Nutritional and hormonal regulation of the TOR effector 4E-binding protein (4E-BP) in the mosquito Aedes aegypti

Abstract

Mosquitoes require blood for egg development, and, as a consequence, they transmit pathogens of devastating diseases. Target of rapamycin (TOR) signaling is a key pathway linking blood feeding and egg development in the mosquito Aedes aegypti. We show that the regulation of the TOR effector translational repressor 4E-BP is finely tuned to the nutritional requirements of the female mosquito, and it occurs at transcriptional and post-translational levels. Immediately after blood feeding, 4E-BP became hyperphosphorylated, suggesting rapid inhibition of its translational repression function. 4E-BP was highly phosphorylated after in vitro incubation of the fat body in the presence of amino acids; this phosphorylation was rapamycin insensitive, in contrast to another TOR target, S6K, phosphorylation of which was rapamycin sensitive. A high level of 4E-BP phosphorylation was also elicited by insulin. Rapamycin and the PI3K inhibitor LY294002 blocked insulin-mediated 4E-BP phosphorylation. RNA-interference depletion of the insulin receptor or Akt resulted in severe reduction of 4E-BP phosphorylation. Phosphorylation and stability of 4E-BP was dependent on its partner eIF4E translation initiation factor. Silencing of 4E-BP resulted in reduction of the life span of adult female mosquitoes. This study demonstrates a dual nutritional and hormonal control of 4E-BP and its role in mosquito egg development.—Roy, S. G., Raikhel, A. S. Nutritional and hormonal regulation of the TOR effector 4E-binding protein (4E-BP) in the mosquito Aedes aegypti.

The eukaryotic translation initiation machinery is composed of eukaryotic initiation factors and controls the recruitment of ribosomes to mediate cap-dependent translation. The initiation of cap-dependent translation is tightly controlled by endogenous and exogenous stimuli (1, 2). The highly conserved target of rapamycin (TOR) pathway, containing hormonal insulin and amino acid (AA) nutritional branches, plays a critical role in regulation of translation by conveying extracellular nutritional conditions (3, 4). Activation of this pathway stimulates translation via phosphorylation of the major downstream effector molecules—the translational activator, RPS6-p70-protein kinase (S6K), and the translational repressor, eukaryotic initiation factor 4E binding protein (4E-BP). The translational repressor 4E-BP is critical in modulating translational events and is dependent on nutritional, developmental, and stress conditions. Under nutritionally favorable conditions, positive stimulation through the TORC1 (TOR complex 1) pathway phosphorylates and inactivates the translational repressor 4E-BP, resulting in its dissociation from the eukaryotic initiation factor 4E (eIF4E) and facilitating the assembly of the eIF4F complex, a protein complex required for the initiation of cap-dependent translation (4). Although the TOR pathway is highly conserved in eukaryotic organisms from yeast to human, fine-tuning of the TOR pathway by either hormonal or nutritional signals permits various organisms to respond rapidly to their unique developmental or nutritional needs.

Mosquitoes transmit numerous dangerous diseases and impose an enormous burden on the health and well-being of humans. Female mosquitoes serve as vectors of disease pathogens, because they need vertebrate blood to reproduce. Ingestion of a protein-rich vertebrate blood meal induces a massive and rapid expression of numerous genes, including those encoding yolk protein precursors (YPPs) in the mosquito fat body, a tissue analogous to vertebrate liver and adipose tissue. Synthesized YPPs are secreted from the fat body and then internalized by developing oocytes (5). Obligatory blood feeding presents a unique requirement for signaling systems to be able to rapidly respond to the appearance of a large amount of nutritional reserves. The activation of the cellular transcriptional and translational machinery constitutes a crucial step for synthesizing massive amounts of YPPs.

The TOR pathway provides an essential signaling link in conveying the appearance of a massive pool of AAs that mosquitoes utilize during their gonadotrophic cycles (6). Blood-meal activation of TOR occurs via several signaling AAs (7, 8) and leads to downstream events, including activation of S6K via its phosphorylation, translation of GATA, and transcription of the major YPP, vitellogenin (Vg; refs. 9, 10). The small GTPase Ras homologue enriched in brain (Rheb) is crucial for mediating AA signaling to TOR in the mosquito fat body (11). RNA-interference (RNAi)-mediated depletion of either S6K or Rheb demonstrated that the TOR pathway is required for mosquito egg development (9, 11). Moreover, insulin-like peptides are involved in the regulatory circuit linking blood digestion and egg maturation in the mosquito Aedes aegypti (12–14). It has been demonstrated for several insects, including A. aegypti, that insulin and TOR signaling serve as an important sensor of nutritional status and are required for initiation of reproductive events, such as Vg synthesis and oocyte maturation (15–18). The insulin/phosphoinositide 3-kinase (PI-3K) pathway controls the Forkhead box factor O (FOXO) transcription factor, inactivating it through phosphorylation; conversely, repression of this pathway results in activation of FOXO (19, 20). FOXO is required for activation of Vg expression in A. aegypti (21).

In this study, we investigated how the TOR-dependent translational repressor 4E-BP is regulated during blood-meal-activated gonadotrophic cycles, and what role it plays in controlling the translational machinery in the A. aegypti fat body. We show that a blood meal augments 4E-BP phosphorylation in adult female mosquitoes, thus promoting its repression. This phosphorylation event could be replicated in vitro in the presence of either AAs or insulin. Starvation induced expression of the A. aegypti 4E-BP (Aa4E-BP) gene. RNAi depletion of the A. aegypti insulin receptor (AaInR), A. aegypti protein kinase B (AaAkt), or A. aegypti Rheb (AaRheb) severely reduced 4E-BP phosphorylation in vitro. RNAi silencing of FOXO via RNAi also inhibited 4E-BP transcription, both in vitro and in vivo. We present data showing that RNAi depletion of 4E-BP affects mosquito longevity. Thus, the translational repressor 4E-BP is critical in modulating translational events that are dependent on nutritional, developmental and stress conditions.

MATERIALS AND METHODS

Mosquito rearing

A. aegypti mosquitoes were maintained in laboratory culture, as described previously (12). To initiate egg production, mosquitoes were fed on white rats. All procedures for using vertebrate animals were approved by the University of California Riverside Institutional Animal Care and Use Committee (A20100016; May 27, 2010).

In vitro fat body culture

The in vitro fat body culture was performed as described previously (7, 12). Fat bodies were incubated either in medium lacking AAs (AA−; containing equimolar amounts of mannitol in place of AAs) or with medium containing all the AAs (AA+). The following reagents were used: bovine insulin solution (Sigma-Aldrich, St. Louis, MO, USA), rapamycin (Sigma-Aldrich; dissolved in ethanol, 2 mM), LY294002 (Calbiochem, Gibbstown, NJ, USA; dissolved in ethanol), cycloheximide (Calbiochem) and 20-hydroxyecdysterone (20E; Sigma-Aldrich; dissolved in ethanol; 1 μM for all experiments).

Cloning and sequencing of Aa4E-BP cDNA

Expressed sequence tag cDNA sequences coding for the Aa4E-BP gene were identified in the MIT BROAD database (http://www.broad.mit.edu/annotation/genome/aedes_aegypti.2/Blast.html), using the Drosophila 4E-BP/Thor protein as the template (tBlastn). Common PCR techniques, using gene-specific primers, were employed for the amplification of the full-length 4E-BP cDNA from the A. aegypti fat body cDNA pool. All PCR products were cloned in pCRII-TOPO vector (Invitrogen, Carlsbad, CA, USA). The full-length cDNA and deduced AA sequences of Aa4EBP were compared with the BLAST tool from the National Center for Biotechnology Information (NCBI; Bethesda, MD, USA). Sequence alignments were performed using the T-Coffee server, with Clustal as the algorithm of choice. Conserved AA sequences between different species were shaded using the BoxShade program. The values for the fraction of sequences that must agree to shading were set for at least 60% identity.

RNA extraction, reverse transcription, and real-time PCR

Total RNA was isolated by means of the 1-step phenol/guanidinium/thiocyanate method (TRIzol; Invitrogen). Aliquots of 1 μg total RNA treated with amplification-grade RNase-free DNase I (Invitrogen) were used in the cDNA synthesis reactions using an Omniscript reverse transcriptase kit (Qiagen, Valencia, CA, USA). Reverse transcription was carried out in 20-μl reaction mixtures containing random primer or oligo(dT) primer at 37°C for 1 h; PCR products were obtained using a HotStar Taq Master Mix kit (Qiagen).

Quantitative real-time PCR (qPCR) reactions were performed using the iCycler iQ system (Bio-Rad, Hercules, CA, USA) in 96-well plates using TaqMan primers and probes, and SYBR green primers for 4E-BP and S7 ribosomal protein (internal control).Quantitative measurements were performed in triplicate and normalized to the internal control of S7 ribosomal mRNA for each sample (except for Aa4E-BP expression profile). Primers and probes were as follows (all TaqMan probes used the Black Hole Quencher and were synthesized by Qiagen): Vg forward, 5′-ATGCACCGTCTGCCATC; Vg reverse, 5′-GTTCGTAGTTGGAAAGCTCG; Texas Red-labeled Vg probe, 5′AAGCCCCGCAACCGTCCGTACT; S7forward, 5′-TCAGTGTACAAGAAGCTGACCGGA; S7 reverse, 5′-TTCCGCGCGCGCTCACT-TATTAGATT; 4E-BP forward, 5′-CCTCCGAACTGCCGATCTCTACTCGT; 4E-BP reverse, 5′-TCAGATTCATCAAGAAGGCACGTTCG.

Antibodies and immunoblot analysis

The polyclonal antibody against A. aegypti 4E-BP was generated by Proteintech group (Chicago, IL, USA). It was produced using a fusion A. aegypti 4E-BP protein expressed in Escherichia coli BL21 (DE3) cells and purified by Qia-express Ni-NTA columns (Qiagen). The following commercially available antibodies were used in this study: the polyclonal antibody recognizing human phosphorylated p70 S6K at Thr 389 and A. aegypti S6K at Thr 388 (Upstate, Lake Placid, NY, USA; ref. 9); the polyclonal antibody against S6K protein (Santa Cruz Biotechnology, Santa Cruz, CA, USA); polyclonal antibodies detecting phospho-4E-BP [Thr37/46 (236B4); 2855, Cell Signaling Technologies Inc., Danvers, MA, USA]; and monoclonal antibody against β-actin (Sigma-Aldrich). For immunoblots, groups of 9 fat bodies were homogenized in 100 μl of a breaking buffer (50 mM Tris, pH 7.4; 1% Igepal; 0.25% sodium deoxycholate; 150 mM NaCl; 1 mM EDTA; 1 mM phenylmethyl-sulfonylfluoride; 1× protease inhibitor mixture; and 1× phosphatase inhibitor mixture, Sigma-Aldrich). Total protein, 50 μg, was boiled in 1× LDS NuPage sample buffer (Invitrogen) with 1× sample reducing agent (Invitrogen) for 5–10 min. Protein samples were resolved on 4–15% gradient SDS-PAGE gels (Bio-Rad) and electrotransferred to PVDF membranes. The membranes were incubated overnight in Starting Block T20 (PBS) blocking buffer (Thermo Scientific, Waltham, MA, USA) at 4°C. After washing with PBS/T (PBS, pH 7.4, with 0.05% Tween20), the membranes were incubated with the primary antibody in the Starting Block T20 overnight at 4°C. After a second washing in PBS/T, the membranes were incubated with secondary antibody in the same blocking buffer and again at 4°C overnight. The bands were visualized with chemiluminescent substrate ECL (Thermo Scientific). Each experiment was performed ≥3 times using different cohorts of female mosquitoes, with similar outcomes.

RNAi-knockdown experiments

For production of double-stranded RNA (dsRNA), PCR primers were synthesized against specific regions of A. aegypti TOR (AaTOR), AaRheb, AaInR, AaAkt, A. aegypti FOXO (AaFOXO) AaeIF4E, A. aegypti eukaryotic initiation factor 4A (AaeIF4A), A. aegypti eukaryotic initiation factor 4 γ (AaeIF4G), A. aegypti eukaryotic initiation factor 2 α (AaeIF2α), A. aegypti eukaryotic initiation factor 3 α (AaeIF3α), and Aa4E-BP, with T7 minimal primers attached to both the 5′ and 3′ ends. Primers were chosen such that the PCR products would be between 500 and 600 nt in size. The PCR products were then concentrated using the Qiagen PCR cleanup and gel extraction kits. Finally, double strands of RNA were produced by means of in vitro transcription with the T7 RNA polymerase using the MEGAscript T7 kit (Ambion, Austin, TX, USA). dsRNA from a small portion of the bacterial gene MaL (dsMaL) was used as a negative control. The control dsRNA Mal was amplified from the plasmid 28iMal (New England Biolabs, Ipswich, MA, USA) and contains a nonfunctional part of the E. coli MalE gene that encodes the maltose-binding protein. dsRNA (1 μg) in 0.3–0.5 μl of distilled water was injected into the thorax of CO₂-anesthetized female mosquitoes at 1 d after emergence. The dsRNA-injected mosquitoes were allowed to recover for 4–5 d and then subjected to further examination. RNAi-depletion efficiency was analyzed by means of RT-PCR. The following primers were used for RT-PCR analysis: 4E-BP forward, 5′-GCAGCGAGCAGTCAGTGTGTGCAGGT; 4E-BP reverse, 5′-GAAGTTGGATGGTTGGTTGGATCAGC; TOR forward, 5′-GCTGAAGA-GCCCCTCGTC; TOR reverse, 5′-CACGTGCATGACGCTTTC; eIF4E forward, 5′-TGGCTGGCAAATCCAACGAA; eIF4E reverse, 5′-ATTGCCCTCGTCTTCCCACA; eIF4A forward, 5′-ATGGACCCGCTGGTATGCAA; eIF4A reverse, 5′-TCCAGTTTT-CCTGCTTCACATCG.

Mosquito survival rate experiments

One-day-old female mosquitoes were injected with dsRNA, as described previously, and maintained on 10% sucrose. After the fifth day, the mosquitoes were provided only water. The number of dead mosquitoes was counted and noted down each day. Twenty mosquitoes were used in evaluation of each time point in a single experiment. The entire longevity time course was repeated 4 times with different cohorts of mosquitoes. The cumulative results were plotted and analyzed in an Excel sheet (Microsoft, Redmond, WA, USA). Statistical analysis was performed using the log-rank test.

RESULTS

Cloning and sequence features of Aa4E-BP

There is only a single 4E-BP gene in the A. aegypti genome, similar to the case of Drosophila melanogaster, and in contrast to mice and humans, which have 2 and 3 4E-BP genes, respectively. The cDNA corresponding to the Aa4E-BP open reading frame encodes for a protein comprised of 118 AAs. Aa4E-BP was rich in Ser/Thr (18.64%) and contains several potential phosphorylation sites (Ser/Thr-Pro). It has a conserved structure with 3 domains: the RAIP motif at the N-terminal end; the TOS motif, consisting of FEMDI (the final 5 AAs at the C-terminal end); and the eIF4E binding site with the consensus sequence of YXXXXLΦ (22). This domain structure was similar in the 4E-BPs of 4 insects (A. aegypti, Anopheles gambiae, D. melanogaster, and Bombyx mori); however, there were several AA substitutions compared to mammalian 4E-BPs (Supplemental Fig. S1).

Expression of the 4E-BP gene in the mosquito fat body during the first vitellogenic cycle

During previtellogenesis, Aa4E-BP mRNA was present at a high level, reaching a peak at 24 h posteclosion but remaining high until blood feeding (Fig. 1A). The 4E-BP mRNA level dramatically dropped within an hour after blood intake, reaching almost undetectable levels between 3 and 12 h post-blood meal (PBM). The level started rising again from 24 h PBM and returned to previtellogenic levels between 48 and 72 h PBM. Hence, our results suggested that 4E-BP gene transcription was suppressed following nutritional stimuli from a blood meal.

Figure 1.

Regulation of Aa4E-BP gene expression in fat bodies of A. aegypti females. A) Aa4E-BP mRNA profile during pre- and postvitellogenic periods. Relative mRNA levels were determined using qPCR. Three groups of 3 fat bodies, dissected from 3 mosquitoes were used per time point. Samples from 3 independent biological replicates were analyzed. For these experiments, mosquitoes were maintained continuously on 10% sucrose solution prior to blood feeding. Timing of sampling is indicated in hours after eclosion and after blood feeding. B) Starvation-induced stimulation of 4E-BP transcription. Three-day-old adult females fed sucrose and water were collected and then maintained on water for 4 d (PV starved). Control mosquitoes (PV+1 h sucrose) were returned to feeding on 10% sucrose for 1 h prior to the qPCR analysis. The experiment was performed in triplicate using separate cohorts of mosquitoes. C) Fat bodies dissected from mosquitoes as in B were incubated in either AA-free (AA−) medium, AA− medium with insulin (Ins), AA+ medium, or AA− medium with 20E for 3 h. Total RNA was collected and analyzed for 4E-BP transcript by means of qPCR. Data was normalized using S7 as the internal control. Data represent means ± se of triplicate samples from 3 unique cohorts of mosquitoes. D) Effect of RNAi depletion of FOXO on 4E-BP transcript level in the fat body. One-day-old adult female mosquitoes were injected with either dsFOXO (iFOXO) or control dsMal (iMal) and allowed to recover for 4 d, with water as their only diet. Fat bodies were then dissected; total RNA was collected and subjected to qPCR for analysis of 4E-BP transcript levels. Data represent ± se of triplicate samples from 3 unique cohorts of mosquitoes. iMal, control dsRNA; S7, S7 ribosomal RNA used as a loading control.

Prior to blood feeding, female mosquitoes feed on nectar. To determine the effect of carbohydrates on transcription of the 4E-BP gene in fat bodies, we maintained previtellogenic female mosquitoes for 3 d on 10% sucrose solution and then on water for 4 d. An additional group of mosquitoes was maintained under the same conditions, but was given 10% sucrose solution for 1 h at the end of the experiment. 4E-BP mRNA level was elevated in fat bodies of starved mosquitoes, but was significantly lower in mosquitoes exposed to sucrose for 1 h (Fig. 1B). Thus, nutrient deprivation resulted in up-regulation of 4E-BP gene expression, while availability of nutrient, both carbohydrate and protein, down-regulated its expression.

Next, we asked what factors regulated 4E-BP gene expression. Fat bodies from female mosquitoes, which were maintained on water for 4 d, were dissected and incubated in AA− medium for 3 h. Other batches of fat bodies from the same pool of mosquitoes were incubated in the AA− medium containing either insulin or 20E, or in the medium containing AAs (Fig. 1C). The 4E-BP mRNA level remained high in fat bodies after AA− incubation. In contrast, its level dropped significantly after exposure to either insulin or AAs when compared with control (AA−). 20E had no effect on 4E-BP mRNA level. Thus, factors implicated in nutritional regulation—AAs and insulin—down-regulated 4E-BP expression.

To confirm involvement of insulin in regulation of 4E-BP expression in the fat body of previtellogenic females, we utilized RNAi depletion of FOXO. The level of 4E-BP transcript was significantly lower in fat bodies from female mosquitoes with RNAi depletion of FOXO incubated in AA− medium than in the iMal control (Fig. 1D). Supplemental Fig. S2A shows the effectiveness of RNAi depletion of FOXO. The latter experiment suggested that 4E-BP gene expression was up-regulated by FOXO and down-regulated by insulin.

Induction of Aa4E-BP phosphorylation in the mosquito fat body after blood feeding

Western blot analysis utilizing anti-phosphor-4E-BP antibodies to detect phosphorylated 4E-BP at position Thr 37/46 showed that the level of Aedes 4E-BP phosphorylation was low in fat bodies of female mosquitoes prior to blood feeding; however, it dramatically increased immediately after blood feeding, within 1 h PBM (Fig. 2A). It declined again to a background level by 12 h PBM. 4E-BP protein level was high before blood feeding, but declined between 6 and 24 h PBM. As a control, we monitored the S6K phosphorylation level, which showed a similar pattern of drastic elevation during the first hours PBM. The latter is in agreement with our previously reported data (9). However, unlike 4E-BP protein level, the level of S6K protein did not changed throughout vitellogenesis (Fig. 2B). These data demonstrate that blood meal stimuli result in rapid and dramatic hyperphosphorylation of the translational repressor 4E-BP, a step required for repression of its activity and increased level of translation. A similarly dramatic increase in 4E-BP phosphorylation occurred in ovaries immediately after ingestion of blood (not shown). The specificity of antibodies against Aedes 4E-BP protein is shown in Fig. 2C; 4E-BP protein is abundant in the fat bodies of Mal-depleted females but not in those from 4E-BP-depleted females.

Figure 2.

Induction of 4E-BP phosphorylation in the mosquito fat body on blood feeding. A) PBM phosphorylation of 4E-BP. Fat bodies from adult female mosquitoes were isolated at indicated time points before and after a blood meal (BM), and the level of 4E-BP phosphorylation was assessed by means of immunoblot analysis with antibodies 4E-BP phosphorylated at Thr37/46 residues. Antibody against nonphosphorylated 4E-BP protein was used to control 4E-BP protein levels. Actin detected by anti-actin antibody was used as a loading control. B) PBM phosphorylation of S6K. Protein samples from the same experiments as in A were subjected to Western blot analysis using antibodies recognizing either phosphorylated S6K or native S6K protein; 9 mosquitoes (3 groups of 3 mosquitoes) were used for each experiment. All the experiments were repeated ≥3 times using 3 unique cohorts of mosquitoes. C) Immunoblot analysis of specificity of antibodies against Aedes 4E-BP protein in fat bodies from Mal (iMal)- and 4E-BP (i4EBP)-depleted female mosquitoes. Fat bodies from previtellogenic females were dissected and incubated for 3 h in the presence of AAs to activate 4E-BP phosphorylation, which was visualized with antibodies recognizing 4E-BP phosphorylation at Thr37/46 residues. Actin was used as loading control.

Regulation of 4E-BP phosphorylation by AAs

Next, we examined whether AAs served as a stimulus activating 4E-BP phosphorylation in the mosquito fat body. A low level of 4E-BP phosphorylation in fat bodies from previtellogenic female mosquitoes was not affected by incubation in the AA− medium. In contrast, there was a very high increase in 4E-BP phosphorylation level in the presence of AAs, reminiscent of that of blood feeding stimulation in vivo (Fig. 3). Rapamycin (150 nM) did not inhibit AA-mediated 4E-BP phosphorylation (Fig. 3). In a dose–response experiment, even 50 μM rapamycin did not inhibit 4E-BP phosphorylation in vitro (Supplemental Fig. S2B). In contrast, 150 nM of rapamycin was effective in completely inhibiting AA-mediated S6K phosphorylation in the mosquito fat body in vitro (8). Addition of cycloheximide to the AA+ medium did not affect 4E-BP phosphorylation, indicating that there was no requirement of de novo protein synthesis for induction of 4E-BP phosphorylation by AAs (Supplemental Fig. S2C).

Figure 3.

In vitro induction of 4E-BP phosphorylation in fat bodies by either insulin or AAs. Fat bodies from previtellogenic females were dissected and incubated for 3 h in either AA− or AA+ medium, which was supplemented with insulin, 20E, the PI3K inhibitor LY294002, or the TOR inhibitor rapamycin. Western blot analyses were performed using anti-phospho-4E-BP antibody. A second blot was probed with the antibody against Aedes 4E-BP protein as loading controls. Concentrations of reagents used: insulin, 17 μM; heat-inactivated insulin, 100 μM; LY294002, 5μM; rapamycin, 150 nM; 20E, 1 μM. All experiments were repeated ≥3 times with 3 unique cohorts of mosquitoes; 9 mosquitoes (3 groups of 3 mosquitoes) were used for each treatment.

Insulin/PI3K pathway involvement in 4E-BP phosphorylation in the mosquito fat body

We examined whether insulin had a stimulatory role in 4E-BP phosphorylation. Fat bodies from previtellogenic female mosquitoes were incubated in the AA− medium; addition of insulin to the AA− medium elicited a strong 4E-BP phosphorylation response, similar to that of blood-meal activation or AA activation (Fig. 3). Heat-inactivated insulin had no effect on 4E-BP phosphorylation, indicating that bovine insulin mimicked the action of natural mosquito insulin-like peptide hormones. 20E failed to activate 4E-BP phosphorylation (Fig. 3). To further validate specificity of insulin action, we employed the PI-3K inhibitor LY294002. This drug blocked insulin-mediated 4E-BP phosphorylation in the fat body in vitro (Fig. 3). The inhibition was dose dependent, with 5 μM being the most efficient concentration (Supplemental Fig. S2D) and also effective in inhibiting insulin-mediated S6K phosphorylation in vitro (12).

To further corroborate involvement of the insulin/PI3K pathway in 4E-BP phosphorylation, we utilized the RNAi approach. Silencing of either InR or Akt diminished insulin-activated phosphorylation of 4E-BP in the fat body (Fig. 4A). Although knockdown of either gene resulted in a significant reduction in 4E-BP phosphorylation, the effect of Akt was much stronger than that of InR.

Figure 4.

Effect of RNAi-mediated knockdowns on 4E-BP phosphorylation. A) Depletion of InR and Akt inhibited insulin-mediated 4E-BP phosphorylation in mosquito fat bodies. One-day-old female mosquitoes were injected with dsAaInR, dsAaAkt, or dsMal (control). Injected mosquitoes were allowed to recover for 5 d. Fat bodies from mosquitoes with depleted InR, Akt, or Mal were incubated in AA− medium supplemented with 17 μM of bovine insulin for 3 h. Total protein was collected and subjected to immunoblot analysis using an antibody against phospho-4E-BP (Thr37/46); duplicate blots were probed with antibody to 4E-BP protein. B) Combination of TOR RNAi and rapamycin effectively blocked AA-mediated 4E-BP phosphorylation in cultured fat bodies. dsRNA injections were administered as described in A. Fat bodies were then dissected and incubated in AA+ medium alone or AA+ medium with 150 nm rapamycin for 3 h. C) Rheb knockdown inhibited AA-mediated 4E-BP phosphorylation. RNAi Rheb depletion, conditions for fat body incubation, and detection of phosphorylated 4E-BP were similar to procedures in A. All experiments were repeated ≥3 times with 3 unique cohorts of mosquitoes. D) Effect of depletion of eIF4E (i4E) or eIF4A (i4A) on 4E-BP phosphorylation. dsRNA injections were performed similarly to procedure described in A; fat bodies were dissected and incubated in AA+ medium for 3 h. Immunoblot analyses were performed to determine levels of 4E-BP phosphorylation. A second blot was probed with native 4E-BP in order to determine total levels of the 4E-BP protein in the mosquito fat body. A third blot was also probed with a monoclonal antibody against actin, which served as a loading control. All experiments were repeated ≥3 times with 3 unique cohorts of mosquitoes. E) Effect of depletion of eIF4E or eIF4A on S6K phosphorylation. dsRNA injections were performed similarly to procedure described in A; fat bodies were dissected and incubated in the AA+ medium for 3 h. Immunoblot analyses were performed to determine the levels of S6K phosphorylation. A second blot was probed with native S6K in order to determine total levels of this protein in the mosquito fat body. All experiments were repeated ≥3 times with 3 unique cohorts of mosquitoes.

Regulation of 4E-BP phosphorylation by TOR

TOR plays a critical role in conveying extracellular nutritional conditions (3, 4). Phosphorylation of 4E-BP in fat bodies of female mosquitoes with RNAi TOR depletion was dramatically reduced compared with iMal control when fat bodies were incubated in AA+ medium (Fig. 4B). 4E-BP phosphorylation was further inhibited when these fat bodies were incubated in the presence of 150 nM rapamycin. Effectiveness of RNAi TOR depletion is shown in Supplemental Fig. S2B. Rheb, a component of the TOR pathway, is required for mammalian TORC1 activation by AAs (4). Likewise, mosquito Rheb is obligatory for AA activation of TOR and S6K phosphorylation (11). RNAi depletion of Rheb significantly reduced 4E-BP phosphorylation in mosquito fat bodies stimulated with AAs when compared with the Mal control (Fig. 4C).

Effect of translational machinery on 4E-BP

We utilized the RNAi approach to test the role of components of translation initiation machinery in regulating 4E-BP phosphorylation status. Phosphorylation of 4E-BP was markedly diminished in fat bodies of mosquitoes with the eIF4E depletion knockdown, while eIF4A RNAi-mediated depletion had no effect when compared with the Mal control (Fig. 4D). eIF4E RNAi depletion resulted in a considerable decrease in 4E-BP protein when compared with Mal and actin controls (Fig. 4D). RNAi depletions of other components of the translational machinery did not affect 4E-BP protein (Fig. 4E and Supplemental Fig. S3A), suggesting that 4E-BP stability is negatively affected by depletion of its heterodimeric partner eIF4E. In contrast to its effect on 4E-BP phosphorylation, depletion of eIF4E had no effect on AA-mediated phosphorylation or stability of S6K, which was similar to Mal and eIF4A, indicating that the effect of eIF4E on 4E-BP was specific (Fig. 4E). RT-PCR analysis showed that RNAi depletion of eiF4A and eIF4E was effective (Supplemental Fig. S3A). RNAi depletion of other genes encoding components of transcriptional machinery—AaeIF4G, AaeIF2α, and AaeIF3α—had no effect on AA-mediated 4E-BP phosphorylation (Supplemental Fig. S3B). RNAi depletion of either eiF4A or eIF4E strongly inhibited AA-mediated activation of Vg transcription in fat bodies incubated in vitro, showing requirement of translation for this process (Supplemental Fig. S3C).

RNAi depletion of 4E-BP decreases longevity of female mosquitoes

We found that transcription of Aa4E-BP was up-regulated during starvation, suggesting a survival strategy via inhibition of translation. Thus, we asked whether depletion of 4E-PB would have any effect on resistance of mosquitoes to nutrient deprivation. One-day-old female mosquitoes were injected with either 4E-BP or Mal dsRNA and kept under identical conditions (10% sucrose solution and water) for 3 d to recover; they were then maintained on water only. Under these conditions, 4E-BP-depleted mosquitoes had significantly reduced longevity compared to those with Mal control (Fig. 5A). In a simultaneous experiment, we investigated the effectiveness of RNAi-mediated 4E-BP knockdown over time in adult female mosquitoes. We collected fat-body samples at 4, 7, 11, and 15 d after dsRNA injection, extracted total RNA, and performed RT-PCR analysis. S7 was used as the internal control gene, while dsRNA against Mal was used as the nonspecific control for RNAi. Results showed that 4E-BP mRNA was maintained at the low level in 4E-BP-depleted mosquitoes during the 7-d assay (Fig. 5B). This experiment demonstrated a long-term effectiveness of 4E-BP depletion and suggested that the observed decrease in survival of 4E-BP-depleted mosquitoes under nutrient deprivation was indeed due to the lack of this translation repressor. Because 4E-BP expression is controlled by FOXO (19, 20), we also conducted a similar survival experiment using mosquitoes with depleted FOXO. These mosquitoes also exhibited markedly reduced longevity (Supplemental Fig. S3D).

Figure 5.

Effect of 4E-BP on longevity of A. aegypti female mosquitoes. A) One-day-old female mosquitoes were injected with dsMal or ds4E-BP and were allowed to recover for 5 d with a normal diet of 10% sucrose solution and water. Then they were maintained on water only and scored for mortality. Results were replicated 5 times with unique cohorts of mosquitoes. Data represent average survival rate per time point. B) RT-PCR analysis of RNA knockdown (4E-BP) at 4 and 7 d. Blots show Aa4E-BP and AaS7 expression in the fat bodies injected with ds4E-BP and dsMal.

DISCUSSION

The TOR-mediated translational repressor 4E-BP is critical in modulating translational events, allowing responses to nutritional, developmental, or stress conditions (1–4). We show here that the regulation of 4E-BP is finely tuned according to the nutritional requirements of the female mosquito as a blood-feeding organism, and it occurs at transcriptional and post-translational levels. Expression of the 4E-BP gene in the female mosquito fat body is regulated by nutritional cues; it is high during the previtellogenic period and drops dramatically after blood feeding. Moreover, starvation of previtellogenic female mosquitoes elevates the level of the 4E-BP transcript. Drosophila 4E-BP mediates stress conditions, including starvation (23, 24). However, 4E-BP gene expression itself is up-regulated under starvation conditions; flies fed either sucrose or AAs had a significantly lower level of 4E-BP mRNA than those that were starved (23). Our experiments suggest that both insulin and AAs are involved in regulating 4E-BP gene expression in the mosquito A. aegypti; exposure of the fat body from starved female mosquitoes to either stimulus resulted in a significant drop in the level of 4E-BP mRNA. Involvement of insulin/PI3K pathway in regulating 4E-BP gene expression was further confirmed by FOXO RNAi depletions in female mosquitoes under starvation conditions, which resulted in a significant decrease of 4E-BP mRNA levels. In Drosophila, 4E-BP gene expression is up-regulated by dFOXO; the insulin/PI3K pathway stimulates Akt, which directly phosphorylates dFOXO and inhibits d4E-BP transcription (19, 20). However, dramatic down-regulation of Aedes 4E-BP expression following blood feeding of a female mosquito represents a unique feature of this organism. Additional reduction of the 4E-BP mRNA level due to its regulated degradation cannot be ruled out.

A spectacular increase occurs in 4E-BP phosphorylation immediately following blood ingestion in multiple tissues of mosquito females (ref. 25 and the present study). 4E-BP has been proposed to function as a metabolic brake under conditions of stress (20). In mosquitoes, however, this drastic shift from hypo- to hyperphosphorylated state of 4E-BP represents a part of normal physiology. S6K also undergoes a similar rapid hyperphosphorylation immediately following blood ingestion in multiple tissues of mosquito females (refs. 9, 11, 25 and the present study). Thus, a simultaneous, blood-meal-dependent, TOR-mediated hyperphosphorylation of a translational repressor 4E-BP and a translational activator S6K signifies an important adaptation of mosquitoes for rapid triggering of the translational machinery, required for the onset of reproductive events in the female mosquito. It is likely that such a mechanism of blood-meal-dependent activation of translation is conserved among hematophagous arthropods.

Our experiments have shown that the TOR-dependent hyperphosphorylation of 4E-BP occurs by a dual mechanism via the insulin/PI3K pathway and AAs. AAs served as a powerful stimulus, activating 4E-BP hyperphosphorylation in the mosquito fat body incubated in vitro. Such phosphorylation of 4E-BP in fat bodies of female mosquitoes with TOR RNAi depletion was dramatically reduced when fat bodies were incubated in the presence of AAs. RNAi depletion of Rheb diminished the fat body's ability for AA-mediated 4E-BP phosphorylation, indicating that Rheb is essential for transducing signals through AA signaling to TOR and its downstream target 4E-BP. Overexpression of Rheb in mammalian cells promotes 4E-BP phosphorylation at several residues, including Thr37/46 (4, 26).

The effect of insulin on 4E-BP phosphorylation in fat bodies from previtellogenic female mosquitoes was similarly strong as those elicited by either AAs or blood feeding. Specificity of insulin action on 4E-BP phosphorylation was confirmed by the lack of response from heat-inactivated insulin and its inhibition with LY294002. Silencing of either InR or Akt diminished insulin-activated phosphorylation of 4E-BP in the fat body. Taken together, these experiments clearly demonstrate involvement of the insulin/PI3K pathway in regulation of 4E-BP activity in the mosquito fat body.

Pharmacological tests with rapamycin suggest that insulin- and AA-mediated 4E-BP phosphorylation events in the mosquito fat body occur via distinct signaling downstream of TOR. AA-mediated 4E-BP phosphorylation at residues Thr37/46 was rapamycin insensitive in contrast to the insulin-mediated one, which was rapamycin sensitive. The above data suggest that the AA-mediated phosphorylation of 4E-BP reflects a rapamycin-insensitive output from TOR. This is in accordance with a report that the 4E-BP phosphorylation at Thr37/46 sites is rapamycin insensitive in some mammalian cell types (27). However, this drug inhibited phosphorylation of S6K in the mosquito fat body, irrespective of the mode of activation, by either AAs or insulin (refs. 9, 11 and the present study). Differential action of rapamycin in inhibition of S6K and 4E-BP has been reported for mammalian TOR (26, 27). Moreover, this effect of rapamycin on S6K is not cell specific (26). In the mosquito midgut, AA-mediated 4E-BP phosphorylation has been reported to have limited sensitivity to inhibition with rapamycin, whereas S6K is rapamycin sensitive (25).

Recognition of cellular mRNA by the eIF4F translation initiation complex is a central event of eukaryotic translation (2, 3). The repressor protein 4E-BP competes with eIF4G for binding to eIF4E, thereby preventing formation of the eIF4F complex and inhibiting cap-dependent translation (2, 15). This translation block caused by 4E-BP is reversible by 4E-BP phosphorylation at several residues, including Thr37/46 (2, 22, 24). RNAi depletion of eIF4E had a severe effect on 4E-BP phosphorylation in the mosquito fat body. However, because the protein level of 4E-BP was also diminished, we consider that eIF4E, being a partner of 4E-BP, is required for stability or/and translation of 4E-BP protein. The effect of eIF4E was 4E-BP specific, because its RNAi depletion had no effect on either total S6K level or its phosphorylation in the presence of AAs. Moreover, depletion of neither eIF4G nor eIF4A had any effect on either 4E-BP phosphorylation or the overall level of 4E-BP. Thus, eIF4E control of 4E-BP protein level represents yet another layer of regulation required for rapid post-blood-feeding onset of translation in the mosquito.

Numerous reports have demonstrated involvement of insulin pathway and FOXO in the control of life span in multiple organisms (28–30). In Drosophila, foxo loss of function leads to reduced survival during AA starvation (31). In this study, we showed that Aedes mosquitoes depleted of FOXO had a reduction in their longevity during starvation. FOXO is also important for maintaining a diapause and survival during overwintering of Culex pipiens mosquitoes (32). A key role of FOXO in proteostasis and prevention of age-related reduction in life span is mediated in a large part by its positive target, 4E-BP (33). 4E-BP is essential for starvation and oxidative stress resistance in Drosophila, and its depletion has deleterious effects on the life span of this organism (23, 24). Overexpression of Akt, which is an inhibitor of FOXO, in midguts of Anopheles stephensi female mosquitoes, reduced their life span (34). In this report, we found that A. aegypti female mosquitoes with depletion of either 4E-BP or FOXO had significantly reduced longevity.

In summary, we show that 4E-BP, a key repressor of translation, is a unique sensor of the nutritional status in female mosquitoes. In addition to responding to stress factors, such as starvation, 4E-BP serves as a specific role characteristic of a blood-feeding organism, providing a metabolic brake prior to intake of vertebrate blood. We demonstrate here that rapid hyperphosphorylation of 4E-BP occurs in response to blood intake by means of distinct outputs from TOR, linked to insulin and AAs. Considering that blood-feeding female mosquitoes are the most sought after vectors of medical importance, detailed understanding of nutritional pathways critical for egg development is of great importance. Manipulation of the TOR-mediated nutritional pathway could lead to the development of novel control strategies for disruption of the mosquito reproductive cycle, reduction of the life span, and consequential prevention of pathogen transmission by these disease vectors.