Co-expression of Dop1R1 and Dop2R genes in adult Drosophila brain.

(A and B) The expression of Dop1R1-T2A-GAL4 and Dop2R-T2A-GAL4 visualized by UAS-mCD8::GFP (green) The neuropil was stained by using Brp::SNAP (red). Maximum-intensity projections of the whole brain.

(C) Schematic of the KCs and the MB-innervating dopamine neurons from the PAM and PPL1 clusters.

(D-G) Double labelling of Dop1R1 and Dop2R gene expression visualized by Dop1R1-T2A-LexA/lexAop-rCD2::GFP (green) and Dop2R-T2A-GAL4/UAS-CD4::tdTomato (red). Dopamine neurons were immunostained with anti-TH antibody (blue). Single optical sections. Cell bodies of the PAM cluster (D) and the PPL1 cluster (E). The ellipsoid body (F) and ensheathing glia (G). Scale bars, 50 µm (A and B), 5 µm (D-G).

Cell type-specific visualization of endogenous proteins with GFP11 tag.

(A) Principle of cell type specific fluorescent labelling of target proteins by GFP11 tag. GFP1-10 and membrane marker CD4::tdTomato are expressed in the target cells by GAL4/UAS system. In the target cells, reconstitution of GFP occurs on the endogenous proteins tagged with GFP11.

(B) As an example, DopEcR::GFP11 is visualized in KCs using MB-GeneSwitch, a ligand-inducible GAL4 driver. A merged image of reconstituted GFP (green) and cellular membrane visualized by UAS-CD4::tdTomato (magenta). Maximum intensity projection of the whole left MB.

(C) The workflow for visualizing subcellular protein enrichment by localization index (LI). A single sagittal section of the MB calyx and peduncle is shown. The ratio of reconstituted GFP to membrane signal in the left image is calculated and normalized by the mean intensity of all voxels to provide LI. In the middle image, LI is color-coded so that red represents local receptor enrichment. In the right image, the LI color is mapped on the membrane signal.

Subcellular localization of Dop1R1 and Dop2R in the Kenyon Cells.

Subcellular localization of Dop1R1 and Dop2R in the KCs is visualized by GFP11-tag. MB-GeneSwitch was used to express GFP1-10 and CD4::tdTomato in the KCs. Receptors in αβ KCs were visualized after 12 hours of RU486 feeding (see Materials and Methods).

(A) Schematic of the distribution of presynapses and postsynaposes in the KCs.

(B-G) Enrichment of Dop1R1 and Dop2R in the MB lobe. Frontal views of the MB (B and D) and sagittal views of the MB (C, E, F and G) of the maximum-intensity projections of the whole left MB are shown. Reconstituted GFP signals for both Dop1R1:: and Dop2R::GFP11 distributed throughout the MB lobe and the calyx (B-E). Visualization by LI showed more pronounced enrichment of Dop2R (G) than Dop1R1 (F) in the lobe as exemplified in the tip (arrows, 10 µm square, magnified in the insertion). Scalebars, 20 µm (B-G).

(H) Mean LI for Dop1R1 and Dop2R in the calyx, the peduncle, the tip of α and β lobes. Student’s t test was performed to compare LI of Dop1R1 and Dop2R in each region (N = 3). Error bars; SEM. * p< 0.05.

Presynaptic localization of Dop1R1 and Dop2R in Kenyon Cells and giant neurons.

(A and B) Double labelling of dopamine receptors (green) and Brp (magenta). Single focal slice at the tip of the vertical lobe. Insertions are the magnified images of white squares in the main panel.

(C) Colocalization analysis showed that Dop2R had significantly higher colocalization with Brp than Dop1R1 (p = 0.0259, Mann-Whitney U test, N = 6). Error bars; SEM.

(D-E) Punctate Brp expression in a giant neuron culture differentiated from cytokinesis-arrested neuroblasts of OK371-GAL4/UAS-mCD8::GFP embryos. Aggregated Brp condensates (magenta) were observed in the neurite terminals of the cells marked with mCD8::GFP (cyan) in (E).

(F-I) Double labelling of dopamine receptors (green) and the AZs (magenta). Dop1R1::Venus

(F and G) or Dop2R::Venus (H and I) was crossed with Brp::SNAP.

Scale bars, 20 µm (A and B), 10 μm (D, F and H), 5 μm (E), 1 μm (G and I).

Dop2R and Dop1R1 at the presynaptic sites of Kenyon Cells in the calyx.

(A-F) Double labelling of dopamine receptors (green) and Brp (magenta). From the single focal slice of the calyx (A and D), white squares including dendritic claws (B and E) and proximal dendrites (E and F) are magnified.

Scalebars, 20 µm (A and D).

Subcellular localization of Dop1R1 and Dop2R in dopamine neurons.

(A-B) Maximum-intensity projection image showing the distribution of presynaptic sites in the PAM neurons. (A) R58E02-GAL4 was used to express UAS-mCD8::GFP (magenta) and UAS-nSyb-CLIP (magenta). (B) Visualization by LI showing enrichment of nSyb signals in the lobe projection of the PAM neurons.

(C) Illustrated projection pattern of the PAM neurons. Red puncta indicate the sparse distribution of presynaptic sites in dendrites.

(D-L) Subcellular localization of GFP11-tagged Dop1R1 and Dop2R in the PAM neurons. R58E02-GAL4 was used to express UAS-GFP1-10and UAS-CD4::tdTomato in the PAM neurons.

(D and E) Reconstituted GFP signals of Dop1R1::GFP11 (D) and Dop2R::GFP11 (E) in the PAM neurons. Maximum-intensity projections of the left hemisphere including the MB lobe, SMP, SIP and SLP (superior medial, intermediate, and lateral protocerebrum, respectively). (F and G) Stronger presynaptic enrichment of Dop2R (G) than that of Dop1R1 (F) in the PAM dopamine neurons visualized by LI. β’1 compartment is magnified in the insertion image.

(H-J) R15A04-GAL4 was used to measure LI in the PAM-β’1 neuron. (H and I) Presynaptic terminals of the PAM-β’1 neurons are shown (dashed line). (H) Mean LI for Dop1R1 and Dop2R in β’1 (p = 0.0056, Mann-Whitney U test, N = 9). Error bars; SEM.

(K and L) A single optical slice of γ5 compartment in the MB lobe obtained using Airyscan. Merged image of reconstituted GFP (green) and CD4::tdTomato (magenta). Insertions are the magnified images of the presynaptic boutons of PAM-γ5 (white squares).

Scale bars, 20 µm (A-B, D-I), 5 µm (K and L).

Bidirectional modification of dopamine receptor expression in dopamine neurons.

(A) Schematic diagram of the starvation protocol. (B) Schematic illustration of the MB projection of the PAM and PPL1 dopamine neurons marked by R15A04-GAL4 and TH-GAL4. (C and D) Dop1R (C) and Dop2R (D) in the presynaptic terminals of PAM-γ5 before and after 48 hours of starvation.

(E) Quantification of dopamine receptor levels in the presynaptic terminals of PAM-γ5 after 0, 10, 24 and 48 hours of starvation (n = 6-13).

(F and G) Reconstituted GFP signals of Dop1R1::GFP11 (F) and Dop2R::GFP11 (G) in the PPL1 neurons. In the MB projections of the PPL1 neurons, Dop1R1 is expressed in only the α3 compartment (F). Dop2R is expressed in all MB projections (G). Maximum-intensity projections of the MB lobe.

(H and I) Dop1R (H) and Dop2R (I) in the presynaptic terminals of PPL1-α3 before and after 24 hours of starvation.

(J) Quantification of the dopamine receptor levels in the presynaptic terminals of PPL1-α3 after 0, 10 and 24 hours of starvation (n = 7-10).

Interaction effects between genotypes and starvation time on protein levels were tested by Two-way ANOVA (E and J).

Scale bar, 10 µm (C, D, H and I), 20 µm (F and G).

Error bars; SEM (E and J). * p< 0.05, ** p< 0.01, *** p< 0.001, n.s. = not significant.

Starvation-dependent change of dopamine receptors in PAM and PPL1.

(A-F) Dop1R (A-C) and Dop2R (D-F) in the presynaptic terminals of PAM-α1 (A and D), PAM-β2 (B and E) and PAM-β’1 (C and F) before and after 48 hours of starvation.

(G and H) Quantification of dopamine receptor levels in the presynaptic terminals of the PAM neurons after 0, 10, 24 and 48 hours of starvation (n = 6-13).

(I-K) Dop2R in the presynaptic terminals of PPL1-α’2 (I), PPL1-γ1pedc (J) and PPL1-γ2 (K) before and after 24 hours of starvation.

(L) Quantification of Dop2R levels in the presynaptic terminals of the PPL1 neurons after 0, 10 and 24 hours of starvation (n = 7-10).

Scale bar, 10 µm (A-F, I-K). Error bars; SEM (G, H and L). * p< 0.05, ** p< 0.01, *** p< 0.001, n.s. not significant.

The dual dopaminergic feedback regulating starved-state dependent expression of appetitive behavior.

(A) A working model showing the role of the dual dopaminergic feedback regulation. In starved state, increased Dop1R1 in PAM neurons and increased Dop2R in PPL1 neurons changes the balance between the synaptic outputs from these DANs to favor appetitive behavior.

(B) According to the model, loss of Dop2R in PPL1 upregulates output from PPL1 to attenuate appetitive behavior in starved flies.

(C) Knockdown of Dop2R in the PPL1 neurons by MB504B-GAL4 reduced 3-hour appetitive memory performance (n=14-15). Error bars; SEM. * p< 0.05.