Articles, Cellular/Molecular

Role of G-Proteins in Odor-Sensing and CO2-Sensing Neurons in Drosophila

C. Andrea Yao and John R. Carlson
Journal of Neuroscience 31 March 2010, 30 (13) 4562-4572; https://doi.org/10.1523/JNEUROSCI.6357-09.2010

Abstract

A central question in insect chemoreception is whether signaling occurs via G-proteins. Two families of seven-transmembrane-domain chemoreceptors, the odor (Or) and gustatory receptor (Gr) families, have been identified in Drosophila (Clyne et al., 1999, 2000; Vosshall et al., 1999). Ors mediate odor responses, whereas two Grs, Gr21a and Gr63a, mediate CO2 response (Hallem et al., 2004; Jones et al., 2007; Kwon et al., 2007). Using single-sensillum recordings, we systematically investigate the role of Gα proteins in vivo, initially with RNA interference constructs, competitive peptides, and constitutively active Gα proteins. The results do not support a role for Gα proteins in odor sensitivity. In parallel experiments, manipulations of Gαq, but not other Gα proteins, affected CO2 response. Transient, conditional, and ectopic expression analyses consistently supported a role for Gαq in the response of CO2-sensing neurons, but not odor-sensing neurons. Genetic mosaic analysis confirmed that odor responses are normal in the absence of Gαq. Gγ30A is also required for normal CO2 response. The simplest interpretation of these results is that Gαq and Gγ30A play a role in the response of CO2-sensing neurons, but are not required for Or-mediated odor signaling.

Introduction

The role of G-proteins is a fundamental issue in insect chemosensory signaling. Chemoreception in vertebrates and Caenorhabditis elegans utilizes seven-transmembrane-domain receptors that signal via G-proteins. Insects also have large families of odor receptor (Or) and gustatory receptor (Gr) proteins that are predicted to contain seven transmembrane domains; however, Or and Gr proteins lack sequence similarity to G-protein-coupled receptors (GPCRs), and the topology of Ors differs from that of GPCRs (Neuhaus et al., 2005; Benton et al., 2006; Lundin et al., 2007).

The role of G-proteins in chemosensory signaling has been examined in Drosophila, but a clear consensus has not been reached. Studies suggest that Gαs and Gγ1 play a role in the response of Gr5a-expressing taste neurons (Ishimoto et al., 2005; Ueno et al., 2006). The requirement for G-proteins in olfaction is unclear. Two recent studies found that an Or and a ubiquitously expressed coreceptor, Or83b, form a ligand-gated ion channel (Sato et al., 2008; Wicher et al., 2008). However, the studies proposed different mechanisms for activation of the channel complex. One study found that G-protein-mediated signaling plays a negligible role in receptor activation (Sato et al., 2008), whereas the other found that the Or complex is activated by cyclic nucleotides and acts as a GPCR (Wicher et al., 2008). Other studies have reported that olfactory receptor neurons (ORNs) require G-proteins: mutations in Gαq were found to reduce the sensitivity of antennal neurons to several odors (Kain et al., 2008), and expression of Gαq RNA interference (RNAi) in ORNs led to decreased odor-evoked behavioral responses (Kalidas and Smith, 2002).

Whereas most antennal ORNs express Ors, which mediate responses to odors, one class of ORNs in the antenna expresses two Grs that mediate a response to CO2 (Jones et al., 2007; Kwon et al., 2007). Many insect species have a similar class of ORNs dedicated to CO2 sensation (Grant et al., 1995; Lu et al., 2007; Syed and Leal, 2007). However, the mechanism of CO2 transduction in insects is unknown.

In this study, we investigate whether G-proteins are required for either Or- or Gr-mediated responses in the antenna. We disrupted signaling of each Drosophila Gα and Gγ protein individually in vivo, via several independent means, and generated null clones of Gαq, Gαo, and Gαs. We also investigated the Gγ subunits. We found that individually disrupting each Gα and Gγ had little or no effect on the Or-mediated responses tested. However, disruption of Gαq or Gγ30A decreased the Gr-mediated CO2 response. Using mosaic analysis, conditional and ectopic expression analyses, and transient expression of constitutively active Gα proteins, we show that Gαq is necessary for both the development and function of the CO2-sensing neuron. Our results support a model in which Gαq and Gγ30A play a role in CO2 response but are not required for Or-mediated olfactory transduction.

Materials and Methods

Drosophila stocks and transgenes.

To generate DGαq-GAL4, 5.1 kb of DNA upstream of the ATG of Gαq-3 was inserted into the pGAL4 vector. The primers used were 5′-AGCGACAAACTGGGGAACAACTG and 5′-GCTAACTAGATTTATAGACGCTTTCGCGTG. To generate the UAS-Gα peptide constructs, PAGE-purified, phosphorylated complementary primers were annealed and cloned into an upstream activation sequence (UAS) expression vector. Each peptide consists of a start codon, a glycine, which has been shown to stabilize the peptide (Gilchrist et al., 2002) and the last 11 amino acids of the specified Gα. Primer sequences were 5′-AATTCATGGGACTGCAATCGAACCTTAAGGAATATAATTTGGTCTAAGC and 5′-GGCCGCTTAGACCAAATTATATTCCTTAAGGTTCGATTGCAGTCCCATG (Gαq), 5′-AATTCATGGGACAAAGGATGCACCTTCGTCAATATGAATTGTTATAGGC and 5′-GGCCGCCTATAACAATTCATATTGACGAAGGTGCATCCTTTGTCCCATG (Gαs), 5′-AATTCATGGGACTCTCGGAGAACGTGTCCAGCATGGGCCTATTCTAGGC and 5′-GGCCGCCTAGAATAGGCCCATGCTGGACACGTTCTCCGAGAGTCCCATG (Gα73B), 5′-AATTCATGGGAATAGCAAACAACCTGCGCGGCTGTGGACTGTACTAAGC and 5′-GGCCGCTTAGTACAGTCCACAGCCGCGCAGGTTGTTTGCTATTCCCATG (Gαo), 5′-AATTCATGGGAATCAAGAACAATCTGAAACAAATTGGCTTATTCTGAGC and 5′-GGCCGCTCAGAATAAGCCAATTTGTTTCAGATTGTTCTTGATTCCCATG (Gαi), and 5′-AATTCATGGGAGTGATTAATATCTCTCTAAACCGGAACTGGATATAAGC and 5′-GGCCGCTTATATCCAGTTCCGGTTTAGAGAGATATTAATCACTCCCATG (Gαq scrambled).

The Gr21a-mCD8-GFP line was from Barry Dickson (Research Institute of Molecular Pathology, Vienna, Austria). The UAS-DGαs RNAi line was from Michael Forte (Vollum Institute, Portland, OR). Gαq221C, UAS-Gαq1F1 RNAi, UAS-Gαq, UAS-GαqGTP, and Gαq-GAL4 lines were from Gaiti Hasan (Tata Institute of Fundamental Research, Bangalore, India). The UAS-GαiGTP line was from Fumio Matsuzaki (RIKEN, Kobe, Japan). The norpA24 line was from William Pak (Purdue University, West Lafayette, IN). The Gr21a-GAL4 line was from Kristin Scott (University of California, Berkeley, Berkeley, CA). Or83b-GAL4 and UAS-Gαq RNAi lines were from Dean P. Smith (University of Texas Southwestern, Dallas, TX). The P{GawB}NP1535 precise excision line was from Teiichi Tanimura (Kyushu University, Fukuoka, Japan). UAS-GαoGTP, Gαo007, and UAS-ptx lines were from Andrew Tomlinson (Columbia University, New York, NY). The concertinaRC10 line was from Eric Wieschaus (Princeton University, Princeton, NJ). Gαq1 and inaD lines were from Charles Zuker (University of California, San Diego, La Jolla, CA). The UAS-Gαo RNAi, UAS-Gα73B RNAi, UAS-Gαs RNAi, UAS-Gγ30A RNAi, and UAS-Gγ1 RNAi lines were obtained from the National Institute of Genetics Fly Stock Center (Shizuoka, Japan). UAS-Gαs, UAS-GαsGTP, yw; PNP1535/CyO, and yw; P{GawB}NP6216/CyO lines were obtained from the Drosophila Genetic Resource Center (Kyoto, Japan). Gβ76C1, tub-GAL80ts, ey-FLP, FRT 42B, GαsR60, FRT 42B, tubP-GAL80/Cyo, FRT 42D, tubP-GAL80/Cyo, UAS-DsRed, UAS-GFP, PBac{RB}Ggamma30Ae00834, hs-GAL4, GαsR60, and Df(2L)ED678/Cyo lines were obtained from the Bloomington Drosophila Stock Center (Bloomington, IN). Or22a-GAL4, UAS-Gr63a, UAS-Gr21a, and Δhalo lines, in which Or22a and Or22b are deleted, were described previously (Dobritsa et al., 2003; Kwon et al., 2007).

Isolation of a Gαq mutant.

The Gαq28 deletion line was generated by P-element excision of the yw; P{GawB}NP6216/CyO strain, which carries the P{GawB} insertion element in the 5′ UTR of Gαq. The P-element was mobilized using Δ2–3. Approximately 400 w− flies were screened by PCR for deletions in the Gαq coding region. Only one mutant, Gαq28, had a deletion that extended into the coding region. The mutation was mapped and confirmed by single-egg PCR analysis using the following primers: 5′-GCTGAACTGCCAGGAA ATCCG and 5′-TTTAAATCTACAAACATTCGCAGT.

Electrophysiology.

All single-unit recordings were performed as described previously (de Bruyne et al., 2001). All recordings were from female antennal sensilla with the exception of the mosaic analysis with a repressible cell marker (MARCM) and empty neuron experiment recordings, which were from males and females. Age- and gender-matched flies were used for the control recordings. All flies tested had constructs that were either injected or backcrossed into w1118 flies, ensuring a similar genetic background. Flies expressing the appropriate GAL4 construct alone were used as the control in all experiments except for the antennal basiconic 1C (ab1C) Gα peptide experiments. Gα peptide constructs were injected into a wild-type strain that appears to have a low CO2 response. Therefore, flies carrying the UAS-Gα peptide alone were used as the control in these experiments. Methyl butyrate, ethyl acetate, and butyric acid were obtained from Sigma. Methyl butyrate and ethyl acetate were dissolved in paraffin oil, whereas butyric acid was dissolved in water. A 500 ms pulse of odorant was puffed into a continuous air stream that was directed at the antenna. CO2 recordings were performed essentially as described previously (Kwon et al., 2007). Briefly, a 25 ml/s air stream was replaced for 500 ms by a 25 ml/s flow of 0.1, 0.5, 1, 2, 5, or 20% CO2 dissolved in N2 (Airgas). The firing rates for both olfactory and CO2 recordings were calculated by counting the number of spikes fired during the 500 ms stimulus period and multiplying by two. Neither background nor control pulse firing rates were subtracted. Statistical analysis of the methyl butyrate and butyric acid responses was performed using two-tailed Student's t test.

Immunohistochemistry.

w; Gq-GAL4/Gr21a-mCD8-GFP; UAS-DsRed and w; DGq-GAL4/Gr21a-mCD8-GFP; UAS-DsRed flies were used to localize Gαq. Sixteen micrometer sections of the antenna were fixed in 4% paraformaldehyde, washed extensively in PBS plus 0.3% Triton-X (PBS-T), and fixed in PBS-T plus 5% normal goat serum. The primary antibodies used were anti-DsRed (1:200; catalog #632496; Clontech) and anti-green fluorescent protein (GFP; 1:200; catalog #11814460001; Roche). Secondary antibodies used were anti-rabbit IgG conjugated to Alexa Fluor 568 (1:250; A10042; Invitrogen) and anti-rabbit IgG conjugated to Alexa Fluor 488 (1:500; A11001; Invitrogen). Samples were mounted in Vectashield (Vector Laboratories) and viewed on a Bio-Rad MRC 1024 confocal microscope.

Generation of null clones.

Gαq, Gαo, and Gαs null clones were generated using the MARCM system (Lee and Luo, 2001). ey-FLP was used to generate clones in the antenna (Hummel et al., 2003). The Gr21a-GAL4 reporter was used to identify clones including a CO2 neuron, whereas the Or22a-GAL4 reporter was used to identify clones including an ab3A neuron. The control recordings in the MARCM experiments were taken from GFP− neurons on the same antennae as the GFP+ neurons as a precaution against genetic background and age effects.

Flies of the following genotypes were tested: (1) w; FRT 42B, Gαq28/FRT 42B, tubP-GAL80; X-GAL4, UAS-GFP/ey-FLP (Gαq); (2) w; FRT 42D, GαsR60/FRT 42D, tubP-GAL80; X-GAL4, UAS-GFP/ey-FLP (Gαs); (3) w; FRT 42D, Gαo007/FRT 42D, tubP-GAL80; X-GAL4, UAS-GFP/ey-FLP (Gαo); (4) w; FRT 42B/FRT 42B, tubP-GAL80; X-GAL4, UAS-GFP/ey-FLP (control).

Flies of the following genotype were used for the Gαs rescue experiment: UAS-Gαs; FRT 42D, GαsR60/FRT 42D, tubP-GAL80; X-GAL4, UAS-GFP/ey-FLP. These flies express Gαs only in ab1C or ab3A that are genotypically null for Gαs.

Using MARCM, we were able to reliably produce Gαq−/− ab3A neurons, Gαs−/− ab3A neurons, and Gαs−/− ab1C neurons at frequencies of approximately five GFP+ neurons per antenna. However, despite using two different Gαq deletion alleles, Gαq221C and Gαq28, in combination with each of two different markers of ab1C neurons, Gr21a-GAL4 and Gr63a-GAL4, few if any GFP+ ab1C neurons were generated. In two cases we observed no GFP+ ab1C neurons, in the third case we found a single GFP+ ab1C neuron in 74 antennae examined, and in the fourth case we observed a very low frequency of GFP+ ab1C neurons. Furthermore, the few GFP+ ab1C neurons generated do not resemble any of the GFP+ cells in the other experiments. These labeled ab1C neurons have much weaker GFP fluorescence, and the fluorescence could often be observed only in the cell body. Because of their low frequency and very weak GFP fluorescence, we are not confident that these GFP+ ab1C neurons are Gαq−/−.

Spatiotemporal expression of Gαq inhibitors.

Temporal and regional gene expression targeting (TARGET) was used to express UAS-GαqGTP or UAS-Gαq1F1 RNAi selectively in adult flies (McGuire et al., 2003). w; UAS-X; Gr21a-GAL4, tub-GAL80ts flies were raised at 18°C and transferred to 29°C 1 day after eclosion. Genetically identical, age-matched flies that were maintained at 18°C rather than transferred to 29°C were used as the control. As an additional guard against leaky GαqGTP or Gαq1F1 RNAi expression, we examined the antennae of 20-d-old adult w; UAS-GFP; Gr21a-GAL4, tub-GAL80ts flies raised and maintained at 18°C. These flies did not show any GFP activity, confirming that tub-GAL80ts suppresses Gr21a-GAL4 at the permissive temperature (data not shown). Finally, we compared the responses of Gr21a-GAL4 flies that carried a UAS construct to those of controls that did not carry a UAS construct. Flies that do not have the UAS construct had similar CO2 responses to the control flies that were not heat shocked.

hs-GAL4 and UAS-GαqGTP or UAS-GαsGTP were used to transiently express GαqGTP or GαsGTP in adult flies. Seven-day-old adult flies reared at 18°C were heat shocked for 30 min at 37°C and allowed to recover for 1 h at 25°C. Siblings that were placed at 25°C for 1.5 h were used as controls.

Statistical analysis.

Each ORN was tested with multiple concentrations of each ligand. Therefore, we compared the responses from different genotypes using a two-way ANOVA with repeated measures in one factor followed by a Holm–Sidak post hoc test to compare individual concentrations. The between-group factor is the genotype of the fly, and the repeated measure is the concentration of the ligand. Two-tailed t tests were used to compare the spontaneous firing rates of neurons.

Results

Knockdown of Gαq reduces CO2 response but not odor response

As a first step toward testing the role of G-proteins in olfactory responses, we identified the Gα genes that are expressed in the antenna. The Drosophila genome contains eight Gα genes. We found evidence for antennal expression of six of them, concertina, Gαq, Gαs, Gα73B, Gαo, and Gαi, by reverse transcriptase (RT)-PCR analysis of adult antennal RNA (data not shown).

Drosophila antennae contain three functional types of large antennal basiconic (ab) sensilla, ab1, ab2, and ab3 (Fig. 1a,b). ab2 contains two ORNs, ab2A and ab2B, and ab3 contains ab3A and ab3B (de Bruyne et al., 2001). The odor response of each of these neurons is mediated by an Or unique to that ORN and Or83b, a broadly expressed coreceptor (Hallem et al., 2004; Larsson et al., 2004). ab1 contains four ORNs. Three of the neurons, ab1A, ab1B, and ab1D, have odor responses that are also mediated by a unique Or and Or83b. The other neuron, ab1C, responds uniquely to CO2. This neuron does not express Or83b; instead, the CO2 response is mediated by two Grs, Gr21a and Gr63a (Jones et al., 2007; Kwon et al., 2007).

Figure 1.
Figure 1.

Experimental systems. a, Drosophila head. The arrowhead indicates the antenna. b, Schematic of antenna. Different colored dots represent different functional types of basiconic sensilla. c, Disruption of G-proteins in the endogenous CO2 neuron. d, Heterologous expression of the CO2 receptors with and without G-proteins. e, Sample recording traces from control and Gαq RNAi-expressing ab1C. Panel a was adapted from Carlson (1996), b was adapted from van der Goes van Naters and Carlson (2006), and c and d were adapted from Kwon et al. (2007).

To investigate the roles of the six Gα proteins in olfactory and CO2 transduction in the antenna, we focused on responses from three Or-expressing ORNs, ab1A, ab2A and ab3A, and the Gr-expressing ORN, ab1C (Fig. 1b). To test whether disruption of Gα affects the function of the ORNs, we performed extracellular electrophysiological recordings (Clyne et al., 1997; de Bruyne et al., 2001) that allowed us to measure the response of a single ORN to either odors or CO2. Among the six Gα genes expressed in the antenna, a homozygous viable mutant was available for one, concertina. We tested this mutant by electrophysiological recording and found that ab1A, ab2A, ab3A, and ab1C responded normally to broad concentrations of odorants and CO2 (n = 6; p > 0.5 for each neuron class tested; data not shown). Therefore, concertina was not investigated further.

We then performed genetic manipulations that resulted in either loss of function or gain of function to investigate the roles of the five remaining Gα genes. We used either Or83b-GAL4 (Kalidas and Smith, 2002) or Gr21a-GAL4 (Scott et al., 2001) to drive expression of RNAi constructs or competitive peptides (loss of function), or constitutively active Gα proteins (gain of function) (Fig. 1c). Since Or83b-GAL4 is expressed exclusively in Or-expressing ORNs and Gr21a-GAL4 is expressed solely in ab1C, we could ask whether the different Gα proteins function in ORNs as opposed to other cell types.

First, we examined the requirement of G-proteins in ORNs by knocking down Gα activity with available Gα RNAi constructs. We found that RNAi against Gαq, Gαs, Gαo, and Gα73B did not reduce the odor responses of ab1A, ab2A, or ab3A (Fig. 2a). Our results might seem to contradict a previous report showing that flies expressing this Gαq RNAi construct had decreased behavioral responses to odors (Kalidas and Smith, 2002). One possible explanation is that Gαq is required for behavioral response in a process that occurs downstream of ligand recognition and transduction, perhaps by coupling to other GPCRs. For instance, recent studies have found that ORNs express GABAB receptors at the axon terminals and that expression of a Gαo inhibitor alters synaptic transmission between the ORN and second-order neurons (Olsen and Wilson, 2008; Root et al., 2008).

Figure 2.
Figure 2.

RNAi. a, RNAi against Gαq, Gαs, Gαo, and Gα73B did not decrease odor responses of ab1A, ab2A, or ab3A (p > 0.05), ORNs whose responses are mediated by Ors (Dobritsa et al., 2003; Hallem et al., 2004; Larsson et al., 2004; Couto et al., 2005). b, Gαq RNAi decreased response to CO2 (p < 0.001), which is mediated by Gr21a/Gr63a (Jones et al., 2007; Kwon et al., 2007). Overexpression of Gαq in a Gαq RNAi background rescued the CO2 response. RNAi against Gαs, Gαo, and Gα73B did not affect CO2 response (p > 0.05). n = 6–17. Error bars indicate SEM.

A second report showed that a different Gαq RNAi construct, Gαq1F1, decreased the electrophysiological odor response of ab2A (Kain et al., 2008). We also tested Gαq1F1 RNAi using the same broad odorant concentration range used to test the previous Gαq RNAi construct and, consistent with our previous results, we did not observe a decrease in odor response in ab1A, ab2A, or ab3A (n = 6; p > 0.1 for each neuron class; data not shown). The discrepancy with the other study may be attributable to the use of different drivers: we used Or83b-GAL4, whereas the previous analysis of ab2A used Gαq-GAL4, which expresses in non-neuronal support cells as well as in ORNs (Talluri et al., 1995; Kain et al., 2008). Therefore, the reduction in olfactory sensitivity in the previous study could have been caused by effects on cells other than ORNs.

In contrast, Gαq RNAi, but not other Gα RNAi constructs, produced a major decrease in CO2 response (Figs. 1e, 2b). The CO2 response was significantly reduced across a wide response range (Fig. 2b). If this reduction in CO2 response is in fact caused by an RNAi-mediated reduction in levels of Gαq, then overexpression of Gαq RNA should rescue the response. Indeed, overexpression of Gαq-3 RNA, the antennal Gαq splice variant (Talluri et al., 1995; Ratnaparkhi et al., 2002; Kain et al., 2008), restored the CO2 response in flies expressing Gαq RNAi (Fig. 2b). To test whether CO2 response depends specifically on the antennal Gαq splice variant, we tested flies with a missense mutation that only affects the other major Gαq splice variant, Gαq-1 (Scott et al., 1995), and found no defect in response to concentrations of CO2 ranging from 0.1 to 20% (n = 9; p > 0.1; data not shown). We also tested the second Gαq RNAi construct, Gαq1F1, for its effects on CO2 response and obtained results similar to those with the first Gαq RNAi construct. Gαq1F1 RNAi decreased CO2 response across a wide concentration range; at concentrations >0.1% there was a 25–30% decrease in CO2 response (n = 12; p < 0.001; data not shown). Again, expression of Gαq-3 RNA rescued the defect (n = 9; data not shown).

Recent studies have suggested that Gαs may play a role in chemosensory transduction (Wicher et al., 2008). Therefore, we tested an additional Gαs RNAi construct, DGαs RNAi, with the same broad concentration range of odorants. Expression of DGαs RNAi in Gr5a-expressing taste neurons reduced Gαs expression levels and reduced electrophysiological responses to sugar (Ueno et al., 2006). However, consistent with our results with the other Gαs RNAi construct, DGαs RNAi did not decrease the odor or CO2 responses of ab1A, ab2A, ab3A, or ab1C (n = 6–12; data not shown). Taken together, our RNAi data suggest that Gαq is required for CO2 response, whereas no Gα proteins are required for Or-mediated responses.

Competitive Gαq peptide reduces CO2 response but not odor response

As a complementary approach to RNAi, we next used competitive peptides to knock down Gα protein activity. Peptides consisting of the 11 C-terminal amino acids of a Gα protein have been shown to bind to GPCRs and compete with the endogenous Gα (Hamm et al., 1988; Gilchrist et al., 1999). Such peptides do not activate downstream signaling pathways and have been shown to decrease the receptor-mediated response to ligands in cell culture and in transgenic mice (Rasenick et al., 1994; Akhter et al., 1998).

We generated UAS lines that express the 11 C-terminal amino acids of Gαq, Gαs, Gαo, Gα73B, and Gαi, and expressed each using Or83b-GAL4 or Gr21a-GAL4. Consistent with the RNAi results, none of the peptides affected the Or-mediated ab1A, ab2A, and ab3A responses (n = 6; p > 0.05 for each neuron class; data not shown). However, the Gαq peptide decreased ab1C response to concentrations of CO2 ranging from 0.5 to 20% (n = 24; p < 0.001; data not shown). Also consistent with the RNAi results, Gαs, Gαo, Gα73B, and Gαi peptides did not affect the ab1C response to any of the five CO2 concentrations tested (n = 9–21; p > 0.05; data not shown). As an additional control for the specificity of the Gαq peptide, we generated a scrambled Gαq peptide that has the same amino acid composition as the Gαq peptide, but in a random order. Flies expressing the Gαq scrambled peptide responded to CO2 at levels comparable to control flies at all concentrations tested (n = 15; p > 0.05; data not shown).

Constitutively active Gαq reduces CO2 response

The previous two methods sought to knock down Gα protein signaling. We next used a gain-of-function manipulation: expression of a constitutively active form of a Gα protein, GαGTP. GαGTP proteins contain a single point mutation that renders them unable to hydrolyze GTP and, therefore, constitutively active (Bourne et al., 1991). If a certain Gα protein mediates olfactory or CO2 transduction, constitutive activation of the G-protein should mimic prolonged exposure to odors or CO2, resulting in ORN adaptation and subsequent decrease in sensitivity.

GαqGTP nearly abolished CO2 response at all tested doses (Fig. 3b). The decreased CO2 response is not attributable to cell death: ab1C expressing GαqGTP and GFP appear morphologically normal, as observed in confocal images (data not shown). The severe reduction of CO2 sensitivity after long-term expression of GαqGTP is consistent with the notion that Gαq mediates CO2 transduction. Mutations in human rhodopsin that result in a constitutively active opsin protein lead to photoreceptor adaptation and loss of cell sensitivity (Jin et al., 2003). Constitutive activation of ab1C may result in a similar loss of CO2 sensitivity.

Figure 3.
Figure 3.

Constitutively active Gα proteins. a, GαqGTP decreased the odor responses of ab1A, ab2A, and ab3A (p ≤ 0.01), whereas GαsGTP, GαoGTP, and GαiGTP did not (p > 0.05). b, GαqGTP dramatically decreased CO2 response (p < 0.001), whereas GαsGTP, GαoGTP, and GαiGTP had no effect (p > 0.05). n = 6–15. Error bars indicate SEM.

The effect of GαqGTP is specific: expression of GαsGTP, GαoGTP, and GαiGTP did not affect CO2 response (Fig. 3b). Thus, the finding that CO2 response is specifically affected by GαqGTP is in agreement with previous results.

Odor responses of ab1A, ab2A, and ab3A were not reduced by expression of GαsGTP, GαoGTP, or GαiGTP (Fig. 3a), consistent with our previous data. However, GαqGTP did reduce the odor responses of ab1A, ab2A, and ab3A (Fig. 3a). Notably, the effect of GαqGTP on CO2 response is much more severe than on odor responses. To better understand the different effects of GαqGTP on the Or- and Gr-mediated responses in these long-term expression studies, we next examined how transient expression of GαqGTP affects the neurons.

Transient expression of GαqGTP increases the spontaneous firing of CO2-sensing neurons but not odor-sensing neurons

As shown above, long-term expression of GαqGTP resulted in decreased responses from both odor-sensitive and CO2-sensitive neurons (Fig. 3a,b). Based on our previous data, which showed that loss of Gαq function affected Gr- and Or-expressing neurons differently, long-term expression of GαqGTP may be decreasing the odor and CO2 responses via different mechanisms. For instance, GαqGTP could decrease responses by causing either adaptation or deterioration in health. To distinguish between these two possibilities and to investigate whether Grs or Ors signal through Gαq, we transiently expressed GαqGTP in adult flies using hs-GAL4 and tested the flies immediately. Non-heat-shocked siblings were used as controls. If Gαq mediates odor or CO2 responses, transient expression of GαqGTP should mimic constitutive receptor activation, even in the absence of a ligand. Consistent with this prediction, we found that transient expression of GαqGTP produced a threefold increase in the spontaneous firing rate of ab1C (Fig. 4a,b). This effect is specific to ab1C; GαqGTP expression in ab3A does not affect its spontaneous firing rate (Fig. 4c,d). The effect is also specific to GαqGTP, as transient GαsGTP expression in neither ab1C nor ab3A affects its spontaneous firing rate (n = 12; p > 0.4 for each neuron class; data not shown).

Figure 4.
Figure 4.

Transient expression of GαqGTP. a, Traces of ab1C spontaneous firing from control flies and flies transiently expressing GαqGTP. Dots mark spikes from ab1C, which can be distinguished by their relative amplitudes; other spikes derive from other neurons in the same sensillum. b, Heat-shock-induced expression of GαqGTP increased the spontaneous firing rate of ab1C threefold. c, Traces of ab3A spontaneous firing from GαqGTP-expressing and control flies. Dots mark spikes from the ab3A neuron. d, The spontaneous firing rate of ab3A was unaffected by GαqGTP expression (p = 0.4). n = 9–17. Error bars indicate SEM.

The increase in spontaneous firing observed in ab1C is consistent with the direct involvement of Gαq in CO2 signaling. Conversely, short-term expression of GαqGTP in ab3A did not affect spontaneous firing. Consistent with our Gαq loss-of-function results, this result suggests that Gαq does not act directly in ab3A signaling. An indirect role, such as a role in neuronal maintenance, may thus account for the effect of long-term GαqGTP expression in ab1A, ab2A, and ab3A (Fig. 3a). The finding that transient expression of GαqGTP increases the spontaneous firing of ab1C but not ab3A further suggests that Gαq is involved directly in CO2 signaling but not odor signaling.

Next, we examined the odor and CO2 responses of ab3A and ab1C neurons that transiently express GαqGTP. Our results thus far suggest that long-term expression of GαqGTP decreases ab1C responses via adaptation of the neuron, whereas GαqGTP may affect the health of ab3A. If this interpretation is correct, we would predict that short-term expression of GαqGTP should affect their responses differently. In accordance with this prediction, ab3A neurons that transiently express GαqGTP have normal responses to odorant dilutions ranging from 10−8 to 10−2 (n = 9; p > 0.8; data not shown). Conversely, ab1C neurons that transiently express GαqGTP have a reduced response to CO2 (n = 12; p < 0.001; data not shown). Moreover, the reduction in CO2 response is much less pronounced than in flies that expressed GαqGTP for longer. In heat shocked flies, the response to 1% CO2 is reduced by 33% compared to 88% for long-term expression (Fig. 3b) (data not shown). The most likely explanation for these results is that the longer expression period of GαqGTP produces a greater degree of adaptation and results in a greater loss of sensitivity.

The results, taken together, are consistent with a model in which G-proteins act directly in Gr21a/Gr63a signaling but not in Or signaling.

Verification of Gαq expression in CO2-sensitive neurons

Above we used three independent means of altering levels of Gαq signaling in ab1C, and all three affected CO2 response. Others have shown that Gαq is expressed in a majority of ORNs, including the large basiconic ORNs tested in this study (Talluri et al., 1995; Kain et al., 2008). However, to confirm that ab1C expresses Gαq, we performed a double-label analysis. To visualize ab1C, we used a Gr21a-GFP direct fusion line (Couto et al., 2005) and observed GFP in a subset of large basiconic sensilla on the proximomedial region of the antenna, consistent with the location of ab1C neurons (de Bruyne et al., 2001).

To determine whether Gαq is expressed in these GFP-expressing cells, we first used an anti-Gαq antibody in a double-label experiment. However, the antibody showed a broad labeling pattern with a signal-to-noise ratio that was not sufficiently high for us to draw confident conclusions about colocalization. As a second means of determining whether the GFP-expressing ab1C cells express Gαq, we used two independently generated GAL4 lines driven by the Gαq promoter, Gαq-GAL4 (Kain et al., 2008) and DGαq-GAL4, which we generated. Flies carrying Gαq-GAL4; UAS-DsRed or DGαq-GAL4;UAS-DsRed both showed expression of DsRed in many cells broadly distributed across the antenna (data not shown), as reported previously in Drosophila (Talluri et al., 1995; Ratnaparkhi et al., 2002; Kain et al., 2008), and consistent with results found with an Anopheles ortholog (Rutzler et al., 2006). Previous reports have shown that Gαq is expressed in a majority of ORNs as well as in non-neuronal support cells (Talluri et al., 1995; Kain et al., 2008). All GFP-positive cells were also positive for DsRed, indicating that all Gr21a-GFP-expressing neurons also express Gαq-GAL4 (data not shown). Thus, these results support the conclusion that the CO2-sensing neurons express Gαq.

Analysis of odor and CO2 responses in Gαq, Gαs, and Gαo null clones

Next, we examined ORNs that are null for Gα proteins. We examined Gαq, Gαs, and Gαo because these three Gα genes have been identified in ORNs in many different species of insects (Talluri et al., 1995; Miura et al., 2005; Rutzler et al., 2006). Null alleles for Gαq (Banerjee et al., 2006), Gαs (Wolfgang et al., 2001), and Gαo (Katanaev et al., 2005) are homozygous lethal. Therefore, we used the MARCM system (Lee and Luo, 2001) to generate homozygous null mutant clones of these Gα genes in the antenna. We used Gr21a-GAL4, UAS-GFP to identify mutant ab1C neurons and Or22a-GAL4, UAS-GFP to identify mutant ab3A neurons.

As a control for efficiency of recombination, we first generated clones in a wild-type background that contained no mutations of G-protein genes. We found that GFP+ ab3A and ab1C were generated consistently: 5.5 ± 1.1 and 4.5 ± 0.8 GFP+ neurons were detected per antenna, respectively (n = 22 antennae; n = 20 antennae), with ∼80% of all antennae examined having at least one GFP+ neuron. This control experiment established criteria by which to evaluate the frequency and appearance of GFP+ cells mutant for G-protein genes.

We then generated null clones of Gαs. The frequencies at which GFP+ ab3A and ab1C were generated (7.0 ± 1.2 neurons/antenna; 5.7 ± 0.7 GFP+ neurons/antenna) were similar to the frequencies in the control experiment. Gαs−/− ab1C showed normal responses to CO2 in 6-d-old flies, consistent with our previous analyses showing that Gαs is not required for CO2 signaling (Fig. 5a). When older flies (∼20 d) were examined, a decrease in CO2 response was observed in Gαs−/− ab1C neurons at all concentrations tested, with a 33% decrease in response at 1% CO2 (n = 16–18; p < 0.001). Since the decrease was observed only in older flies, it may reflect a progressive deterioration of neuronal health, rather than a specific defect in the forward pathway of CO2 signaling. The simplest interpretation of these results is that Gαs is not required for CO2 signaling.

Figure 5.
Figure 5.

MARCM analysis. a, Gαs−/− ab1C had normal CO2 responses (p = 0.19). b, c, Gαs−/− ab3A had a decreased response to near-saturating concentrations of methyl butyrate (b) and butyric acid (c; p < 0.01 for responses to 0.1% and higher concentrations of methyl butyrate and 5% and 10% butyric acid). d, e, Expression of Gαs selectively in Gαs−/− ab3A rescues the responses to all concentrations of methyl butyrate (d) and butyric acid (e). f–i, Gαq−/− ab3A neurons generated using either the Gαq28 (f, g) or Gαq221C (h, i) allele had normal responses to methyl butyrate and butyric acid (p > 0.5 for both alleles). n = 8–24. Error bars indicate SEM.

The absence of Gαs in ab3A did not affect its response to low concentrations of methyl butyrate (0.05% and lower) or butyric acid (1% and lower) (Fig. 5b,c). However, at near-saturating concentrations of both odorants tested, there was a small (≤17.0%) decrease in the odor response regardless of the age of the fly. These abnormalities were rescued by driving a copy of the Gαs gene specifically in ab3A, suggesting that the role of Gαs is cell autonomous (Fig. 5d,e). Thus, Gαs may contribute to ab3A odor responses, but its contribution is very modest.

For Gαq, we generated a null allele, Gαq28. This allele was isolated by imprecise excision of a P-element in the 5′ UTR of Gαq-3. Gαq28 contains a 5428 bp deletion that removes the first eight exons of Gαq-3. Gαq28 yielded GFP+ ab3A at a frequency similar to that of the control experiment (5.3 ± 1.3 vs 5.5 ± 1.1 GFP+ neurons/antenna; n = 14 and n = 22 antennae, respectively). Consistent with our previous results, the absence of Gαq in ab3A did not affect olfactory responses to the two odors tested (Fig. 5f,g). A previous report showed that mosaic analysis with another Gαq null allele, Gαq221C, resulted in decreased electroantennogram responses, which represent a decrease in the summed receptor potentials of populations of antennal neurons (Kain et al., 2008). To address this discrepancy, we tested the null allele used in this report (Gαq221C) and obtained results indistinguishable from those with Gαq28 both in terms of the frequency of clones generated and the phenotype of the Gαq221C clones (Fig. 5h,i). Taken together, the simplest interpretation of our results is that Gαq is not required for olfactory transduction in ab3A.

Efforts to produce Gαq−/− ab1C yielded very different results from those designed to produce Gαq−/− ab3A or control clones with marked ab1C neurons. First, GFP+ ab1C neurons were either not generated or generated at a very low frequency in the Gαq MARCM experiments (see Materials and Methods). Despite using two different Gαq deletion alleles, Gαq28 and Gαq221C, and two different markers of ab1C neurons, Gr21a-GAL4 and Gr63a-GAL4, few if any GFP+ ab1C neurons were generated, implicating a crucial role for Gαq in ab1C development and/or survival.

No Gαo−/− ab1C or ab3A neurons were identified. This is consistent with findings by others that Gαo is required for asymmetrical cell division and nervous system development (Katanaev et al., 2005; Katanaev and Tomlinson, 2006).

In summary, Gαs−/− ab1C had normal responses to CO2, Gαs−/− ab3A had small reductions in odor response, Gαq−/− ab3A was normal in response to two odorants tested, and efforts to determine the physiological effect of a complete loss of Gαq function in ab1C were inconclusive. The simplest interpretation of the repeated difficulty in obtaining Gαq−/− ab1C is that Gαq, like Gαo, plays a major role in ab1C development or survival.

Pertussis toxin, an inhibitor of Gαo, does not affect CO2 or odor responses

The mosaic analysis described above supports a role for Gαo in olfactory system development. To test the possibility of a role for Gαo in ORN function, we analyzed the effect of pertussis toxin (PTX) on odor and CO2 response. PTX is a selective inhibitor of Gαo, since Gαo is the only G-protein in Drosophila that contains the PTX recognition site (Ferris et al., 2006). We found that expression of ptx in ab1A, ab2A, ab3A, and ab1C did not affect odor or CO2 responses across the broad concentration range tested (n = 6; p > 0.2 for each neuron class; data not shown).

These results argue against a role for Gαo in the physiology of odor or CO2 signaling, consistent with our results from experiments with RNAi, competitive peptides, and GαoGTP.

Disruption of Gαq signaling in adults decreases CO2 response

Our results thus far suggest that although Gαq is not required for Or-mediated responses, it is required for the Gr-mediated CO2 response. To test whether Gαq plays a role in the physiology of the CO2 neuron, as opposed to an exclusively developmental role, we used the TARGET system (McGuire et al., 2003). TARGET allowed us to interfere with Gαq signaling selectively in adult flies: thus, expression of Gαq1F1 RNAi or GαqGTP in ab1C was repressed during development and expressed only after eclosion.

Expression of Gαq1F1 RNAi or GαqGTP selectively in adulthood was sufficient to decrease CO2 response (n = 9–15; p < 0.001). In fact, for both constructs, a 30–83% decrease in response was observed at all CO2 concentrations tested, which ranged from 0.1 to 5%. The level of reduction observed in these flies was similar to that observed when the constructs were expressed with Gr21a-GAL4 alone (Fig. 3b) (data not shown). This similarity is not surprising given that the Gr21a-GAL4 driver is likely expressed late in development: although the developmental time course of Gr21a expression is unknown, Or genes in the antenna are first detected during the late pupal stage (Clyne et al., 1999; Elmore and Smith, 2001; Larsson et al., 2004).

Gαq increases CO2 response but not odor response in an in vivo expression system

Coexpression of Gr21a and Gr63a in an in vivo expression system, the “empty neuron system” (Dobritsa et al., 2003), conferred a CO2 response to ab3A, which normally does not respond to CO2 (Fig. 1d) (Kwon et al., 2007). However, the response was modest compared to the endogenous ab1C response, perhaps because of limiting quantities of another component in ab3A essential for a full CO2 response. We coexpressed Gαq and found that it dramatically increased the CO2 response (Fig. 6). The response increased by a factor of five, to a level comparable to that of the endogenous neuron, at 0.1, 0.5, and 1% CO2 (n = 17–21; p > 0.1; data not shown). This effect is specific to Gαq: coexpression of Gαs with Gr21a and Gr63a did not increase the CO2 response (Fig. 6). The effect of Gαq is also specific to the CO2 response: as a control, we expressed Gαq in wild-type ab3A neurons and found that overexpression of Gαq did not increase responses to a broad concentration range of the two odorants tested, arguing against the possibility that Gαq overexpression causes a general increase in ab3A excitability (n = 12; p > 0.05; data not shown).

Figure 6.
Figure 6.

Ectopic expression of the CO2 receptors in the empty neuron. a, Sample traces of responses to 5% CO2 from neurons expressing Gr21a and Gr63a alone or with Gαq or Gαs. The bar indicates a 500 ms pulse of 5% CO2. b, Coexpression of Gαq with Gr21a and Gr63a increased CO2 response. Coexpression with Gαs had no effect (p = 0.6). n = 17–27. Error bars indicate SEM.

Gγ30A is also required for CO2 response

Our data support a role for Gαq in CO2 response in ab1C. In Drosophila phototransduction, Gγ30A is the binding partner of Gαq-1 (Schulz et al., 1999). Given that Gγ30A RNA was also detected using RT-PCR from adult antennal RNA (data not shown), we wondered whether Gγ30A is also the binding partner of Gαq in ab1C.

In functional tests we found that Gγ30A RNAi decreased CO2 response (Fig. 7a). Consistent with this result, flies homozygous for a 6 kb insertion in Gγ30A have a reduced CO2 response (Fig. 7a). Flies with one copy of the insertion in heterozygous combination with a deletion spanning the Gγ30A gene region had a response to CO2 that was even further reduced (Fig. 7a). In contrast, most Or-mediated responses appeared normal in flies with this insertion (Fig. 7b). ab1A had a slight decrease in odor response. However, the effect was modest relative to the decrease in CO2 response.

Figure 7.
Figure 7.

Gγ. a, RNAi against Gγ30A decreased the CO2 response (p < 0.001). Flies homozygous for an insertion in Gγ30A had a decreased CO2 response relative to heterozygotes used as a control for the genetic background (p < 0.001). Flies with one copy of the insertion allele and one copy of a deletion allele spanning the Gγ30A region had a further decreased response to CO2 (p < 0.001). RNAi against Gγ1 did not affect the CO2 response (p = 0.8). Gγ1 mutant flies also had normal CO2 responses (p = 0.6). b, Insertions in Gγ30A or Gγ1 had little or no effect on odor responses (p > 0.1 for all combinations tested except for the ab1A response of Gγ30A insertion flies; for details, see Results). n = 6–18. Error bars indicate SEM.

There is only one other Gγ gene in the Drosophila genome, Gγ1. Gγ1 RNA was also detected using RT-PCR from adult antennal RNA (data not shown). Gγ1 has been found to play a role in sugar sensing by Gr5a-expressing gustatory neurons in the labelum (Ishimoto et al., 2005). A Gγ1 RNAi construct (Ishimoto et al., 2005) that decreased responses of gustatory neurons to sugar did not affect the CO2 response (Fig. 7a). An insertion in the 5′ UTR of Gγ1, which decreases Gγ1 expression by ∼90% (Ishimoto et al., 2005), also did not affect CO2 response (Fig. 7a) or the odor responses of ab1A, ab2A, or ab3A (Fig. 7b).

In summary, these results suggest that Gγ30A is required along with Gαq in the response of ab1C to CO2. The results also suggest that Or-mediated odor responses do not require Gγ30A or Gγ1.

Other visual transduction genes are not required for CO2 response

The involvement of Gαq and Gγ30A in both vision and CO2 sensing suggested that these two processes might share other signaling elements. We tested the phototransduction mutants Gβ76C (Gβ), inaD (a scaffolding protein), and norpA (phospholipase C). Null mutant flies of each of these three genes had normal responses to CO2 (data not shown). This finding raises the possibility that although CO2 response requires Gαq and Gγ30A, these G-proteins may not signal through canonical G-protein-signaling pathways. A recent study proposed that Ors are both GPCRs and ligand-gated ion channels, and that G-protein activation results in current through the Or channel (Wicher et al., 2008). Our data suggest that Gαq plays a direct role in CO2 signaling. However, we cannot determine whether Gαq activates conventional downstream signaling components or if it acts on a channel formed from Gr21a and Gr63a.

Discussion

Recent studies in heterologous systems have provided evidence that insect odor receptors can act as ligand-gated ion channels (Sato et al., 2008; Wicher et al., 2008). Canonical members of the Or family form a complex with the coreceptor Or83b (Larsson et al., 2004; Neuhaus et al., 2005; Benton et al., 2006), and the heteromeric receptor appears to play a direct role in olfactory signaling as an ion channel. Conversely, for Grs, there is evidence that taste neurons expressing Gr5a require G-proteins (Ishimoto et al., 2005; Ueno et al., 2006). These findings have raised an intriguing question: to what extent does signaling by the Or–Gr superfamily of insect chemoreceptors differ from that of chemoreceptors in vertebrates and C. elegans? Odor and taste receptors in these organisms are believed to signal directly through G-proteins. We have systematically examined the G-proteins of Drosophila using both loss-of-function and gain-of-function experiments for roles in chemosensory response. All experiments in this study were performed in vivo in Drosophila ORNs, as opposed to in heterologous systems.

We found that disruption, knockdown, or even complete elimination of G-proteins had little or no effect on Or-mediated signaling in vivo. Our results indicate that CO2 response, mediated by two Grs, requires Gαq and Gγ30A. Further investigation using constitutively active Gαq, mosaic analysis, and conditional expression showed that Gαq is required for both the survival and function of ab1C.

Reduction of G-protein function has little or no effect onOr-mediated odor response

Recent studies have proposed conflicting models for Or signaling. One study suggested in particular that Gαq is required by ab2A for full sensitivity to low concentrations of odors (Kain et al., 2008). However, the study used a pan-neuronal driver to mark Gαq−/− cells and examined a narrow response range. Using ORN class-specific GAL4 drivers and examining a broad response range of odor concentrations, we tested three ORN classes, including ab2A, that had been manipulated in multiple ways to cause a loss of Gα function. We found that knockdown of Gαq in ab1A, ab2A, and ab3A did not decrease odor responses. Most strikingly, ab3A neurons that were genotypically Gαq−/−, as determined with an ab3A-specific driver, showed normal responses across a broad response range for the odors tested. Thus, our findings do not support the model that Gαq is required for Or signaling.

A second recent study proposed that Or83b can act as a cyclic-nucleotide-gated ion channel that depends on Gαs (Wicher et al., 2008). Our results suggest that Gαs is not required for the primary response to odors. Gαs−/− ab3A has a decreased response to higher concentrations of odors; however, this effect is modest and suggests that the primary pathway of olfactory transduction in ab3A neurons does not require Gαs.

The third study proposed that the Or–Or83b complex is a ligand-gated ion channel whose activation does not require G-proteins (Sato et al., 2008). Our results support this proposition that G-proteins are not required for primary Or signaling. A dual-activation model was recently proposed to reconcile the different models of Or signaling (Nakagawa and Vosshall, 2009). In this model, Or activation results in a primary ionotropic response as well as a slow metabotropic component that modulates the Or complex. Our results could be interpreted to support this model, with the Gαs pathway potentiating the main response. However, we note that manipulation of Gαs signaling with RNAi, competitive peptides, and GαsGTP had no effect on the response of ab1A, ab2A, or ab3A. Deletion of the Gαs gene produced at most a 17.1% reduction in response. Furthermore, the decreased response was only seen at the highest odor concentrations, at which the ORN is near or at its maximum response rate for the given odor. If the role of Gαs were to potentiate receptor function, one might expect that it would boost low signals or signals across multiple concentrations, rather than only at the highest concentrations.

Gr-mediated CO2 response requires G-proteins

In contrast to our results with Ors, we found that multiple, complementary approaches consistently revealed a role for G-proteins—Gαq and Gγ30A—in Gr-mediated CO2 response. Knockdown of Gαq via various genetic manipulations all resulted in reduced CO2 response in ab1C. One might hypothesize that Gαq plays an indirect role, such as in neuronal maintenance or in modulation of the response. However, this hypothesis is difficult to reconcile with the finding that transient expression of GαqGTP increases the spontaneous firing rate of ab1C. The simplest interpretation of the transient expression analysis is that Gαq plays a direct role in CO2 signaling, although the mechanism by which it acts is not revealed by our analysis.

Different requirements for G-proteins in different chemosensory neurons

Signaling in ab1C differs from that in ab1A, ab2A, and ab3A in its requirements for G-proteins. This difference is consistent with other differences between ab1C and the other ORNs. ab1C expresses Gr members of the Or–Gr insect chemoreceptor superfamily, whereas most other ORNs express Or members. ab1A, ab2A, and ab3A express the coreceptor Or83b; ab1C does not express Or83b, and a Gr gene analogous to Or83b has not been identified. Or83b-expressing ORNs have GABAB receptors, which signal through Gαo, at their axon terminals, thereby allowing for inhibitory modulation of synaptic transmission (Olsen and Wilson, 2008; Root et al., 2008). ab1C does not express GABAB receptors and does not exhibit presynaptic inhibition (Root et al., 2008).

Different ORNs also differ in their developmental requirements for G-proteins. We found that Gαq plays a role in the development of ab1C but not ab3A. These results are consistent with differences in morphology between ab1C and the other ORNs tested (Shanbhag et al., 1999). Moreover, a microRNA, miR-279, and its target, Nerfin-1, specifically regulate the development of CO2-sensitive neurons (Cayirlioglu et al., 2008).

There may be additional differences in the requirements for G-proteins among chemosensory neurons that express Gr genes. There is evidence that Gαs and Gγ1 each play a role in the sugar response of Gr5a-expressing taste neurons, but not in the bitter response of Gr66a-expressing taste neurons (Ishimoto et al., 2005; Ueno et al., 2006). The functional significance of the heterogeneity in G-protein requirements merits investigation.

Receptor promiscuity and the evolution of insect olfactory transduction: a unifying model

It is intriguing and perhaps revealing that despite the different requirements for G-proteins, receptors of the Or–Gr family are able to function in quite distinct contexts. The ab1C receptors Gr21a and Gr63a can also function in ab3A, although they are limited by the availability of Gαq (Fig. 6), which is not required for the function of the endogenous ab3A receptor, Or22a. Reciprocally, the Ors Or43a and Or83b can function when coexpressed in ab1C (Benton et al., 2006). Moreover, coexpression of Or22a and Or83b in either Gr66a- or Gr5a-expressing taste neurons confers odor responses to both (Hiroi et al., 2008).

Do receptors signal through alternate pathways when expressed in alternate contexts? Coexpression of Or43a and a promiscuous Gα, without Or83b, was found to confer an odor response to Xenopus oocytes (Wetzel et al., 2001). The Bombyx mori receptor BmOR1 did not respond to its ligand when expressed alone in Xenopus oocytes, but was capable of responding when coexpressed with either the Or83b ortholog, BmOR2, or with BmGαq (Nakagawa et al., 2005). The response was more robust with BmOR2, which is reminiscent of our finding that the CO2 response is greater when Gr21a and Gr63a are coexpressed with Gαq, a gene required for the endogenous CO2 response.

One simple, unifying model consistent with all of these results taken together is that Ors and Grs are capable of acting via multiple mechanisms, one requiring Or83b and one requiring G-proteins, efficiently in one case and inefficiently in the other. Ors are believed to represent a relatively recent branch of an ancient Gr family (Robertson et al., 2003). Perhaps Ors evolved to rely exclusively on a ligand-gated ionotropic signaling mechanism, whereas Grs maintained the ability to signal via a G-protein-dependent mechanism. If so, a limited ability of Ors and Grs to each signal through another mechanism might reflect this evolutionary transition.

Footnotes

  • This work was supported by a National Science Foundation predoctoral fellowship (C.A.Y.) and grants from the National Institutes of Health (J.R.C.). We thank J. Árnadóttir, K. Menuz, and C.-Y. Su for insightful discussions and helpful comments on this manuscript.

  • Correspondence should be addressed to Dr. John R. Carlson at the above address. john.carlson{at}yale.edu

References

  1. ↵
    1. Akhter SA,
    2. Luttrell LM,
    3. Rockman HA,
    4. Iaccarino G,
    5. Lefkowitz RJ,
    6. Koch WJ
    (1998) Targeting the receptor-Gq interface to inhibit in vivo pressure overload myocardial hypertrophy. Science 280:574–577.
    OpenUrlAbstract/FREE Full Text
  2. ↵
    1. Banerjee S,
    2. Joshi R,
    3. Venkiteswaran G,
    4. Agrawal N,
    5. Srikanth S,
    6. Alam F,
    7. Hasan G
    (2006) Compensation of inositol 1,4,5-trisphosphate receptor function by altering sarco-endoplasmic reticulum calcium ATPase activity in the Drosophila flight circuit. J Neurosci 26:8278–8288.
    OpenUrlAbstract/FREE Full Text
  3. ↵
    1. Benton R,
    2. Sachse S,
    3. Michnick SW,
    4. Vosshall LB
    (2006) Atypical membrane topology and heteromeric function of Drosophila odorant receptors in vivo. PLoS Biol 4:e20.
    OpenUrlCrossRefPubMed
  4. ↵
    1. Bourne HR,
    2. Sanders DA,
    3. McCormick F
    (1991) The GTPase superfamily: conserved structure and molecular mechanism. Nature 349:117–127.
    OpenUrlCrossRefPubMed
  5. ↵
    1. Carlson JR
    (1996) Olfaction in Drosophila: from odor to behavior. Trends Genet 12:175–180.
    OpenUrlCrossRefPubMed
  6. ↵
    1. Cayirlioglu P,
    2. Kadow IG,
    3. Zhan X,
    4. Okamura K,
    5. Suh GS,
    6. Gunning D,
    7. Lai EC,
    8. Zipursky SL
    (2008) Hybrid neurons in a microRNA mutant are putative evolutionary intermediates in insect CO2 sensory systems. Science 319:1256–1260.
    OpenUrlAbstract/FREE Full Text
  7. ↵
    1. Clyne P,
    2. Grant A,
    3. O'Connell R,
    4. Carlson JR
    (1997) Odorant response of individual sensilla on the Drosophila antenna. Invert Neurosci 3:127–135.
    OpenUrlCrossRefPubMed
  8. ↵
    1. Clyne PJ,
    2. Warr CG,
    3. Freeman MR,
    4. Lessing D,
    5. Kim J,
    6. Carlson JR
    (1999) A novel family of divergent seven-transmembrane proteins: candidate odorant receptors in Drosophila. Neuron 22:327–338.
    OpenUrlCrossRefPubMed
  9. ↵
    1. Clyne PJ,
    2. Warr CG,
    3. Carlson JR
    (2000) Candidate taste receptors in Drosophila. Science 287:1830–1834.
    OpenUrlAbstract/FREE Full Text
  10. ↵
    1. Couto A,
    2. Alenius M,
    3. Dickson BJ
    (2005) Molecular, anatomical, and functional organization of the Drosophila olfactory system. Curr Biol 15:1535–1547.
    OpenUrlCrossRefPubMed
  11. ↵
    1. de Bruyne M,
    2. Foster K,
    3. Carlson JR
    (2001) Odor coding in the Drosophila antenna. Neuron 30:537–552.
    OpenUrlCrossRefPubMed
  12. ↵
    1. Dobritsa AA,
    2. van der Goes van Naters W,
    3. Warr CG,
    4. Steinbrecht RA,
    5. Carlson JR
    (2003) Integrating the molecular and cellular basis of odor coding in the Drosophila antenna. Neuron 37:827–841.
    OpenUrlCrossRefPubMed
  13. ↵
    1. Elmore T,
    2. Smith DP
    (2001) Putative Drosophila odor receptor OR43b localizes to dendrites of olfactory neurons. Insect Biochem Mol Biol 31:791–798.
    OpenUrlCrossRefPubMed
  14. ↵
    1. Ferris J,
    2. Ge H,
    3. Liu L,
    4. Roman G
    (2006) G(o) signaling is required for Drosophila associative learning. Nat Neurosci 9:1036–1040.
    OpenUrlCrossRefPubMed
  15. ↵
    1. Gilchrist A,
    2. Bunemann M,
    3. Li A,
    4. Hosey MM,
    5. Hamm HE
    (1999) A dominant-negative strategy for studying roles of G proteins in vivo. J Biol Chem 274:6610–6616.
    OpenUrlAbstract/FREE Full Text
  16. ↵
    1. Gilchrist A,
    2. Li A,
    3. Hamm HE
    (2002) G alpha COOH-terminal minigene vectors dissect heterotrimeric G protein signaling. Sci STKE 2002:PL1.
    OpenUrlPubMed
  17. ↵
    1. Grant AJ,
    2. Wigton BE,
    3. Aghajanian JG,
    4. O'Connell RJ
    (1995) Electrophysiological responses of receptor neurons in mosquito maxillary palp sensilla to carbon dioxide. J Comp Physiol A Neuroethol Sens Neural Behav Physiol 177:389–396.
    OpenUrlPubMed
  18. ↵
    1. Hallem EA,
    2. Ho MG,
    3. Carlson JR
    (2004) The molecular basis of odor coding in the Drosophila antenna. Cell 117:965–979.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Hamm HE,
    2. Deretic D,
    3. Arendt A,
    4. Hargrave PA,
    5. Koenig B,
    6. Hofmann KP
    (1988) Site of G protein binding to rhodopsin mapped with synthetic peptides from the alpha subunit. Science 241:832–835.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    1. Hiroi M,
    2. Tanimura T,
    3. Marion-Poll F
    (2008) Hedonic taste in Drosophila revealed by olfactory receptors expressed in taste neurons. PLoS One 3:e2610.
    OpenUrlCrossRefPubMed
  21. ↵
    1. Hummel T,
    2. Vasconcelos ML,
    3. Clemens JC,
    4. Fishilevich Y,
    5. Vosshall LB,
    6. Zipursky SL
    (2003) Axonal targeting of olfactory receptor neurons in Drosophila is controlled by Dscam. Neuron 37:221–231.
    OpenUrlCrossRefPubMed
  22. ↵
    1. Ishimoto H,
    2. Takahashi K,
    3. Ueda R,
    4. Tanimura T
    (2005) G-protein gamma subunit 1 is required for sugar reception in Drosophila. EMBO J 24:3259–3265.
    OpenUrlCrossRefPubMed
  23. ↵
    1. Jin S,
    2. Cornwall MC,
    3. Oprian DD
    (2003) Opsin activation as a cause of congenital night blindness. Nat Neurosci 6:731–735.
    OpenUrlCrossRefPubMed
  24. ↵
    1. Jones WD,
    2. Cayirlioglu P,
    3. Kadow IG,
    4. Vosshall LB
    (2007) Two chemosensory receptors together mediate carbon dioxide detection in Drosophila. Nature 445:86–90.
    OpenUrlCrossRefPubMed
  25. ↵
    1. Kain P,
    2. Chakraborty TS,
    3. Sundaram S,
    4. Siddiqi O,
    5. Rodrigues V,
    6. Hasan G
    (2008) Reduced odor responses from antennal neurons of Gqα, phospholipase Cβ, and rdgA mutants in Drosophila support a role for a phospholipid intermediate in insect olfactory transduction. J Neurosci 28:4745–4755.
    OpenUrlAbstract/FREE Full Text
  26. ↵
    1. Kalidas S,
    2. Smith DP
    (2002) Novel genomic cDNA hybrids produce effective RNA interference in adult Drosophila. Neuron 33:177–184.
    OpenUrlCrossRefPubMed
  27. ↵
    1. Katanaev VL,
    2. Tomlinson A
    (2006) Dual roles for the trimeric G protein Go in asymmetric cell division in Drosophila. Proc Natl Acad Sci U S A 103:6524–6529.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    1. Katanaev VL,
    2. Ponzielli R,
    3. Semeriva M,
    4. Tomlinson A
    (2005) Trimeric G protein-dependent frizzled signaling in Drosophila. Cell 120:111–122.
    OpenUrlCrossRefPubMed
  29. ↵
    1. Kwon JY,
    2. Dahanukar A,
    3. Weiss LA,
    4. Carlson JR
    (2007) The molecular basis of CO2 reception in Drosophila. Proc Natl Acad Sci U S A 104:3574–3578.
    OpenUrlAbstract/FREE Full Text
  30. ↵
    1. Larsson MC,
    2. Domingos AI,
    3. Jones WD,
    4. Chiappe ME,
    5. Amrein H,
    6. Vosshall LB
    (2004) Or83b encodes a broadly expressed odorant receptor essential for Drosophila olfaction. Neuron 43:703–714.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Lee T,
    2. Luo L
    (2001) Mosaic analysis with a repressible cell marker (MARCM) for Drosophila neural development. Trends Neurosci 24:251–254.
    OpenUrlCrossRefPubMed
  32. ↵
    1. Lu T,
    2. Qiu YT,
    3. Wang G,
    4. Kwon JY,
    5. Rutzler M,
    6. Kwon HW,
    7. Pitts RJ,
    8. van Loon JJ,
    9. Takken W,
    10. Carlson JR,
    11. Zwiebel LJ
    (2007) Odor coding in the maxillary palp of the malaria vector mosquito Anopheles gambiae. Curr Biol 17:1533–1544.
    OpenUrlCrossRefPubMed
  33. ↵
    1. Lundin C,
    2. Kall L,
    3. Kreher SA,
    4. Kapp K,
    5. Sonnhammer EL,
    6. Carlson JR,
    7. Heijne G,
    8. Nilsson I
    (2007) Membrane topology of the Drosophila OR83b odorant receptor. FEBS Lett 581:5601–5604.
    OpenUrlCrossRefPubMed
  34. ↵
    1. McGuire SE,
    2. Le PT,
    3. Osborn AJ,
    4. Matsumoto K,
    5. Davis RL
    (2003) Spatiotemporal rescue of memory dysfunction in Drosophila. Science 302:1765–1768.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Miura N,
    2. Atsumi S,
    3. Tabunoki H,
    4. Sato R
    (2005) Expression and localization of three G protein alpha subunits, Go, Gq, and Gs, in adult antennae of the silkmoth (Bombyx mori) J Comp Neurol 485:143–152.
    OpenUrlCrossRefPubMed
  36. ↵
    1. Nakagawa T,
    2. Vosshall LB
    (2009) Controversy and consensus: noncanonical signaling mechanisms in the insect olfactory system. Curr Opin Neurobiol 19:284–292.
    OpenUrlCrossRefPubMed
  37. ↵
    1. Nakagawa T,
    2. Sakurai T,
    3. Nishioka T,
    4. Touhara K
    (2005) Insect sex-pheromone signals mediated by specific combinations of olfactory receptors. Science 307:1638–1642.
    OpenUrlAbstract/FREE Full Text
  38. ↵
    1. Neuhaus EM,
    2. Gisselmann G,
    3. Zhang W,
    4. Dooley R,
    5. Stortkuhl K,
    6. Hatt H
    (2005) Odorant receptor heterodimerization in the olfactory system of Drosophila melanogaster. Nat Neurosci 8:15–17.
    OpenUrlCrossRefPubMed
  39. ↵
    1. Olsen SR,
    2. Wilson RI
    (2008) Lateral presynaptic inhibition mediates gain control in an olfactory circuit. Nature 452:956–960.
    OpenUrlCrossRefPubMed
  40. ↵
    1. Rasenick MM,
    2. Watanabe M,
    3. Lazarevic MB,
    4. Hatta S,
    5. Hamm HE
    (1994) Synthetic peptides as probes for G protein function. Carboxyl-terminal G alpha s peptides mimic Gs and evoke high affinity agonist binding to beta-adrenergic receptors. J Biol Chem 269:21519–21525.
    OpenUrlAbstract/FREE Full Text
  41. ↵
    1. Ratnaparkhi A,
    2. Banerjee S,
    3. Hasan G
    (2002) Altered levels of Gq activity modulate axonal pathfinding in Drosophila. J Neurosci 22:4499–4508.
    OpenUrlAbstract/FREE Full Text
  42. ↵
    1. Robertson HM,
    2. Warr CG,
    3. Carlson JR
    (2003) Molecular evolution of the insect chemoreceptor gene superfamily in Drosophila melanogaster. Proc Natl Acad Sci U S A 100(Suppl 2):14537–14542.
    OpenUrlAbstract/FREE Full Text
  43. ↵
    1. Root CM,
    2. Masuyama K,
    3. Green DS,
    4. Enell LE,
    5. Nassel DR,
    6. Lee CH,
    7. Wang JW
    (2008) A presynaptic gain control mechanism fine-tunes olfactory behavior. Neuron 59:311–321.
    OpenUrlCrossRefPubMed
  44. ↵
    1. Rutzler M,
    2. Lu T,
    3. Zwiebel LJ
    (2006) Gα encoding gene family of the malaria vector mosquito Anopheles gambiae: expression analysis and immunolocalization of AGαq and AGαo in female antennae. J Comp Neurol 499:533–545.
    OpenUrlCrossRefPubMed
  45. ↵
    1. Sato K,
    2. Pellegrino M,
    3. Nakagawa T,
    4. Nakagawa T,
    5. Vosshall LB,
    6. Touhara K
    (2008) Insect olfactory receptors are heteromeric ligand-gated ion channels. Nature 452:1002–1006.
    OpenUrlCrossRefPubMed
  46. ↵
    1. Schulz S,
    2. Huber A,
    3. Schwab K,
    4. Paulsen R
    (1999) A novel Gγ isolated from Drosophila constitutes a visual G protein gamma subunit of the fly compound eye. J Biol Chem 274:37605–37610.
    OpenUrlAbstract/FREE Full Text
  47. ↵
    1. Scott K,
    2. Becker A,
    3. Sun Y,
    4. Hardy R,
    5. Zuker C
    (1995) Gq[alpha] protein function in vivo: Genetic dissection of its role in photoreceptor cell physiology. Neuron 15:919–927.
    OpenUrlCrossRefPubMed
  48. ↵
    1. Scott K,
    2. Brady R,
    3. Cravchik A,
    4. Morozov P,
    5. Rzhetsky A,
    6. Zuker C,
    7. Axel R
    (2001) A chemosensory gene family encoding candidate gustatory and olfactory receptors in Drosophila. Cell 104:661–673.
    OpenUrlCrossRefPubMed
  49. ↵
    1. Shanbhag SR,
    2. Muller B,
    3. Steinbrecht RA
    (1999) Atlas of olfactory organs of Drosophila melanogaster 1. Types, external organization, innervation and distribution of olfactory sensilla. Int J Insect Morphol Embryol 28:377–397.
    OpenUrlCrossRefPubMed
  50. ↵
    1. Syed Z,
    2. Leal WS
    (2007) Maxillary palps are broad spectrum odorant detectors in Culex quinquefasciatus. Chem Senses 32:727–738.
    OpenUrlAbstract/FREE Full Text
  51. ↵
    1. Talluri S,
    2. Bhatt A,
    3. Smith DP
    (1995) Identification of a Drosophila G protein alpha subunit (dGq alpha-3) expressed in chemosensory cells and central neurons. Proc Natl Acad Sci U S A 92:11475–11479.
    OpenUrlAbstract/FREE Full Text
  52. ↵
    1. Ueno K,
    2. Kohatsu S,
    3. Clay C,
    4. Forte M,
    5. Isono K,
    6. Kidokoro Y
    (2006) Gsα is involved in sugar perception in Drosophila melanogaster. J Neurosci 26:6143–6152.
    OpenUrlAbstract/FREE Full Text
  53. ↵
    1. van der Goes van Naters W,
    2. Carlson JR
    (2006) Insects as chemosensors of humans and crops. Nature 444:302–307.
    OpenUrlCrossRefPubMed
  54. ↵
    1. Vosshall LB,
    2. Amrein H,
    3. Morozov PS,
    4. Rzhetsky A,
    5. Axel R
    (1999) A spatial map of olfactory receptor expression in the Drosophila antenna. Cell 96:725–736.
    OpenUrlCrossRefPubMed
  55. ↵
    1. Wetzel CH,
    2. Behrendt HJ,
    3. Gisselmann G,
    4. Stortkuhl KF,
    5. Hovemann B,
    6. Hatt H
    (2001) Functional expression and characterization of a Drosophila odorant receptor in a heterologous cell system. Proc Natl Acad Sci U S A 98:9377–9380.
    OpenUrlAbstract/FREE Full Text
  56. ↵
    1. Wicher D,
    2. Schafer R,
    3. Bauernfeind R,
    4. Stensmyr MC,
    5. Heller R,
    6. Heinemann SH,
    7. Hansson BS
    (2008) Drosophila odorant receptors are both ligand-gated and cyclic-nucleotide-activated cation channels. Nature 452:1007–1011.
    OpenUrlCrossRefPubMed
  57. ↵
    1. Wolfgang WJ,
    2. Hoskote A,
    3. Roberts IJ,
    4. Jackson S,
    5. Forte M
    (2001) Genetic analysis of the Drosophila Gs(alpha) gene. Genetics 158:1189–1201.
    OpenUrlPubMed
Back to top

In this issue

Email
Print
View Full Page PDF
Citation Tools
Respond to this article
Role of G-Proteins in Odor-Sensing and CO2-Sensing Neurons in Drosophila
C. Andrea Yao, John R. Carlson
Journal of Neuroscience 31 March 2010, 30 (13) 4562-4572; DOI: 10.1523/JNEUROSCI.6357-09.2010
Twitter logo Facebook logo Mendeley logo

Jump to section

Responses to this article

Respond to this article

Jump to comment:

No eLetters have been published for this article.