Anti-diuretic action of a CAPA neuropeptide against a subset of diuretic hormones in the disease vector Aedes aegypti

The mosquito Aedes aegypti functions as the primary vector for several viruses; most notably dengue, chikungunya, yellow fever and Zika (Laurence et al., 2005; Murray et al., 2013; Ng and Ojcius, 2009). Like other blood-feeding arthropods, female A. aegypti transmit these pathogenic species during the secretion of saliva into the host during a blood meal (Ribeiro et al., 1984).

Adult A. aegypti exclusively oviposit in freshwater, which is hypotonic to larval haemolymph. Thus, larvae are susceptible to the continual gain of freshwater from drinking and also by osmotic flux of fluid from the external environment across the body surface (Bradley, 1987; Patrick et al., 2006). Concurrently, larvae are subject to the constant diffusional loss of predominant haemolymph ions to the external media (Patrick et al., 2006). As such, larvae possess distinct osmoregulatory mechanisms allowing them to achieve and maintain ionic homeostasis by excreting a dilute urine and maximizing the active uptake of ions into the haemolymph from the external media (Bradley, 1987).

Post-eclosion, adult A. aegypti are subject to disparate osmoregulatory challenges. Under non-feeding circumstances, A. aegypti must conserve haemolymph volume to prevent desiccation and the concentration of their haemolymph ions (Patrick et al., 2006). Initially upon blood feeding by adult females, there is a significant water and NaCl load from the blood plasma, and a subsequent K⁺ load during erythrocyte lysis (Coast, 2009). Therefore, a carefully controlled mechanism must exist which can selectively regulate the absorption and secretion of ions to maintain homeostasis following a blood meal.

A number of specialized organs are responsible for ionoregulation and osmoregulation in A. aegypti, including the midgut, Malpighian tubules (MTs) and the hindgut, as well as the anal papillae in larval stages (Bradley, 1987). The five functionally and structurally homologous MTs are responsible for the formation of primary urine (Beyenbach et al., 1993). The MTs are composed of two cell types that form a simple epithelium; large principal cells and thin stellate cells (Beyenbach et al., 2010). The principal cells facilitate active transport of Na⁺ and K⁺ into the lumen of the MTs from the haemolymph, while stellate cells aid in the transepithelial secretion of Cl⁻ (O'Connor and Beyenbach, 2001). The active driving force responsible for ion transport is the V-type H⁺-ATPase (proton pump) (Wieczorek et al., 1989, 2009), which is localized on the brush-border membrane of principal cells (Weng et al., 2003). The movement of protons into the Malpighian tubule (MT) lumen creates a proton gradient, which energizes exchange of alkali cations (via Na⁺/H⁺ and/or K⁺/H⁺ antiporters) across the apical membrane (Wieczorek, 1992; Wieczorek et al., 1991). A basolaterally localized bumetanide-sensitive Na⁺:K⁺:2Cl⁻ cotransporter has been identified in the MTs of A. aegypti, transporting Na⁺, K⁺ and Cl⁻ from the haemolymph into the principal cells (Hegarty et al., 1991). Transport of KCl and NaCl provides the osmotic gradient that drives water movement into the MTs via aquaporins (Klowden, 2013). The fluid secreted by the MTs is received by the hindgut, and is further modified via the selective reabsorption of solutes and water (Larsen et al., 2014).

The MTs do not receive nervous system innervation; thus, regulation of epithelial ion transport and fluid secretion occurs via circulating messengers in the haemolymph (Beyenbach, 2003). Diuretic hormone regulation of the MTs has been extensively studied in A. aegypti. One of the factors known to influence fluid secretion by MTs is the biogenic amine serotonin (5-hydroxytryptamine, 5-HT). In larval A. aegypti, 5-HT titre in the haemolymph increases in response to salinity and this hormone increases fluid and ion secretion by MTs (Clark and Bradley, 1996, 1997). In adult A. aegypti, 5-HT is also known to increase ion and fluid secretion rates, albeit without eliciting pronounced natriuretic activity (Veenstra, 1988). Upon blood meal engorgement by adult females, mosquito natriuretic peptide (MNP) is released into the haemolymph and stimulates diuresis and natriuresis (Beyenbach and Petzel, 1987; Petzel et al., 1986; Wheelock et al., 1988). MNP acts via cAMP to stimulate the secretion of Na⁺-rich primary urine by the MTs (Beyenbach, 2003; Petzel et al., 1987), which has been identified as a calcitonin-like diuretic hormone 31 (DH₃₁) in both A. gambiae and A. aegypti (Coast et al., 2005). In comparison, the corticotropin-releasing factor (CRF)-related diuretic peptide (DH₄₄) stimulates non-specific transport of both cations (Coast et al., 2005), and also increases levels of cAMP in MTs (Cady and Hagedorn, 1999). Kinin-like diuretic peptides have been shown to influence the activity of insect MTs by increasing fluid secretion and Cl⁻ conductance into the lumen of the MTs through signalling via inositol 1,4,5-trisphosphate and elevating intracellular calcium levels through mobilization of IP₃-sensitive calcium stores (Cady and Hagedorn, 1999; O'Donnell et al., 1996). Members of the CAPA peptide family are also known to elicit control over the MTs, whereby either diuretic or anti-diuretic actions have been established in different insects (Pollock et al., 2004; Paluzzi et al., 2008; Ionescu and Donini, 2012; Quinlan et al., 1997). CAPA peptides, which are also referred to as periviscerokinins, are linked to the neurohaemal system of the abdominal ganglia (Predel and Wegener, 2006), and evidence from the triatomine bug Rhodnius prolixus indicates these peptides are released as hormones to coordinate the cessation of diuresis in this insect (Paluzzi and Orchard, 2006).

Although there has been extensive research investigating the neuroendocrine control of A. aegypti MTs, most studies have focused on diuretic hormones (Cady and Hagedorn, 1999; Clark and Bradley, 1996, 1998; Clark et al., 1998a,b; Coast et al., 2005; Donini et al., 2006; Veenstra, 1988). Limited studies examining anti-diuretic control of the mosquito MTs include an exogenous anti-diuretic factor from the beetle Tenebrio molitor, which was shown to decrease fluid secretion by adult MTs through the intracellular second messenger cGMP (Massaro et al., 2004). In addition, low concentrations of an endogenous CAPA neuropeptide were also shown to inhibit larval fluid secretion through a cGMP signalling mechanism, although supraphysiological concentrations of either CAPA or cGMP showed diuretic activity (Ionescu and Donini, 2012).

In the current study, we investigated the anti-diuretic activity of AedaeCAPA-1 on MTs from fourth instar larvae and adult female A. aegypti stimulated by different diuretic factors. Specifically, we utilized concentrations of AedaeCAPA-1 similar to those that were previously found to have anti-diuretic activity on unstimulated MTs in larval A. aegypti (Ionescu and Donini, 2012). We examined the activity of AedaeCAPA-1 on transepithelial fluid transport in both fourth instar larvae and adult females and determined the concentration and transport of the primary cations (Na⁺ and K⁺) present in the secreted fluid of MTs isolated from adult female A. aegypti. Our results reveal that CAPA peptides are selectively anti-diuretic against a subset of diuretic factors, in both larval and adult stage mosquitoes, but do not influence the proportion of ions transported under the stimulation of each specific diuretic hormone. The proposed second messenger cGMP mimics the selective anti-diuretic actions against diuretics known to target the principal cells and, furthermore, is also capable of inhibiting kinin-stimulated diuresis targeting stellate cells. These findings indicate that both principal and stellate cell-mediated diuresis are reduced through signalling mechanisms involving cGMP, but only the CAPA anti-diuretic hormone acting on principal cells is currently known while the stellate cell-directed anti-diuretic hormone remains elusive.

Eggs of Aedes aegypti (Liverpool strain) were collected from an established laboratory colony maintained in the Department of Biology at York University, Toronto, ON, Canada, and hatched in double-distilled water in an environmental chamber (26°C, 12 h:12 h light:dark cycle). Larvae were fed daily with several drops of a solution composed of 2% (w/v) Argentine beef liver powder (NOW Foods, Bloomingdale, IL, USA) and 2% (w/v) brewer's yeast. Adult female A. aegypti were routinely fed on sheep's blood in Alsevers solution (Cedarlane Laboratories, Burlington, ON, Canada) using an artificial feeding system described previously (Rocco et al., 2017). All adults were fed 10% sucrose solution ad libitum.

Serotonin hydrochloride (5-hydroxytryptamine, 5-HT) was purchased from a local supplier (Sigma-Aldrich, Oakville, ON, Canada), dissolved in ultrapure water to make a 1 mmol l⁻¹ stock solution and aliquots were frozen (−20°C) and stored in the dark. AedaeCAPA-1 (GPTVGLFAFPRV-NH₂) was commercially synthesized (Genscript, Piscataway, NJ, USA) at a purity of >98% and 1 mmol l⁻¹ stock solutions were prepared in cell culture-grade ultrapure water (Wisent Corporation, St Bruno, QC, Canada). We utilized peptide orthologues from other species for the three families of mosquito diuretic peptides tested: Drosophila melanogaster calcitonin-related peptide (DromeDH₃₁) was used in lieu of mosquito natriuretic peptide (MNP), and was a generous gift from Michael J. O'Donnell (McMaster University, Hamilton, ON, Canada); Rhodnius prolixus corticotropin releasing factor-related (CRF-related) diuretic peptide (RhoprDH) was used in lieu of the mosquito CRF-related diuretic peptide, and was a generous gift from Ian Orchard (University of Toronto, Mississauga, ON, Canada); Culex depolarizing peptide (CDP) was used as a representative mosquito kinin-related peptide (Hayes et al., 1994) and was obtained from a commercial supplier (Bachem, Bubendorf, Switzerland).

Dosages of diuretic hormones were adapted from previous studies and selected based on the dosage that most closely achieves half-maximal effective concentration (EC₅₀) of isolated MTs. Specifically, following preliminary experimentation, a dose of 50 nmol l⁻¹ and 10 nmol l⁻¹ RhoprDH in larvae and adult MTs, respectively, was found to produce an intermediate level of stimulation, which is similar to concentrations of a mosquito CRF-related diuretic peptide previously used on A. aegypti MTs (Clark et al., 1998a). A dose of 25 nmol l⁻¹ DromeDH₃₁ peptide was used for both larvae and adults, based on the dose–response determinations for A. gambiae MNP (Coast et al., 2005). A 10 nmol l⁻¹ and 50 nmol l⁻¹ dosage of CDP for larvae and adults, respectively, was selected as an intermediate titre, as previous research determined a kinin-related peptide with an EC₅₀ of 15 nmol l⁻¹ in stimulating fluid secretion by isolated A. aegypti MTs (Schepel et al., 2010). Following preliminary examination and referring to previous investigations on the dose dependency of 5-HT secretory activity on A. aegypti MTs (Clark and Bradley, 1998; Veenstra, 1988), an intermediate dose of 100 nmol l⁻¹ 5-HT was selected to stimulate fluid secretion by the MTs. Finally, concentrations of 0.1–1 fmol l⁻¹ (10⁻¹⁵–10⁻¹⁶ mol l⁻¹) were selected for AedaeCAPA-1 to challenge the different diuretic factors, as this was within the range of concentrations previously found to have significant anti-diuretic activity in larval A. aegypti (Ionescu and Donini, 2012).

In order to determine secretion rates, a modified Ramsay secretion assay (Ramsay, 1954) was performed on isolated MTs of fourth instar A. aegypti larvae and 3–6 day old female adults. Dissections were performed under physiological saline adapted from Petzel et al. (1987) that contained (in mmol l⁻¹): 150 NaCl, 25 Hepes, 3.4 KCl, 7.5 NaOH, 1.8 NaHCO₃, 1 MgSO₄, 1.7 CaCl₂ and 5 glucose, and titrated to pH 7.1. Immediately before use, the physiological saline was diluted 1:1 with Schneider's Insect Medium (Sigma-Aldrich). MTs were transferred to a Sylgard-lined Petri dish containing 20 µl saline bathing droplets submerged beneath hydrated mineral oil to prevent evaporation. The proximal end of the MT was removed from the bathing saline and suspended in the mineral oil by looping the end around a Minuten pin to allow for secretion measurements.

In the interest of determining whether AedaeCAPA-1 antagonizes the effects of the known diuretic factors (5-HT, RhoprDH, CDP and DromeDH₃₁), secretion rates were monitored for larval MTs over 30 min treatments and adult MTs over 60 min treatments. Following the incubations, the size of the secreted droplets was measured via the use of a microscope eyepiece micrometer and fluid secretion rate (FSR) was calculated as described previously (Donini et al., 2008). For experimental treatments, the MTs were either treated with the diuretic factor alone or in combination with AedaeCAPA-1 at a final concentration of 0.1 fmol l⁻¹ (for larvae) or 1 fmol l⁻¹ (for adult female). Basal (unstimulated) secretion rates were measured for 60–120 min in adult MTs treated with saline to determine control rates of secretion and confirm the activity of the diuretic hormones used in this study.

Dose–response activity of a membrane-permeable analogue of cyclic guanosine monophosphate, 8-bromo-cGMP (henceforth referred to as cGMP) (Sigma-Aldrich, Oakville, ON, Canada), was determined on basal fluid secretion rates of adult MTs using concentrations ranging from 10 nmol l⁻¹ to 10 µmol l⁻¹. Maximal inhibition of MT fluid secretion by cGMP was observed at 100 nmol l⁻¹ and, consequently, this concentration of cGMP was used against diuretic hormone-stimulated MTs from adult female A. aegypti.

The concentrations of Na⁺ and K⁺ were measured by using ion-selective microelectrodes (ISME). Microelectrodes were pulled from glass capillaries (TW-150-4, World Precision Instruments, Sarasota, FL, USA) using a Sutter P-97 Flaming Brown pipette puller (Sutter Instruments, San Raffael, CA, USA). Next, the microelectrodes were silanized with N,N-dimethyltrimethylsilylamine (Fluka, Buchs, Switzerland). The Na⁺ microelectrode was backfilled with 100 mmol l⁻¹ NaCl and the K⁺ microelectrode with 100 mmol l⁻¹ KCl. The tip of the electrode was filled with Na⁺ ionophore (sodium ionophore II cocktail A, Fluka) for Na⁺ measurements and K⁺ ionophore (potassium ionophore I cocktail B, Fluka) for K⁺ measurements. The electrode tips were then coated with ∼3.5% (w/v) polyvinyl chloride (PVC) dissolved in tetrahydrofuran, to avoid displacement of the ionophore cocktail when submerged in paraffin oil (Rheault and O'Donnell, 2004). Reference electrodes (1B100F-4, World Precision Instruments) were backfilled with 500 mmol l⁻¹ KCl, which were used for recording both Na⁺ and K⁺ concentrations. Ion concentrations were only measured in the adult MTs, as our focus was primarily on this developmental stage. The actions of AedaeCAPA-1 on fluid secretion rates of larval MTs stimulated with different diuretic factors were examined to determine whether similar anti-diuretic actions were evident.

Data was compiled using Microsoft Excel and transferred to Graphpad Prism software v.7 to create figures and conduct all statistical analyses. Data were analysed accordingly using either unpaired t-tests or one-way ANOVA and Bonferroni post-test, with differences between treatments considered significant at P<0.05.

In order to determine the effects of AedaeCAPA-1 on stimulated larval A. aegypti MTs, we measured the production of primary urine using Ramsay secretion assays. Application of 0.1 fmol l⁻¹ AedaeCAPA-1 along with 100 nmol l⁻¹ 5-HT led to significantly lower fluid secretion rates of larval MTs compared with those following application of 5-HT alone (Fig. 1A). However, co-application of 0.1 fmol l⁻¹ AedaeCAPA-1 and 50 nmol l⁻¹ RhoprDH or 10 nmol l⁻¹ CDP to larval MTs resulted in no difference in secretion values relative to those of MTs treated with RhoprDH alone (Fig. 1B) or CDP alone (Fig. 1C). Similar to the activity observed for larval MTs treated with 5-HT, the rate of fluid secretion by MTs stimulated with 25 nmol l⁻¹ DromeDH₃₁ was significantly inhibited when AedaeCAPA-1 was co-applied (Fig. 1D).

Application of 100 nmol l⁻¹ 5-HT to adult female MTs resulted in an approximately 6-fold significant increase in fluid secretion compared with unstimulated MTs (Fig. 2A). In contrast, co-application of 1 fmol l⁻¹ AedaeCAPA-1 with 5-HT in adult MTs resulted in a reduced fluid secretion rate, not different from that of unstimulated MTs (Fig. 2A). Moreover, fluid secretion rates of MTs treated with AedaeCAPA-1 and 5-HT were significantly lower than those of MTs treated with 5-HT alone (unpaired t-test, P=0.01). To examine the influence of AedaeCAPA-1 on transepithelial cation transport, Na⁺ and K⁺ concentrations were measured using ISME. MTs treated with 5-HT alone or together with AedaeCAPA-1 did not exhibit any difference in the concentration of Na⁺ (Fig. 2B) or K⁺ (Fig. 2C) in the secreted fluid. However, relative to unstimulated MTs, both treatments led to a significant decrease in Na⁺ concentration and, correspondingly, a significant increase in K⁺ concentration. The transepithelial transport rate of both Na⁺ (Fig. 2D) and K⁺ (Fig. 2E) was approximately tripled in MTs treated with 5-HT alone compared with unstimulated controls. Co-application of AedaeCAPA-1 led to a similar K⁺ flux which was not different from that of MTs treated with 5-HT alone; however, Na⁺ transport rate was significantly reduced when AedaeCAPA-1 was co-applied compared with the effect of 5-HT alone (unpaired t-test, P=0.02). Overall, AedaeCAPA-1 did not influence the relative proportions of cations transported (Fig. 2F), with 5-HT treatment leading to a similar transport ratio to that of unstimulated MTs.

Adult female MTs treated with both 1 fmol l⁻¹ AedaeCAPA-1 and 10 nmol l⁻¹ RhoprDH did not exhibit changes to fluid secretion compared with MTs receiving RhoprDH alone (Fig. 2A), with both experimental treatments resulting in secretion rates significantly greater than those of unstimulated controls. MTs treated with RhoprDH alone or together with AedaeCAPA-1 did not have any difference in the concentration of Na⁺ (Fig. 2B) or K⁺ (Fig. 2C) in the secreted fluid, although compared with unstimulated MTs, a decrease in the Na⁺ concentration was observed. As expected, because of the lack of inhibition of AedaeCAPA-1 on RhoprDH-stimulated fluid secretion, transepithelial transport of Na⁺ (Fig. 2D) was indifferent between treatments involving RhoprDH alone or together with AedaeCAPA-1 or when compared with unstimulated controls. The transport rate of K⁺ in RhoprDH-stimulated MTs was similar to that of unstimulated controls (Fig. 2E) whereas a small but significant increase in K⁺ flux was observed when AedaeCAPA-1 was co-applied with RhoprDH compared with unstimulated controls. Overall, AedaeCAPA-1 did not influence the ratio of cations transported (Fig. 2F), with RhoprDH leading to a roughly equimolar transport rate of each cation comparable to that of unstimulated control MTs.

The secretion rates of adult MTs treated with CDP alone or co-applied with 1 fmol l⁻¹ AedaeCAPA-1 were similar (Fig. 2A), with both experimental treatments having secretion rates significantly greater (∼12-fold) than those of unstimulated controls. Addition of 1 fmol l⁻¹ AedaeCAPA-1 to adult MTs treated with CDP had no influence on Na⁺ (Fig. 2B) and K⁺ (Fig. 2C) concentrations, although a decrease in the Na⁺ concentration of the secreted fluid was observed compared with unstimulated controls. Similarly, AedaeCAPA-1 did not influence transepithelial cation transport in CDP-stimulated MTs (Fig. 2D,E) with both cations showing an increase in transport rate relative to unstimulated controls. As a result, AedaeCAPA-1 had no effect on the ratio of cation transport in CDP-stimulated MTs (Fig. 2F) and neither treatment differed from unstimulated controls.

Adult MTs treated with 25 nmol l⁻¹ DromeDH₃₁ and 1 fmol l⁻¹ with AedaeCAPA-1 had a significantly lower secretion rate by over 7-fold (unpaired t-test, P<0.0001) compared with that of MTs stimulated with DromeDH₃₁ alone (Fig. 2A). Notably, MTs treated with DromeDH₃₁ alone had secretion rates significantly greater than those of unstimulated controls whereas MTs receiving DromeDH₃₁ combined with AedaeCAPA-1 had rates of secretion that were not different from those of unstimulated controls. Fluid secreted by adult MTs receiving DromeDH₃₁ combined with AedaeCAPA-1 (but not alone) had a significant increase in Na⁺ concentration compared with unstimulated controls (Fig. 2B). The two DromeDH₃₁ experimental treatments also had a significantly different Na⁺ concentration in the secreted fluid (unpaired t-test, P=0.007), with higher Na⁺ titre in the MTs co-treated with AedaeCAPA-1. Comparatively, K⁺ concentrations in the secreted fluid was significantly reduced compared with unstimulated controls when DromeDH₃₁ was tested alone or together with AedaeCAPA-1 (Fig. 2C). As expected, considering the influence of AedaeCAPA-1 on the DromeDH₃₁-stimulated fluid secretion rate, the transepithelial transport rate of both cations was also affected. Specifically, compared with unstimulated MTs, DromeDH₃₁ led to significantly greater transport rate of Na⁺ whereas AedaeCAPA-1 abolished this increase (Fig. 2D). In particular, the two experimental treatments were also significantly different, with AedaeCAPA-1 significantly reducing Na⁺ transport rate. Compared with unstimulated controls, no significant change was observed in the K⁺ transport rate in MTs treated with DromeDH₃₁ alone or together with AedaeCAPA-1 (Fig. 2E), although these two experimental treatments were found to be significantly different (unpaired t-test, P=0.001) such that AedaeCAPA-1 significantly reduced K⁺ transport rate. Thus, although the ratio of cations was modified by treatment with DromeDH₃₁, which led to an approximate 2-fold increase in Na⁺ transport compared with unstimulated controls, the elevated Na⁺:K⁺ transport ratio was sustained in DromeDH₃₁-stimulated MTs challenged with AedaeCAPA-1 (Fig. 2F).

CAPA peptides have been shown to elicit their effects on insect MT fluid secretion through second messenger pathways that involve cGMP (Ionescu and Donini, 2012; Kean et al., 2002; Pollock et al., 2004; Quinlan et al., 1997). Thus, we investigated the activity of this prospective second messenger in adult A. aegypti MTs to determine its actions on basal and diuretic hormone-stimulated fluid and ion secretion. Application of 100 and 10 nmol l⁻¹ cGMP resulted in a significant inhibition of basal secretion rates, with maximal inhibition leading to a 10-fold decrease, observed with treatment of 100 nmol l⁻¹ cGMP (Fig. 3A). cGMP did not influence the Na⁺ and K⁺ concentrations in the secreted fluid at any of the tested concentrations (Fig. 3B,C); however, cation transport rate was reduced (Fig. 3D,E), with a significant decrease in Na⁺ flux when 100 and 10 nmol l⁻¹ cGMP was used compared with unstimulated MTs. Overall, cGMP did not change the relative ratio of cation transport in the secreted fluid compared with unstimulated controls (Fig. 3F).

In order to determine whether cGMP is involved in the anti-diuretic function of CAPA peptides in adult MTs, we investigated the effects of 100 nmol l⁻¹ cGMP, which we determined was a maximally inhibitory concentration on basal secretion rates. Similar to the effects of AedaeCAPA-1, 100 nmol l⁻¹ cGMP significantly decreased the fluid secretion rate of 5-HT- and DromeDH₃₁-stimulated MTs, whilst having a small inhibitory effect on CRF-stimulated MTs (Fig. 4A). While all diuretic factors elicited significantly greater secretion rates relative to unstimulated MTs, a significant decrease in fluid secretion was observed in 5-HT-stimulated MTs (unpaired t-test, P=0.007) and DromeDH₃₁-stimulated MTs (unpaired t-test, P=0.003) when co-applied with cGMP, but not when MTs were stimulated with the CRF-related peptide RhoprDH. Interestingly, and rather unexpectedly, cGMP also resulted in a substantial (∼9-fold) decrease in fluid secretion rate in CDP-stimulated MTs (unpaired t-test, P=0.004). Although the cation concentrations in most treatments were not influenced by cGMP (Fig. 4B,C) and were similar to unstimulated controls, CDP-stimulated MTs had a significantly higher Na⁺ concentration in the secreted fluid when cGMP was co-applied compared with unstimulated MTs and those treated with CDP alone (unpaired t-test, P=0.013). The transepithelial transport rate of both Na⁺ and K⁺ decreased in DromeDH₃₁-stimulated (unpaired t-test, P=0.004 for Na⁺ and P=0.006 for K⁺) and CDP-stimulated MTs (unpaired t-test, P=0.006 for Na⁺ and P=0.012 for K⁺) when they were treated with cGMP (Fig. 4D,E), while no effect on RhoprDH-stimulated cation transport was observed when cGMP was co-applied. In 5-HT-stimulated MTs, only the Na⁺ transport rate was significantly reduced (unpaired t-test, P=0.014) when cGMP was co-applied. Consistent with the effect of application of AedaeCAPA-1 to MTs stimulated with the various diuretics, cGMP had no overall effect on the relative proportions of the two primary cations in the secreted fluid (Fig. 4F).

Transepithelial transport of ions and osmotically obliged water by insect MTs is regulated by at least five types of hormones derived from the nervous system (see Fig. 5). These include: (i) biogenic amines, such as 5-HT, (ii) CRF-related peptides, (iii) CDPs, or kinin-related peptides, (iv) cardioacceleratory peptides (CAPA) and (v) calcitonin-like (DH₃₁) peptides (Clark and Bradley, 1998; Schepel et al., 2010). These peptides all elicit diuretic activity in A. aegypti MTs, whereas only CAPA peptides have been shown to function additionally as an anti-diuretic in larval A. aegypti MTs. Specifically, CAPA peptides elicit a diuretic function in larval A. aegypti at high concentrations, while functioning as an anti-diuretic at low concentrations (Ionescu and Donini, 2012). Similarly, the diuretic activity of CAPA peptides has been shown to be conserved for a variety of Dipteran species, but is lacking in non-Dipteran insects (Pollock et al., 2004).

The present study examined the effect of AedaeCAPA-1 on fluid secretion and ion transport by A. aegypti MTs stimulated with various diuretic factors. Considering these diuretics target unique receptors, involving distinct signalling cascades and, in some cases, have dissimilar cellular targets (i.e. principal versus stellate cells), we predicted AedaeCAPA-1 would not have a ubiquitous inhibitory action against all diuretic factors. Given its high structural similarity to the endogenous calcitonin-like mosquito natriuretic peptide, DromeDH₃₁ was expected to have both diuretic and natriuretic activities, functioning to stimulate fluid secretion whilst favouring transepithelial Na⁺ transport (Coast et al., 2005). Previous studies have shown mosquito DH₃₁ to be the main hormone eliciting natriuresis in A. aegypti and A. gambiae (Coast et al., 2005). Here, we found that AedaeCAPA-1 inhibits fluid secretion by MTs of larval and adult stage mosquitoes stimulated by DromeDH₃₁ and, moreover, inhibits transport of both primary cations without influencing the natriuretic activity of DromeDH₃₁ on adult MTs. Anti-diuretic activity of AedaeCAPA-1 has been reported previously on larval A. aegypti MTs, with low doses of CAPA being effective against 5-HT-stimulated secretion (Ionescu and Donini, 2012). Similar results were observed in the current study, with AedaeCAPA-1 inhibiting fluid secretion stimulated by 5-HT in both larval and adult stage mosquitoes. In addition, the kaliuretic action of 5-HT measured in adults was not influenced, as transport of both cations was inhibited and the secreted fluid remained K⁺ rich compared with that for unstimulated MTs. Studies performed on the kissing bug, Rhodnius prolixus, established a dose-dependent inhibitory (i.e. anti-diuretic) action of native and structurally related CAPA peptides (Paluzzi and Orchard, 2006; Paluzzi et al., 2008). RhoprCAPA-α2 reduces the natriuresis stimulated by both endogenous diuretic hormones, namely 5-HT and the CRF-related RhoprDH, which revealed that natriuretic activity could be hindered without influencing RhoprDH-stimulated diuresis (Paluzzi et al., 2012). Thus, inhibitory activity of CAPA peptides in these distinct blood-feeding insects may be driven by two unique mechanisms as natriuretic activity of DromeDH₃₁ was unchanged.

Notably, the present study examined the effect of AedaeCAPA-1 against DH₃₁-stimulated secretion for the first time in insects. Based on previous studies, the DH₃₁-like MNP acts via its second messenger, cAMP, activating the V-type H⁺-ATPase and selectively driving an increase in Na⁺ secretion into the lumen (Coast et al., 2005). The additional Na⁺, as well as Cl⁻ as a counter-ion, is secreted with osmotically obliged water, increasing secretion ∼7-fold and the Na⁺:K⁺ concentration ratio ∼10-fold (Coast et al., 2005). Our data indicate lumen-directed secretion of Na⁺ ions is maintained in the presence of AedaeCAPA-1, albeit at a significantly lower transport rate. Similar to the actions of the DH₃₁-related peptide, 5-HT induces diuresis via the activation of adenylate cyclase to increase intracellular concentrations of cAMP (Cady and Hagedorn, 1999; Clark and Bradley, 1998). It is believed that cAMP promotes the assembly of the V-type H⁺-ATPase in the apical membrane to initiate proton and cation secretion into the MT lumen (Baumann and Walz, 2012; Rein et al., 2008).

Similar to the effects observed with AedaeCAPA-1, cGMP strongly inhibited fluid secretion and cation transport in both 5-HT- and DH₃₁-stimulated adult MTs (Fig. 5). In Tenebrio molitor, two structurally unrelated anti-diuretic peptides were identified (Tenmo-ADFa and Tenmo-ADFb) that are potent inhibitors effective in the femtomolar and picomolar range (Eigenheer et al., 2002, 2003), which is comparable to the femtomolar concentrations used in the present study. Both ADFa and ADFb exert their effects on MTs through cGMP, reducing cAMP to inhibit fluid secretion by T. molitor MTs (Eigenheer et al., 2002, 2003). Similarly, studies have shown that cGMP inhibits fluid secretion in R. prolixus and probably does so by reducing intracellular levels of cAMP (Quinlan et al., 1997), which is elevated after stimulation with 5-HT and the CRF-related peptidergic diuretic hormone (Gioino et al., 2014; Te Brugge et al., 2002). This was further supported when the addition of cAMP, at high concentrations, reversed the inhibitory effects of cGMP (Quinlan and O'Donnell, 1998). Similarly, previous research on A. aegypti revealed that cGMP is probably the second messenger, leading to the anti-diuretic action of CAPA peptides in larvae (Ionescu and Donini, 2012) as well as an exogenous factor, T. molitor ADFa (Massaro et al., 2004), with similar inhibitory actions on adult MTs.

In contrast to the anti-diuretic activity on fluid secretion stimulated by 5-HT and the DH₃₁-related peptide, diuresis stimulated by a CRF-related peptide (RhoprDH) was not affected by AedaeCAPA-1 treatment in either larval or adult MTs. In addition, cation concentrations and transport rates remained unchanged by AedaeCAPA-1 in adult MTs stimulated by RhoprDH. Similarly, application of cGMP had only a minor effect on secretion rate and did not influence cation transport in RhoprDH-stimulated MTs, suggesting cGMP may play a role in the anti-diuretic function of CAPA peptides. In mosquitoes, the DH₃₁-related peptide is natriuretic whereas the CRF-related diuretic peptide elicits non-selective transport of cations (Coast et al., 2005), and our observations are consistent with these findings. In insects, CRF-like peptides have been shown to initiate diuresis in MTs via the transcellular and paracellular pathways, suggesting that multiple receptors and second messenger systems may be involved. Specifically, low nanomolar concentrations of a CRF-related peptide were linked to the stimulation of the paracellular pathway only (Clark et al., 1998b), mediating this action through intracellular Ca²⁺ as a second messenger (Clark et al., 1998a). In contrast, high nanomolar concentrations of a CRF-related peptide were shown to influence both paracellular and transcellular transport, increasing intracellular Ca²⁺ and cAMP (Clark et al., 1998b). Thus, this suggests the mild inhibitory action of cGMP on RhoprDH-stimulated fluid secretion may be linked to signalling via cAMP as a result of the mid-nanomolar concentration of RhoprDH utilized in this study.

Insect kinin-related peptides, including the CDP utilized here, are known to activate a Cl⁻ conductance pathway independently of cAMP, utilizing inositol 1,4,5-trisphosphate (IP₃) as a second messenger to stimulate the release of intracellular Ca²⁺ (Hayes et al., 1994; O'Donnell et al., 1998; Yu and Beyenbach, 2002) following activation of the kinin receptor (Lu et al., 2011; Radford et al., 2002). Thus, there is probably no direct interaction between CAPA and CDP signalling pathways. Additionally, the CAPA receptors are localized to the principal cells in D. melanogaster (Terhzaz et al., 2012), whereas the kinin receptor has been immunolocalized to the stellate cells of the MTs, and functions to regulate anion permeability of the epithelium (Lu et al., 2011; Radford et al., 2002). Interestingly, our results found that application of cGMP resulted in an inhibition of CDP-stimulated fluid secretion and cation transport, similar to inhibitory effects on diuresis stimulated by 5-HT and DromeDH₃₁ (Fig. 5). In D. melanogaster, cGMP was shown to inhibit depolarization induced in kinin-stimulated MTs, suggesting an anti-diuretic effect (Ruka et al., 2013) as fluid secretion rates were not directly measured. Thus, in agreement with observations made in another dipteran insect, our results support the presence of an additional anti-diuretic hormone that signals through cGMP and reduces the activity of diuretic factors targeting stellate cells (Ruka et al., 2013).

Synthesized by neurosecretory cells in the central nervous system, CAPA peptides in D. melanogaster act on the principal cells in the MTs by activating L-type voltage-gated Ca²⁺ channels, increasing intracellular Ca²⁺ levels (Kean et al., 2002). The rise in Ca²⁺ levels activates nitric oxide synthase (NOS) to initiate production of nitric oxide (NO). Finally, NO activates a soluble guanylate cyclase, resulting in the production of cGMP, and consequent effects on secretion (Dow and Davies, 2003). In A. aegypti, it has been proposed that low levels of CAPA peptides lead to activation of protein kinase G, via elevated levels of cGMP (Ionescu and Donini, 2012), and resulting in its anti-diuretic activity in larval A. aegypti. When tested on D. melanogaster MTs, CAPA peptides stimulated fluid secretion in the 1 mmol l⁻¹ to 10 nmol l⁻¹ range (Pollock et al., 2004). However, CAPA peptides tested using lower concentrations that are anti-diuretic in mosquitoes do not activate calcium signalling in D. melanogaster MTs expressing an aequorin transgene (Davies et al., 2013), indicating species-specific activities of this neuropeptide family, which is reasonable considering their highly different diets and lifestyles.

Similar to effects seen in larval A. aegypti (Ionescu and Donini, 2012), our results suggest AedaeCAPA-1 increases cGMP levels in principal cells, as this second messenger mimicked the anti-diuretic actions in 5-HT- and DromeDH₃₁-stimulated MTs in adult females. Antagonistic effects between cGMP and cAMP, the latter being the second messenger of DH₃₁ and 5-HT, were illustrated previously in R. prolixus, whereby cGMP was proposed to activate a cAMP-specific phosphodiesterase, reducing levels of the second messenger activating diuresis (Quinlan and O'Donnell, 1998). Similar antagonistic effects of cAMP and cGMP are emerging in A. aegypti MTs, given the present and previous accounts of cGMP being an inhibitory second messenger (Ionescu and Donini, 2012; Massaro et al., 2004). Notably, however, both these second messengers have been shown to lead to stimulation of diuresis in D. melanogaster principal cells (Davies et al., 1995; Kean et al., 2002; Kerr et al., 2004). In stellate cells, cGMP signalling was shown to stimulate diuresis following activation of the mammalian atrial natriuretic peptide receptor expressed specifically in stellate cells (Kerr et al., 2004), while it was inferred to have an anti-diuretic effect on kinin-stimulated MTs (Ruka et al., 2013). Nonetheless, as the natriuretic activity of DromeDH₃₁ is unaffected in response to AedaeCAPA-1, it is possible that the inhibitory mechanism differs somewhat to that in R. prolixus, where diuretic hormone-induced natriuresis is attenuated by the endogenous CAPA peptide (Paluzzi et al., 2012). Taken together, the inhibition of fluid secretion by MTs stimulated by select diuretic factors (i.e. DH₃₁-related peptide and 5-HT) and the absence of modulation of the relative proportions of the primary cations transported highlights the importance of AedaeCAPA-1 solely as an anti-diuretic hormone in A. aegypti. Thus, natriuretic and kaliuretic activity, elicited by the DH₃₁-related peptide and 5-HT, respectively, is maintained while the rate of diuresis is slowed, which could have implications for downstream processes such as reabsorption in the hindgut.

Receptors for some of these regulators of the MTs have been identified in A. aegypti, including a kinin receptor expressed in stellate cells (Lu et al., 2011; Pietrantonio et al., 2005) and a DH₃₁-related peptide receptor that shows a peculiar distal to proximal gradient of expression in principal cells of the MTs (Kwon and Pietrantonio, 2013), as well as two CRF-related peptide receptors, one of which is highly enriched in the renal MTs (Jagge and Pietrantonio, 2008). Other receptors, such as the endogenous 5-HT receptor expressed within the A. aegypti MTs, remain elusive, as a 5-HT7 receptor isoform transcript was localized to the tracheolar cells associated with the MTs but was not localized to MT epithelium, suggesting that another receptor variant must be present (Pietrantonio et al., 2001). Similarly, expression of the CAPA receptor in A. aegypti has not been identified (Pollock et al., 2004), although orthologues have been described in other Dipteran species and found to be enriched in the MTs (Iversen et al., 2002; Olsen et al., 2007; Park et al., 2002), with localization to the principal cells (Terhzaz et al., 2012). Thus, it is apparent that more research is required to fully understand the complex interaction and cross-talk between all the hormonal regulators of A. aegypti MTs as well as those found in other insects. Further understanding of the role of each specific hormone family, including both diuretic and anti-diuretic factors, will help resolve this complex regulatory network. Given the central importance of the MTs in insect biology, these insights will be useful considering the need for novel strategies or compounds which more specifically and efficiently reduce the burden of disease vector species as well as insect pests of agriculture.

The authors would like to thank Lesia Szyca for her contributions towards scientific illustrations. The authors are also grateful to Professors Ian Orchard (University of Toronto Mississauga) and Michael J. O'Donnell (McMaster University) for providing aliquots of RhoprDH and DromeDH₃₁ used in this study.