Altered striatal dopamine regulation in ADGRL3 knockout mice

Abstract

Dopaminergic signaling is essential for regulating movement, learning, and reward. Disruptions in this system are linked to neuropsychiatric disorders such as ADHD. ADGRL3, an adhesion G protein-coupled receptor highly expressed in the brain, is genetically associated with increased ADHD risk. ADGRL3 knockout in animals alters expression of dopaminergic markers and induces dopamine-related behavioral changes. However, its precise role in modulating dopamine signaling remains unclear. We investigated how ADGRL3 knockout affects striatal dopamine release in mice using ex vivo fast-scan cyclic voltammetry and in vivo fiber photometry with a dopamine sensor. Ex vivo measurements showed increased electrically-evoked dopamine release across the striatum. Conversely, in vivo recordings revealed reduced task-induced dopamine signals in the nucleus accumbens during an operant fixed interval task. This reduction was not due to impaired dopamine availability, as amphetamine-evoked release was unchanged. These findings suggest ADGRL3 modulates dopamine release in complex ways via different pre- and postsynaptic mechanisms.

Figure 1.. ADGRL3 knockout mice show increased horizontal locomotion and decreased rearing activity across development.

(a) Weight of wild-type (WT) and ADGRL3 knockout (KO) mice. (b) Horizontal locomotor activity, (c) rearing activity, and (d) time in center for WT and ADGRL3 KO mice across development (N=9 WT, 17 KO). A two-way repeated measures ANOVA was performed to analyze the effect of age and genotype on each dependent variable (weight, basic movements, rearing, or time in center). This was followed by ídák’s multiple comparisons test (*, p < 0.05; **, p<0.01; ***, p<0.001; ****, p < 0.0001).

Figure 2.. ADGRL3 knockout mice exhibit increased activity in PiezoSleep boxes during the dark cycle compared to controls.

Activity levels and sleep/wake cycles in wild-type (WT) and ADGRL3 knockout (KO) mice were monitored for five consecutive days using the Piezo-Sleep mouse behavioral tracking system (Signal Solutions LLC). (a) Average activity scores for WT and ADGRL3 KO mice presented in Zeitgeber time (N=4 WT, 4 KO). (b) Average change in the amplitude of activity for WT and ADGRL3 KO mice. (c) Percentage of time spent asleep and (d) average duration of sleep bouts during the dark cycle. (e) Percentage of time spent asleep and (f) average duration of sleep bouts during the light cycle. (g) Interdaily stability was calculated to quantify the rest-activity rhythms for each mouse between different days. A value closer to 1 indicates stronger coupling to a zeitgeber with a period length of 24 hours (50). (h) Intradaily variability was calculated to quantify the fragmentation of the rest-activity pattern. This value converges to 0 for a perfect sine wave and approaches two for Gaussian noise (50). All statistical analyses were conducted using an Unpaired t-test (****, p < 0.0001; ***, p < 0.001; **, p < 0.01; and *, p < 0.05).

Figure 3.. ADGRL3 knockout mice show increased lick responses but similar taste preference for an evaporated milk reward compared to WT mice.

Taste preference for and licking behavior toward an evaporated milk reward were measured for WT and ADGRL3 knockout (KO) mice using a “Davis Rig” gustometer (N=6 WT, 8 KO). (a) Average weights for each genotype over the course of the study. (b) Number of trials presented, (c) number of completed and incomplete trials, (d) and total number of licks across the 30 min session. Statistical analysis consisted of an unpaired t test (***, p<0.001; **, p<0.01). (e) Average latency until first lick by genotype. Statistical analysis consisted of an unpaired t test (****, p<0.0001). (f) Average licks by each genotype for 6 different concentrations of evaporated milk (0%, 6.25%, 12.5%, 25%, 50%, and 100%). Statistical analysis consisted of a Two-way ANOVA followed by ídák’s multiple comparison test (****, p<0.0001 for % of evaporated milk). No statistical difference was detected for genotype or reward.

Figure 4.. ADGRL3 KO mice have greater levels of evoked dopamine (DA) and slower DA reuptake in the striatum.

Coronal sections that included striatum were obtained from wild-type (WT) and ADGRL3 knockout (KO) mice aged P102-P183. (a) Schematic of location for fast-scan cyclic voltammetry (FSCV) recordings from the dorsal striatum. Slices were stimulated with a bipolar concentric electrode and DA was detected using a carbon fiber electrode. (b) FSCV recordings of evoked DA release from the dorsal striatum (N=6 WT, N=8 KO; 31 slices WT, 44 slices KO). (c) DA reuptake from the dorsal striatum (N=5 WT, N=7 KO; 25 slices WT, 47 slices KO). (d) Averaged dopamine transients in response to a single electrical pulse in the dorsal striatum from all tested mice, illustrating signal kinetics. (e) Schematic of location for FSCV recordings from the ventral striatum. (f) FSCV recordings of evoked DA release from the ventral striatum (N=5 WT, N=7 KO; 24 slices WT, 36 slices KO). (g) DA reuptake from the ventral striatum (N=4 WT, N=7 KO; 19 slices WT, 42 slices KO). (h) Averaged dopamine transients in response to a single electrical pulse in the ventral striatum from all tested mice, illustrating signal kinetics. All quantal release events were analyzed in IgorPro using a publicly available GitHub-based analysis program (53) Peak parameters included amplitude (Imax, pA) and event duration (Tau1, ms). All statistical analyses were performed using an unpaired t-test (*, p<0.05).

Figure 5.. Performance of WT and ADGRL3 KO mice in a continuous reinforcement task (CRF).

(a) (Left) Experimental timeline illustrating progression from viral injection and optic fiber implantation to fiber photometry. (Middle) Optic fibers were unilaterally implanted into the nucleus accumbens. A representative coronal section shows dLight1.2 expression in the striatum and the location of the optic fiber, with dashed lines indicating the fiber shaft. (Right) WT and ADGRL3 knockout (KO) mice were trained on a fixed ratio 1 schedule of reinforcement, where each lever press was rewarded with evaporated milk from a dipper. Training occurred over seven sessions on separate days. (b) Percentage of lever presses relative to total possible lever presentations. (c) Percentage of trials in which mice lever pressed and consumed the evaporated milk reward. Statistical analysis in B and C consisted of a two-way ANOVA followed by ídák’s multiple comparison test (*, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001). (N=14 WT, 13 KO). (d) Following training, WT and ADGRL3 KO mice completed a continuous reinforcement task where they could earn up to 60 rewards per session. Dopamine levels were monitored via fiber photometry during this operant task. (e) Average dLight1.2 traces aligned to lever extension for WT and ADGRL3 KO mice. Shaded area (yellow) indicates the timeframe used for AUC calculation. (f) Area under the curve (AUC) and (g) peak height for the dLight1.2 traces. Statistical analysis consisted of an unpaired t test. (N=9 WT, 10 KO).

Figure 6.. Performance of WT and ADGRL3 KO mice in a fixed interval task.

(a) Mice were tested across five fixed intervals (2–24 seconds), with three sessions per interval. During each trial, the lever was extended for the designated interval, after which it remained extended until pressed. A successful press delivered a 10-second evaporated milk reward. Each session included up to 60 rewarded trials, with a variable intertrial interval. (b) Total lever presses per interval. No significant genotype differences. (N=10 WT, 10 KO). (c) Latency to first lever press, (d) average latency to lever press, and (e) latency to reward across sessions. All statistical analyses were performed using a two-way ANOVA with ídák’s multiple comparison test (**, p<0.01; ***, p<0.001; ****, p<0.0001). (N=8 WT, 10 KO). (f, g) Average dLight1.2 traces aligned to lever extension for WT and ADGRL3 KO mice. (h) Area under the curve (AUC) from 0–1.5 secs. Statistical analysis was performed using a two-way ANOVA with ídák’s multiple comparison test (*, p<0.05; **, p<0.01). (N=8 WT, 10 KO). (i, j) Average dLight1.2 traces aligned to dipper extension for WT and ADGRL3 KO mice. (k) AUC from 0–1.0 secs. Statistical analysis was performed using a two-way ANOVA followed by ídák’s multiple comparison test (**, p<0.01). (N=8 WT, 10 KO).

Figure 7.. Open field test with amphetamine challenge suggests phasic dopamine release capacity is similar between WT and ADGRL3 KO mice.

(a) Mice injected with AAV5-hSyn-dLight1.2 and implanted with an optic fiber in the nucleus accumbens were used to measure dopamine responses to amphetamine in an open field test. Following a 30-minute habituation period, mice received a saline injection to assess the effects of injection stress on dopamine signaling, followed by amphetamine administration (1.5, 5, 10, or 20 mg/kg subcutaneous) on separate days. Dopamine activity was recorded using fiber photometry, and locomotor behavior was tracked with AnyMaze software (Stoelting Co., Wood Dale, IL). (b) Difference in average distance travelled by WT and ADGRL3 KO mice injected with saline versus varying doses of amphetamine. Statistical analysis was performed using a two-way ANOVA followed by ídák’s multiple comparison test (**, p<0.01). ****, p<0.0001 for genotype and ***, p<0.001 for dose. (N=10 WT, 10 KO). Average dLight1.2 traces aligned to amphetamine injection for (c) WT and (d) ADGRL3 KO mice. (e) Mean dF/F, (f) slope, and (g) latency to reach plateau for WT and ADGRL3 KO mice following amphetamine injection at the four doses. Statistical analysis was performed using a two-way ANOVA (*, p<0.05; **, p<0.01; ****, p<0.0001). (N=10 WT, 10 KO).