Mutants of norpA, encoding phospholipase Cβ (PLCβ), and itpr, encoding inositol (1,4,5)-trisphosphate receptor(IP3R), both attenuate response to diuretic peptides of Drosophila melanogaster renal (Malpighian) tubules. Intact tubules from norpA mutants severely reduced diuresis stimulated by the principal cell- and stellate cell-specific neuropeptides, CAP2b and Drosophila leucokinin (Drosokinin), respectively, suggesting a role for PLCβ in both these cell types. Measurement of IP3production in wild-type tubules and in Drosokinin-receptor-transfected S2 cells stimulated with CAP2b and Drosokinin, respectively, confirmed that both neuropeptides elevate IP3 levels.
In itpr hypomorphs, basal IP3 levels are lower,although CAP2b-stimulated IP3 levels are not significantly reduced compared with wild type. However,CAP2b-stimulated fluid transport is significantly reduced in itpr alleles. Rescue of the itpr90B.0 allele with wild-type itpr restores CAP2b-stimulated fluid transport levels to wild type. Drosokinin-stimulated fluid transport is also reduced in homozygous and heteroallelic itpr mutants.
Measurements of cytosolic calcium levels in intact tubules of wild-type and itpr mutants using targeted expression of the calcium reporter,aequorin, show that mutations in itpr attenuated both CAP2b- and Drosokinin-stimulated calcium responses. The reductions in calcium signals are associated with corresponding reductions in fluid transport rates.
Thus, we describe a role for norpA and itpr in renal epithelia and show that both CAP2b and Drosokinin are PLCβ-dependent, IP3-mobilising neuropeptides in Drosophila. IP3R contributes to the calcium signalling cascades initiated by these peptides in both principal and stellate cells.
The accepted paradigm for hormonally stimulated increases in intracellular cytosolic calcium concentration ([Ca2+]i) in non-excitable cells centres on G-protein-coupled activation of phospholipase Cβ (PLCβ) upon ligand—receptor binding, resulting in an increase in intracellular levels of the second messengers diacylglycerol (DAG)and inositol (1,4,5)-trisphosphate (IP3). IP3 binds to its receptor on the endoplasmic reticulum (ER), IP3R, which functions as a calcium-release channel. Calcium is released from intracellular stores, with associated calcium entry via plasma membrane calcium channels by an unknown mechanism termed `capacitative calcium entry'(Berridge, 1997; Putney, 1997).
The processes of calcium signalling in vivo have been intensely investigated in Drosophila phototransduction, where novel calcium channels and associated signalling proteins have been discovered in forward genetic screens (Hardie and Raghu,2001). Importantly, studies in Drosophila have subsequently revealed vertebrate and human homologues of such signalling complexes (Harteneck et al.,2000). Light-driven phototransduction has been shown to be PLC-dependent; as such, most studies of PLC function in vivo in Drosophila have utilised the photoreceptor model. Two genes encode PLCβ in Drosophila: norpA and Plc21C. Mutants in norpA severely reduce phototransduction(Pak et al., 1970). Molecular cloning of the gene, together with biochemical analyses of PLC activity in the eye of norpA mutants showed that norpA encodes a retinal-specific PLC similar to bovine brain PLC(Bloomquist et al., 1988). By contrast, Plc21C is more generally expressed, with a 7 kb transcript in the eye and central nervous system (CNS) and a 5.6 kb transcript in heads and bodies (Shortridge et al.,1991).
Mutagenic analysis of the IP3R has also been informative in insects. In Drosophila, a single gene, itpr-83, encodes IP3R (Hasan and Rosbash,1992; Yoshikawa et al.,1992). Drosophila IP3R has most similarity to type I IP3R in vertebrates. Embryonic expression of itprhas been documented, and a delayed-development phenotype has been identified in mutants (Venkatesh and Hasan,1997). In the adult, expression has been shown in photoreceptors,brain and antennae, with expression also documented in the eye, which suggests a role for IP3R in chemosensation and in visual processes. However,in spite of the documented role of PLC in the eye, recent work has shown that IP3 signalling is unnecessary for phototransduction to occur(Hardie and Raghu, 1998; Venkatesh and Hasan,1997).
While norpA and itpr have been assigned neural roles— indeed, norpA is generally considered to be visual system specific — it is likely that much more general roles for these genes exist but have yet to be documented due to lack of an informative physiological phenotype. The Drosophila Malpighian tubule is an ideal genetic model for transporting epithelia and provides a robust phenotype for the integrative physiology of cell signalling and transport genes(Dow and Davies, 2001). Previous work has established that ion transport and cell signalling process are compartmentalised into different tubule cell subtypes: the principal and stellate cells (Dow and Davies,2001). Furthermore, direct measurements of cell-specific intracellular calcium signalling mechanisms using targeted aequorin show a direct modulation of fluid transport by agents that mobilise intracellular calcium (O'Donnell et al.,1998; Rosay et al.,1997; Terhzaz et al.,1999; MacPherson et al.,2001). The Drosophila neurohormones capa-1 (a member of the CAP2b family; Kean et al.,2002) and Drosophila leucokinin (Drosokinin; Terhzaz et al., 1999) have been shown to stimulate fluid transport rates, which are associated with increases in cytoplasmic calcium concentration in principal and stellate cells, respectively, in tubule main segment. The CAP2b response is dependent on extracellular calcium (Rosay et al., 1997); furthermore, it has been recently demonstrated that TRP/TRPL (transient receptor potential/TRP-like) and L-type calcium channels play a role in this response (MacPherson et al., 2000, 2001). However, the relative contribution of intracellular calcium stores to these neurohormone-induced responses is still unclear. Use of the ER Ca2+-ATPase inhibitor thapsigargin in the presence of extracellular calcium results in elevation of intracellular calcium levels in both principal and stellate cells and increased fluid transport rates (Rosay et al., 1997). However, in the absence of extracellular calcium the response is abolished in principal cells but remains in stellate cells. Thus,it appears that the contribution of ER calcium stores, and thus of calcium signalling via IP3R to capacitative calcium entry in tubule cells, is cell-type specific. Alternatively, the putative thapsigargin-sensitive pool in principal cells is very small and is emptied too rapidly to monitor. It is thus of interest to try to dissect the different contributions of calcium signalling genes in the context of principal and stellate cell function.
In this study, we show that mutations in norpA reduce stimulation of fluid transport by both CAP2b and Drosokinin. Furthermore,CAP2b and Drosokinin both elevate IP3 levels. Thus,PLCβ and IP3 are involved in stimulated fluid transport; we have thus utilised itpr mutants to define the contribution of IP3R to neurohormone-mediated increases in epithelial fluid transport. Genetic blockade of IP3R function results in inhibition of neuropeptide-fluid transport rates associated with either CAP2bor Drosokinin. In itpr mutants, downregulation of fluid transport is associated with reductions in neuropeptide-stimulated intracellular calcium levels. Thus, PLCβ-mediated IP3 signalling plays a functional role in calcium signalling and epithelial fluid transport in Drosophila, confirming that the important norpA and itpr genes are not neural specific.
Materials and methods
Coelenterazine was purchased from Molecular Probes (Leiden, The Netherlands) and dissolved in ethanol before use. Schneider's medium was obtained from GIBCO Life Technologies (Invitrogen Ltd, Paisley, UK). Neuropeptides CAP2b (pyroELYAFPRV-amide; Davies et al., 1995) and Drosokinin (NSVVLGKKQRFHSWG-amide; Terhzaz et al., 1999) were synthesised by Research Genetics, Inc.(Invitrogen Ltd). All other chemicals were obtained from Sigma (Pool, UK).
Drosophila melanogaster (Meigen) were maintained on a 12h: 12h L:D cycle on standard cornmeal—yeast—agar medium at 25°C. Oregon R(OrR) wild-type flies used were those described previously(Dow et al., 1994). Mutant lines of norpA (Bloomquist et al.,1988; Pearn et al.,1996) and itpr(Venkatesh and Hasan, 1997)have been described previously. itpr lines were used as out-crossed lines to OrR to rule out the effects of balancer chromosomes on the tubule phenotype. Multiple alleles were utilised in this study in order to control for any effects of genetic background in individual lines. Choice of alleles was based on the health of the lines. Lines used were as follows: norpAH52/norpAH52 (temperature-sensitive allele); norpAP24/norpAP24 (kind gifts of R. C. Hardie, University of Cambridge); itprXR12/+ (X-ray inversion); itpr90B.0/+, itpr1664/+, itpr1664/itprXR12,itpr1664/itpr90B.0 and itpr1664/itpr1664 (P-element insertions); itprWC361/itprUG3 (EMS alleles). itprlines were slow-growing due to eclosion defects(Venkatesh and Hasan, 1997). Additionally, the following lines were also utilised:hsGAL4;itpr90B.0 and UAS-itpr(Venkatesh et al., 2001). Temperature-sensitive (ts) and hsGAL4 lines were heat-shocked at 37°C for 30 min and left to recover at 23°C before experimentation.
To produce flies in which tubule calcium measurements could be made using the calcium reporter aequorin (Rosay et al., 1997; Terhzaz et al.,1999), it was necessary to generate itpr lines in an aequorin background under control of an hsGAL4 promoter. These were based on an X-chromosome insertion of UAS-aeq, and an hsGAL4 construct on chromosome 2,leaving chromosome 3 free for itpr alleles. The following lines were generated and utilised for this study:aeq;hsGAL4;itprXR12/+,aeq;hsGAL4;itpr90B.0/+,aeq;hsGAL4;itpr1664/+,aeq;hsGAL4;itpr1664/itpr1664 and aeq;hsGAL4;itprWC361/itprUG3. Extremely poor viability of the itpr1664/itprXR12 and itpr1664/itpr90B.0 heteroallelics in the aequorin background did not allow use of these lines for calcium measurements. Verification of aequorin expression was achieved at several stages of the crossing procedure by measuring total light output by dissected, intact tubules after lysis in Triton/CaCl2 as described below. The presence of the appropriate itpr allele was verified in progeny of aeq;itpr flies by RT-PCR. Maintenance of the itpr phenotype was assessed by fluid transport assays in the presence of either CAP2b or Drosokinin.
Transport (fluid secretion) assays
Flies were cooled on ice and then decapitated prior to isolation of whole tubules. Malpighian tubules were isolated into 10 μl drops of a 1:1 mixture of Schneider's medium and Drosophila saline (NaCl, 117.5 mmoll-1; KCl, 20 mmoll-1; CaCl2, 2 mmoll-1; MgCl2, 8.5 mmoll-1;NaHCO3, 10.2 mmoll-1; NaH2PO4, 4.3 mmoll-1; Hepes, 15 mmoll-1; glucose 20 mmoll-1) under liquid paraffin, and fluid secretion rates were measured as described previously (Dow et al., 1994) under the different conditions described in the text. CAP2b and Drosokinin were added as solutions in assay medium at 30 min.
Heterologous expression of the Drosokinin receptor
The recently characterised Drosokinin receptor(Radford et al., 2002) was used to assay Drosokinin-stimulated IP3 production. In order to transfect S2 cells with Drosokinin receptor, the protocol described below was adopted.
Primers were designed for the Drosokinin receptor (CG1062; Radford et al., 2002) to allow amplification from the start to the stop codon of the coding sequence. Forward primers were designed to include a 5′ Kozak translational initiation sequence (G/ANNATGG). Amplification was carried out with EXPAND High Fidelity Polymerase (Roche Diagnostics Ltd, Lewes, UK)according to manufacturers instructions. Forward (GACATGGACTTAATCGAGCAGGAG)and reverse (TTAAAGTGGTTGCCACAAGGAC) primers were used to generate a fragment of 1626bp. OrR cDNA was used as template. The PCR product was purified by gel extraction and directly cloned into the pMT-V5/His TOPO TA inducible expression vector (Invitrogen). The construct was verified by restriction enzyme digestion and sequencing to ensure that no mutations had been induced during cloning.
S2 cell culture
S2 cells were maintained in DES (Drosophila expression system)medium (Invitrogen) supplemented with 10% heat-inactivated foetal calf serum(FCS; Invitrogen). Cells were grown in suspension at an initial density of 2-4×106 cells ml-1 at 23°C. S2 cells were transiently transfected at a density of 1×106 cells ml-1 using calcium phosphate (Invitrogen), according to manufacturer's instructions. Cells were transfected with 20μg of the Drosokinin receptor expression construct and were used 24h post-induction of the metallothionein promoter with Cu2+ ions.
Mass measurement of inositol (1,4,5)-trisphosphate (IP3)levels
IP3 levels in tubules were measured by a quantitative radioligand-binding assay as described elsewhere(Palmer et al., 1989) using an IP3-binding protein preparation derived from bovine adrenal gland.
Tubules (20 per sample) were dissected from wild-type and itprmutants into 9μl of Schneider's medium. Samples were stimulated with CAP2b (10-7 moll-1) for 0 s (control), 2 s or 5 s in a final sample volume of 10μl and performed in triplicate. Initial experiments showed that IP3 levels peaked at 5 s post-stimulation(data not shown); this time was used for all subsequent experiments.
S2 cell preparations
To stimulate S2 cells, Drosokinin(Terhzaz et al., 1999) was diluted to working concentration in DES medium/FCS, then added to 5×104 cells (approximating to 5000 transfected cells) in DES medium/FCS to a final concentration of 10-7 moll-1 for the appropriate time. Initial experiments showed that peak IP3generation occurred at 10 s after peptide stimulation (data not shown). Cells were co-transfected with an eGFP (enhanced green fluorescent protein) control plasmid in order to measure transfection efficiency using a haemocytometer. The same transfection batch was used for all samples in any one data set;stimulations were performed in duplicate.
For both tubule and S2 cell preparations, reactions were terminated with 10% (v/v) ice-cold perchloric acid and samples were homogenised using a Polytron homogeniser on ice. Cellular debris was removed by centrifugation and the supernatants neutralised with 1.5 moll-1 KOH/60 mmoll-1 Hepes in the presence of 2μl Universal Indicator. Precipitated salts were spun down and the supernatants transferred to fresh tubes. Additions of 2500 d.p.m. [3H]Ins(1,4,5)P3([3H]IP3; specific activity 370-1850 GBq mmoll-1; Amersham Biosciences UK Ltd, Little Chalfont, UK),incubation buffer and binding protein were made [final concentrations: 25 mmoll-1 Tris-HCl (pH 9); 5 mmoll-1 NaHCO3; 1 mmoll-1 EDTA; 1 mmoll-1 EGTA; 0.25 mmoll-1dithiothreitol (DTT); 1 mg ml-1 bovine serum albumin (Fraction V);0.4 mg ml-1 binding protein] to a final volume of 400μl, and the samples were incubated on ice for 45 min prior to centrifugation at 12 000g (4°C) for 1 min. Supernatants were removed by aspiration, the pellets dissolved in 1 ml of scintillation fluid and the radioactivity therein determined by scintillation counting. A standard curve,using 0-40 pmol IP3 per sample, was generated in parallel. Non-specific binding was determined using 100 pmol IP3. Standard curves were plotted as %B/Bo versus pmol of unlabelled IP3, where B is the specific binding of[3H]IP3 (at a given concentration of unlabelled IP3), and Bo is the maximal specific binding of[3H]IP3 (at 0 pmol of unlabelled IP3). A similar calculation was made using values of specific binding of[3H]IP3 of tissue samples and the IP3 content therein determined using the standard curve.
Protein concentrations in tubule samples were assessed by Lowry assays. Three replicate samples were pooled for assay of IP3 content in order to obtain measurable levels of IP3. Duplicate samples were assayed for each experimental sample.
Measurements of [Ca2+]i using an aequorin transgene under heat-shock control
hsGAL4;aeq (Rosay et al.,1997) were used as control animals, and protocols used were essentially those previously described. For each assay, 20-40 tubules from 4-14-day-old adults were dissected in Schneider's medium 2 h after heat-shock(37°C for 30 min). Tubules were pooled in 160μl of the same buffer and aequorin reconstituted with the cofactor coelenterazine (final concentration,2.5 μmoll-1). Bioluminescence recordings were made with a luminometer (LB9507; Berthold, Pforzheim, Germany); recordings were made every 0.1 s for each tube. Each tube of 20 tubules was used for a single data point:after recording [Ca2+]i levels, tissues were disrupted in 350μl lysis solution [1% (v/v) Triton X-100/100 mmoll-1CaCl2], causing discharge of the remaining aequorin and allowing estimation of the total amount of aequorin in the sample. Calibration of the aequorin system and calculation of [Ca2+]i were performed as previously described (Rosay et al., 1997). Mock injections with Schneider's medium were applied to all samples prior to treatment with neuropeptides.
PLCβ contributes to CAP2b and Drosokinin-induced fluid transport
Two norpA alleles of differing severity were available for study. norpAP24 eyes do not express norpA protein, as assessed by western blots, and show markedly reduced PLC activity(Pearn et al., 1996); as expected, norpAP24 shows no electroretinogram response to any light stimulus. This allele, therefore, would be expected to display a severe phenotype if norpA acted in tubule function. In norpAH52, however, some norpA transcript is detectable at the restrictive temperature, although this is severely reduced compared with wild type (Bloomquist et al.,1988). As such, any epithelial phenotype would not be expected to be as severe as that of norpAP24.
Both norpA mutants manifest an epithelial phenotype(Fig. 1). No significant change in basal secretion rate was observed repeatedly in norpA mutants;however, both alleles, norpAH52 and norpAP24, severely attenuated CAP2b-induced secretion to similar extents (Fig. 1A). By contrast, for Drosokinin-stimulated fluid transport, the norpA alleles were distinguishable(Fig. 1B); in the eye, norpAP24 displays a much more severe phenotype compared with norpAH52. Kinetics of the fluid secretion response in all lines were similar.
Thus, the data show that fluid transport induced by both CAP2band Drosokinin requires a PLCβ-dependent signalling pathway; as such, norpA function is not confined to phototransduction.
CAP2b and Drosokinin stimulate IP3production
CAP2b and Drosokinin have both been shown to act through intracellular calcium; so, if their action relies on PLC(Fig. 1), they should also each act to raise IP3 in tubules. Table 1 shows that CAP2b, which acts only on principal cells, increases IP3levels in intact wild-type tubules. This, together with the data in Fig. 1, shows that CAP2b acts via a phosphoinositide (PI)-PLC-dependent mechanism.
|Ligand .||Tissue .||Control .||Stimulation .|
|CAP2b||Oregon R Malpighian tubules||0.75±0.11 nmol μg protein-1||1.26±0.10* nmol μg protein-1|
|Drosokinin||Drosokinin-receptor-transfected S2 cells||1.08±0.08 pmol 5000 cells-1||9.45±2.39* pmol 5000 cells-1|
|Ligand .||Tissue .||Control .||Stimulation .|
|CAP2b||Oregon R Malpighian tubules||0.75±0.11 nmol μg protein-1||1.26±0.10* nmol μg protein-1|
|Drosokinin||Drosokinin-receptor-transfected S2 cells||1.08±0.08 pmol 5000 cells-1||9.45±2.39* pmol 5000 cells-1|
Significant differences between inositol (1,4,5)-trisphosphate(IP3) content in control and stimulated samples are denoted by * (P<0.05, Student's t-test, unpaired samples, N=4).
Aedes leucokinins have been shown to stimulate IP3production in mosquito tubules (Cady and Hagedorn, 1999). However, as Drosokinin exerts its effects solely via stellate cells in Drosophila tubules(Terhzaz et al., 1999), and as Drosophila tubules each contain only 22 stellate cells(Sozen et al., 1997),Drosokinin-stimulated IP3 levels could not be reliably quantified in intact tubules (data not shown). Accordingly, an in vitro approach was used. Drosophila S2 cells transfected with the Drosokinin receptor have been shown to display increased cytosolic intracellular calcium when stimulated with Drosokinin (Radford et al., 2002). Data in Table 1 show that Drosokinin increases IP3 content in these cells by approximately eightfold. As the Drosokinin receptor has been localised to only stellate cells in vivo(Radford et al., 2002), it is probable that Drosokinin stimulates IP3 production in stellate cells in intact tubules. As with CAP2b, Drosokinin action requires activation of PI-PLC.
Resting and CAP2b-stimulated levels of IP3are significantly reduced in itpr mutants
Intracellular signalling is frequently regulated by feedback. If IP3 signalling contributes to the biological effects of neuropeptides in tubules, lesions in the IP3R might feed back to downregulate the IP3 signalling molecule in vivo.
Measurement of basal IP3 levels in wild-type and itprtubules shows that disruption of IP3R results in a reduction in resting IP3 levels. Significant reductions in basal IP3levels are observed in heterozygous alleles, homozygous itpr1664/itpr1664 and heteroallelic lines(Fig. 2A).
IP3 levels were also measured in neurohormone-stimulated intact tubules. As Drosokinin-stimulated IP3 production was not measurable in intact tubules, experiments were conducted only on CAP2b-stimulated wild-type and itpr tubules(Fig. 2B). IP3content is increased in all CAP2b-stimulated itpr tubules,although to a lesser extent in homozygous itpr1664/itpr1664 and heteroallelic itprWC361/itprUG3 lines compared with wild type. However, these differences are not statistically significant. Thus,tubules from itpr mutants are able to generate IP3 in response to CAP2b, suggesting that PI-PLC-dependent signalling is not compromised in these lines.
Diuresis is inhibited by disruption of itpr
As mutations in norpA result in an epithelial phenotype, we investigated the possibility of uncovering such phenotypes in itprmutants. We show that CAP2b stimulation of fluid transport is inhibited in heterozygous, homozygous and heteroallelic itpr alleles(Fig. 3A). Significant inhibition is observed in itpr90B.0/+, itpr1664/itprXR12 and itpr1664/itpr90B.0. However,CAP2b-stimulated fluid transport is almost completely abolished in the homozygous itpr1664/itpr1664 and heteroallelic itprWC361/itprUG3 lines. It is possible that the non-correlation between the severity of the alleles(itpr1664/itpr90B.0 versus itpr1664/itpr1664) and CAP2b-stimulated transport in these alleles is due to either differences in genetic background or by impact of the mutations on other interacting signalling pathways that are activated by CAP2b.
By contrast, Drosokinin-stimulated fluid transport is not affected by either heterozygous itprXR12/+, itpr90B.0/+ or itpr1664/+ alleles(Fig. 3B). Significant inhibition is only observed in the heteroallelics itpr1664/itprXR12,itpr1664/itpr90B.0 and itprWC361/itprUG3 and in the homozygous itpr1664/itpr1664. Furthermore, there are marked differences in the severity of inhibition between Drosokinin- and CAP2b-stimulated fluid transport, especially in itpr1664/itpr1664 and itprWC361/itprUG3 alleles.
Thus, itpr acts in both principal and stellate cells to transduce CAP2b and Drosokinin diuretic signals.
Rescue of itpr90B.0restores CAP2b-stimulated fluid transport levels
Sometimes, phenotypes ascribed to mutant loci in Drosophila are subsequently found to be caused by second-site mutations or other genetic accidents incidental to the original study. It is thus desirable to confirm that mutant effects are genuinely due to the locus of interest. Crosses were established between heterozygous hsGAL4;itpr90B.0 and UAS-itpr homozygotes to allow rescue of itpr90B.0. Tubules from flies with the w/UAS-itpr;+/hsGAL4;+/itpr90B.0 genotype were used in transport assays. Control lines used were OrR and UAS-itpr. Non-heat-shocked w/UAS-itpr;+/hsGAL4;+/itpr90B.0were not used as controls, as the heat-shock promoter is `leaky' and is transcribed at 25°C (G. Hasan, unpublished).
Fig. 4 shows that expression of itpr restores CAP2b-stimulated fluid transport to levels indistinguishable from wild-type. Although disruption of itprmay compromise associated signalling pathways resulting in downregulation of CAP2b-stimulated transport, successful rescue of itpr90B.0/+ with USA-itpr provides strong evidence that the epithelial phenotype observed in this line is associated with mutated itpr.
CAP2b-induced calcium signalling is impaired initpr mutants
Previous work has shown that capa peptides(Kean et al., 2002) increase intracellular calcium in tubule main segment principal cells(Rosay et al., 1997) via plasma membrane calcium channels (MacPherson et al., 2000, 2001). However, release from intracellular stores is not measurable in this cell type(Rosay et al., 1997), thus calling into question the role of IP3R-sensitive stores in principal cells. If release of calcium from IP3-sensitive internal stores does occur upon CAP2b stimulation, changes in this response may be observed in itpr mutants.
Using targeted aequorin, cytosolic calcium measurements in wild-type and itpr tubules stimulated with CAP2b were performed; typical traces are shown in Fig. 5A. Fig. 5Ai shows the biphasic rise, consisting of a rapid primary peak followed by a slow secondary rise in cytosolic calcium in CAP2b-stimulated wild-type aeq;hsGAL4 tubules. This calcium signature is also observed in capa/CAP2b-stimulated wild-type tubules (Kean et al.,2002). The response is reduced in heterozygous itpralleles, aeq;hsGAL4;itprXR12/+ and aeq;hsGAL4;itpr90B.0/+(Fig. 5B), which may result in the transport phenotype observed. Interestingly, in heterozygous itpr1664/+, the primary response is unaffected, with only the secondary response being reduced; in this line, stimulated fluid transport is not significantly different from control(Fig. 3A). By contrast, the alleles that display the most severe transport phenotype(itpr1664/itpr1664 and itprWC361/itprUG3) also display an attenuated calcium response to CAP2b (Fig. 5Av,vi,B). Both reduction in the primary peak and loss of the secondary rise are observed in these lines. However, none of the itpralleles completely abolish CAP2b-stimulated calcium signalling. This is consistent with itpr being an essential gene and with viable alleles all being hypomorphs rather than nulls.
These results thus suggest that IP3R-mediated calcium release from intracellular stores contributes significantly to calcium signalling, and consequent diuresis, in principal cells.
itpr mutants reduce Drosokinin-induced calcium signals
We have shown previously that leucokinin(Rosay et al., 1997; O'Donnell et al., 1998) and endogenous Drosophila leucokinin(Terhzaz et al., 1999) elevate intracellular calcium levels in stellate cells. Furthermore, experiments in calcium-free medium show that stellate cells display emptying of intracellular calcium stores in the presence of the ER-calcium ATPase inhibitor,thapsigargin (Rosay et al.,1997). Thus, Drosokinin-stimulated calcium increases should be reduced in itpr mutants.
Data in Fig. 6A show typical traces of Drosokinin-stimulated increases in cytosolic calcium in wild-type and itpr tubules. A rapid response is observed in wild-type tubules(Fig. 6Ai; Terhzaz et al., 1999; Radford et al., 2002). This response is severely reduced in itpr1664/itpr1664 and itprWC361/itprUG3 flies(Fig. 6Av,vi,B). Drosokinin-stimulated fluid transport is also reduced in these lines(Fig. 3B); thus, functional IP3R contributes to Drosokinin-stimulated calcium signalling and fluid transport.
The role of PLC in vivo has been most studied in Drosophila phototransduction. Experimental evidence suggested that norpA and plc21 were expressed in the eye or CNS(Bloomquist et al., 1988; Shortridge et al., 1991);furthermore, the dependence of phototransduction on PI-PLC supported specific roles for these genes in the eye. However, RT-PCR data show that norpA and plc21 are both expressed in tubules (M. R. MacPherson and V. P. Pollock, unpublished). We have shown here that norpA mutants display an epithelial phenotype, which is revealed upon neuropeptide stimulation. Furthermore, the severity of the epithelial response for each allele correlates with that observed in the eye and also with the levels of expression of norpA and biochemical activity of PLC. However, while we have established an epithelial role for norpA, it has been difficult to assess the contribution of plc21 to epithelial transport and signalling due to our lack of genetic tools and suitable antibodies.
IP3 and IP3R have been shown to be critical for calcium signalling in non-excitable cells via a mechanism of IP3-induced release of calcium from internal stores. Previous work has shown that IP3R is expressed in Malpighian tubules(Pollock et al., 2000; Blumenthal, 2001). We now ascribe functional significance to this finding and show that CAP2bstimulates IP3 production in intact Drosophila tubules. Furthermore, we show that Drosokinin mobilises IP3 in Drosokinin receptor-transfected S2 cells.
A role for IP3R in neurohormone-stimulated epithelial transport has been demonstrated using an allelic series of itpr mutants. These itpr lines have been extensively investigated previously in order to determine the role of IP3R in vivo. itprXR12/+(inversion), itpr90B.0/+ (null) and itpr1664/+ (hypomorph) display delayed moulting and reduced expression of the ecdysone-inducible gene E74. In terms of severity, itpr90B.0 and itprXR12 are the most severe alleles (Venkatesh and Hasan, 1997). These lines, as well as itprWC361/itprUG3, have also been used in studies of olfaction, where it has been shown that, in the most severe alleles, olfactory adaptation is not maintained(Deshpande et al., 2000). CAP2b-stimulated epithelial transport is reduced in the null mutant(itpr90B.0) compared with wild-type tubules, while the most severe phenotypes are observed in itpr1664homozygotes and the heteroallelics. Rescue of the itpr90B.0 transport phenotype is demonstrated with UAS-itpr, which strongly suggests that the tubule phenotype is associated with disruption in IP3R. Interestingly, while the null mutant displays a phenotype upon CAP2b stimulation, only the most severe itpr1664 homozygotes and the heteroallelics affect Drosokinin-stimulated secretion. Therefore, reduced CAP2b-stimulated fluid transport in itpr90B.0may be due to principal cell-specific signalling processes associated with the null mutation.
Intriguingly, resting levels of IP3 in tubules are significantly reduced in all itpr mutants apart from itpr1664/itprXR12, suggesting a possible feedback mechanism between receptor (IP3R) and second messenger(IP3). However, this is not associated with an epithelial phenotype, as no difference in basal rate is observed in tubules from itpr lines as compared to wild-type flies.
We have previously shown that the neuropeptide CAP2b induces a rise in cytosolic calcium in only principal cells that is dependent on extracellular calcium; these calcium signalling events, which may mediate NO/cGMP signalling, correlate with CAP2b-stimulated fluid transport. The CAP2b-induced response is abolished by L-type calcium channel inhibitors and also in mutants for the plasma membrane calcium channels, TRP and TRPL (MacPherson et al., 2000, 2001). Thus,CAP2b-induced calcium signalling occurs via multichannel mechanisms. We demonstrate here that IP3R plays a role in CAP2b-induced calcium signalling(Fig. 5). Extensive work using norpA mutants has shown that norpA-encoded PLC plays a critical role in rhabdomeres. Thus, PLCβ, which is required for the cleavage of phosphatidylinositol (4,5)-bisphosphate (PIP2) to IP3 and diacylglycerol (DAG), is necessary for phototransduction. Interestingly, however, using the itpr null and itpr1664 mutants, IP3 signalling has been shown to be unnecessary to activate light-activated conductance(Raghu et al., 2000a). This response is, however, dependent on calcium entry via TRP/TRPL channels. DAG has been shown to activate native TRP and TRPL channels in photoreceptors and recombinant TRPL channels(Chyb et al., 1999). Furthermore, recent studies have supported the role of DAG in TRP/TRPL action:DAG kinase mutants display constitutive activation of TRP and TRPL channels(Raghu et al., 2000b), and, in vertebrate cells, TRPC3 (transient receptor potential-like channel 3) has been shown to be activated by DAG independently of IP3R(Venkatachalam et al., 2001). Thus, PLC is involved in regulation of TRP/TRPL plasma membrane calcium channels without a requirement for IP3. It is possible, then, that PLC-activated DAG generation may regulate TRP/TRPL channels in tubules, which may contribute significantly to the `multichannel' mode of action of CAP2b. Furthermore, if DAG/TRP/TRPL signalling were compromised in some itpr lines, this may explain the significant impact of itpr alleles (for example, itpr null) on CAP2b-stimulated, but not Drosokinin-stimulated, fluid transport and calcium signalling (Figs 3, 5).
Previous work has shown that leucokinin and Drosokinin increase cytosolic calcium concentration in only stellate cells in tubules, which express the Drosokinin receptor (Rosay et al.,1997; Terhzaz et al.,1999; Radford et al.,2002). This suggests that IP3-mediated calcium signalling may occur in stellate cells. Activation of calcium signalling may be linked to chloride shunt conductance, which is also confined to stellate cells (O'Donnell et al.,1998). It is thus possible that Drosokinin stimulates PLCβ-dependent, IP3-mediated calcium signalling in stellate cells in vivo, which increases chloride conductance, resulting in increased fluid transport.
Thus, Malpighian tubules, in contrast to photoreceptors, require both functional PLCβ and IP3R for neuropeptide-activated signal transduction in principal and stellate cells. The expression of plc21in tubules, however, and the role of DAG on plasma membrane calcium channels suggest that extremely complex mechanisms of signalling are used by tubule cells.
In summary, we have demonstrated non-neuronal, epithelial phenotypes for norpA (PLCβ) and itpr (IP3R) and have correlated cell-specific signalling events for both IP3 and calcium to transport phenotypes.
This work was supported by the BBSRC.