Mating status-dependent dopaminergic modulation of auditory sensory neurons in Drosophila

Abstract

Mating status often modulates responses to courtship sounds in animals. The neural mechanisms underlying this modulation, however, have not been well clarified. Here, we show that dopaminergic signals are involved in modulating the responses of auditory sensory neurons in Drosophila melanogaster females depending on their mating status. These neurons abundantly express three types of dopamine receptors, with some having direct synaptic connections with dopaminergic neurons. Of these receptors, suppressing the expression of Dop1R2 reduces sound responses of auditory sensory neurons in unmated but not mated females. Moreover, the expression of Dop1R2 in auditory sensory neurons enhances the song response behavior of unmated females, manifested by copulation receptivity when exposed to songs. Our research suggests that dopaminergic modulation via Dop1R2 is involved in mating the state-dependent regulation of auditory sensory processing.

Keywords: Behavioral neuroscience; Biological sciences; Natural sciences; Neuroscience; Sensory neuroscience.

Graphical abstract

Figure 1

Dopamine receptor expression patterns in JO neurons (A) Expression levels of dopamine receptors in JO neurons. Dots represent the normalized expression level of each gene in single JO neurons obtained from the single-nucleus RNA-seq data. (B–I) Expression patterns of dopamine receptors in JO (B–E) and AMMC in the brain (F–I). DsRed marker was expressed by T2A-Gal4 knock-in lines to visualize putative dopamine-receptor-expressing cells (magenta). Dop1R1-Gal4 (B and F), Dop1R2-Gal4 (C and G), Dop2R-Gal4 (D and H), and DopEcR-Gal4 (E and I) (Kondo et al.³⁵) were used. Grayscale images of the top panels are shown in the lower panels. Elav (maker of neuronal cell bodies) immunolabeling was used to label JO neurons (green) within the second antennal segment (black square in the left diagram) (B–E). The rCD2::GFP marker expression was driven by the NV0114 strain, which labels most JO neurons (green) (F–I). JO neurons in each subgroup innervate into five subregions of the AMMC in the brain (yellow dotted lines and the left diagram in F–I). D, dorsal; L, lateral (the same in the following figures). JO, Johnston’s organ; AMMC, antennal mechanosensory and motor center. Scale bar = 10 μm. Genotypes of flies are listed in Table S1 (the same as in the following figures). See also Figures S1–S3, and Tables 1, 2, and S1.

Figure 2

JO neurons have synaptic connections with dopaminergic neurons (A) trans-Tango analysis. trans-Tango signals are induced in postsynaptic neurons. (B and C) TH-Gal4 -induced trans-Tango signals in JO (B) and AMMC (C). tdTomato was used to visualize trans-Tango signals, labeled with anti-DsRed antibodies (magenta). In the JO, cell bodies of JO neurons were labeled with anti-Elav antibodies (green; marker of neuronal cell bodies). In the AMMC, synapses were labeled with nc82 antibodies (green). Scale bar = 10 μm. See also Figure S4 and Tables 3, 4, 5, and S1.

Figure 3

Dop1R2 expression increases JO-A neuron responses in unmated females (A) Experimental setup of calcium imaging. Unmated females were stabilized ventral side up. Artificial courtship songs (35 ms and 75 ms IPIs) and 100-Hz pure tone were played through a loudspeaker. (B) The region of interest (ROI, outlined in white) for the analysis. Two ROIs were set at the axon bundle of JO-A and JO-B neurons in the AMMC as indicated. A, anterior. Scale bar = 50 μm. (C and D) Calcium responses of JO-A (C) and JO-B (D) in unmated females with Dop1R2 knockdown in JO neurons. Welch’s t-tests corrected with the Benjamini and Hochberg method; n = 10–11 per genotype. Left, Time traces of raw ΔF/F responses. Thin and bold lines show time traces of the response in each individual and the average of all individuals, respectively. Gray-shaded areas indicate the time window of sound playback. Right, peak calcium responses. Dots and bars show peak responses in each individual and the average of all individuals, respectively (the same in the following figures). (E–G) Estimated increase and decrease rates in JO-A neurons. Welch’s t-tests corrected with the Benjamini and Hochberg method; n = 10–11 per genotype. A differential equation that incorporates three factors is fitted to the time series data of calcium responses. $r$, estimated increase rate during the stimulus (E); $d_1$, decrease rate during the stimulus (F); $d_2$, decrease rate after the stimulus (G). See also Figures S5–S7, S14 and Tables 6, S1. n.s., p > 0.05; ∗, p < 0.05; ∗∗, p < 0.01.

Figure 4

Dop1R2 expression does not affect JO-A responses in mated females (A) Experimental scheme for calcium imaging in mated females. Females are maintained with males (blue) for 12 to 24 h before being subjected to the calcium imaging. (B) Calcium responses in JO-A axons of mated females. Welch’s t-tests corrected with the Benjamini and Hochberg method; n = 11 per genotype. (C) Calcium responses in JO-A axons of pseudo-unmated females (mated with sex peptide null males). Welch’s t-tests corrected with the Benjamini and Hochberg method; n = 16 per genotype. See also Figures S8–S13, S14 and Tables 6, 7, 8, 9, S1. n.s., p > 0.05.

Figure 5

Dop1R2 knockdown in JO neurons decreases the song response behavior of unmated females (A) Female copulation assay. An unmated female was paired with a wing-clipped wild-type male in a chamber. An artificial pulse song was played back through a loudspeaker. The experiment was also conducted under a no-sound condition (See Figures S15 and S16). (B) F-Gal4 driver strain was used to express mCD8::GFP markers (green). nc82 antibodies were used for counter-labeling (magenta). Scale bar = 50 μm. (C–E) Cumulative copulation rates during song playback. Females with Dop1R1 (C), Dop1R2 (D), or DopEcR (E) knockdown in JO neurons were tested. RMST tests corrected with the Benjamini and Hochberg method; n = 40–48 per genotype. (F) Evaluation of female song response behavior with Restricted mean time lost (RMTL). F-Gal4 driver strain was used to express each RNAi construct. Dot and vertical line represent the estimated value and standard error, respectively. RMST tests corrected with the Benjamini and Hochberg method; n = 40–48 per genotype. See also Figures S14–S17 and Tables 10, 11, 12, and S1. n.s., p > 0.05; ∗∗∗, p < 0.001.

Figure 6

Dop1R2 knockdown in JO-AB decreases the song response behavior of unmated females (A–C) Cumulative copulation rates during song playback. Unmated females with Dop1R2 knockdown were tested. RNAi-mediated knockdown was induced in JO-AB (A), JO-A (B), and JO-B neurons (C). Gal4 driver strains that selectively label JO-AB neurons, JO-A neurons, and JO-B neurons (JO15, R74C10, and R91G04 strains, respectively) were used to express the Dop1R2 RNAi construct. RMST tests corrected with the Benjamini and Hochberg method; n = 40–52 per genotype. (D) Comparison of female song response behavior using RMTL. RMST tests corrected with the Benjamini and Hochberg method; n = 40–52 per genotype. Dot and vertical line represent the estimated value and standard error, respectively. (E) Dop1R2 knockdown effect. ΔRMTL represents the difference in RMTL between experimental and control groups in each condition. ΔRMTLs of Dop1R2 knockdown in all subgroups of JO neurons, JO-AB neurons, JO-A neurons, and JO-B neurons are shown. Dot and vertical line represent the estimated value and 95% Confidence Interval, respectively. See also Figures S14–S16 and Tables 10, 11, 12, and S1. ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.001.