Role of G-proteins in odor-sensing and CO2-sensing neurons in Drosophila

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 Galpha proteins in vivo, initially with RNA interference constructs, competitive peptides, and constitutively active Galpha proteins. The results do not support a role for Galpha proteins in odor sensitivity. In parallel experiments, manipulations of Galpha(q), but not other Galpha proteins, affected CO2 response. Transient, conditional, and ectopic expression analyses consistently supported a role for Galpha(q) 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 Galpha(q). Ggamma30A is also required for normal CO2 response. The simplest interpretation of these results is that Galpha(q) and Ggamma30A play a role in the response of CO2-sensing neurons, but are not required for Or-mediated odor signaling.

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 Gα_q 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).

Figure 2.

RNAi. a, RNAi against Gα_q, Gα_s, Gα_o, 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, Gα_q RNAi decreased response to CO₂ (p < 0.001), which is mediated by Gr21a/Gr63a (Jones et al., 2007; Kwon et al., 2007). Overexpression of Gα_q in a Gα_q RNAi background rescued the CO₂ response. RNAi against Gα_s, Gα_o, and Gα73B did not affect CO₂ response (p > 0.05). n = 6–17. Error bars indicate SEM.

Figure 3.

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

Figure 4.

Transient expression of Gα_q^GTP. a, Traces of ab1C spontaneous firing from control flies and flies transiently expressing Gα_q^GTP. 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 Gα_q^GTP increased the spontaneous firing rate of ab1C threefold. c, Traces of ab3A spontaneous firing from Gα_q^GTP-expressing and control flies. Dots mark spikes from the ab3A neuron. d, The spontaneous firing rate of ab3A was unaffected by Gα_q^GTP expression (p = 0.4). n = 9–17. Error bars indicate SEM.

Figure 5.

MARCM analysis. a, Gα_s^−/− ab1C had normal CO₂ responses (p = 0.19). b, c, Gα_s^−/− 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 Gα_s selectively in Gα_s^−/− ab3A rescues the responses to all concentrations of methyl butyrate (d) and butyric acid (e). f–i, Gα_q^−/− ab3A neurons generated using either the Gα_q²⁸ (f, g) or Gα_q^221C (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.

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 Gα_q or Gα_s. The bar indicates a 500 ms pulse of 5% CO₂. b, Coexpression of Gα_q with Gr21a and Gr63a increased CO₂ response. Coexpression with Gα_s had no effect (p = 0.6). n = 17–27. Error bars indicate SEM.

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.