Functional characterisation of the Anopheles leucokinins and their cognate G-protein coupled receptor

The Malpighian (renal) tubules of insects are mission-critical organs(Beyenbach, 2003b; Dow and Davies, 2001) and as such are an excellent target tissue for the development of novel insecticides against agriculturally and biomedically relevant pest species. Not only do they regulate water and ion homeostasis, but they have a major role in detoxification. Thus, modulation of fluid transport by Malpighian tubules is an effective method for novel pest control strategies.

Insect leucokinins elicit potent diuretic effects on the Malpighian tubules of various insect species (Gade,2004) and have also been implicated in a number of other physiological functions, such as the contraction of the hindgut(Holman et al., 1999; Howarth et al., 2002; Veenstra et al., 1997) –hence the term `myokinin'. In addition, recent studies have suggested the involvement of leucokinins in dietary regulation and energy mobilisation(Nachman et al., 2002; Seinsche et al., 2000). As such, the insect leucokinins have attracted a great deal of interest as lead molecules for novel pesticides, including the development of peptidase-resistant analogues of this family of peptides(Teal et al., 1999). Given the diverse roles of insect leucokinins, elucidation of the mode of action of these peptides via their cognate G protein-coupled receptors (GPCRs) is of importance. Furthermore, as leucokinins have only been found in invertebrates,it is likely that careful design of leucokinin antagonist or agonist analogues will avoid interactions with mammalian species.

While identification of leucokinins and their cognate receptors has been successfully undertaken in some insects(Holman et al., 1984; Holman et al., 1999; Veenstra et al., 1997),including the genetically tractable Dipteran, Drosophila melanogaster(Radford et al., 2002; Terhzaz et al., 1999), less progress has been made in studies of leucokinin signalling in biomedically relevant insects.

The malaria mosquito, Anopheles gambiae, is one such insect. Initial attempts to curb the spread of malaria involved the use of larvicides and insecticides, against the mosquito vectors, and also the use of chloroquine, which halts the progression of the disease in patients. Despite these efforts, resistance has evolved in both the mosquitoes and in the malaria parasites. Thus malaria, as other vector-borne diseases, is now classed as a re-emerging disease (Gubler,1998). However, the sequencing of the A. gambiae genome(Holt et al., 2002) provides a fresh direction for anti-malarial research. The action of leucokinins on the Malpighian tubules of the yellow fever mosquito, Aedes aegypti, has already been studied in great detail(Beyenbach, 2003b; Veenstra et al., 1997). Thus in Anopheles, it is likely that leucokinins will also play an important role in the regulation of water and ion homeostasis. An initial survey of the completed Anopheles genome identified a single leucokinin-like gene (Riehle et al.,2002). Approximately 37 neurohormone receptor-like-encoding sequences were also identified in a survey of the GPCR repertoire of the Anopheles genome (Hill et al.,2002). It is likely that one of these will represent a receptor for Anopheles leucokinins.

In this study we identified three leucokinin peptides (Anophelesleucokinins I-III) from the A. gambiae genome and demonstrated effects on calcium signalling via a putative cognate GPCR-coupled receptor cloned from its close relative, A. stephensi. All Anophelesleucokinins increase intracellular calcium in a dose-dependent manner. Furthermore, the Anopheles leucokinin receptor is responsive to D. melanogaster Drosophila leucokinin, while the D. melanogaster leukokinin-like receptor (LKR) is only sensitive to Anopheles leucokinins I and II.

Identification of Anopheles gambiae Miles leucokinins: TBLASTN analysis was carried out on the completed A. gambiae genome using the predicted protein product of the Aedes aegypti L. preproleucokinin transcript (Veenstra et al.,1997) (GenBank accession no. AAC47656). The BLOSUM62 matrix(default settings) was used for all BLAST analysis. A 333 bp sequence was identified, which potentially encoded a protein with homology to the Aedes preproleucokinin (E-value 2×10^–29). This region plus the surrounding 20 kb of genomic sequence either side was then analysed with the Softberry FgenesH gene prediction program(www.softberry.com).

Identification of A. gambiae leucokinin receptor: TBLASTN analysis was carried out on the completed Anopheles gambiae genome using the predicted protein product of the Drosophila leucokinin receptor,CG10626 (Radford et al.,2002). The protein sequences of the Lymnaea stagnalis L.(Cox et al., 1997) and Boophilus microplus Canestrini(Holmes et al., 2003)leucokinin-like receptors were also used to confirm the sequence match. The BLOSUM62 matrix (default settings) was used for all BLAST analysis as above. This identified a sequence within the Anopheles genome that encoded a putative protein of 377 amino acids (GenBank accession no. agCP10499, E-value 2×10^–59). No other good sequence matches were identified in the completed genome sequence. However, this sequence only represented a portion of a presumed GCPR, encompassing only the strictly conserved TM domains. Efforts were made to identify a full-length transcript by the use of the FgenesH gene prediction program, although subsequent attempts to amplify the putative open reading frame (ORF) region by polymerase chain reaction (PCR) from Anopheles cDNA failed. Therefore, a RACE approach was undertaken in order to identify the correct cDNA sequence.

A. stephensi Liston and its close relative A. gambiae are malaria-carrying anopheline mosquitoes. For reasons of availability, A. stephensi was used as a source of cDNA in this study. Non-infective,sugar-water-fed adults were a kind gift from Dr L. Ranford-Cartwright,University of Glasgow, UK. Female animals were used upon receipt. If mosquitoes were not used immediately, they were maintained over a 12 h:12 h L:D 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.

For cDNA preparations, total RNA was extracted (Sigma Tri-reagent,Gillingham, Dorset, UK) from whole A. stephensi and reverse transcribed with Superscript II (Invitrogen). 1 μl of the reverse transcription reaction was used as a template for PCR containing the gene-specific primer pairs given below. Additionally, to control against genomic contamination in cDNA preps, primers were used that had been designed around intron/exon boundaries of the predicted A. gambiae leucokinin receptor gene. Use of such primers verified the quality of the cDNA used in PCR reactions. Further controls were performed, which included non-reverse transcribed template (i.e. no cDNA). The primers used were:GGAATCTGCCCGAGTTTATGTG and GTTCTTCAGCATCGTAATGTCGC. PCR cycle conditions for reactions were as follows: 93°C 3 min; 36 cycles of (93°C 30 s,59°C 30 s, 72°C 1 min); 72°C (1 min). PCR products obtained from such RT-PCR experiments were cloned into pCRII-TOPO™ using the Invitrogen Topoisomerase (TOPO TA Cloning) system. Cloned plasmids were purified using Qiagen kits and sequenced to confirm their identity.

Poly(A)⁺ RNA was purified from whole fly total RNA using the magnetic Dynabeads^® mRNA purification kit(Dynal^® Bromborough, UK) according to the manufacturers'instructions. The RACE procedure was carried out using the SMART™ RACE cDNA Amplification kit (Clontech, Oxford, UK). This kit provides a method for performing both 5′- and 3′-RACE. 5′- and 3′-RACE-ready cDNAs are generated as separate cDNA samples, using 1 μg poly(A)⁺ mRNA as starting material for each of the 5′- and 3′-RACE-ready cDNAs.

SMART™ RACE PCR reactions were carried out according to the manufacturers' instructions using Advantage 2 Polymerase Mix (Clontech). Both 5′- and 3′-RACE reactions were set up according to the protocol,using 200 nmol l^–1 gene-specific primer and 2.5 μl RACE-ready cDNA in the appropriate reaction mix. Gene-specific primers were carefully designed in such a way that they had the following characteristics:23–28 nucleotides, 50–70% GC, T_m≥70°C. To perform 5′-RACE PCR, an antisense primer was designed, and for 3′-RACE PCR a sense primer was designed. Primers were situated as close as possible to the end of known cDNA sequence in order to keep the size of RACE products to a minimum.

Designing primers with a T_m≥70°C allowed the use of touchdown PCR to improve the specificity of the amplification. This method uses an annealing temperature during the initial PCR cycles that is higher than the T_m of the universal primer, allowing only gene-specific synthesis during these cycles. Cycling was performed in thin-walled dome-topped 0.2 ml PCR tubes in a Hybaid PCR Express-Gradient thermocycler. This was performed as follows: 94°C, 3 min; 5 cycles of 94°C 5 s, 72°C 3 min; 5 cycles of 94°C 5 s, 70°C 10 s,72°C, 3 min; 20–25 cycles of 94°C 5 s, 68°C 10 s, 72°C 3 min. Note that the extension time is dependent on the length of the fragment being amplified; 3 min is suitable for cDNA fragments of 2–4 kb.

RACE products were then separated by agarose gel electrophoresis under standard conditions and individual products gel-purified. RACE products were then directly cloned into pCRII-TOPO™ vector and individual clones analysed by restriction enzyme digestion and automated sequencing.

The ORF of the A. stephensi leucokinin-like receptor was amplified using the primers GCCCAGAAGAAATCATGCAAGCAACAG and GCAAAACAGCTCACAGTTAATACACATTGCTCG, and A. stephensi whole fly cDNA as template (see Fig. 3). This was cloned into the pMT/V5-His TOPO^® vector, and the correct orientation determined by restriction enzyme digestion. Constructs were then sequenced to confirm error free cloning of the ORF. The amplification product included the native stop codon to prevent inclusion of the C-terminal V5-His peptide in the expressed protein. S2 cells, cultured under standard conditions(Radford et al., 2002) were transiently transfected with the apoaequorin ORF(Radford et al., 2002) and the A. stephensi leucokinin-like receptor ORF constructs, and expression induced using Cu²⁺ (Radford et al., 2002).

The three putative Anopheles leucokinins identified in this work were synthesised as C-terminally amidated peptides (Research Genetics/Invitrogen Inc.). Peptides were dissolved in H₂O to a concentration of 1 mmol l^–1 and then diluted to the required working concentration in Schneider's medium supplemented with 10% foetal calf serum (FCS; Invitrogen Inc.).

Transfected S2 cells were harvested and incubated with 2.5 μmol l^–1 coelenterazine in the dark at room temperature (RT) for 1–2 h (Radford et al.,2002). 25,000 cells were then placed in 135 μl Schneider's medium supplemented with 10% FCS in a well of a white polystyrene 96-well plate (Berthold Technologies, Redbourn, UK). Bioluminescence recordings were carried out using a Mithras LB940 automated 96-well plate reader (Berthold Technologies) and MikroWin software. 15 μl of each of the Anopheles leucokinin peptides was applied at the required concentration. At the end of each recording samples were disrupted by the addition of 100 μl lysis solution, and the Ca²⁺ concentrations calculated as previously described (Rosay et al., 1997).

Rabbit anti-peptide antibodies were raised against the epitope PHPDSGGESGGDGE (residues 531–543; Genosphere Technologies, Paris,France). An N-terminal cysteine residue was incorporated to permit conjugation to bovine serum albumin (BSA). The antiserum to Anopheles leucokinin receptor showed some background immunoreactivity and, therefore, was purified on a HiTrap Protein A HP column (Amersham Pharmacia Biotech, Little Chalfont,UK) according to the manufacturer's instructions. The protocol used for immunohistology was as described previously(Radford et al., 2002). Briefly, the IgG to Anopheles leucokinin receptor was diluted 1:1000 or the pre-immune serum diluted 1:500. Primary antibody incubations were performed overnight. A Texas Red-conjugated affinity-purified goat anti-rabbit antibody (Jackson Immunologicals, Westgrove, PA, USA) was used at a dilution of 1:1000 for visualization of the primary antiserum. Prior to mounting on slides, tubules were stained with 1 μg ml^–1 of 4′,6′-diamidino-2-phenylindole hydrochloride (DAPI; Sigma-Aldrich,Gillingham, UK). Slides were viewed using a Zeiss 510 META confocal microscope and images were processed with a Zeiss LSM 5 Image Browser.

Protein samples were prepared from tubule or head tissues by homogenization in ice-cold Tris lysis buffer (20 mmol l^–1 Tris, pH 7.5, 250 mmol l^–1 sucrose, 2 mmol l^–1 EDTA, 100 mmol l^–1 NaCl, 50 mmol l^–1β-mercaptoethanol, 2% (w/v) SDS) with protease inhibitor cocktail(P-8340, Sigma). Samples were centrifuged for 10 min at 13 000 g at 4°C to remove debris. Supernatants were removed to a clean tube and assayed for protein concentration (Lowry protein assay). 15μg of each sample were run on SDS-PAGE and blotted according to standard methods. The filter was blocked for 3 h in PBS with 0.1% Tween 20 and 10%non-fat dry milk and washed in PBS/Tween 20 once for 5 min. The filter was incubated for 3 h at RT with IgG to Anopheles leucokinin receptor,diluted 1:1000 (or the pre-immune serum diluted 1:500) in PBS/Tween 20/milk,washed in PBS/Tween 20 three times for 10 min, and incubated for 1 h with secondary antibody (1:5000 horseradish peroxidase-labelled anti-rabbit IgG antibody; Amersham Biosciences) diluted in PBS/Tween 20/milk. The filter was then washed in PBS for 1 h and protein bands visualized using enhanced chemiluminescence (ECL, Amersham Biosciences).

Where appropriate, statistical significance was assessed using Student's t-test for unpaired samples, taking P<0.05 as the critical value.

The Softberry FgenesH gene prediction program predicted a gene contained within a 2179 bp genomic region, including a single intron of 629 bp. Further inspection revealed a transcript of 1545 bp in length, containing an ORF of 891 bp, a 265 bp 5′-UTR and a 3′-UTR of approximately 385 bp(Fig. 1). This would appear to be a full mature transcript, as it contains a perfect consensus sequence for the initiation of transcription (–2 to +4, TCAGTT), and a polyadenylation signal consensus sequence in the 3′-UTR(1508–1513, AATAAA). There is also a TATA box motif upstream of the initiation of transcription site (–30 to –34, TATAA). There is a single presumed ATG start codon at positions 266–268, preceded by four in-frame stop codons.

The ORF of the A. gambiae leucokinin gene encodes a predicted protein of 296 amino acids (Fig. 1). Analysis using the PROSITE program(Bairoch et al., 1997)identified a putative 24-amino-acid signal peptide. Within the remaining protein sequence there are three leucokinin peptides predicted, which have been named according to their similarity to the three known Aedes aegypti leucokinins. The Anopheles leucokinins I and III(Anopheles leucokinins) are flanked by dibasic proteolytic cleavage sites, and in all three peptides a C-terminal Gly is present, which is predicted to be processed into a C-terminal amide group in the mature peptides. A different proteolytic cleavage site is present at the N terminus of Anopheles leucokinin II, consisting of a single Arg with a Lys at residue –8. Four Cys residues are also present in the protein region before to the leucokinin peptide sequences. The position of the four Cys residues is identical in Aedes, suggesting that these residues may play an important role in the function of the precursor protein, perhaps in the formation of disulphide bridges(Veenstra et al., 1997). It has been proposed that these residues are responsible for paraldehyde-fuchsin staining observed in the leucokinin-immunoreactive neuroendocrine cells of the abdominal ganglion in hemimetabolous insects(Veenstra et al., 1997). Owing to the conservation of this staining between insect species, it was suggested that this region of the precursor protein would be conserved between species. However, alignment of the Anopheles and Aedes leucokinin precursors revealed very little other sequence conservation within this region.

Using the same technique, a leucokinin-like precursor gene was also identified in the available sequence of the Drosophila pseudoobscuragenome. The D. pseudoobscura sequence encodes a putative protein of 176 amino acids, containing only a single leucokinin peptide sequence,identical to Drosophila leucokinin from D. melanogaster. A full-length mature transcript could not be reliably identified for D. pseudoobscura, although alignment of the protein with that from D. melanogaster (data not shown) suggested that the entire ORF had been identified.

Anopheles leucokinin I is 15 residues in length, equal to Drosophila leucokinin, the longest leucokinin known to date(Fig. 2). It is also similar in sequence, being identical to Drosophila leucokinin in the bioactive C-terminal pentapeptide–Phe–His–Ser–Trp–Gly–amide. Unsurprisingly, Anopheles leucokinin is most similar across its entire length to Aedes leucokinin I, being identical at 10 residues. Significant similarity can also be seen to another mosquito leucokinin,culekinin III. Similar to Aedes aegypti and Culex salinarius, there are three leucokinins present in Anopheles. The shortest of the three, Anopheles leucokinin II (7 residues), is identical to culekinin I, with only one residue different from Aedesleucokinin II (Fig. 2). In addition, the C-terminal pentapeptide is identical to Drosophilaleucokinin and Anopheles leucokinin I. The third peptide, Anopheles leucokinin III, is 10 residues in length, 1 longer than Aedes leucokinin III, but equal to culekinin II(Fig. 2). The C-terminal core is more divergent in Anopheles leucokinin III, although it retains the essential–Phe–X₁–X₂–Trp–Gly–amide motif, the His being replaced by a Tyr and the Ser being replaced by a Pro. This is identical to the C-terminal cores of Aedes leucokinin III and culekinin II, although less similarity is seen in the more N-terminal residues.

This characterisation of a leucokinin peptide-encoding gene within the A. gambiae genome shows that, as in the yellow fever mosquito, Aedes aegypti (Veenstra et al.,1997), three leucokinin peptides are encoded by a single transcript. Three leucokinins have also been identified in Culex salinarius (Meola et al.,1998), suggesting that this may be a conserved feature among mosquitoes. By contrast, only one leucokinin is found in Drosophila melanogaster (Terhzaz et al.,1999), Drosophila pseudoobscura (this work) and Musca domestica (Coast et al.,2002). This division reflects Dipteran taxonomy: Anopheles,Aedes and Culex are all members of the family Culicidae, whereas Drosophila and Musca are both Schizophora. The two groups thus diverge at the suborder level (Nematocera and Brachycera, respectively);so it is possible that both at the gene organisation level and their proposed modes of action (Beyenbach,1998; Dow and Davies,2003), the leucokinins may plausibly differ between mosquitoes and Drosophila.

In non-Dipteran insects, the numbers of known leucokinins vary widely:three leucokinins have been isolated from the moth Helicoverpa zea;eight from Leucophaea maderae and five from Acheta domesticus and Periplaneta Americana(Torfs et al., 1999). It would be interesting to determine whether the leucokinin peptides from non-mosquito species are also contained within one precursor protein.

Having identified a putative leucokinin receptor from the A. gambiae genome, RT-PCR primers were designed within regions of high sequence similarity to CG10626 and PCR carried out on A. stephensi cDNA. Successful PCR showed bands of expected product sizes for A. gambiae cDNA and genomic DNA (470 bp and 632 bp, respectively;data not shown). Cloning and sequencing of these bands confirmed that the identified Anopheles sequence is expressed. Primers were then designed within the identified gene sequence to carry out 5′-RACE and 3′-RACE analysis to determine the full transcript sequence of the putative leucokinin receptor gene. Three primers were designed for each direction and RACE-ready cDNAs were prepared from A. stephensi whole fly poly(A)⁺ mRNA. Discrete products were amplified using the AnLKR 5′ RACE 1 (∼1.2 kb), 2 (∼1.4 kb) and 3 (∼1.9 kb), and 3′ RACE 2 (∼1.1 kb) and 3 (∼1.8 kb) primers (data not shown). Amplified products were gel extracted, cloned into the pCRII-TOPO^® vector and sequenced in full. From this information the full transcript of the putative A. stephensi leucokinin receptor gene was assembled.

Analysis of the sequence identified from 5′-RACE and 3′-RACE experiments suggests that the sequence of the full mature transcript for this gene has been identified. It is a 2684 bp transcript containing the coding sequence of 1722 bp, a 962 bp 5′-UTR and a 181 bp 3′-UTR. The sequence (bases 1–4) contains a portion of a consensus sequence for the initiation of transcription. The lack of known genomic sequence for A. stephensi precluded the analysis of upstream sequence in order to identify further regulatory sequences. However, no downstream promoter element(DPE)-dependent promoter sequences could be identified in the transcript sequence, as has been found in the Drosophila genomic sequence for LKR (Radford et al., 2002). A single ATG start codon was identified, beginning at base 963 and terminating at a stop codon at position 2682. Upstream of this presumed start codon there are 17 in-frame stop codons. A polyadenylation signal is also present within the 3′-UTR of the sequence (AATAAA, 18 bp from the polyadenylation site). By alignment of the A. stephensi transcript with the A. gambiae genome it is likely that at least six introns are contained within this gene. There is less conservation of the nucleotide sequence within the 5′-UTR, and so the presence of additional introns within this region cannot be ruled out. Again, by inference from the A. gambiae genome,the transcript is thought to be contained within approximately a 7.5 kb genomic region.

The ORF of the A. stephensi leucokinin-like receptor transcript encodes a 574 amino acid protein, which has an estimated molecular mass of 65 kDa. Analysis using the TMHMM program(Krogh et al., 2001) suggests that this predicted protein exhibits the conserved 7 TM domain structure,consistent with it being a functional GPCR(Fig. 3). Other conserved GPCR motifs are also present, such as a triplet motif Asp-Tyr-His at residues 136–138, just downstream of the putative third TM domain. Also two conserved Cys residues, Cys¹¹² and Cys²⁰¹, located in the first and second extracellular loops respectively, are predicted to form a disulphide bond. There are also four potential N-glycosylation sites within the protein sequence, Asn¹⁴ and Asn¹⁸ in the N-terminal region, and Asn¹⁹⁰ and Asn¹⁹⁵ in the putative second extracellular loop. Interestingly, a difference exists between the C-terminal domain of the A. stephensi leucokinin receptor and that of the D. melanogaster LKR: it does not contain the epitope used to raise the anti-CG10626 (D. melanogaster LKR) antibody(Radford et al., 2002).

The protein sequences of the known leucokinin-like receptors, the Drosophila LKR (CG10626; Radford et al., 2002), the lymnokinin receptor (GenBank accession AAD11810; Cox et al., 1997),the B. microplus receptor (AAF72891; Holmes et al., 2003) and the putative A. stephensi receptor(Fig. 3) were aligned using the CLUSTAL X program (Thompson et al.,1994). The sequence alignment was annotated using BioEdit(Hall, 1999)(Fig. 4). The alignment demonstrates that there is considerable similarity between the four protein sequences, particularly within the TM domain-containing regions, with the N-and C-terminal regions being more divergent(Fig. 4). However, the sequence similarity is not as high within TM domains IV and V. The size and spacing of the TM domains is also consistent between the proteins, with only the second extracellular loop being variable in size. Interestingly, the first extracellular loop also appears highly conserved within these proteins,suggesting possible involvement in ligand binding. Several key residues are also conserved. A typical GPCR triplet motif is present immediately after the third TM domain as either an Asp–Arg–Tyr or Asp–Arg–His sequence. Cys residues, thought to form a disulphide bridge in GPCRs, are also present in the first and second extracellular loops of all but the lymnokinin receptor. Similarly putative N-glycosylation sites in the second extracellular loop are present in all proteins except the lymnokinin receptor. Although there is a great deal of sequence diversity within the C-terminal domains, several Ser and Thr residues appear conserved, representing possible sites of phosphorylation.

TBLASTN analysis using the putative A. stephensi receptor was also used to identify a similar sequence in the A. gambiae genome sequence. In addition, Drosophila LKR was used to identify a similar sequence within the partially sequenced D. pseudoobscura genome. Without experimental confirmation the C-terminal domains could not be reliably predicted for the A. gambiae and D. pseudoobscura proteins. Therefore, the putative seven TM domain-containing regions of all the protein sequences were determined using the TMHMM program, and then aligned as before. From the resulting output dnd file a dendrogram was created using the TREEVIEW program (Page, 1996)(Fig. 5). The percentage identity and similarity of each were also calculated using BioEdit and were scored on the BLOSUM62 matrix (Table 1). This was carried out for the TM domain-containing regions and for the four known full-length proteins.

The dendrogram of the TM domain regions of the leucokinin-like receptors reflects the phylogeny of the species concerned(Fig. 5). The D. melanogaster and D. pseudoobscura sequences are closely related,as are the A. stephensi and A. gambiae sequences. These four sequences are more closely related to each other than to the Boophilus sequence, with the molluscan Lymnaea sequence being the least similar. This ancestral relationship is verified by the identity and similarity values for each sequence comparison(Table 1).

Having identified both leucokinins and a leucokinin-like receptor within Anopheles, it was important to establish that they are a functional receptor–ligand pairing. S2 cells were transiently transfected with the apoaequorin ORF and the A. stephensi leucokinin-like receptor ORF constructs, and their expression induced, as previously described (Radford et al.,2002). The S2 cells were then subsequently assayed for agonist-dependent activation by monitoring [Ca²⁺]_ilevels. An agonist-dependent response in [Ca²⁺]_i level was observed for each of the three Anopheles leucokinin peptides,with an order of potency of I>II>III for this particular concentration(Fig. 6).[Ca²⁺]_i levels increased from basal levels of 50 nmol l^–1 to a peak concentration of 365 nmol l^–1,325 nmol l^–1 and 300 nmol l^–1, respectively,upon addition of Anopheles leucokinin I, II or III, representing a 6-to 7.3-fold increase. The [Ca²⁺]_i responses were biphasic in nature, with a primary Ca²⁺ spike followed by a secondary wave that peaked at approximately 175 nmol l^–1, for all three peptides, 20–30 spost-stimulation.

Dose–response curves were then generated for the action of each Anopheles leucokinin on the A. stephensi leucokinin receptor. The receptor responds to all three Anopheles leucokinins in a dose-dependent manner (Fig. 7). Anopheles leucokinin I appears to be slightly more potent at stimulating the receptor, with an EC₅₀ value of 2.0 nmol l^–1, compared to values of 7.4 nmol l^–1 and 8.4 nmol l^–1 for the action of Anopheles leucokinin II and III, respectively. The EC₅₀ values for the actions of these peptides on the A. stephensi receptor are considerably higher than the value for the action of Drosophila leucokinin on the Drosophila LKR, 56.5 pmol l^–1(Radford et al., 2002). Similar EC₅₀ values were determined for the effect of the eight known leucokinins on Leucophaea maderae hindgut contraction(Cook et al., 1989; Cook et al., 1990). The existence of a higher affinity leucokinin receptor within Anophelescannot be ruled out, although it is likely that this sequence would also have been identified from the genomic sequence. Nonetheless, it should be remembered that the action of Aedes leucokinins on Aedestubules were consistent with the existence of more than one receptor(Veenstra et al., 1997); and that the broad concentration range of Drosophila leucokinin on Drosophila tubule was also taken as suggestive of multiple receptor classes (Terhzaz et al.,1999).

The effects of the Anopheles leucokinins on S2 cells expressing the Drosophila LKR, CG10626(Radford et al., 2002) were also established. As this was assessing cross-specific activity, relatively high concentrations of peptide were used (10^–6 mol l^–1 and 10^–7 mol l^–1). The application of both Anopheles leucokinin I (15 amino acids) and Anopheles leucokinin II (7 amino acids) produced a concentration-dependent increase in [Ca²⁺]_i in the S2 cells (Fig. 8). Anopheles leucokinin III (10 amino acids) did not produce any observable [Ca²⁺]_i response at either concentration tested. This is probably because Anopheles leucokinin III possesses a C-terminal pentapeptide, which diverges from the–Phe–His–Ser–Trp–Gly–amide present in Drosophila leucokinin. Both Anopheles leucokinin I and II contain a C-terminal pentapeptide identical to that of Drosophilaleucokinin. The fact that only the Anopheles receptor responds to the divergent Anopheles leucokinin III peptide suggests that the Anopheles receptor has a broader specificity than the Drosophila receptor. The only extracellular regions of these proteins that are considerably different in sequence are the short N-terminal domain and the second extracellular loop. It is tempting to speculate that differences in these regions may define the specificity of the receptor-ligand interaction. Although responses to Anopheles leucokinin I and II were seen at both 10^–6 mol l^–1 and 10^–7 mol l^–1, the response to 10^–6 mol l^–1 was significantly larger. At this concentration, [Ca²⁺]_i levels were seen to increase from basal levels of 50–60 nmol l^–1 to a peak concentration of 250 nmol l^–1 (Anopheles leucokinin I) and 208 nmol l^–1 (Anopheles leucokinin II),approximately a four- and fivefold increase, respectively. It was not possible to determine whether these were maximal responses because the high concentrations required meant that dose–response curves could not be generated. For both peptides the [Ca²⁺]_i response was biphasic in nature, with a primary Ca²⁺ spike and evidence of a sustained secondary wave that peaked at approximately 130 nmol l^–1 20–30 s post-stimulation. Although the primary[Ca²⁺]_i responses to 10^–7 mol l^–1Anopheles leucokinin I and II were different,the secondary responses were similar.

The effects of Drosophila leucokinin(Terhzaz et al., 1999) on S2 cells expressing the A. stephensi leucokinin receptor were also ascertained. Drosophila leucokinin was found to stimulate the A. stephensi receptor in a similar manner to the Anophelesleucokinins, displaying an EC₅₀ value of 1.1 nmol l^–1, very similar to that of Anopheles leucokinin I(Fig. 9).

Antisera against the Anopheles leucokinin receptor identified a band of the predicted size of 65 kDa on western blots, together with a heavier band of approximately 72 kDa (Fig. 10). A similar doublet was observed in Drosophila, and was shown to be due to N-glycosylation of the receptor(Radford et al., 2002). Consistent with this, four potential N-glycosylation sites are present within the receptor sequence.

Immunocytochemistry of adult Anopheles tubule revealed staining specific to the stellate cells (Fig. 11), as has previously been reported for the Drosophilaleucokinin receptor.

This paper identifies and characterises a cognate pairing of the Anopheles leucokinins and their receptor in a genus containing major human and animal disease vectors. By comparison with the abundant knowledge of the leucokinin family in insects, it is now possible to distinguish significant differences in the numbers of leucokinins and their potencies across the Order Diptera. This may help to explain a radical difference between the diuretic actions of leucokinins in Diptera. In the Drosophila tubule, the Drosophila leucokinin receptor is found only in the type II (stellate) cells, and (using a transgenic aequorin calcium reporter) the peptide is known to raise calcium only in these cells(Radford et al., 2002). By contrast, in the Aedes tubule, the leucokinins are thought to act on principal cells to regulate paracellular permeability(Beyenbach, 2003a; Yu and Beyenbach, 2004). Given the relatively divergent taxonomy of these two insects, and the differences in prepropeptide structure and in receptor C-terminal sequence, the functional differences might not be so surprising. However, in Anopheles, which is phylogenetically much closer to Aedes, all aspects of leucokinin signalling, including the receptor localisation to stellate cells in the tubule, appear much closer to Drosophila than to Aedes. It will thus be of great interest to locate the homologous receptor within the Aedes tubule.

This work was supported by the Biotechnology and Biological Sciences Research Council, UK and by a Wellcome Trust PhD scholarship to J.C.R.