Transient Dysregulation of Dopamine Signaling in a Developing Drosophila Arousal Circuit Permanently Impairs Behavioral Responsiveness in Adults

Abstract

The dopamine ontogeny hypothesis for schizophrenia proposes that transient dysregulation of the dopaminergic system during brain development increases the likelihood of this disorder in adulthood. To test this hypothesis in a high-throughput animal model, we have transiently manipulated dopamine signaling in the developing fruit fly Drosophila melanogaster and examined behavioral responsiveness in adult flies. We found that either a transient increase of dopamine neuron activity or a transient decrease of dopamine receptor expression during fly brain development permanently impairs behavioral responsiveness in adults. A screen for impaired responsiveness revealed sleep-promoting neurons in the central brain as likely postsynaptic dopamine targets modulating these behavioral effects. Transient dopamine receptor knockdown during development in a restricted set of ~20 sleep-promoting neurons recapitulated the dopamine ontogeny phenotype, by permanently reducing responsiveness in adult animals. This suggests that disorders involving impaired behavioral responsiveness might result from defective ontogeny of sleep/wake circuits.

Keywords: D1 receptor; genetics; ontogeny; schizophrenia; sleep; visual.

Figure 1

Transient activation of dopamine during development decreases behavioral responsiveness to mechanical stimuli in adults. (A) Timeline of experiment. Th-Gal4/UAS-TrpA1 flies were exposed to elevated temperatures (31°C) during their late pupal stage, which activates dopaminergic neurons specifically. Behavioral experiments were then performed on adult males at room temperature, using the Drosophila ARousal Tracking (DART) system. (B) Adult flies were placed in individual tubes with access to food, and their responsiveness to mechanical stimuli (vibrating motors) was monitored hourly over 3 days and nights using DART. (C) Average speed (mm/s) of Th-Gal4/UAS-TrpA1 flies (N = 32) to hourly mechanical vibrations for day (light gray) and night (dark gray). (D) Average speed of identically treated UAS-TrpA1/+ genetic controls (N = 32). (E) Average speed of identically treated Th-Gal4/+ genetic controls (N = 32). (F) Average daytime responsiveness of treated Th-Gal4/UAS-TrpA1 animals (maroon) compared to genetic controls (gray). (G) Average responses are compared to each other by zeroing the baseline (pre-stimulus) speed, and summarized average daytime responsiveness (mm/s ± SEM) is shown in the histogram. (H) Average nighttime responsiveness of treated Th-Gal4/UAS-Gal4 animals (maroon) compared to genetic controls (gray). (I) Average nighttime responsiveness (mm/s ± SEM) for the three strains. ***P < 0.001, by one-way ANOVA, adjusted for multiple comparisons by Post Hoc Tukey’s test.

Figure 2

Transient activation of dopamine during pupal development does not alter waking activity or sleep duration. (A) Average daytime (light-gray background) and nighttime (dark-gray background) pre-stimulus speed (mm/s ± SEM) for treated Th-Gal4/UAS-TrpA1 animals (maroon, N = 32) compared to UAS-TrpA1/+ (dark gray, N = 32) and Th-Gal4/+ (light gray, N = 32). Flies are the same as in Figure 1. (B) Average daytime (light-gray background) and nighttime (dark-gray background) wake activity (mm/s speed per waking minute ± SEM, see Materials and Methods) for treated Th-Gal4/UAS-TrpA1 animals (maroon) compared to genetic controls (gray). (C) Average daytime (light-gray background) and nighttime (dark-gray background) sleep duration in minutes of sleep/hour ± SEM for treated Th-Gal4/UAS-TrpA1 animals (maroon) compared to genetic controls (gray). (D) Average daytime sleep intensity (% immobile flies that reacted to the stimulus ± SEM). Data [same color scheme as in (A–C)] are divided into three groups, depending on how long flies were immobile prior to the stimulus event. (E) Average nighttime sleep intensity (% reactive ± SEM). *P < 0.05, **P < 0.01; ***P < 0.001, by one-way ANOVA, adjusted for multiple comparisons by Post Hoc Tukey’s test.

Figure 3

Transient activation of dopamine during pupal development decreases visual responsiveness in adults. Top: timeline of experiment, exactly the same as in Figure 1. (A) Left panel: the visual arena, consisting of a circular platform surrounded by a moat of water surrounded by six LED arrays displaying virtual objects, two dark bars on a blue background. Right panel: fixation directed to the vertical bars is calculated as the median angle of deviation {(α₉₀₀ + 1) ÷ 2}^th value (see Materials and Methods). (B) Left panel: example of a fixation trace for a typical treated genetic control (Th-Gal4/+, black) and a typical treated Th-Gal4/UAS-TrpA1 fly (maroon). Right panel: average angle of deviation (±SEM) for Th-Ga4/UAS-TrpA1 (maroon, N = 30), UAS-TrpA1/+ (light gray, N = 30), and for Th-Gal4/+ (dark gray, N = 28). *P < 0.05, by MANOVA between grouped means, adjusted for multiple comparisons by a Post Hoc Tukeys test. (C) Left panel: the same visual arena as in (A), but with a moving grating displayed on the LEDs. Right panel: optomotor responsiveness is calculated as median optomotor index (OI) = {(γ₉₀₀ + 1) ÷ 2}^th value (see Materials and Methods) (D) Left panel: example optomotor trace for a treated genetic control (Th-Gal4/+, black) and a treated Th-Gal4/UAS-TrpA1 fly (maroon). Right panel: average OI (±SEM) for Th-Ga4/UAS-TrpA1 (maroon, N = 18), UAS-TrpA1/+ (light gray, N = 20), and for Th-Gal4/+ (dark gray, N = 18). (E) Timeline of experiment where dopamine neurons are transiently activated in adult flies. (F) Average angle deviation (±SEM) of adult flies treated as in E for Th-Ga4/UAS-TrpA1 (red, N = 10), UAS-TrpA1/+ (light gray, N = 10), and for Th-Gal4/+ (dark gray, N = 10). (G) OI (±SEM) of adult flies treated as in (E) for Th-Ga4/UAS-TrpA1 (red, N = 44), UAS-TrpA1/+ (light gray, N = 48), and for Th-Gal4/+ (dark gray, N = 50).

Figure 4

Transient activation of dopamine during development induces hypolocomotion in adults only under specific visual conditions. (A) Timeline of experiment, exactly the same as in Figure 1. (B) Timeline of experiment where dopamine neurons are transiently activated in adult flies. (C) Average speed (mm/s ± SEM) for developmentally treated Th-Ga4/UAS-TrpA1 (left panel, maroon) and adulthood-treated Th-Ga4/UAS-TrpA1 (right panel, red) compared to genetic controls (gray) in response to rotating stimuli. (D) Number of pauses (±SEM) for developmentally treated Th-Ga4/UAS-TrpA1 (maroon) and adulthood-treated Th-Ga4/UAS-TrpA1 (red) compared to genetic controls (gray) in response to rotating stimuli. (E) Total distance traveled (m ± SEM) for developmentally treated Th-Ga4/UAS-TrpA1 (maroon) and adulthood-treated Th-Ga4/UAS-TrpA1 (red) compared to genetic controls (gray) in response to rotating stimuli. (F) Average speed (mm/s ± SEM) for developmentally treated Th-Ga4/UAS-TrpA1 (left panel, maroon) and adulthood-treated Th-Ga4/UAS-TrpA1 (right panel, red) compared to genetic controls (gray) in response to stationary objects. (G) Number of pauses (±SEM) for developmentally treated Th-Ga4/UAS-TrpA1 (maroon) and adulthood-treated Th-Ga4/UAS-TrpA1 (red) compared to genetic controls (gray) in response to stationary objects. (H) Distance traveled (m ± SEM) for developmentally treated Th-Ga4/UAS-TrpA1 (maroon) and adulthood-treated Th-Ga4/UAS-TrpA1 (red) compared to genetic controls (gray) in response to stationary objects. Developmentally treated: Th-Ga4/UAS-TrpA1 (maroon, N = 36), UAS-TrpA1/+ (light gray, N = 39), and for Th-Gal4/+ (dark gray, N = 37). Adult treated: Th-Ga4/UAS-TrpA1 (red, N = 10), UAS-TrpA1/+ (light gray, N = 10), and for Th-Gal4/+ (dark gray, N = 10). **P < 0.01, ***P < 0.001, by MANOVA between grouped means, adjusted for multiple comparisons by a Post Hoc Tukey’s test.

Figure 5

Dopamine levels in pupae and adults following developmental and adult manipulation of dopamine activity. (A) High-performance liquid chromatography (HPLC) was performed on heads collected from pupae or adults, following a heat treatment in the pupal stage (T1 and T2) or following a heat treatment in the adult stage (T3). (B) Developmental manipulation. Left: average pupal dopamine levels ([DA] ± SEM) zeroed to wild type levels [N = 7 (35 flies)], for Th-Ga4/UAS-TrpA1 [maroon, N = 6 (30 flies)], UAS-TrpA1/+ [dark gray, N = 6 (30 flies)] and Th-Gal4/+ [light gray, N = 6 (30 flies)]. Right: average adult dopamine levels ([DA] ± SEM) zeroed to wild type [N = 9 (45 flies)] for Th-Ga4/UAS-TrpA1 [maroon, N = 8 (40 flies)], UAS-TrpA1/+ [gray, N = 8 (40 flies)] and Th-Gal4 [N = 8 (40 flies)] levels. **P < 0.01 by one-way ANOVA, adjusted for multiple comparisons by a Post Hoc Tukey’s test. (C) Adult manipulation. Average adult dopamine levels ([DA] ± SEM) zeroed to wild type [N = 7 samples (35 flies)] levels, for Th-Ga4/UAS-TrpA1 [maroon, N = 7 samples (35 flies)], UAS-TrpA1/+ [gray, N = 11 samples (55 flies)] and Th-Gal4/+ [light gray N = 6 samples (30 flies)]. Significance tested by one-way ANOVA, adjusted for multiple comparisons by a Post Hoc Tukey’s.

Figure 6

Transient pan-neuronal knockdown of Dop1R1 and Dop1R2 during development recapitulates arousal defects in adult animals. (A) Timeline of experiment, as in Figure 1, except that the heat treatment produces knockdown of D1 or D2 receptors. (B) Average responsiveness (white bars, mm/s ± SEM) or sleep duration (black bars, min ± SEM) during the day and night for treated nSyb-Gal4/UAS-Dop1R1/R2 RNAi or D2 RNAi; tubulin (tub)-Gal80TS animals (N = 132, 66, and 75, respectively) compared to nSyb-Gal4/+; tub-Gal80TS genetic controls, set as zero (N = 228). (C) Daytime sleep intensity (% reactive ± SEM) for the same flies as in (B). (D) Nighttime sleep intensity (% reactive ± SEM) for the same flies as in (B). (E) Relative gene expression (±SEM) for Dop1R1 in pupae and adults following developmental knockdown. (F) Relative gene expression (±SEM) for Dop1R2 in pupae and adults following developmental knockdown. Fold change compared to either genetic control (1 or 2) is shown. ***P < 0.001; **P < 0.01; *P < 0.05, by one-way ANOVA, adjusted for multiple comparisons by a Post Hoc Tukey’s test.

Figure 7

Manipulating sleep-promoting neurons impairs behavioral responsiveness. (A) 22 Gal4 circuits were activated with UAS-NachBac and resulting adult progeny were behaviorally characterized. Average nighttime responsiveness (white bars, mm/s ± SEM) or sleep duration (black bars, min ± SEM) for each strain is shown relative to the Gal4 genetic control. Different neuronal categories are indicated with symbols. N = 51 for all genotypes, including each respective Gal4 genetic control. (B) Schema of central brain regions associated with C5-Gal4, 23E10-Gal4, and 201y-Gal4 expression. dFB, dorsal fan-shaped body; vFB, ventral fan-shaped body. (C) Average daytime and nighttime responsiveness and sleep duration (±SEM) of treated C5-Gal4/UAS-Dop1R1; tubulin (tub)-Gal80TS animals, 23E10-Gal4/UAS-Dop1R1; tub-Gal80TS animals, and 201y-Gal4/UAS-Dop1R1; tub-Gal80TS animals (N = 56, 81, and 56, respectively), normalized to their corresponding Gal4 control (C5-Gal4/+; N = 108, 23E10-Gal4/+; N = 159, 201y-Gal4/+; N = 41), and compared to both Gal4 and RNAi (UAS-Dop1R1; tub-Gal80TS/+, N = 81, not shown) genetic controls. (D) Average daytime and nighttime responsiveness and sleep duration (±SEM) of treated C5-Gal4/UAS-Dop1R2; tub-Gal80TS animals, 23E10-Gal4/UAS-Dop1R2; tub-Gal80TS animals and 201y-Gal4/UAS-Dop1R2; tub-Gal80TS animals (N = 34, 66, and 30, respectively), normalized to their corresponding Gal4 control (C5-Gal4/+; N = 108, 23E10-Gal4/+; N = 159, 201y-Gal4/+; N = 41) and compared to both Gal4 and RNAi (UAS-Dop1R2; tub-Gal80TS/+, N = 32, not shown) genetic controls. *P < 0.05, by one-way ANOVA, adjusted for multiple comparisons by a Post Hoc Tukey’s test.

Figure 8

Transient 23E10-Gal4 driven knockdown of D1 receptors during development increases daytime sleep intensity in adult flies. (A) Daytime (left) and nighttime (right) sleep intensity (% reaction proportion ± SEM) of treated C5-Gal4/UAS-Dop1R1; tubulin (tub)-Gal80TS animals (light gray, N = 56) and C5-Gal4/UAS-Dop1R2; tub-Gal80TS (dark gray, N = 34) compared to C5-Gal4/+ (black, N = 108) and RNAi (UAS-Dop1R1; tub-Gal80TS/+, N = 81, not shown) genetic controls. (B) % reaction proportion (±SEM) of treated 23E10-Gal4/UAS-Dop1R1; tub-Gal80TS animals (light gray, N = 81), and 23E10-Gal4/UAS-Dop1R2; tub-Gal80TS (dark gray, N = 66) during the day (left) and night (right) compared to 23E10-Gal4/+ (black, N = 159). *P < 0.05, **P < 0.01, decreased% reaction proportion compared to both genetic controls, by one-way ANOVA, adjusted for multiple comparisons by a Post Hoc Tukey’s test. (C) % reaction proportion (±SEM) of treated 201y-Gal4/UAS-Dop1R1; tub-Gal80TS animals (white, N = 56) and 201y-Gal4/UAS-Dop1R2; tub-Gal80TS (dark gray, N = 30) compared to 201y-Gal4/+ (black, N = 41) and RNAi (UAS-Dop1R2; tub-Gal80TS/+, N = 32, not shown) genetic controls.