Figure 1.
L-Canavanine activates DmXR and N-methyl-l-arginine (NMA) inhibits DmXR.
(A) l-Canavanine (Cana) is an agonist of DmXR. l-Canavanine at a concentration of 10 mM (black) activated DmXR in HEK cells, but had no effect on the unique fly metabotropic glutamate receptor DmGluA, which was activated by 1 mM glutamate (checkered). Note that 10 mM l-canavanine did not activate the DmX ligand binding pocket mutant receptor (DmXRT176A). (B) Dose-response curve of l-canavanine on DmXR. IP production by DmX receptors activated with increasing concentrations of l-canavanine (half-maximal effective concentration [EC50] of l-canavanine = 0.5±0.2 mM). (C) N-methyl-l-arginine (NMA) inhibits DmXR. The activation of DmXR by 1 mM l-canavanine (grey) was completely inhibited by 1 mM NMA (diagonally hatched on grey background). Thus NMA, which has no effect by itself (diagonally hatched on white background), is a potent antagonist of DmXR. l-arginine (1 mM) did not activate DmXR (horizontally hatched on white background) nor antagonize 1 mM l-canavanine effect on DmXR (horizontally hatched on grey background). (D) Dose-response curve of NMA antagonistic effect on DmXR. IP production by DmX receptors activated with 2 mM l-canavanine (l-cana) in the presence of increasing concentrations of the antagonist NMA (half-maximal inhibitory concentration [IC50] of NMA = 0.2±0.2 mM). For (A–D), data are expressed as the IP (inositol triphosphate) production in HEK cells coexpressing a chimeric G-protein α-subunit Gqi9 and the indicated receptor in presence of drugs relatively to IP production in the basal conditions. The vertical bars represent the standard error of the mean (SEM of triplicate determinations from typical experiments). Asterisks indicate significant differences by t-test (p<0.001).
Figure 2.
L-Canavanine triggers a chemosensory repulsive effect.
Two-choice feeding tests allowed the measure of the preference between the blue and the red solutions of wild-type (WT) and chemosensory defective pox-neuro (poxn) mutant flies. A preference index (PI) value of 1 or 0 indicates a complete preference or aversion for the blue solution, respectively. A value of 0.5 indicates no preference. See Materials and Methods for the calculation of the PI. In our conditions, WT flies prefer the blue solution (white). When l-canavanine (Cana, black) was added at the indicated concentration (1 to 40 mM), it inhibited the intake of the blue solution in a concentration-dependent manner, being very aversive at 30–40 mM. In contrast, poxn flies fed equally on the blue solution with (black) or without (dotted) 40 mM l-canavanine, indicating a chemosensory action for the plant amino acid. Error bars indicate SEM. The single and double asterisks indicate significant differences by t-test (p<0.05 and p<0.001, respectively) between WT flies that fed on the blue solution without (white) or with (black) l-canavanine.
Figure 3.
The L-canavanine–induced repulsive behavior requires the DmXR function.
(A) Pharmacological inhibition of DmXR by the NMA antagonist reduces l-canavanine repulsive effect. Histograms show the preference index (PI) for the blue solution from two-choice feeding assays in which drugs have been added to the blue solution. There was no significant difference in the behavior of WT flies (WT) that fed similarly on the blue control solution (white) and on the blue solution containing either 30 mM NMA (diagonally hatched) or 30 mM l-arginine (Arg) (horizontally hatched). Significantly more flies fed on the blue solution containing 20 mM l-canavanine (Cana)+30 mM NMA (dotted) than on the blue solution containing 20 mM l-canavanine (grey). l-arginine (30 mM) had no effect on 20 mM l-canavanine repulsive effect (checkered). Error bars indicate SEM. Asterisks indicate significant differences by t-test (p<0.001). (B) mtt mutant flies are insensitive to l-canavanine. Histograms show the preference index (PI) for the blue solution from two-choice feeding assays by using the blue solution without (white) or with (black) 30 mM l-canavanine. WT and the genetic background control f01266/f01266 flies are strongly repulsed by l-canavanine and avoid feeding it. In contrast, mttf06268/mttf06268 and mttf06268/DfEx7096 mutant flies are insensitive to l-canavanine, as they fed similarly on the blue solution with or without l-canavanine. Error bars indicate SEM. Asterisks indicate significant differences between the intakes of the blue solution without or with l-canavanine by t-test (p<0.001).
Figure 4.
The f06268 and the DfEx7096 lines are mtt null mutants.
(A) Schematic diagram showing the structure of the mtt gene and the insertion sites of the transposon lines used in this study. Black boxes show coding regions of mtt and CG12780 cDNAs. Black arrow shows the most distal transcription start site (+1) of mtt (predicted by Flybase). Arrowhead indicates the orientation of transcription of the CG12780 gene. The f06268 line carries a piggyBac transposon inserted 35 bp downstream the third mtt exon. The f01266 piggyBac transposon and the NP4288-GAL4 enhancer trap lines are inserted 1.7 Kb downstream and 3.1 Kb upstream, respectively, from the transcription start site of mtt. The Df(2R)Exel7096 line corresponds to a small deficiency that completely removes the mtt locus and some adjacent genes (CG8697 to CG2397). (B) QRT-PCR analysis on mtt mutants. The expression levels of mtt RNA on homozygous wild-type (WT), f06268, Df(2R)Ex7096, and f01266 adults flies were measured by quantitative real-time reverse-transcriptase polymerase chain reaction (QRT-PCR). WT and f01266 homozygous flies show comparable levels of mtt RNA, whereas no amplification is obtained from the f06268 and the Df(2R)Ex7096 homozygous lines. Error bar indicates SEM.
Figure 5.
mtt is present in chemosensory sensilla.
(A–D) and (E) show mtt in situ hybridization on WT and Df(2R)Ex7096 mtt mutant labellum, respectively. (A) In WT labellum, single cells (arrows indicate two single cells in focus) are labeled by the mtt riboprobe. (A) is a composite image of one image focusing on the labeled cells and one image focusing on the chemosensory bristles. (B) and (C) are high-magnification views of one sensilla visible in (A) (upper arrow). (B) focuses on the labeled cell. (C) is a composite image of (B) and a different focal plane in which the nearby chemosensory bristle is visible. Note that one single neuron-like cell (indicated by an arrow) is labeled near the chemosensory bristle visible in (C). This cell is likely a neuron, as an axon is weakly visible (indicated by an arrowhead). (D) shows another sensilla with a different Nomarski setting illustrating that the single labeled cell (indicated by a black arrow) is present in a chemosensilla housing at least five cells. (E) No signal is detected on Df(2R)Exel7096 mutant labellum. (E) is a composite image of one image focusing within the labellum and one image focusing on the chemosensory bristles. (F–I) Labellum of NP4288-GAL4/UAS-nlsGFP showing the GFP-expressing taste neurons. (F) In one labial palp, around 28 GRNs are present (see also Figure S3). (I) is a composite image of the fluorescent and bright-field images shown in (G) and (H), respectively. Arrowhead in (G and I) indicates the dendrite innervating the taste bristle sensilla. (J and K) NP4288-GAL4/UAS-mCD8GFP double homozygous foreleg showing that taste neurons are labeled in the tarsi. Arrowhead in (J) indicates the dendrite of the taste neuron. Note that only one of the two labeled neurons is visible at this focal plane. (K) is a composite image of the fluorescent and bright-field images.
Table 1.
Expression patterns of the GAL4 lines described in this study.
Figure 6.
L-Canavanine does not affect the occurrence of PER, but induces a DmXR-dependent retraction of the proboscis after the reflex.
For behavioral analyses, the solutions (100 mM sucrose in white and 100 mM sucrose+40 mM l-canavanine in black) were put in contact with the leg tarsi during the 5-s assay. l-Canavanine (Cana) does not affect the percentage of proboscis extension reflex (PER) in WT and mtt mutant (mttf06268/mttf06268 and mttf06268/DfEx7096) flies. After the PER response, l-canavanine triggers the proboscis retraction (PR) in WT, but not in mtt mutant flies. Error bars indicate SEM. Asterisks indicate significant differences by t-test (p<0.001).
Figure 7.
mtt is required in Gr66a-GRNs.
(A) Knockdown of mtt expression by RNAi in NP4288-GRNs and Gr66a-GRNs suppresses the l-canavanine (Cana)-induced PR gustatory phenotype. Histograms show the percentage of PR of controls (NP4288/+, mttf06268/+;UAS-mtt RNAi/+, UAS-mtt RNAi/+, and Gr66a:mttf06268/+) and mtt heterozygous flies expressing the mtt RNAi construct in NP4288-positive GRNs (NP4288/mttf06268;UAS-mtt RNAi/+) or in Gr66a-GRNs (Gr66a:mttf06268;UAS-mtt RNAi/+). Compared to controls, the down-regulation of mtt in NP4288-GRNs or in Gr66a-GRNs suppresses the l-canavanine–induced PR. For all genotypes, l-canavanine does not significantly affect the percentage of PER (see Figure S4A). Behavioral analyses were performed as described in Figure 6. Error bars indicate SEM. Double asterisks indicate significant differences by t-test (p<0.001). (B) Expression of mtt in bitter-sensitive Gr66a-GRNs rescues the PR mutant gustatory phenotype of mtt mutant flies. Histograms show the percentage of PR of control flies (mttf06268/+;UAS-mtt/+), mtt mutant flies carrying one copy of each GRN GAL4 (Gr66a:mttf06268/mttf06268, Gr5a:mttf06268/mttf06268, and NP1017/+;mttf06268/mttf06268) and mtt mutant flies expressing mtt in bitter-, sugar-, and water sensitive GRNs (Gr66a:mttf06268/mttf06268;UAS-mtt/+, Gr5a:mttf06268/mttf06268;UAS-mtt/+, and NP1017/+;mttf06268/mttf06268;UAS-mtt/+, respectively). Note that only mtt mutant flies expressing mtt in Gr66a-GRNs show a rescue of the l-canavanine–induced PR phenotype.
Figure 8.
DmXR is localized to GRN dendrites and is dispensable for caffeine sensitivity.
(A) Subcellular localization of the DmX receptor. Anti-HA immunostaining on labellum of Gr66a-GAL4/+;UAS-HAmtt/+ flies. HAmtt is a HA-tagged version of the DmX receptor. When DmXR is expressed in Gr66a-GRNs, it is principally targeted to the dendrite. Note the strong localization of the receptor at the tip of the dendrite (arrowhead). (B) The aversive effect of caffeine is not affected in mtt mutant flies. Behavioral analyses were performed as described in Figure 6 except that 100 mM sucrose (white) and 100 mM sucrose+50 mM caffeine (grey) solutions were used. Histograms show the percentage of PER and PR of control (WT) and mtt mutant (mttf06268/mttf06268) flies. For both control and mtt mutant flies, the caffeine inhibits the sucrose-induced PER. This effect is not complete as some flies still performed a PER in the presence of caffeine. Most of these flies retract their proboscis in response to caffeine, similar to what is observed with l-canavanine. Error bars indicate SEM. Double asterisks indicate significant differences by t-test (p<0.001).