Flies were maintained on standard cornmeal agar medium at 25°C with 50% relative humidity under the 12 h light/12 h dark illumination condition. Canton-S was used as a wild-type strain. The DopEcR mutant used in this study is the insertion mutant DopEcR^c02142 (also known as DopEcR^PB1; hereafter der) generated by the Gene Disruption Project (FlyBase Consortium, 2003; Thibault et al., 2004) and has been previously described (FlyBase Consortium, 2003; Inagaki et al., 2012; Petruccelli et al., 2016). der was obtained from the Bloomington Stock Center (stock no. 10847) and backcrossed with Cantonized w¹¹¹⁸ for six generations, and then the X chromosome was replaced with that of Canton-S to remove the w¹¹¹⁸ mutation. elav-GAL4 (stock no. 8765), c739-GAL4 (stock no. 7362), c305a-GAL4 (stock no. 30829), UAS-mCD8-GFP (stock no. 5137) and PTRiP.JF03415 (stock no. 31981; FlyBase Consortium, 2003; Perkins et al., 2015) flies were obtained from the Bloomington Stock Center; NP1131-GAL4 from Dr. Dubnau (Stony Brook University School of Medicine, Stony Brook, NY, USA); fru^NP21-GAL4 from Dr. Yamamoto (Tohoku University, Sendai, Japan); NP225-GAL4 from Dr. Thum (University of Konstanz, Konstanz, Germany); tub-GS-GAL4 from Dr. Kitamoto (University of Iowa, Iowa City, IA, USA); and MB-GS-GAL4 from Dr. Roman (University of Houston, Houston, TX, USA). We have previously described MB247-GAL4 and MB247-GAL4, GAL80^ts (Kim et al., 2007, 2013). DopEcR cDNA containing the open reading frame (Srivastava et al., 2005) was cloned under UAS in the gateway vector pTW (Akbari et al., 2009). The cloned receptor was injected into w¹¹¹⁸ embryos, and germ-line transformed lines were outcrossed with Cantonized w¹¹¹⁸ for six generations to normalize the genetic background and to remove potential second site mutations. Individual transgenes were placed in the der mutant background for rescue experiments. We previously reported the dDA1 (D1) mutant dumb¹ and dumb² (Kim et al., 2007) and the damb mutant defective in DAMB (D5; Cassar et al., 2015). For conditional rescue experiments involving the gene switch lines MB-GS-GAL4 and tub-GS-GAL4, 10 mM RU486 (Mifepristone, M8046, Sigma-Aldrich, Saint Louis, MO, USA) was made in 80% ethanol and added to fly food to the final concentration of 500 μM. Flies were reared on the food containing RU486 for 1 day before and between ethanol exposures. All genotypes used for behavioral analyses including the controls (Canton-S and der mutants carrying only GS-GAL4) were fed with RU486 or vehicle for comparison.

The polyclonal DopEcR antibody was made commercially in a New Zealand white rabbit against the peptide GEPIHDKEYATALAEN that corresponds to the third cytoplasmic loop of the receptor (Pacific Immunology Corp, Ramona, CA, USA). Immunostaining was performed as previously described (Kim et al., 2013; Lim et al., 2014). Briefly, 4–5 day-old male brains were dissected in phosphate buffered saline (PBS) where the trachea around the brain was removed. Dissected brains were individually fixed with 4% PFA (paraformaldehyde and 0.04 M Lysine in PBS) at 4°C for 3 h and then rinsed three times in PBHT containing 0.5% Triton X-100 for 10 min each. Brains were solubilized in 1% Triton X-100 in PBHT for 1 h, incubated in the blocking solution (5% normal goat serum in PBHT) for 2 h and then incubated with the anti-DopEcR antibody (1:100 diluted in the blocking solution) at room temperature overnight. Brains were washed four times in PBHT for 1 h at room temperature and then overnight at 4°C before incubation with the goat Alexa 488-conjugated anti-rabbit IgG (Molecular Probes, Carlsbad, CA, USA) at room temperature for 2 h. After washes in PBHT, PBS and 0.12 M Tris-HCl, pH 7.4 (three times in each solution), brains were mounted in the VECTASHIELD medium (Vector Labs, Burlingame, CA, USA). Images were taken using the Zeiss LSM 700 confocal microscope (Carl Zeiss, Thornwood, NY, USA) and analyzed using the ImageJ software (NIH).

One to two-day-old males were collected under carbon dioxide (CO₂) and aged in food vials for 2–3 days before tests. A group of 33 males was used as one data point in all behavioral tests. Ethanol exposure was performed in the Flypub consisting of a plastic chamber (57 mm D × 103 mm H) with the clear ceiling for videotaping behavior and the open bottom for administering ethanol as previously described (Lee et al., 2008). Flies were acclimated to the chamber for 10 min before ethanol exposure. A small petri dish containing a cotton pad applied with 1 ml of 95% ethanol was inserted to the bottom opening and flies were exposed to ethanol vapor till they were sedated. Four to six Flypubs were recorded together using a HD video camera (Q2F-00013 Microsoft LifeCam Studio, Redmond, WA, USA). The recorded movie files were used to score courtship activity. Flies were exposed to ethanol every 24 h for six consecutive days and were kept in food vials between exposures. The sedative effect of ethanol was measured by counting every 2 min the number of flies lying on their back or immobile for over 10 s. To obtain the mean sedation time (MST), the total sedation time, i.e., ∑(the number of sedated flies at each time interval × each time interval after ethanol administration, e.g., 2, 4, 6 and etc.), was divided by the total number of flies (Lee et al., 2008). Courtship activity consisting of singing (unilateral wing vibration), licking or attempted copulation (Baker et al., 2001) was monitored during 30 s (1 block) and the maximum number of flies engaged in courtship at a given time was scored. The average of 10 consecutive blocks (i.e., 5 min) giving the highest value was used to represent the percentage of males engaged in active intermale courtship per Flypub (Lee et al., 2008). Our earlier study (Lee et al., 2008) has shown that the maximal level of ethanol-induced courtship disinhibition is achieved on the exposure 4 or 5 and then maintained steady. Thus, we focus on exposure 1 for the initial level of ethanol-induced disinhibition, exposure 2 for sensitization induction and exposure 6 for maintenance in this study. The genotypes were blinded to the experimenters conducting ethanol exposure and scoring courtship or sedation.

Statistical analyses were performed using Minitab 16 (Minitab, State College, PA, USA) and JMP 13 (SAS, Cary, NC, USA). All data are reported as mean + or ± standard error of means (SEM). Normality was determined by the Anderson Darling goodness-of-fit test. Normally distributed data were analyzed by a two-tailed Student’s t-test or analysis of variance (ANOVA) with post hoc Tukey-Kramer HSD or Dunnett’s tests. Non-normally distributed data were analyzed by Kruskal-Wallis and post hoc Mann-Whitney tests.

To investigate the roles of D1 family receptors in chronic ethanol effects, we employed the Flypub for mild ethanol delivery (Lee et al., 2008). We first measured the sedative effect of ethanol. Compared to the control Canton-S males, it took longer for der mutant males to get sedated (p < 0.0001; Figure 1A), demonstrating that der males defective in DopEcR have decreased sensitivity to the sedative effect of ethanol. This corroborates the finding by Petruccelli et al. (2016). In contrast, dumb and damb males defective in dDA1 (D1) and DAMB (D5) receptors, respectively, exhibited normal sensitivity (p > 0.05, Figure 1B). When MSTs of dumb, damb and der males were examined during daily ethanol exposures, all mutants developed tolerance similar to Canton-S (Canton-S: F_(3,101) = 35.9762, p < 0.0001; der: F_(3,90) = 7.4871, p = 0.0002; dumb¹, p < 0.001; dumb², p < 0.0001; p < 0.0001, damb; Figures 1C,D). This indicates that D1 family receptors are not important for tolerance to the sedative effect of ethanol.

Drosophila males typically court females and rarely court males. Under daily ethanol exposure, however, Canton-S males display the escalated levels of intermale courtship (R² = 0.7289, F_(2,48) = 64.5414, p < 0.0001; Figure 2A), which require dopamine neuronal activity (Lee et al., 2008). To explore the mechanism by which dopamine regulates behavioral disinhibition and sensitization, we examined the D1 family receptor mutants’ courtship behavior under the influence of ethanol. Both dumb and damb males developed behavioral sensitization to the disinhibition effect of ethanol (dumb¹: R² = 0.8966, F_(3,24) = 69.3456, p < 0.0001; dumb²: R² = 0.7936, F_(3,24) = 30.7652, p < 0.0001; damb: R² = 0.9316, F_(3,20) = 90.8291, p < 0.0001; Figure 2B). der males, on the other hand, exhibited the substantially reduced levels of intermale courtship on all exposures compared to Canton-S males (p < 0.0001; Figure 2A). This suggests that DopEcR is required for behavioral sensitization to the disinhibition effect of ethanol.

To identify the neural structure where DopEcR regulates behavioral sensitization, we employed the GAL4/UAS binary system and RNA interference (RNAi) for cell type-specific knockdown of DopEcR expression. In this study we used an additional control line carrying UAS-GFP and UAS-DopEcR RNAi since courtship behavior could be sensitive to the mw in a transgenic construct (Lee et al., 2008). To establish effectiveness of DopEcR RNAi, we used the pan-neuronal driver elav-GAL4 to express double-stranded DopEcR RNA for RNAi in all neurons. Like der mutants, the flies with pan neuronal DopEcR knockdown showed severe impairment in behavioral sensitization (p < 0.0001; Figure 3A). We reasoned that the neural substrate for the DopEcR’s function in behavioral sensitization could be the neurons regulating courtship behavior or high order brain structures mediating learning and memory. Fruitless-expressing neurons control male courtship behavior (Manoli et al., 2005; Stockinger et al., 2005) thus represent a potential neural site for the DopEcR’s function. The projection neurons are another candidate for the DopEcR’s function because they have dendrites in the antennal lobes and axons at the lateral horn and the MB calyx that are high order brain centers for pheromone information processing, learning and memory (Thum et al., 2007; Grosjean et al., 2011). When DopEcR was knocked down in Fruitless neurons, we did not observe a significant change in behavioral sensitization (p > 0.05; fru-GAL4 in Figure 3A) while DopEcR knockdown in the projection neurons resulted in slightly increased sensitization (p = 0.0186; NP225-GAL4). Above all, we observed markedly reduced sensitization in the flies with DopEcR knockdown in the MB neurons (p < 0.0001; MB247-GAL4 in Figure 3B). The MB consists of αβ, α’β’ and γ neurons where MB247-GAL4 is expressed in αβ and γ neurons. We next asked whether DopEcR in each MB substructure is sufficient for behavioral sensitization. When DopEcR RNAi was induced only in αβ, α’β’ or γ neurons via the c739-, c305a- or NP1131-GAL4 driver, respectively, the flies developed normal behavioral sensitization (p > 0.05). This suggests that DopEcR in the αβ and γ, but not αβ or γ alone, is needed for behavioral sensitization to the disinhibition effect of ethanol.

DopEcR is expressed throughout development and adulthood (FlyBase Consortium, 2003; Inagaki et al., 2012; Ishimoto et al., 2013; Petruccelli et al., 2016). To test whether the sensitization phenotype is caused by developmental or physiological DopEcR deficiency, we adopted two approaches, TARGET and Gene Switch (GS) for temporally restricted reinstatement of DopEcR expression in the MB neurons of der mutants. TARGET (McGuire et al., 2004) is the GAL4/UAS combined with GAL80^ts that confers temporally restricted expression of a transgene downstream of UAS, which we used successfully in the study of dDA1 in olfactory memory formation (Kim et al., 2007). Briefly, GAL80^ts is active as a GAL4 repressor at 20°C but inactive at 30°C, allowing GAL4 activity thereby UAS activation. The der mutants carrying tub-GAL80^ts, MB247-GAL4 and UAS-DopEcR cDNA were reared at 30°C throughout development but maintained at 20°C right after eclosion to induce DopEcR expression only during development (Figure 4A). To induce DopEcR only during adulthood, possibly at the time of ethanol exposure, the der mutants carrying tub-GAL80^ts, MB247-GAL4 and UAS-DopEcR cDNA were reared at 20°C throughout development but maintained at 30°C 2 days after eclosion. Canton-S and der mutant carrying tub-GAL80^ts and MB247-GAL4 but not UAS-DopEcR cDNA were treated with the same temperature manipulation to serve as controls. As shown in Figure 4A, the der males with DopEcR expression only during development exhibited impaired behavioral sensitization thus there was no rescue (F_(2,16) = 23.2, p < 0.0001). In contrast, the der males with DopEcR expression only during adulthood fully reinstated behavioral sensitization (p > 0.05 compared to Canton-S; Figure 4B). This indicates the role of DopEcR during adulthood for disinhibition sensitization.

We observed that the flies with the temperature manipulation displayed highly variable ethanol sensitivity and sensitization. Thus as a complementary approach, we used the GS system in which GAL4 is fused to the progesterone receptor. Only in the presence of the steroid RU486, GAL4 can activate UAS for downstream gene expression (Roman et al., 2001). We tested the der mutants carrying UAS-DopEcR cDNA and tub-GS-GAL4 or MB-GS-GAL4 for ubiquitous or MB expression of DopEcR, respectively, at the time of ethanol exposure. When treated with RU486, the der males with DopEcR expression in all cells or MB neurons displayed the level of sensitization substantially higher than that of the der males carrying only tub-GS-GAL4 or MB-GS-GAL4, but comparable to the Canton-S level (F_(6,24) = 43.2375, p < 0.0001; Figure 4C). The der males carrying the same transgenes (i.e., UAS-DopEcR-cDNA and tub-GS-GAL4 or MB-GS-GAL4) that were not fed with RU486 exhibited impaired sensitization similar to the der mutants carrying tub-GS-GAL4 or MB-GS-GAL4 (p > 0.05 by post hoc Tukey-Kramer HSD test; Figure 4C). These observations together demonstrate that DopEcR expression during adulthood is sufficient for sensitization, supporting the physiological role of DopEcR at the time of ethanol exposure for this behavioral plasticity.

The study of DopEcR enhancer-GAL4 shows that DopEcR is expressed in the MB αβ and γ neurons (Ishimoto et al., 2013). It is however unclear where DopEcR is localized in the MB. To address this, we used immunohistochemical analysis. We made the fusion construct of Glutathione S-transferase and the third cytoplasmic loop of DopEcR as we have previously characterized the dDA1 and DAMB expression patterns (Han et al., 1996, 1998). We also made the antibody against the peptide corresponding to part of the third cytoplasmic loop. The antibodies made against the fusion protein in rabbits and mice did not provide reliable staining; however, the antibody made against the peptide revealed consistent staining in the MB neuropil. It is worth mentioning that the antibody did not penetrate inside the brain under numerous conditions that we tried and also strongly stained the cell membrane of nearly all neurons and glia in both Canton-S and der (Figure 5; Supplementary Movie Files). Nonetheless, DopEcR immunoreactivity was clearly visible in the MB calyx (dendritic structure; Figure 5A, Supplementary Figure S1), α lobe core and β lobe (axonal structure) in the Canton-S brain (Figure 5C, Supplementary Movie 1). DopEcR immunoreactivity in the γ lobe was also detectable but at a very low level (Figure 5C and Supplementary Movie 1). On the contrary, DopEcR immunoreactivity in all MB neuropil was barely detectable in the der brain (Figures 5B,D, Supplementary Figure S1 and Supplementary Movie 2). These observations suggest that the site of DopEcR’s function for sensitization is the MB dendrites in the calyx or axons in the α, β or γ lobe, or both locations.