Conservation of capa peptide-induced nitric oxide signalling in Diptera

The insect renal system is composed of Malpighian tubules that vary in number and structure in different species. In all insects, however, its osmoregulatory and homeostatic functions are thought to be critical to life. Neuropeptide control of secretion by Malpighian tubules has been studied in many insect species including Drosophila melanogaster(Dow and Davies, 2003), Musca domestica (Coast,2001a), Rhodnius prolixus(O'Donnell and Spring, 2000; Te Brugge et al., 2002), Periplaneta americana (Kay et al., 1992), Tenebrio molitor(Wiehart et al., 2002), Locusta migratoria, Schistocerca gregaria(Schoofs et al., 1997), Acheta domesticus (Spring and Clark, 1990) and Formica polyctena (Laenen et al., 1999, 2001). Diuresis in the disease vector species has been most studied in Aedes aegypti(Beyenbach, 2003; Pullikuth et al., 2003);however, less is known about tubule function in the malaria mosquito, Anopheles gambiae, or the tsetse fly, Glossina morsitans.

Drosophila melanogaster Malpighian tubules are now accepted as a genetic model of transporting epithelia(Dow and Davies, 2003). In the development of this renal model, different techniques have been developed to assess tubule function: fluid transport rates(Dow et al., 1994a),electrophysiological responses (Davies et al., 1995), ion transport(Dow, 1999) and calcium signalling using aequorin transgenes(Rosay et al., 1997). This battery of physiological assays, in combination with the powerful genetic tools associated with Drosophila, has allowed rapid, organotypic analysis of the cell-specific control of tubule function(Dow and Davies, 2003). Given the conserved and critical role of the tubule in insect life, findings from the Drosophila tubule may usefully be applied to those insect species with less developed genomic resources but greater economic or medical significance (Dow and Davies,2003). In particular, findings from Drosophila might be useful in studies of other Diptera; for example, Aedes, Anopheles and Glossina.

Diuresis in Drosophila tubules has been shown to be directly stimulated by exogenous guanosine 3′, 5′-cyclic monophosphate(cGMP), which enters tubule cells via a cyclic nucleotide transporter(Riegel et al., 1998), and by nitric oxide (NO; Dow et al.,1994b). NO/cGMP signalling is compartmentalised to principal cells in the main, fluid-secreting segment of tubules, containing the electrogenic vacuolar H⁺-ATPase (V-ATPase) pump(Dow, 1999), which energises fluid transport. Furthermore, electrophysiological studies suggest that cGMP signalling modulates V-ATPase activity(Davies et al., 1995),suggesting that cGMP signalling may regulate ion transport in tubules.

NO/cGMP signalling is also activated by a nitridergic family of neuropeptides, capa, which comprise the only known insect NO/cGMP-mobilising peptides. Capa-1 and capa-2 are encoded by the capa gene in Drosophila (Kean et al.,2002). Both capa-1 and capa-2, as well as the closely related Manduca sexta CAP_2b, induce diuresis and stimulate NO/cGMP signalling and intracellular calcium increases in Drosophila tubule principal cells (Davies et al., 1995, 1997; Kean et al., 2002; Rosay et al., 1997). To date,secretion by Aedes and Anopheles tubules has not been shown to be stimulated by NO or cGMP, although the A. stephensi gene encoding nitric oxide synthase (NOS) has been cloned(Luckhart et al., 1998); also, A. gambiae tubules have been shown to express NOS transcripts(Dimopoulos et al., 1998). All known insect NOS-encoding genes are very similar(Davies, 2000), resulting in virtually identical sequences for NOS protein; as such, conservation of function at the physiological level may be anticipated. Recently, data mining of the A. gambiae genome has identified capa peptides in this species(Riehle et al., 2002). Although capa-like signalling beyond the Diptera can be inferred from the existence of the cardinal CAP_2b in the Lepidoptera, other reports have suggested that cGMP is antidiuretic in other insect orders, for example Hemiptera (Quinlan et al.,1997) and Coleoptera (Eigenheer et al., 2002, 2003), or that CAP_2b is without effect in Orthoptera(Coast, 2001b). It is thus of great interest to assess the phylogenetic scope of the highly unusual autocrine capa/NOS/NO/cGMP signalling model beyond Drosophila. Furthermore, the application of knowledge of tubule function in D. melanogaster to those of insect disease vectors will advance understanding of tubule physiology in the context of specific cell types and tubule regions in these animals.

Our results show that, whereas all insect tubules so far studied contain NOS (and thus have the machinery to respond to capa), only the dipteran species studied show functional responses. The scope of action of this peptide may thus be general within, but limited to, certain endopterygote orders.

Wild-type Oregon R flies (OrR) flies were maintained on standard Drosophila diet over a 12 h:12 h photoperiod at 55% humidity at 22°C.

These were obtained as non-infective, sugar-water-fed adults from a colony maintained by Professor E. Devaney, University of Glasgow. Female animals were used upon receipt.

Non-infective, sugar-water-fed, adults were provided as a kind gift of Dr L. Ranford-Cartwright, University of Glasgow. Female animals were used upon receipt. If mosquitoes were not used immediately, they were maintained over a 12 h:12 h photoperiod at 55% humidity at 22°C, on 5% sucrose (v/v)solution ad libitum for a maximum of 3 days before use in experiments.

Non-infective adults were provided by Dr S. Welburn, University of Edinburgh, and by Professor D. Barry, University of Glasgow. Animals were used immediately upon receipt.

These were obtained from Bugs Direct (Well Cottages, Devon, UK) and either used immediately or maintained on grass over a 12 h:12 h photoperiod at 55%humidity at 22°C for a maximum of 3–4 days. All insects were cold-anaesthetised and decapitated prior to dissection to isolate intact tubules.

Capa neuropeptides used in this study are shown in Table 1. Of the Drosophila capas, capa-1 (GANMGLYAFPRVamide) was used here, because of its identical mode of action to, but slightly greater potency than, capa-2(Kean et al., 2002). Both A. gambiae capa peptides were synthesised: QGLVPFPRVamide(AngCAPA-QGL) and GPTVGLFAFPRVamide (AngCAPA-GPT). All peptides were synthesised by Invitrogen Corp. (Renfrew, UK). A. gambiae capa peptides were identified by data mining the A. gambiae genome. While this study was in progress, identical sequences for Anopheles capa peptides were published elsewhere(Riehle et al., 2002).

Analysis of capa receptor expression in dipteran tubules was carried out by RT-PCR according to standard protocols(Dow et al., 1994b) from cDNA templates prepared from Drosophila melanogaster, Anopheles stephensiand Anopheles gambiae tubules. The capa receptor has been identified in D. melanogaster; searching the A. gambiae genome reveals a possible candidate for the Anopheles capa receptor.

For each cDNA preparation, 20 tubules were dissected, poly(A)⁺RNA extracted (Dynal mRNA direct kit; Dynal Biotech UK, Wirral, UK) and reverse transcribed with Superscript Plus (Gibco BRL, Invitrogen Ltd, Paisley,Renfrewshire, UK). 1 μl of the reverse transcription reaction was used as a template for PCR, containing the following gene-specific primer pairs: Drosophila capaR Forward,5′-GCGGCCGCCTAAAATGAATTCATCGACCG-3′; Drosophila capaRReverse, 5′-GTCTAGAGCCTCGTGCTTAAATACAAG-3′; putative A. gambiae capaR Forward, 5′-TGTTGACCGTGTTGAAGTGTTGC-3′;putative A. gambiae capaR Reverse,5′-CTGTTCTTTGCCTTTCCAATGCTC-3′. Additionally, to control against genomic contamination in cDNA preps, primers that had been designed around intron/exon boundaries of the capa receptor gene were used. Use of such primers verified the cDNA quality used in PCR reactions. Further controls were performed that included nonreverse transcribed template (i.e. no cDNA).

PCR cycle conditions for reactions with Drosophila cDNA template were as follows: 93°C (3 min), 36 cycles of [93°C (30 s), 54.3°C(30 s), 72°C (1 min)] and 72°C (1 min).Conditions were similar for A. gambiae and A. stephensi cDNA templates except that the annealing temperature used was 59°C. PCR products obtained from such RT-PCR experiments were cloned using the Invitrogen Topoisomerase (TOPO TA Cloning) system (Renfrew, Scotland). Cloned plasmids were purified using Qiagen kits (Crawley, UK) and sequenced to confirm their identity.

Very few A. gambiae were available for study; thus, for all following experiments, A. stephensi was used.

Immunocytochemistry to fixed, intact tubules from all insect species was performed using a universal anti-NOS (anti-uNOS) antibody according to previously published protocols (MacPherson et al., 2001), as described in the legend to Fig. 2. The anti-uNOS antibody is an affinity-purified rabbit universal anti-NOS antibody, used at 1:100 dilution and specified for Drosophila use (anti-uNOS; PA1-039;Affinity BioReagents, via Cambridge BioScience, Cambridge, UK). This antibody is directed against an epitope that is closely conserved in mammalian, insect and even crustacean NOS peptides(Table 2).

This antibody has been used previously in Drosophila(Broderick et al., 2003; Dow and Davies, 2001; Gibbs and Truman, 1998). Specificity of this antibody has been demonstrated by immunoblotting DNOS(Drosophila nitric oxide synthase) protein expression (∼150 kDa protein) in tubules from dNOS transgenic lines; overexpression of DNOS results in increased protein by western analysis, which correlates with increased NOS enzyme activity by direct assays. Also, this antibody has successfully been used for immunocytochemistry in both wild-type and dNOS transgenic Drosophila tubules and in eye tissue(Broderick et al., 2003; Dow and Davies, 2001; Gibbs and Truman, 1998). Staining was visualised using a fluorescein-labelled goat anti-rabbit antibody(Diagnostics Scotland, Edinburgh, UK), used at 1:250 dilution. In order to visualise principal cell nuclei, tubules were counterstained in 1 μg ml^-1 4′, 6′-diamidino-2-phenylindole hydrochloride(DAPI; Sigma Aldrich, Gillingham, UK) for 2.5 min(Broderick et al., 2004). Stained tubules were mounted in VectaShield (Vector Labs, Peterborough, UK). Staining in whole-mount tubules was detected by immunofluorescence using an Axiocam imaging system (Zeiss, Welwyn Garden City, UK).

An assay for NOS-associated NADPH diaphorase activity in Drosophila tubule extracts(Broderick et al., 2003; Kean et al., 2002) was modified for analysis in 96-well plates. Intact tubules were dissected from animals (lines as described in legend to Fig. 3). For each species,either six tubules (Drosophila), five tubules (Aedes and Anopheles) or two tubules (Glossina and Schistocerca) were used. For each sample, tubules were placed in 93μl of 50 mmol l^-1 Tris HCl, pH 7.4, 1% Triton X100 and 5 μl of 10 mmol l^-1 XTT[2,3-bis(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide],sodium salt in 96-well plates. Samples were incubated at 25°C for 20 min. For peptide stimulations, either 1 μl of each peptide (D. melanogaster capa-1, AngCAPA-QGL or AngCAPA-GPT to a final concentration of 10^-7 mol l^-1) or 1 μl phosphate-buffered saline (PBS; control) were added for a further 10 min. To each sample, either 1 μl of 100 mmol l^-1 NADPH or 1 μl PBS(for no-substrate controls) was added, and samples were incubated at 25°C for 7 min. For each species, replicate samples were prepared under the following conditions: tubules only, tubules + NADPH, tubules + peptide,tubules + peptide + NADPH. Samples were homogenised and colorimetric analysis performed for all samples by spectrophotometry at 450 nm (Berthold Mithras plate reader; Berthold Technologies, Redbourn, UK). Blanks were prepared from incubation buffer. Blank samples, `tubules only' samples and `tubules +peptide' samples gave very similar readings. The overall mean of readings for blanks, `tubules only' samples and `tubules + peptide' samples within each experiment was subtracted from results of `tubules + NADPH' and `tubules +peptide + NADPH', respectively. In order to normalise data for all species,results were expressed as % increase over unstimulated tubules (± s.e.m.; N=4–6); i.e. (corrected values for `tubules+ peptide + NADPH' minus mean of corrected values for `tubules + NADPH'/mean of corrected values for `tubules + NADPH')×100%.

To further verify that stimulated NADPH diaphorase activity was due to NOS activation, we utilised the NOS inhibitor N^G-nitro-l-arginine-methyl ester(l-NAME; Calbiochem, Beeston, UK) in the assays above. For these samples, l-NAME was added to samples prepared as described above,at a concentration of 2 μmol l^-1 for 20 min prior to stimulation with capa peptides. Results were expressed as % change (± s.e.m.; N=4) of both peptide-stimulated samples and capa +l-NAME samples compared with controls (samples without either capa or l-NAME).

Cyclic GMP levels were measured in pooled samples of tubules dissected from insects, as detailed in Fig. 4legend, by radioimmunoassay (Amersham Biotrak Amerlex M; Amersham Biosciences,Chalfont St Giles, UK), as previously described(Dow et al., 1994b). Tubules were pre-incubated with the cGMP-specific phosphodiesterase inhibitor Zaprinast (Calbiochem) at 10^-5 mol l^-1 for 10 min. For peptide stimulations, either 1 μl of each peptide (D. melanogastercapa-1, AngCAPA-QGL or AngCAPA-GPT to a final concentration of 10^-7 mol l^-1) or 1 μl PBS (control) were added for a further 10 min. Incubations were terminated with ice-cold ethanol and homogenised. The ethanol was evaporated and samples were resuspended in 0.05 mol l^-1 sodium acetate buffer (Amersham Biosciences) and processed for cGMP content according to manufacturer's protocol. Data were normalised across insect species by expressing results as fmol cGMP μg^-1protein (± s.e.m.; N=4–6). Protein concentrations were determined by Bradford assay.

Tubule secretion was measured according to standard procedures. Intact Malpighian tubules were isolated into 9 μl drops of a freshly prepared mixture of Schneider's medium (Gibco BRL, Invitrogen Ltd) and Drosophila saline (1:1, v/v) under liquid paraffin, and fluid secretion rates measured in tubules as detailed elsewhere(Dow et al., 1994a). Briefly,one end of the tubule was wrapped around a metal pin and the rest of the tubule bathed on the saline drop. A nick was made near the ureter, a drop of secreted fluid collected every 10 min, and the diameter measured using an eyepiece micrometer. The volume of each droplet was calculated as 4/3πr³, where r is the radius of the droplet,and secretion rates plotted against time. Secretion was measured under basal conditions to establish a steady rate of secretion prior to stimulation with peptide(s).

For other insect species, procedures were as for Drosophila, with appropriate modifications to accommodate the widely differing sizes of the tubules. However, apart from Drosophila, all insect tubules were left in the saline bubble in the paraffin dish for at least 15 min after dissection, then wound around the pin and left for another 15 min. Tubules were then nicked to allow bubbles to form; experimental readings commenced 10 min after this. Basal rates were measured for 30 min prior to stimulation with peptides (D. melanogaster capa-1, AngCAPA-QGL or AngCAPA-GPT).

Although the diuretic nature of capa signalling has been well established in Drosophila, its phylogenetic scope has not been explored. Table 1 illustrates the sequences of the capa members identified from the lepidopteran Manduca sexta (Huesmann et al.,1995) and the dipterans D. melanogaster(Kean et al., 2002) and A. gambiae (present study; Riehle et al., 2002). All the peptides share the kinin-like Lx(A/P)FPRVamide motif (Kean et al., 2002),which may confer specificity of action. At present, there are no close matches for capa-like peptides in sequence data for non-insect organisms.

A gene encoding a Drosophila capa receptor has been identified and functionally characterised (Iversen et al., 2002; Park et al.,2002). There is a single clear homologue for the CAPA-R in the published Anopheles genome (BlastP; P=5×10^-90). This is encoded by a gene with GenBank no. XP_312952. At present, there are no compelling matches in available Aedes or other insect sequences; the nearest match in the Aedes genome has been annotated as a 5-HT7 receptor. Nonetheless, if a CAPA-R homologue were found to be expressed in the tubule of another insect, it would strengthen the case for functional conservation of capa signalling. By RT-PCR with intron-spanning primers, it was possible to show that the Anopheles homologue was indeed expressed in the Anopheles tubule(Fig. 1), although the non-degenerate Anopheles primers did not identify a match in the Aedes tubule.

Capa acts on Drosophila tubules to activate NOS viaintracellular calcium: any nitridergic action of capa in other insects would thus require the presence of NOS in tubules. Accordingly, the distribution of NOS in the Malpighian tubules of other species was investigated by immunocytochemistry for NOS (Fig. 2). A universal anti-NOS antibody was utilized for these experiments; this antibody has been previously shown to be specific for Drosophila NOS both in the eye(Gibbs and Truman, 1998) and in tubules by immunocytochemistry(Broderick et al., 2003; Dow and Davies, 2001) and by western blotting (Broderick et al.,2003). Previous work has shown that NOS is expressed in only principal cells of D. melanogaster tubules(Broderick et al., 2003; Davies, 2000). Here, we show clear NOS immunoreactivity only in the main, fluid-transporting segment of the tubule (Fig. 2Aii, region marked `m').

In mosquito tubules, NOS immunoreactivity is observed only in the cytoplasm of principal cells (examples of unstained stellate cells marked by arrows in Fig. 2Bii,iii,Cii,iii). In both A. aegypti and A. stephensi, counterstaining of cell nuclei with DAPI shows the smaller nuclei of the stellate cells (arrows in Fig. 2Biv,Civ), as in Drosophila (Broderick et al.,2004), of which the cytoplasm remains unstained. In contrast to Drosophila, however, the entire length of the tubule is stained using anti-NOS antibody in both mosquito species.

Fig. 2Dii,iii shows NOS immunoreactivity in G. morsitans tubules. Interestingly, in this dipteran species, staining with the anti-NOS antibody appears in all cells. Close inspection of DAPI-stained G. morsitans tubules(Fig. 2Div) does not reveal cell nuclei of different sizes, as in D. melanogaster(Broderick et al., 2004), A. gambiae or A. stephensi(Fig. 2Biv,Civ); it appears that, unlike other Diptera, Glossina does not have obvious stellate cells. Furthermore, staining is also observed throughout the tubule, rather than merely in the main segment.

In the orthopteran out-group, S. gregaria, high background staining is observed in the control tubules(Fig. 2Ei). However, increased staining is observed with the anti-NOS antibody throughout the tubule(Fig. 2Eii) at the membrane and in the cytoplasm, suggesting that expression of NOS occurs in these tubules. This is consistent with a previous report of NAPDH diaphorase activity in orthopteran tubules (Locusta migratoria; M. Elphick, personal communication).

It is thus clear that all the insects studied have at least some of the machinery (NOS) to produce a nitridergic response to capa.

NADPH diaphorase staining is an obligate correlate of NOS activity, both in vertebrates and in insects (Elphick,1997; Davies,2000). We have previously adapted this assay for measurements in vitro (Kean et al.,2002), allowing quantification of NOS-associated NADPH diaphorase activity, which accurately reflects NOS activity(Broderick et al., 2003).

Results in Fig. 3A show that tubules stimulated with all capa peptides tested (i.e. D. melanogaster capa-1, AngCAPA-QGL and AngCAPA-GPT)increase NADPH diaphorase activity across the Diptera. Interestingly, capa-1 is at least as effective as, if not better than, the A. gambiaepeptides in raising NOS activity, at least at the concentration tested, which was based on the maximum response of D. melanogaster tubules to capa-1 as shown in previous work (Kean et al., 2002). By contrast, although S. gregaria tubules both contain NOS immunoreactivity and display similar resting levels of NADPH diaphorase activity to dipteran tubules (results not shown), none of the capa peptides tested elevated NADPH diaphorase activity in this orthopteran species. In each case, l-NAME inhibited the increase in NADPH diaphorase activity to control (unstimulated) levels, confirming the association between increased NADPH diaphorase and NOS activation in these species (Elphick, 1997).

Previous work has shown that NO release, induced by M. sexta and D. melanogaster capa peptides, increases cGMP content in Drosophila tubules (Davies et al.,1995; Kean et al.,2002). Radioimmunoassay for cGMP content showed that all three capa peptides stimulated an increase in cGMP content in (only) dipteran tubules (Fig. 4). G. morsitans tubules were the most responsive to both AngCAPA-QGL and AngCAPA-GPT. Also, while capa-1 was the most effective at increasing tubule NOS activity (Fig. 3), this was not the case for the cGMP assay(Fig. 4). Finally, capa peptides do not increase cGMP in S. gregaria tubules [data in fmol cGMP μg^-1 protein (± s.e.m.; N=4):unstimulated tubules: 0.097±0.003; capa-1 stimulated tubules,0.093±0.008; AngCAPA-QGL, 0.096±0.008; AngCAPA-GPT, 0.106±0.003].

Previous work has shown that fluid secretion is potently stimulated by capa-1 in D. melanogaster (Kean et al., 2002); Fig. 5 shows such stimulation of fluid transport by D. melanogaster tubules with capa-1 at an EC₅₀ value of between 10^-7 and 10^-8 mol l^-1. Capa-1 also stimulates fluid transport by A. aegypti, A. stephensi and G. morsitanstubules. However, still higher rates of secretion occur at very high concentrations of peptide, between 10^-3 (A. stephensi) and 10^-4 mol l^-1 (A. aegypti). Furthermore, G. morsitans tubules are only stimulated to 50% over basal levels at all concentrations of capa-1. In S. gregaria, capa-1 has either no significant effect on secretion or is inhibitory (10^-5,10^-6, 10^-8 mol l^-1). Similarly, capa-1 does not stimulate fluid secretion by tubules from the dictyopteran roach Periplaneta americana (data not shown). Thus, of the species sampled to date, the stimulatory effects of capa-1 on tubules are confined to the Diptera.

Figs 6, 7 show the first demonstration of the physiological effects of A. gambiae capa peptides on tubule fluid secretion in both mosquito and other Diptera. Drosophila tubule secretion is stimulated in a dose-dependent manner in response to AngCAPA-QGL, with an apparent EC₅₀ of 10^-5 mol l^-1 (Fig. 6). All other dipteran tubules tested also respond to AngCAPA-QGL and are more sensitive to the peptide compared with Drosophila tubules,especially at low concentrations [10^-6, 10^-7,10^-8 mol l^-1 (G. morsitans)]. Apart from at 10^-3 mol l^-1, tubules from A. aegypti and A. stephensi show similar responses at all concentrations tested. Also, G. morsitans tubules show a similar pattern of response to both mosquito species. At 10^-7 mol l^-1AngCAPA-QGL,stimulation of secretion rates in A. aegypti, A. stephensi and G. morsitans is identical. By contrast, tubule secretion rates in S. gregaria tubules are not significantly altered at any concentration of AngCAPA-QGL.

Similarly to capa-1 and AngCAPA-QGL, Drosophila tubules respond to all concentrations of AngCAPA-GPT tested, although are most responsive at concentrations of ≥10^-5 mol l^-1,with the maximal response occurring at 10^-3 mol l^-1(Fig. 7). However, responses of all other dipteran tubules tested are similar at 10^-5 and 10^-6 mol l^-1. Maximal response of A. aegyptitubules occurs at 10^-5 mol l^-1, of A. stephensitubules at 10^-6 mol l^-1 and of G. morsitanstubules at 10^-7 mol l^-1AngCAPA-GPT. Interestingly, no stimulation of secretion was observed with A. aegypti tubules at 10^-7 mol l^-1. Note also that the secretion response of these tubules to 10^-6 mol l^-1AngCAPA-GPT is very low; these results are reproducible(N>30). As with the other capa peptides, no significant response is obtained from S. gregaria tubules.

This paper extends our detailed understanding of the unique nitridergic capa pathway from the genetic model Drosophila to four further species of insect in two widely spaced orders, making it one of the most wide-ranging experimental surveys of insect endocrinology to date.

In this work, we show that NOS immunoreactivity is observed in principal cells throughout A. aegypti, A. stephensi and G. morsitanstubules. By contrast, immunoreactivity for NOS is observed in all tubule cells in S. gregaria. For these experiments, an anti-NOS antibody to an epitope contained in all insect NOS sequences known to date was used(Table 2). Although we cannot assert that this antibody is specific to NOS alone, it faithfully reports increased NOS expression via an inducible NOS transgene in Drosophila tubules (Broderick et al., 2003) and is consistent with other measures of NOS activity reported here.

We also demonstrate that D. melanogaster and A. gambiaecapa peptides all stimulate NOS activity, increase cGMP production and elicit an increase in fluid secretion rates in several dipteran species. Thus, this suggests that not only are conserved features of the capa peptide sequences functionally important but that conservation of the sequence and function of the capa receptors must also exist within the Diptera. In particular, we have identified a likely Anopheles homologue of the DrosophilaCAPA-R, which is abundantly expressed in Anopheles tubule. Importantly, none of the capa peptides tested activate NO/cGMP signalling or elevate fluid secretion in S. gregaria. Indeed, capa-1 may be anti-diuretic at some concentrations (Fig. 5), although this is not linked to an increase in cGMP content(Fig. 4). The data are supported by work on L. migratoria, which shows that M. sexta CAP_2b does not affect fluid secretion by these tubules(Coast, 2001b; see Wegener et al., 2002). We have thus demonstrated, for the first time, physiological roles for A. gambiae capa peptides and that capa-stimulated fluid secretion is confined to a range of dipteran insects. We have also measured neuropeptide-stimulated secretion rates in G. morsitans tubules for the first time. Measurement of fluid secretion in the tsetse fly was first published nearly 30 years ago (Gee, 1976a,b). More recent work has re-visited cAMP-stimulated fluid secretion by G. morsitans tubules (Isaacson and Nicolson, 1994). However, our recent development of Glossina tubule physiology will allow study of a critical tissue in a disease vector. The demonstration of conservation of capa signalling in medically important insect vectors suggests new possibilities for novel insecticide targets for pest control.

Importantly, we extend the phylogenetic scope of diuretic cGMP signalling beyond Drosophila. It is apparent that cGMP can act as an anti-diuretic signal in some insects. For example, in T. molitor, two anti-diuretic hormones that act via cGMP have been isolated(Eigenheer et al., 2002, 2003). However, the existence of anti-diuretic, cGMP-mobilising hormones in some insects need not point to a universal mode of action by cGMP in insect tubules. Rather, this suggests a critical distinction in the use of cGMP by different animals and, more than that, a relevant role of cell or tissue concentration of cGMP in physiology.

Locust tubules contain NOS but do not respond to capa. This result does not, however, rule out nitridergic signalling in nondipteran tubules. NOS-encoding genes have been characterized from multiple orders of insect(Davies, 2000), and all contain well-conserved calmodulin-binding domains, implying that, like Drosophila NOS, they are calcium/calmodulin regulated. It is thus probable that any neuropeptide that elevates calcium in Schistocerca(or indeed any insect) tubule will activate NOS to generate NO. The capa peptides perform such a role in Diptera, but our evidence suggests that they do not in Orthoptera. Consistent with this argument, calcium has been shown to be important in L. migratoria tubule stimulation by a partially purified hormone, and cGMP has been shown to be diuretic(Morgan and Mordue, 1985). Of course, the generation of NO in a tissue does not imply that it will be sensed by soluble guanylate cyclase in the same tissue. In the future, it will be of interest to follow the phylogenetic distribution of NO-sensing in insect tubules, in particular those from nondipteran species, including orthopteran insects.

This work was funded by a Biotechnology and Biological Sciences Research Council (BBSRC, UK) GAIN initiative grant to J.A.T.D., S.-A.D. and I.M.M. J.M. is funded by a BBSRC Committee Studentship.