Mutants in Drosophila TRPC Channels Reduce Olfactory Sensitivity to Carbon Dioxide

Background Members of the canonical Transient Receptor Potential (TRPC) class of cationic channels function downstream of Gαq and PLCβ in Drosophila photoreceptors for transducing visual stimuli. Gαq has recently been implicated in olfactory sensing of carbon dioxide (CO2) and other odorants. Here we investigated the role of PLCβ and TRPC channels for sensing CO2 in Drosophila. Methodology/Principal Findings Through behavioral assays it was demonstrated that Drosophila mutants for plc21c, trp and trpl have a reduced sensitivity for CO2. Immuno-histochemical staining for TRP, TRPL and TRPγ indicates that all three channels are expressed in Drosophila antennae including the sensory neurons that express CO2 receptors. Electrophysiological recordings obtained from the antennae of protein null alleles of TRP (trp343) and TRPL (trpl302), showed that the sensory response to multiple concentrations of CO2 was reduced. However, trpl302; trp343 double mutants still have a residual response to CO2. Down-regulation of TRPC channels specifically in CO2 sensing olfactory neurons reduced the response to CO2 and this reduction was obtained even upon down-regulation of the TRPCs in adult olfactory sensory neurons. Thus the reduced response to CO2 obtained from the antennae of TRPC RNAi strains is not due to a developmental defect. Conclusion These observations show that reduction in TRPC channel function significantly reduces the sensitivity of the olfactory response to CO2 concentrations of 5% or less in adult Drosophila. It is possible that the CO2 receptors Gr63a and Gr21a activate the TRPC channels through Gαq and PLC21C.


Introduction
Carbon dioxide (CO 2 ), a green house gas, has context dependent effects on behavior of specific insect species. The moth Manduca sexta uses CO 2 as a cue to evaluate flowers during foraging [1,2] and ovipositioning [3]. Dipterans like the malaria mosquito, Anopheles gambiae, detect their host by following plumes of the host's volatile emissions which contain CO 2 [4]. The role of CO 2 in determining Drosophila behavior in the wild is more complicated. CO 2 was identified as one of the major components of the Drosophila stress odorant released by flies under stressful conditions (dSO) [5]. Other studies have shown that concentrations of CO 2 as low as 0.1% act as a repellant for larval and adult Drosophila [6]. This repulsion can be masked by the presence of low concentrations of food and other odorants, a response presumably mediated by the need to reach fermenting food sources that also exude CO 2 [6,7]. The mechanisms by which Drosophila detects and responds to CO 2 are therefore likely to be complex.
Low concentrations of CO 2 (,10%) are sensed by two receptors, Gr21a and Gr63a, which co-express in the ab1C class of neurons housed in the large basiconic sensilla present on the third antennal segment of Drosophila. Flies lacking either of these receptors lose both electrophysiological and behavioral responses to CO 2 [5,8,9]. The two Drosophila CO 2 receptors, have corresponding homologues in mosquitoes referred to as GPRGR22 and GPRGR24, which co-express in the mosquito maxillary palps [9][10][11]. Thus, understanding the mechanism of sensory transduction downstream of CO 2 receptors is of wide significance. The heterotrimeric G-protein Gaq has been implicated in the transduction of CO 2 stimuli for concentrations of 5% or less [12]. The effectors downstream of Gaq in CO 2 sensing neurons however remain elusive. One of the possible candidates could be the members of the canonical Transient Receptor Potential channel family (TRPC) which, from studies in Drosophila phototransduction have been known to act downstream of Gaq [13].
The TRP superfamily include a large number of cation channels [14] many of which are implicated in the detection and transduction of sensory information across a range of species (reviewed in [15]). In Drosophila, members of this superfamily have been implicated in the detection of a range of sensory stimuli including light [16][17][18], temperature [19][20][21][22], pain [23][24][25] mechanical stimuli [26], taste [27] and chemosensation [28]. Quite recently, a transient receptor potential channel was found to be involved in male-male courtship behavior in Drosophila [29]. In the Drosophila genome, the TRPC subfamily consists of TRP, TRPL and TRPc, encoded by the genes trp, trpl and trpc respectively. Of these the activity of TRP and TRPL are required to generate the light induced conductance in photoreceptors [30,31]. In addition, hypomorphic alleles of trp (trp 301 ) appear to have a defect in adaptation during responses to isoamylacetate and benzaldehyde [32]. In addition to TRP and TRPL, the Drosophila genome encodes a third member of the TRPC subfamily, namely TRPc [33].
In Drosophila photoreceptors, the G-protein coupled receptor rhodopsin transduces photon absorption into the activation of TRP and TRPL channels. This transduction process requires the activity of the Ga subunit Gaq [13]. The activation of PLCb (encoded by norpA) [34] by Gaq is an essential step in the activation of TRP and TRPL. While the subsequent steps in the mechanism of activation remain unresolved (reviewed in [35]), the requirement for G-protein coupled activation of PLCb in TRP and TRPL channel activation can also be recapitulated in heterologous expression systems [36,37]. Although the endogenous receptor and in vivo activation mechanisms of TRPc remain unknown, when expressed in heterologous systems, TRPc is reported to be activated downstream of receptors that trigger G-protein coupled PLC activity [33]. Thus the activation mechanism of Drosophila TRPC channels appears to have a conserved requirement for G-protein coupled PLCb activity. In this study we investigated the possible role of genes encoding TRPC channels in Drosophila CO 2 chemosensation.

Fly Stocks
All flies were maintained at 25uC on standard corn meal agar medium unless specified otherwise. Canton S was used as the wild type strain. Other stocks used were UAS Gq 1F1 RNAi [38], plc21c P319 and Df(2L)p60A obtained from S. Leevers, UK [39], Gr21aGAL4 on 3 rd chromosome received from Barry Dickson (Vienna, Austria), Gr63aGAL4 on 2 nd chromosome, GAL80 ts , Gr63a 1 (null allele of Gr63a), Elav C155 GAL4 on 1 st and UAS RedStinger on 3 rd from Bloomington stock centre, UAS trpl RNAi (VDRC 35571) and UAS trpc RNAi (VDRC 9338) from Vienna Drosophila RNAi Center, UAS H2bRFP [40] from Boris Egger. trp 343 /trp 343 , trpl 302 /trpl 302 are published [41]. The UAS trpl + ) strain was made by Amit Nair as follows. The trpl cDNA has been described earlier [31]. It was obtained as an EcoRI digested fragment from the parent plasmid and sub-cloned, into the Drosophila transformation vector pUAST [42]. Recombinant pUAStrpl + was used for generating stable transformants by standard procedures for microinjection of Drosophila embryos.
Immunohistochemistry UAS H2bRFP was driven in Gr21a receptor expressing cells in order to mark them. Frozen sections of the fly head (14 mm) were taken and stained with antibodies as previously described by Kain et al. [43]. The following primary antibodies were used; chick anti-RFP (1:1000, Millipore), rat anti-TRP (1:20). The antibody against Figure 1. Disruption of plc21C gene leads to impaired CO 2 sensing. Behavior analysis with 3 to 4 days old flies using the Y-maze with 5% CO 2 and air shows reduced CO 2 avoidance in flies homologous for the plc21C insertion allele plc21C P319 (plc21C P319 /plc21C P319 ) and heterologous with plc21C deficiency mutant plc21C p60A (plc21C P319 /plc21C p60A ). Heterozygous controls show normal behavioral avoidance (p,0.0001; two tailed student's t test). Gr63a2/2 flies were used as a negative control. Error bars indicate SEM. doi:10.1371/journal.pone.0049848.g001 Drosophila TRP was generated in house. The C-terminal 300 amino acids of TRP (aa 975-1275) were expressed as a His tagged fusion protein in E. coli and purified using Ni affinity chromatography. Purified antigen was used to immunize rats and generate a polyclonal antiserum. The specificity of the antiserum was tested using both Western blotting as well as immunohistochemistry using the trp 343 null allele as a control. Rabbit anti-TRPL (1:100, catalog number AB5912 from Chemicon international) and rabbit anti-TRPc (1:300, obtained from Shireen A. Davies, University of Glasgow UK; [44]). Monoclonal antibody 22C10 (1:5; DSHB) was used to mark the antennal sensory neurons. Secondary antibodies used were anti-chick, anti-mouse and anti-rabbit IgG conjugated to either Alexa 488 or Alexa 568 (1:200; from Molecular Probes). Labeled samples were mounted in 70% glycerol or in an antifading agent, Vectashield (Vector labs) and examined in Olympus FV1000, at 1 mm slice intervals; data was processed using Image J, Confocal Assistant 4.2 and Adobe Photoshop 5.5. Whole antennal mounts were prepared using Vectashield (Vector labs) after fixing the antennae in 0.4% paraformaldehyde for 10 min followed by two washes in Phosphate buffered saline (PBS) of 10 minutes each. The samples were examined as stated above and the data was processed using FV10-ASW 3.0 viewer and Fiji (Image JA 1.45b) and Adobe Photoshop CS3 Extended.

Electrophysiology
Extracellular field recordings were acquired from the large basiconica rich region on the third antennal segment of the fly antenna [8] using DIGIDATA 1322A 16-Bit Data Acquisition System (Axon Instruments) connected to a DAM 50 Differential amplifier (World Precision Instruments) using borosilicate glass electrodes of 30-35 MV resistance (GC100F-10; Harvard Apparatus Ltd.) containing 0.8% NaCl and a 0.250 mm silver wire (AGW1010; World Precision Instruments). The stimulus was delivered as a 500 ms pulse at a flow rate of 1L per minute. Three different concentrations of CO 2 , 0.5%, 2.5% and 5% were achieved by diluting 100% CO 2 in air and the concentrations were confirmed using a CO 2 sensor (Type-IR-CO 2 gas tester; Heraeus). Air was used as a negative control in addition to being flushed along the delivery tube between each concentration shift to minimize CO 2 accumulation. Flies were allowed to rest for one minute between concentration shifts to avoid adaptation effects. Electroretinograms were recorded from the eyes of flies using 5 s pulses of green light. All traces were analyzed using Clampfit Version 9.0.1.07 software (Axon Instruments). All flies used for electrophysiology were 3 to 4 days old females. A minimum of 10 flies per genotype were tested.

Behavioral Analysis
The Y-maze set up, as described by Das et al. [45], was used to carry out behavioral assays and the Mean avoidance index was calculated as described [9] as the number of flies in the CO 2 arm subtracted from the number of flies in the air arm divided by the total number of flies in both arms. Flies that did not choose either arm were not taken into consideration. The concentration of CO 2 used was 5%. Each experimental set contained 25 to 30 flies of 3 to 4 days of age and ten experimental sets were used per genotype. All genotypes tested were double blinded. The orientation of the arms of the Y-maze was alternated to avoid any side bias.

Data Analysis
Two tailed student's t test was used to compare heterozygous controls with their corresponding homozygous knockout and knock down lines in all molecular, electrophysiological and behavioral experiments.

Relative Quantitation of Gene Expression
250 mg of RNA was extracted from 10 Drosophila 3 rd instar larval brains per sample set with 6 sample sets in total per genotype. Reverse transcription PCR (RT-PCR) was performed as described in [46]. Real-time quantitative PCR (qPCR) was performed on 1:10 dilution of the total cDNA with duplicates per sample set using rp49 primers as internal control and primers specific to the gene of interest (trpc) on the 7500 Fast Real-Time PCR System (Applied Biosystems) operated with 7500 software v2.0.5 using MESA GREEN qPCR TM Master Mix Plus for SYBRH Assay -dTTP (Eurogentec, Belgium).

Experiments with Temperature Sensitive GAL80
Flies of the appropriate genotypes were maintained at the restrictive temperature of 18uC until eclosion and then transferred to the permissive temperature of 29uC when the RNAi was allowed to express. These flies were then used to carry out electrophysiological recordings after ageing for 3 to 4 days.

Results
Phospholipase Cb Encoded by plc21C is Required for Normal Avoidance Behavior of CO 2 The Drosophila genome contains two genes encoding phospholipase Cb referred to as norpA [34] and plc21C [47]. Previous studies have shown that norpA is not required for either behavioral or physiological responses to CO 2 [12]. When two mutants of plc21C, plc21C P319 (an insertion allele) and plc21C p60A (a deficiency line), were tested for their response to 5% CO 2 in a Y-maze behavioral assay both showed reduced avoidance ( Fig. 1, p,0.0001). Canton S (CS) flies were used as positive controls and these showed normal avoidance towards 5% CO 2, while null alleles for the CO 2 receptor, Gr63a, showed complete impairment in CO 2 sensing as demonstrated previously [9] (Fig. 1). As expected, knock down of Gaq in CO 2 sensory neurons (Gr21aGAL4.UASGaq 1F1 ) by a previously tested RNAi construct also reduced the avoidance response of adult Drosophila towards CO 2 [12,38]. The specificity of the behavioral response was further verified by testing the response of Orco null mutants to 5% CO 2 . These flies showed normal levels of avoidance towards CO 2 (as previously observed by Turner et al. [7] Fig. 1). The Orco gene product is a highly conserved atypical member of the olfactory receptor family and serves as a co-receptor for olfactory receptors in Drosophila [48]. It is expressed in a majority of olfactory sensory neurons of the antenna but not in CO 2 sensing neurons [8,48].
Electrophysiological responses were obtained from the region of the third antennal segment housing the large basiconic sensilla containing the ab1C neurons [8]. Both plc21C mutant alleles tested, plc21C P319 /plc21C P319 and plc21C P319 /plc21C p60A showed lowered sensitivity to all the three CO 2 concentrations, thus corroborating the results observed during behavior analysis ( Fig. 2A and B, p,0.0001).
reduced electrophysiological responses to the three concentrations of CO 2 tested as compared to the wild type flies (n = 10, p,0.0001). Gr63a null mutants (Gr63a2/2) and an RNAi knockdown of Gaq in CO 2 sensitive neurons (Gr21aGAL4.UASGaq 1F1 ) were included as test controls. B) Quantification of the field recordings for the genotypes tested (n = 10; p,0.0001). Error bars indicate SEM. doi:10.1371/journal.pone.0049848.g002 From previous studies in Drosophila photoreceptors, it is known that TRPC channels TRP and TRPL are activated by Gaq stimulation of PLCb [13,34]. Therefore, the presence of TRPCs was ascertained in the third segment of adult Drosophila antennae, which host a majority of olfactory sensory neurons including those for CO 2 . The expression of each of these channels was determined in adult Drosophila antennae by immuno-staining with antibodies specific for each TRPC protein. As shown in Fig. 3A, Drosophila TRP, TRPL and TRPc were indeed expressed in the third antennal segment of the fly. Their presence in CO 2 sensory neurons was confirmed by marking the nuclei of these with a Histone2b Red Fluorescent Protein (H2bRFP) fusion construct [40]. Cellular localization of TRPCs appeared to be on cell membranes of neurons with H2bRFP expressing nuclei. Thus the CO 2 receptor neurons of adult Drosophila express TRP, TRPL and TRPc. Null mutants of trp and trpl were used as negative controls to validate the specificity of the antibodies (Fig. 3B).

Null Mutants of trp and trpl Show a Reduced Behavioral Response Towards CO 2
To understand the functional role of TRPCs in olfactory responses to CO 2, protein null mutants in the trp (trp 343 ) and trpl (trpl 302 ) genes were studied. Homozygous trpl 302 when tested for their avoidance to 5% CO 2 in a Y-maze gave a mean avoidance index of just 0.04, as compared to 0.46 obtained for trpl 302 heterozygotes (Fig. 4A, p,0.0001). In order to confirm that the reduced response to CO 2 in trpl 302 flies is indeed due to the mutation in the trpl locus, a wild type trpl transgene [UAS trpl + ] was expressed in the CO 2 receptor neurons of trpl 302 /trpl 302 null flies.
The behavioral avoidance towards CO 2 was restored back to 0.45 in trpl 302 /trpl 302 ;Gr21aGAL4/UAS trpl + animals (Fig. 4A). Interestingly, the behavior of null mutants of trp (trp 343 /trp 343 ) towards 5% CO 2 was also found to be reduced (0.13) although slightly higher than that observed for the trpl null mutants (Fig. 4B, p,0.0001). The behavioral phenotype of trpl 302 /trpl 302 ;trp 343 /trp 343 double mutants was also measured. This was not significantly different from the individual null mutants (Fig. 4B,  In trp and trpl null mutants, the altered behavior towards CO 2 could arise from either a reduction in CO 2 sensing by peripheral sensory neurons or by changes in central brain circuits responsible for the CO 2 avoidance behavior. While rescue by expression of UAS trpl + in the CO 2 sensory neurons suggested that the primary defect was in the periphery, this was further tested by measuring electrophysiological responses, to varying concentration of CO 2, from the antenna. A consistent reduction in the amplitude of electro-antennogram responses of trp 343 /trp 343 and trpl 302 /trpl 302 flies was observed in comparison to wild type and heterozygous controls. Reduced responses were observed for all three concentrations of CO 2 ( Fig. 5A and B, p,0.0001). Expression of the UAS trpl + transgene with Gr21aGAL4 in the trpl 302 /trpl 302 flies rescued the electrophysiological phenotype significantly, further confirming a role for trpl in CO 2 sensory neurons ( Fig. 5A and B). Consistent with the behavioral results, the electrophysiological responses for trpl 302 /trpl 302 ;trp 343 /trp 343 double mutant flies were similar to that of the individual null alleles (Fig. 5A and B). These data suggest that the lowered sensitivity to CO 2 is indeed due to a reduction in CO 2 sensing by the peripheral sensory neurons and A) The mean avoidance index towards 5% CO 2 in a Y-maze behavioral assay is shown for the indicated genotypes. The ability of trpl null homozygotes (trpl 302 /trpl 302 ), to discriminate between 5% CO 2 and air is significantly reduced (p,0.0001) as compared to the heterozygous control. The phenotype of the null mutant is rescued by expressing a wild type trpl transgene in Gr21a receptor neurons (trpl 302 /trpl 302 ; Gr21a GAL4/UAS trpl + ) (p,0.0001). B) Null mutant of trp (trp 343 /trp 343 ) has reduced avoidance to 5% CO 2 in the Y-maze assay. The avoidance response of the double null mutant (trpl 302 /trpl 302 ; trp 343 /trp 343 ) is also reduced but not significantly different from the single null homozygotes (p.0.05). Error bars indicate SEM in A and B. doi:10.1371/journal.pone.0049848.g004 not by changes in central brain circuits. It is also evident from the data presented that TRP and TRPL are not the only channels that function in response to CO 2 in Gr63a and Gr21a positive sensory neurons.
To confirm the genotypes of the TRPC mutants, electroretinogram responses (ERGs) were measured as described in materials and methods. For each genotype, the responses obtained were similar to the published data where it has been shown that a null mutant of trp shows only a transient response to prolonged light stimulus and a null mutant of trpl has oscillations superimposed on its response (Fig. 5C) [49]. Importantly, there was no response seen in trpl 302 /trpl 302 ;trp 343 /trp 343 to the light stimulus [31]. In contrast, the residual responses to CO 2 observed in trpl 302 /trpl 302 ;trp 343 / trp 343 animals suggest that the physiological role of the two TRPC channels, TRP and TRPL, in CO 2 sensing neurons is different from what has been observed in the photoreceptors [31,49].

Down-regulation of trpc and trpl in CO 2 Receptor Neurons Leads to Impaired CO 2 Sensing
Next the effect of down-regulating TRPc, the third TRPC channel in Drosophila was assessed on CO 2 driven behavior and electrophysiology. For this purpose we used the Gr63a GAL4 strain to drive expression of UAS driven RNAi lines for trpc and trpl, so as to knock down these genes specifically in CO 2 sensory neurons. Flies with down-regulation of either trpc or trpl in Gr63a expressing neurons were relatively indifferent to 5% CO 2 (Fig. 6A). In both cases the responses were significantly different from the controls (p,0.0001). The mean avoidance index of trpl knockdown flies was 0.1 while trpc was 0.15. These values are comparable to the avoidance index of trpl null mutants in figure 1. In all cases the avoidance index of controls was equal to or greater than 0.45. In order to validate the RNAi line for trpc, qRT-PCR was carried out on RNA extracted from third instar larval brain samples of the UAS trpc RNAi line driven by a pan neuronal GAL4 (Elav C155 GAL4) as described in the materials and methods. The RNAi line for trpc showed ,45% reduction for the trpc cDNA when compared to its control (Fig 6B, p,0.05). Direct validation of the efficacy of the RNAi line was not possible in the CO 2 receptor neurons due to their low count (25-35 neurons) within the antennae.
Electrophysiological field recordings from trpc and trpl knockdown strains confirmed their inability to sense CO 2 at the same sensitivity as wild-type or control flies. (Fig. 6C, D). At all three concentrations of CO 2 , electrophysiological responses were significantly reduced (p,0.0001). Thus the ability of adult Drosophila to sense and respond to CO 2 in the environment depends to a significant extent on the three TRPC channels, TRP, TRPL and TRPc. Figure 5. Electrophysiological responses to various concentrations of CO 2 obtained from antennae of wild type and mutants. The response from trpl null homozygotes (trpl 302 /trpl 302 ), trp null homozygotes (trp 343 /trp 343 ) and the double null mutants (trpl 302 / trpl 302 ;trp 343 /trp 343 ) appear reduced towards all three CO 2 concentrations tested as compared to wild type (+/+) responses. A) Representative traces of field recordings obtained as described above. B) Quantification of the field recordings for various mutant and control genotypes tested. Both trpl null (trpl 302 /trpl 302 ) and trp null mutants (trp 343 /trp 343 ) along with their double mutant (trpl 302 /trpl 302 ; trp 343 / trp 343 ) show reduced electroantennogram responses (n = 10; p,0.0001). The reduced response of the trpl null is rescued by the expression of wild type TRPL in Gr21a receptor neurons (n = 10; p,0.0001). Gr63a2/2 served as the negative control with no EAG response. C) Photoresponses of the indicated trpl and trp mutants. n = 10 for all genotypes shown. Error bars indicate SEM. doi:10.1371/journal.pone.0049848.g005

Reduced Responses to CO 2 is not a Developmental Defect
To test if the reduced sensitivity to CO 2 is a consequence of TRPC channel function in adult sensory neurons or due to unidentified developmental changes in the olfactory circuit for CO 2 , expression of the UAS RNAi lines for trpl and trpc was limited to adult sensory neurons with the help of a temperature sensitive GAL80 transgene which renders GAL4 inactive at 18uC. At the non-permissive temperature of 29uC the GAL80 can be inactivated thus enabling GAL4 to drive the RNAi [50]. Electrophysiological responses of flies in which UAS trpl RNAi and UAS trpc RNAi were expressed after eclosion showed reduced responses to CO 2 as compared to controls and flies grown exclusively at 18uC (Fig. 7A, p,0.0001). These data confirm that reduced CO 2 responses in flies with RNAi knockdown of trpl and trpc occurs due to the reduction of the individual TRPC proteins in adult antennal sensory neurons and is not due to developmental changes.
Furthermore, adult antennal CO 2 sensory neurons were quantified in different mutant backgrounds (plc21C P319 /plc21C P319 and trpl 302 /trpl 302 ) by driving UAS RedStinger in Gr21a receptor neurons to mark their nuclei. The CO 2 sensory neuron counts were found to be within the normal range (approximately 25 to 30 neurons) [5,8] and similar to the wild type control ( Fig. 7B and C, Figure 6. Down regulation of trpl and trpc in CO 2 receptor neurons results in reduced sensitivity to CO 2 as observed in the responses from the Y-maze behavioral assay with 5% CO 2 in A (p,0.0001). B) qRT-PCR data showing the fold change of trpc gene expression in the UAS trpc RNAi line relative to its control as determined by the comparative DDCt method (N = 6; p,0.05). C) Representative traces of field recordings obtained as described above. Individual genotypes are indicated. D) Quantification of the field recordings for the genotypes tested (n = 10; p,0.0001). Error bars indicate SEM. doi:10.1371/journal.pone.0049848.g006 p value not statistically significant). These observations show that reduced CO 2 sensing by the various mutant lines is not due to a reduction in CO 2 sensing neurons during development. These observations further implicate TRPC channels as components that determine the high sensitivity of adult Drosophila CO 2 sensory perception.

Discussion
The role for TRPC channels in maintaining the high sensitivity of CO 2 detection is important in multiple contexts. Detection of low concentrations of CO 2 (5% or less) shares several similarities with odor detection. Receptors for low concentrations of CO 2, despite belonging to the gustatory class of insect chemosensory receptors, are located within olfactory sensillae on the third antennal segment. Moreover, mutants in dgq, the gene that encodes the a subunit of the heterotrimeric G-protein Gaq, reduce the physiological response recorded from sensory neurons in both cases [12,43]. We now show that mutants of the ubiquitously expressed allele of PLCb, plc21C [47] reduce the response to CO 2 similar to the observation for odors [43] unlike mutants of norpA allele which is expressed strongly in the eyes and is required for phototransduction [34] but not for CO 2 sensing [12]. In olfactory sensory neurons it has been proposed that the physiological response to odorants is a combination of ionotropic and metabotropic receptor signaling. The olfactory receptor and olfactory co-receptor (Or/Orco) complex forms an odor-activated ion channel [51,52] in heterologous systems and is therefore thought to be an ionotropic component, while the olfactory receptor coupling to a G-protein, like Gaq, could initiate the metabotropic component through as yet un-determined ion channels. Unlike olfactory sensory neurons, ab1C, the CO 2 sensing neurons do not express the olfactory receptor and olfactory co-receptor (Or/Orco) complex. Therefore, in these neurons it is possible that the ionotropic component is absent. Our data suggest that TRPCs, which are known to function downstream of Gq/ Plcb signaling [13,33,34,36,37] may contribute to metabotropic signaling in ab1C neurons but our data does not allow us to state this conclusively. However it is evident that the TRPC channels are required for the normal functioning of CO 2 sensing ab1C neurons in adult Drosophila. The presence of a basal response in individual knock outs and knock downs of trp, trpl and trpc and double null mutants of trp and trpl as compared to the complete lack of response in Gr63a null flies suggests that the CO 2 sensing ab1C neurons are not solely dependent on the TRPC channels for function. While it is formally possible that the remaining response in trpl 302 ;trp 343 double nulls is due to trpc, we do not favor this idea primarily because, the response of double mutant nulls was no worse than that of single mutants. The triple mutant combination of trpl 302 ;trp 343 with the trpc RNAi line was poorly viable and hence could not be tested directly.
The consequences of this finding are relevant for Drosophila behavior. Unlike other insect species like moths and mosquitoes, Drosophila are innately repelled by low concentrations of CO 2 presumably because it is an indicator of stress due to a potential threat to naïve flies. However, in conditions where CO 2 is present along with food odorants this repulsion needs to be suppressed. Our data suggest that TRPC channels are a component of this dual sensitivity. Repression of Gq/PLCb signaling and/or TRPCs through mechanisms yet to be identified might reduce the sensitivity to CO 2 and alter the behavior from repulsion to attraction. Interestingly, food odors that can reduce CO 2 responses from ab1C neurons have been identified [7]. Whether these odorants act through repression of TRPCs needs to be Figure 7. Reduced sensitivity to CO 2 is not a developmental defect. A) RNAi lines grown at the restrictive temperature of 18uC (active GAL80) show normal electrophysiological responses to CO 2, since the CO 2 receptor neuron specific GAL4 remains inactive (absence of RNAi expression; p.0.05). RNAi lines grown at the permissive temperature of 29uC (inactive GAL80) show reduced electrophysiological responses to CO 2 due to active GAL4 and RNAi expression (n = 10; p,0.0001). The RNAi heterozygotes in the absence of Gr63aGAL4 show normal responses to CO 2 at 29uC. Error bars indicate SEM. B) Whole antennal mounts showing CO 2 sensory neurons marked using UAS RedStinger driven by Gr21aGAL4 in wild type, plc21C P319 /plc21C P319 and trpl 302 /trpl 302 mutant lines. C) Quantification of CO 2 sensory neurons in the adult antennae of the same lines (n = 14; p value not statistically significant). doi:10.1371/journal.pone.0049848.g007 determined. Thus it appears that the three Drosophila TRPC channels TRP, TRPL and TRPc can act as amplifiers of the signal downstream of a channel yet to be identified while playing redundant roles in this amplification process. The requirement for redundancy might stem from an evolutionarily conserved need to escape stress and or the necessity to find food.