], suggesting that it may impact these processes in the mosquito brain. Finally, p53 has been linked to the regulation of life span in Drosophila[48]. Overexpression of p53 in the female nervous system results in increased life span in females. Thus, it is possible that female-specific upregulation of p53 expression in the A. aegypti female brain may contribute to the relatively longer lifespan of A. aegypti females with respect to males.
Ion transport
A wide variety of ion transporters are upregulated in male pupal heads (Figure 2B, Table 4), which suggests that sexually dimorphic differences in the developing mosquito brain may result from differential ion channel expression. As discussed by Yamamoto and Lopez-Bendito [49], the interaction of neural activity and genetic programs specifies neural circuit composition and organization during development. Changes in electrical activity regulate neural growth, axonogenesis, and the branching of axons. The sex-specific differences in ion transport-related genes noted here are therefore of potential interest with respect to the development of sexually dimorphic neural circuitries in mosquitoes. Furthermore, the activation and inactivation properties of ion channels in axons determine the short-term dynamics of axonal propagation and synaptic transmission [50].
Our microarray experiments detected upregulated male-specific expression of glutamate receptor 7 (Table 4), an ortholog of D. melanogaster ionotropic receptor (IR) 25a. In relation to this, glutamate metabolism was also noted as a significant pathway among DETs upregulated in males (Figure 3; Additional file 3). In Drosophila, IRs, which have evolved from ionotropic glutamate receptors (iGluRs), are expressed in a combinatorial fashion in sensory neurons and respond to many distinct odors [51]. Sensory neurons expressing IR84a in fruit flies innervate a glomerulus that expresses male-specific isoforms of the sex determination gene fruitless (fru), one of three sexually dimorphic glomeruli in the antennal lobe. Mutation of IR84a reduces male courtship (reviewed by [52],[53]). These observations suggest that differential expression of glutamate receptors and differential glutamate metabolism may impact sex-specific behaviors in mosquitoes.
Proteasome
DETs upregulated in females were significantly associated with the proteasome (Figure 3, Additional file 3), which is well known for its role in the intracellular degradation of ubiquitinated proteins. Protein synthesis and degradation are particularly important to neuronal development and function because of the large distances between synapses and soma. The ubiquitin proteasome system influences synaptic protein levels through the regulation of protein function and localization, as well as endocytosis (reviewed by [54]). It should be noted that synj (Figures 4G, 6C,G, 7G,H, 11B, Additional file 1), which also regulates endocytosis at the Drosophila synapse [55], is dimorphically expressed in the A. aegypti pupal brain. Elements of the ubiquitin proteasome system have also been found to regulate synaptic strength, homeostatic plasticity, axon growth and guidance, and dendrite morphogenesis (reviewed by [54]). Thus, differential expression of ubiquitin proteasome pathway components in female and male mosquitoes may have significant impacts on neural development and function that contribute to the development of sexually dimorphic behaviors.
Biosynthetic and catabolic processes
Significantly upregulated DETs in the female pupal head were also associated with biosynthetic and catabolic processes (Figure 3, Additional file 3). For example, glycosphingolipid biosynthesis is significantly over-represented among DETs upregulated in females. The functions of glycosphingolipids have been well studied in the context of development and neurobiology (reviewed by [56],[57]). For example, gangliosides, sialic acid-containing glycosphingolipids, are abundant in the nervous system, where they function in cell-cell recognition, adhesion, and signal transduction [57]. Work in Drosophila has shown that sphingolipid regulators affect cell survival, growth, and specification, as well as the control of lipid storage and responses to nutrient availability [56]. Thus, the differential regulation of sphingolipids in the developing male and female brain could have significant impact on the development of sex-specific behaviors in A. aegypti.
DETs significantly upregulated in female pupal heads are also associated with the glycosaminoglycan (GAG) degradation pathway (Figure 3, Additional file 3). In A. aegypti, cell surface GAGs have primarily been studied in relation to their potential roles as dengue virus receptors (reviewed by [58]). It will therefore be interesting to determine if there is differential GAG expression in other tissues, such as the midgut, a subject for future investigations. In reference to the brain, extracellular proteoglycans, which have one or more covalently attached GAGs, are known to impact many aspects of neural development and CNS maintenance. For example, during development, proteoglycans regulate cell adhesion, neurite formation, axon growth and guidance, and synapse formation. Proteoglycans have also been implicated in axon regeneration and sprouting following injury [59],[60]. Thus, differential GAG degradation could impact multiple aspects of neural development and function in A. aegypti females vs. males.
Dsx is a regulator of dimorphic gene expression in the A. aegypti pupal nervous system
Our search of the A. aegypti genome for the Luo et al. [28] Dsx consensus sequence uncovered 732 putative Dsx binding sites (Additional file 2), 48 of which correspond to genes identified in our microarray experiments (Additional file 3). Our search algorithm allowed for one mismatch in the consensus binding site (see Methods). This was useful to us since it permitted the most rigorous selection of putative targets prior to pursuing the dsx knockdown studies, which are labor-intensive. However, it is quite possible/probable that more mismatches are permissible and that additional binding sites are present in the genome. To more fully pursue the identification of Dsx binding sites throughout the genome, future studies might include a ‘wet lab’ approach, for example chromatin immunoprecipitation with anti-Dsx antibodies, or perhaps through use of the DamID approach utilized by Luo et al. (2011) to assess D. melanogaster Dsx binding sites.
Few of the targets identified in our study correspond to known Dsx transcriptional targets in D. melanogaster[28], suggesting that the targets of Dsx have evolved rapidly within the Diptera and highlighting the need to pursue analysis of dsx function in mosquitoes. To this end, we used siRNA-mediated gene targeting to examine the function of dsx in the developing pupal brain (Figure 11). These experiments indicated that Dsx is a key regulator of sex-specific gene expression during A. aegypti neural development. The detection of Dsx consensus binding sites associated with the p53, synj, geko, cdk4/6, and rab6 genes (Additional file 2), all of which have sexually dimorphic expression in the pupal brain (Additional file 1, Figures 4, 6, 7, 11) that is disrupted by dsx knockdown (Figure 11A-E), suggests that Dsx directly regulates expression of these genes. However, Dsx of course may function to regulate expression of these genes and its other targets through direct and/or indirect mechanisms, a question for future studies. Moreover, studies of the regulation of yolk protein genes, which are direct targets of Dsx in D. melanogaster, suggested that DsxF acts as a transcriptional activator, while DsxM functions as a repressor [61],[62]. Interestingly, none of our dsx knockdown experiments in A. aegypti resulted in upregulation of putative target gene expression in males or females (Figure 11), but it will be interesting to explore this subject further in future investigations.
The roles of dsx in development of the nervous system have been explored in D. melanogaster. In flies, Dsx regulates differences in programmed cell death, which is known to contribute to sexually dimorphic nervous system development [9]. It is therefore possible that control of p53 expression by Dsx in the A. aegypti CNS (Figure 11A) may be required for regulation of sex-specific cell death patterns in the mosquito brain, an interesting subject for future investigations. Dsx has also been associated with dimorphic cell proliferation in the Drosophila CNS, which is interesting in light of the observation that Dsx regulates expression of the cell cycle regulator cdk4/6 in the developing A. aegypti brain (Figure 11D). Moreover, Rideout et al. [63] detected sex-specific differences in the numbers, axonal projections, and synaptic density of Dsx-expressing neurons in Drosophila, and Dsx may regulate these aspects of neural development in mosquitoes. Finally, a recent study demonstrated that in female fruit flies, the neural circuitry associated with female post-mating behavior is specified by Dsx function. During copulation, this female circuitry senses male sex peptides and relays the signal to higher order circuits in the brain that generate post-mating physiological and behavioral output responses in females [64]. It will be interesting to determine if Dsx plays similar roles in mosquitoes.
We note that in comparison to D. melanogaster, in which sex-specific expression of dsx is detected in only several subsets of neurons [63],[65], dsx is expressed much more broadly within the A. aegypti female and male pupal brain (Figures 8A, 9, 10). For example, although dsx expression is not detected in the optic lobe of the D. melanogaster pupal brain, A. aegypti dsx is expressed in the male and female pupal optic lobes (Figures 8A, 9A,D, 10A,B). Moreover, we detect sex-specific differences in dsx expression in the antennal lobe (Figures 9C,F, 10C,D) and mushroom bodies (Figure 9A,D, 10E,F) of A. aegypti, differences which have not been reported in D. melanogaster pupae. These findings suggest that Dsx may play a prominent role in the regulation of sex-specific neural development in A. aegypti, in which its dimorphic expression pattern suggests that it may contribute to sex-specific differences in the visual and olfactory systems, the processing of sensory information, as well as learning and memory.
Previous studies have demonstrated that Dsx and Fru act in the same neurons to generate neuronal wiring and behaviors [63],[66],[67]. Fru targets, which were recently identified in D. melanogaster[68],[69], include many genes that regulate neural processes, including neurotransmission, ion-channel signaling, and neuron development. Furthermore, Neville et al. [68], who detected over-representation of known Dsx targets in their FruM target data set, speculate that Drosophila Dsx and Fru act together, either in a physical complex or through coregulation of target genes, to specify sex-specific neural development. Sex-specific Fru splice forms have been detected in A. aegypti[70], but the expression patterns of these transcripts have not yet been reported in the developing brain, and fru function has not been investigated in mosquitoes. In future studies, it will be interesting to functionally assess the roles of fru isoforms during A. aegypti female and male neural development.
Sexual dimorphism in the visual system
Differential expression of a number of the transcripts detected in the sex-specific pupal head transcriptome localized to the optic lobe of the brain (Figures 4 and 11), where high levels of the female and male dsx splice forms were also detected (Figures 8A, 9A,D, 10A,B). Moreover, many of the over-represented GO categories among genes with Dsx consensus binding sites are related to the compound eye or compound eye development (Table 5). Combined, these findings suggest that male and female compound eyes and higher order visual processing may differ in A. aegypti. Such differences could potentially contribute to female-specific behaviors, such as detection of blood meal hosts or oviposition sites. It is also possible that dependence on vision for mating success differs in males and females. This has been observed in D. melanogaster, in which males have a strong dependence on visual function for finding and tracking females, while female receptivity is based on responses to non-visual cues, such as male courtship song quality [71]. Exploration of these topics will prove interesting in future investigations.
Sexually dimorphic visual development has been particularly well-studied in stalk-eyed flies, which as a result of their elongated and sexually dimorphic eye-stalks are a model system for investigating sexual selection [72]. Wilkinson et al. [72] compared sex-specific global patterns of gene expression in Teleopsis dalmanni, in which they identified 415 female-biased and 482 male-biased transcripts associated with dimorphic eyestalk development. Many of the genes upregulated in developing female eyestalks are associated with cell differentiation and patterning, which was also the case for genes flanking Dsx consensus binding sites in A. aegypti (Table 5). DETs upregulated in males are disproportionately associated with Teleopsis dalmanni growth, which is to be expected given the male-specific eye stalk elongation observed in this species. Genes flanking Dsx consensus sites in A. aegypti (Additional file 2) are also significantly associated with growth (Table 5), and as discussed above, DETs upregulated in females are associated with the cell cycle (Figure 2). It is likely that these genes are associated with overgrowth of female tissues in A. aegypti given that female pupae of this species are larger than males. In support of this notion, cdk4, a positive regulator of cellular growth in D. melanogaster[43],[44], is upregulated in the A. aegypti female pupal brain.
The results of this investigation indicate that rab6, which is upregulated in females, is a target of Dsx (Figure 11E). In Drosophila, Rab6 functions as a GTP-binding protein regulator of vesicular trafficking that has been implicated in the transport of rhodopsin [73]. It will be interesting to determine if Rab6 expression is dimorphic in the eye and in adult A. aegypti, in which it could potentially regulate Rhodopsin transport. The A. aegypti genome project uncovered 10 Rhodopsins [1]. Based on studies in D. melanogaster, one would expect that changes in Rab6 levels would differentially impact transport of the 10 Rhodopsin proteins [73]. If this is the case in A. aegypti, differential Rab6-mediated transport of the 10 Rhodopsin proteins might be of significant consequence to male and female vision. This question, as well as functional analysis of many other DETs in the developing visual system, will be explored in future investigations.