Automated real-time quantification of group locomotor activity in Drosophila melanogaster

Abstract

Recent advances in neurogenetics have highlighted Drosophila melanogaster as an exciting model to study neural circuit dynamics and complex behavior. Automated tracking methods have facilitated the study of complex behaviors via high throughput behavioral screening. Here we describe a newly developed low-cost assay capable of real-time monitoring and quantifying Drosophila group activity. This platform offers reliable real-time quantification with open source software and a user-friendly interface for data acquisition and analysis. We demonstrate the utility of this platform by characterizing ethanol-induced locomotor activity in a dose-dependent manner as well as the effects of thermo and optogenetic manipulation of ellipsoid body neurons important for ethanol-induced locomotor activity. As expected, low doses of ethanol induced an initial startle and slow ramping of group activity, whereas high doses of ethanol induced sustained group activity followed by sedation. Advanced offline processing revealed discrete behavioral features characteristic of intoxication. Thermogenetic inactivation of ellipsoid body ring neurons reduced group activity whereas optogenetic activation increased activity. Together, these data establish the fly Group Activity Monitor (flyGrAM) platform as a robust means of obtaining an online read out of group activity in response to manipulations to the environment or neural activity, with an opportunity for more advanced post-processing offline.

Figure 1

flyGrAM apparatus construction. (a) The fully assembled flyGrAM apparatus consists of a behavioral chamber, underlighting array, two LED arrays for optogenetic stimulation experiments (680 nm), an Arduino Uno, a USB camera, and a scaffold for securement. (b) A blown-out schematic of the behavioral chamber consisting of four laser cut acrylic layers. (c) A fully assembled behavioral chamber with loaded flies contained by 2 separate clear acrylic covers. Each individual arena was connected to an air and vacuum source via separate tubes to facilitate airflow. (d) Wiring schematic for infrared LED illumination array. (e) Detailed wiring diagram for Arduino control of optogenetics LED stimulation.

Figure 2

flyGrAM software: (a) flyGrAM User Interface where user can specify their experimental parameters (b) Video frame from an activity experiment, which highlights detected activity with a blue overlay around moving flies from each ROI. (c) Real-time analysis portion of flyGrAM software that tracks active flies and generates real-time activity plots for each ROI.

Figure 3

Ethanol dose effects on group locomotor activity. (a) Group activity counts for each dose were binned over 10 second periods, averaged across biological replicates of 10 flies each (n = 12), and plotted against time for different ethanol doses. Ethanol was delivered over a 10-minute period starting at 300 s as denoted by the gray shaded region. Lines depict mean +/− standard error. Repeated Measures ANOVA with planned contrasts at 140 s, 300 s, 350 s, 600 s, 800 s, and 1100 s (b–g) indicate a significant interaction between group activity and ethanol dose (F(27.27, 342.85) = 21.02, p = 0.000). Mauchly’s test indicated that the assumption of sphericity had been violated (χ²(14) = 67.226, p = 0.000), therefore multivariate tests are reported (ε = 0.779). All posthoc analyses were performed with Bonferroni corrections. (b) Group activity responses during baseline was not significantly different (p = 1.000) (c), Group activity at ethanol onset was not significantly different across doses, except for 80:35, which was significantly different from 40:75 (p = 0.020) and 60:55 (p = 0.020) (d) Group activity at 350 s showed significant differences between ethanol doses: 40:75 and 50:65 were not significantly different from each other (p = 1.000) but were significantly different from the remaining doses (p = 0.010–0.000). 60:55 was significantly different from all ethanol doses (p = 0.002–0.035), except for 40:75 (p = 0.051). 70:45, 80:35 were not significantly different from each other (p = 1.000) but were different from all of the lower doses and the highest dose (110:05, p = 0.026–0.000) 90:25 and 100:15 were not significantly different from each other or 70:45, 80:35, and 110:05 (p = 0.421–1.000) but were significantly different from the lower doses (40:75, 50:65, 60:55, p = 0.001–0.000). Finally, 110:05 was not significantly different from 90:25 and 110:15 (p = 0.421–1.000) but was significantly different from the lower doses (40:75, 50:65, 60:55, 70:45 and 80:35, p = 0.026–0.000). (e) Group activity at 600 s were not significantly different across doses (p = 0.11–1.00), except for 70:45 and 40:75, which were significantly different from each other (p = 0.015). (f) Group activity at 900 s for 40:75, 50:65, 60:55, 70:45, and 80:35 were not significantly different from each other (p = 1.000), however 90:25 was significantly different from all other doses (p = 0.000) and 100:15 and 110:05 were not significantly different from each other (p = 1.000) but were significantly different from the remaining doses (p = 0.000). (g) Group activity at 1100 s was not significantly different for nearly all doses (40:75, 50:65, 60:55, 70:45, 80:35, and 90:25, p = 1.000), however, 100:15 and 110:05 were significantly different from each other and the remaining doses (p = 0.027–0.000). *p < 0.05, ***p < 0.001.

Figure 4

Ethanol affects discrete behavioral features in individual flies. Activity of flies were tracked and analyzed using Ctrax software from two videos during thirty seconds surrounding three timepoints selected from the 60:55 ethanol dose in Fig. 3 (b, d, and e, n = 20). Data was averaged across 1 second bins and plotted against time. Lines depict mean +/− standard error. Repeated Measures ANOVA were performed on featural data that was averaged across flies for each Timepoint (boxplot inset). (a) Angular velocity significantly increased following 5 minutes of ethanol exposure (Timepoint E) as compared to Timepoints B and D (Repeated Measures ANOVA F(2, 38) = 49.872, p = 0.000; Posthoc using Bonferroni correction p = 0.000). (b) Sideways speed also significantly increased at Timepoint E as compared to Timepoints B and D (Repeated Measures ANOVA F(2, 38) = 29.219, p = 0.000); Posthoc using Bonferroni correction p = 0.000). (c) Speed significantly increased at Timepoint E as compared to Timepoints B and D (Repeated Measures ANOVA F(2, 38) = 22.188, p = 0.000; Posthoc using Bonferroni correction p = 0.000). (d) The numbers of flies close and (e) distance between flies didn’t significantly change, but did significantly decrease in variability (Mauchly’s Test of Sphericity χ²(2) = 13.534, p = 0.001, Mauchly’s Test of Sphericity χ²(2) = 9.244, p = 0.01, respectively). (f) Proximity to the wall of the enclosure also significantly decreased variability and distance with ethanol exposure (Mauchly’s Test of Sphericity χ²(2) = 8.649, p = 0.013; Repeated Measures ANOVA with Greenhouse-Geisser correction F(1.448, 27.506) = 5.500, p = 0.017) Posthoc with Bonferroni correction revealed that Timepoint E was significantly different from baseline (Timepoint B p = 0.046). *p < 0.05, **p < 0.01.

Figure 5

Characterization of wildtype responses to ethanol and red-light stimulation. (a) Group activity counts for each genotype were binned over 10 second periods, averaged across biological replicates (w − CS n = 8, w + CS n = 8, w + B n = 9, w + O n = 7) of 10 flies each and plotted against time. Ethanol was delivered over a 10-minute period starting at 300 s as denoted by the gray shaded region. Lines with shaded ribbon depict mean +/− standard error. Group activity during baseline was averaged across flies and plotted by genotype. Repeated Measures ANOVA with planned contrasts at 140 s, 600 s, and 1100 s revealed a significant interaction between group activity and genotype (F(6, 56) = 14.499, p = 0.00; (a) w + B and w + O were not significantly different from each other, however, w − CS and w + CS responses were significantly different from each other and the remaining wildtype groups (Posthoc with Bonferroni Corrections p = 0.033–0.000). (b) Wildtype responses to 680 nm of red light at 30 Hz and 33 ms (w − CS n = 9, w + CS n = 9, w + B n = 9, w + O n = 9). (c) Wildtype responses to 680 nm of red light at 40 Hz and 25 ms (w − CS n = 9, w + CS n = 9, w + B n = 9, w + O n = 9). (d) Wildtype response to 680 nm of red light 50 Hz and 20 ms (w − CS n = 8, w + CS n = 8, w + B n = 8, w + O n = 8). (e–h) Percent different in group activity from light onset (120–150 s) as compared to baseline (40–70 s). (e) w − CS group activity in response to 30 Hz and 33 ms and 50 Hz and 20 ms of light was significantly different from baseline (One-Sample 2-tailed T-Test t(8) = 4.576 p = 0.002, t(8) = 4.548 p = 0.002, respectively). (f) w + CS group activity in response to 30 Hz and 33 ms and 50 Hz and 20 ms of light was significantly different from baseline (One-Sample 2-tailed T-Test t(8) = 4.206 p = 0.003, t(7) = 11.333, p = 0.000, respectively). (g) w + B group activity in response to 50 Hz and 20 ms of light was significantly different from baseline (One-Sample 2-tailed T-Test t(7) = 3.618, p = 0.009. (h) w + O group activity in response 30 Hz and 33 ms and 50 Hz and 20 ms of light was significantly different from baseline (One Sample 2-tailed T-Test t(8) = 3.552, p = 0.007, t(7) = 2.746, p = 0.029). *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 6

Group locomotor activity in response to ellipsoid body thermogenetic inactivation. (a) Anterior view of expression pattern of 4–67-GAL4. Scale bar is 50 μm. (b) Posterior view of expression pattern of 4–67-GAL4. Scale bar is 50 μm. (c) Neuropil expression pattern of 4–67-GAL4 specific to the ellipsoid body. Scale bar is 50 μm. (d,e) Group activity counts for each genotype were binned over 10 second periods, averaged across biological replicates (+/4–67 n = 12, +/shi^ts n = 12, shi^ts/4–67 n = 12) of 10 flies each and plotted against time. Ethanol was delivered over a 10-minute period starting at 300 s as denoted by the gray shaded region. Lines with shaded ribbon depict mean +/− standard error. Group activity during baseline was averaged across flies and plotted by genotype. (d) Group activity at restricted temperatures was significant reduced in experimental flies (Repeated Measures ANOVA with Greenhouse-Geisser correction F(1.530, 50.485) = 41.896, p = 0.000, Posthoc with Bonferroni correction p = 0.000). (e) Experimental flies exhibited normal ethanol-induced locomotor activity at permissive temperatures. ***p < 0.001.

Figure 7

Group locomotor activity in response to ellipsoid body optogenetic stimulation. Group activity counts for each genotype were normalized to 30 seconds following the onset of light and/or ethanol, binned over 10 second periods, averaged across biological replicates of 10 flies each and plotted against time. Red light and/or ethanol was delivered over a 5- minute period starting at 120 s. Lines with shaded ribbon depict mean +/− standard error. Group activity during baseline was averaged across flies and plotted by genotype. (a) +/4–67 n = 16, pBDP/Chrimson n = 16, 4–67/Chrimson n = 16. Group activity during stimulation (200–410 s) was significantly increased in experimental flies (Repeated Measures ANOVA with a Greenhouse-Geisser correction F = (18.894, 425.106) = 3.254, p = 0.000, Posthoc with Bonferroni correction p = 0.001–0.000). (b) +/467 n = 12, pBDP/Chrimson n = 12, 4–67/Chrimson n = 12. Group activity during ethanol exposure (200–410 s) was not significantly different across genotypes (Repeated Measures ANOVA with a Greenhouse-Geisser correction F = (15.607, 249.712) = 1.949, p = 0.018, Posthoc with Bonferroni correction p = 0.063–1.000). (c) +/4–67 n = 12, pBDP/Chrimson n = 12, 4–67/Chrimson n = 12. Group activity during ethanol exposure and optogenetic stimulation (200 s-410 s) was significantly increase in experimental flies (Repeated Measures ANOVA with a Greenhouse-Geisser correction F = (19.432, 310.913) = 1.690, p = 0.035, Posthoc with Bonferroni correction p = 0.008–0.000). (d) Group activity as compared to controls significantly increased in response to optogenetic stimulation and ethanol exposure as compared to stimulation alone or ethanol alone (One-way ANOVA F(2, 37) = 5.098, p = 0.011, Posthoc with Bonferroni correction, p = 0.033–0.019). *p < 0.05, ***p < 0.001.

Figure 8

Model of ellipsoid body (EB) role in ethanol-induced locomotor activity. (a) Heterogeneity exists in EB that balances excitation and inhibition of locomotor activity. (b) Activation of EB R2/R4 neurons results in the activation of a population of EB which shifts balance towards excitation thereby promoting locomotor activity. (c) Activation of EB R2/R4 neurons and low levels of ethanol results in the same activation of EB neurons, however the presence of low levels of ethanol decrease neural excitability of subpopulations of EB neurons likely reducing inhibition allowing for an even greater shift in excitation thereby promoting enhanced locomotor activity.