Bmal1 ablation in dopaminergic neurons results no alteration in circadian period

To generate Bmal1 conditional knockout mice (Bmal1-cKO; Dat-Cre+/Bmal1-flox+/+) and control mice (Dat-Cre−/Bmal1-flox+/+), we crossed the Bmal1-flox carrying strain with the Dat-Cre carrying strain. Then, we performed immunofluorescence analysis to validate the knockout efficiency of BMAL1 in dopamine neurons. Tyrosine hydroxylase (TH), a key enzyme in dopamine synthesis, is expressed in nearly all dopamine neurons. Therefore, we identified dopamine neurons located at substantia nigra pars compacta (SNc) and ventral tegmental area (VTA) via TH immunofluorescence (Fig. 1A-D). In control mice, TH-positive dopamine neurons co-expressed BMAL1 (Fig. 1A and B). Whereas in cKO mice, TH-positive dopamine neurons lacked BMAL1 expression (Fig. 1C and D). Quantitative analysis revealed a significant decrease in BMAL1 fluorescence intensity in cKO mice compared to controls (60.06 ± 3.15 versus 18.8 ± 1.68 A.U., p < 0.001, unpaired Student’s t test; Fig. 1E). Notably, no reduction in BMAL1 expression was observed in brain regions beyond the SNc and VTA, including the prefrontal cortex (PFC) and dorsal striatum, where Dat-Cre-mediated recombination is not anticipated (Supplementary Fig. 1A and B). These results confirm the specific and efficient knockout of BMAL1 in TH-positive neurons in Bmal1-cKO mice.

As a core circadian gene, Bmal1 is essential for normal circadian behavior and its expression varies within a daily cycle. However, the impact of BMAL1 on circadian rhythm in dopamine neurons is still unknown. Here, we monitored the wheel-running activity of Bmal1-cKO mice and their littermate controls. Animals were housed individually in cages equipped with running wheels. Wheel-running activities were continuously recorded for 2–3 weeks under a 12-hour light/dark (LD) cycle, followed by a switch to constant darkness (DD) for 3–4 weeks. We observed no significant difference in the free-running periods under DD between control and cKO mice (23.51 ± 0.07 versus 23.5 ± 0.03 h, p = 0.183, unpaired Student’s t test), indicating that BMAL1 in dopamine neurons has a negligible impact on circadian period (Fig. 1F-H). We also analyzed the expression of core circadian genes Clock, Per2, and Cry1 in midbrain tissue and found no significant differences between cKO and control mice (Supplementary Fig. 1C), suggesting that Bmal1 deletion in dopaminergic neurons does not broadly disrupt the expression of other core clock genes.

Fig. 1

Conditional knockout of Bmal1 in the midbrain and the circadian rhythm in cKO mice. (A) Representative immunofluorescence image of BMAL1 expression in SNc and VTA brain regions of control (Ctrl) mice. Green represents TH protein, and red represents BMAL1 protein. Scale bar: 300 μm. (B) High magnification images showing BMAL1 expression in TH-positive neurons in the SNc and VTA of Ctrl mice. Scale bar: 10 μm. (C) Representative immunofluorescence image of BMAL1 expression in SNc and VTA brain regions of cKO mice. Green represents TH protein, and red represents BMAL1 protein. Scale bar: 300 μm. (D) High magnification images showing BMAL1 expression in TH-positive neurons in the SNc and VTA of cKO mice. Scale bar: 10 μm. (E) The quantification of fluorescence intensity of BMAL1 was analyzed by ImageJ. ***P < 0.001, unpaired Student’s t test. n = 5 per genotype. (F) Representative double-plotted actograms of Ctrl and cKO mice. (G) The free-running periods of Ctrl (n = 8) and cKO (n = 13) mice; P > 0.05, unpaired Student’s t test. (H) Average wheel-running (RW) activity of Ctrl (n = 8) and cKO (n = 13) mice recorded for 5 days plotted in 30-min bins. Data are presented as mean ± SEM

Knockout of Bmal1 in dopamine neurons causes hyperactivity and cognitive deficits

Regardless of the negligible impact on circadian period in Bmal1-cKO mice, we next investigated the behavioral consequences of selective Bmal1 ablation in dopaminergic neurons using other behavioral tests to assess locomotor activity, cognitive function and social behavior [14]. Previous report indicates that BMAL1 expression in ventral midbrain is elevated at night and reduced during the daytime. Therefore, we performed a series of behavioral tests on Bmal1-cKO mice and their littermate controls during two different zeitgeber time (ZT) periods: at the end of the light phase (ZT09-12) and dark phase (ZT21-24).

Firstly, we used an open field test (OFT) to measure spontaneous locomotor activity. During the ZT09-12 period, cKO mice exhibited significantly greater movement, as evidenced by longer distances traveled and increased zone crossings, compared to control mice (Distance traveled: 3880.45 ± 239.77 versus 5771.14 ± 478.66 cm, p < 0.001, two-way ANOVA test; Zone crossings: 11.56 ± 0.92 versus 15.8 ± 1.53 times, p = 0.030, two-way ANOVA test). Meanwhile, no obvious differences were observed (Distance traveled: 4619.22 ± 184.30 versus 4988.75 ± 347.61 cm, p = 0.853, two-way ANOVA test; Zone crossings: 14.62 ± 1.42 versus 17.09 ± 2.10 times, p = 0.327, two-way ANOVA test) when the test was performed at ZT21-24 (Fig. 2A-C). These results indicate that Bmal1 deficiency in dopaminergic neurons causes hyperactivity in a circadian time-dependent manner. In addition, Bmal1 global knockout mice are known to show deficits in habituation to novel environment in OFT [15]. To dissociate true hyperactivity from slowed habituation of Bmal1 cKO mice, we performed the OFT for 2 consecutive days at ZT09-12. Unlike Bmal1 global knockout mice, our result shows that Bmal1-cKO mice display normal habituation to the novel environment (Supplementary Fig. 2A-D).

Previous studies have shown that global knockout of BMAL1 impair short-term memory and lead to despair-like phenotypes [16, 17]. Therefore, we conducted a Y-maze assay which evaluates attention, learning and memory. In this assay, cKO mice showed significantly reduced correct rate than control mice during both ZT09-12 (72.04 ± 3.71 versus 56.52 ± 3.28 alternation%, p = 0.041, two-way ANOVA test) and ZT21-24 (67.26 ± 2.20 versus 56.84 ± 2.36 alternation%, p = 0.013, two-way ANOVA test) periods (Fig. 2D), suggesting deficits in working memory. We then analyzed despair-like behaviors by using the tail suspension test (TST). In this test, cKO mice exhibited significantly less immobility time than control mice during both time periods (ZT09-12: 138.53 ± 8.32 versus 81.71 ± 13.03 s, p < 0.001; ZT21-24: 149.36 ± 7.35 versus 110.44 ± 10.34 s, p = 0.006, two-way ANOVA test), indicating reduced despair-like behavior (Fig. 2E).

To evaluate social behaviors, we also performed reciprocal social interaction test in cKO mice and their controls. During the ZT09-12 period, cKO mice spent significantly more time engaging in social interactions compared to control mice (50.45 ± 6.12 versus 81.11 ± 2.32 s, p = 0.007, two-way ANOVA test). However, no obvious difference was observed during the ZT21-24 period (12.96 ± 2.32 versus 19.25 ± 5.27 s, p = 0.909, two-way ANOVA test; Fig. 2F).

Recent studies reported a bright-light stimulation test to assess the attention to environmental stimuli of cKO mice. The appearance of bright light simulates the sudden emergence of a natural predator, which can automatically induce exogenous attention [18, 19]. We then preformed the bright-light stimulation test at ZT09-12 to test the attention to environmental stimuli of cKO mice (Fig. 2G). In the test, cKO mice showed increased locomotion in the first dark phase compared to control mice. However, cKO mice response less to the light stimulation (Fig. 2H, the change of distance traveled from 4 to 5 min). Together with the locomotion change between light phase and dark phase 1(58 ± 12.67 versus 7.58 ± 3.26%, p = 0.049, unpaired Student’s t test), these results indicate that cKO mice exhibit impaired attention (Fig. 2I).

Taken together, these findings reveal several behavioral abnormalities in Bmal1-cKO mice, including hyperactivity, impaired working memory, attention deficits, altered despair-like behavior, and abnormal social behavior. Notably, the hyperactivity, as well as impaired attention and working memory, closely resemble the hallmark symptoms of ADHD, a highly heritable neurodevelopmental disorder.

Fig. 2

Behavioral consequences of selective Bmal1 ablation in dopaminergic neurons. (A) Representative movement traces of animals in the OFT. (B) cKO mice travelled significantly longer distances than Ctrl mice in the OFT at ZT09-12; ***P < 0.001, two-way ANOVA test. ZT09-12, n = 22 versus 10 mice; ZT21-24, n = 13 versus 11 mice. (C) The number of zone crossings in the OFT was significantly higher in cKO mice at ZT09-12; *P < 0.05, two-way ANOVA test. ZT09-12, n = 22 versus 10 mice; ZT21-24, n = 13 versus 11 mice. (D) Spontaneous alternation in the Y-maze test was significantly reduced in cKO mice compared to Ctrl mice; *P < 0.05, two-way ANOVA test. ZT09-12, n = 3 versus 7 mice; ZT21-24, n = 15 versus 11 mice. (E) Immobility time in TST was significantly shorter in cKO mice; *P < 0.05, ***P < 0.001, two-way ANOVA test. ZT09-12, n = 17 versus 14 mice; ZT21-24, n = 11 versus 9 mice. (F) Social interaction time in the reciprocal social interaction test was significantly longer in cKO mice at ZT09-12; **P < 0.01, two-way ANOVA test. ZT09-12, n = 12 versus 9 mice; ZT21-24, n = 9 versus 8 mice. (G) The schematic diagram of bright-light stimulation test. (H) Distance traveled in bright-light stimulation test. cKO group mice showed less response in ambulation in response to light stimulation. n = 5 versus 5 mice. (I) The percentage of change in locomotion. n = 5 versus 5 mice, **P < 0.01, unpaired Student’s t test. Data are presented as mean ± SEM

Increase in dopamine release in Bmal1 conditional knockout mice

Dopamine released by mesencephalic dopamine neurons plays an important role in regulating motor control, motivation, reward and cognitive functions. The ADHD-like behaviors observed in Bmal1-cKO mice in our study may therefore result from dysregulated of dopamine release. To test this hypothesis, we investigated both the homeostasis and release dynamics of dopamine.

At the cellular level, the promoter of monoamine oxidase A (MaoA), an enzyme involved in dopamine catabolism, is regulated by the circadian clock components including BMAL1 [20]. Thus, we examined the expression level of MaoA in the midbrain of cKO mice and their littermate controls using immunofluorescence analysis. We observed a significant decrease of MaoA expression in TH-positive neurons (Fig. 3A), indicating that BMAL1 positively regulates MaoA expression, consistent with previous cellular experiments. The fluorescence intensity of MaoA has been calculated (26.94 ± 2.74 versus 19.86 ± 0.86 A.U., p = 0.039, unpaired Student’s t test; Fig. 3B). Furthermore, ELISA analysis revealed a significant increase of dopamine level (7.78 ± 0.89 versus 16.99 ± 3.33 pg/ml, p = 0.047, unpaired Student’s t test) in the downstream nuclei of midbrain in cKO mice (Fig. 3C). These results suggest that the ablation of BMAL1 in dopaminergic neurons downregulates MaoA expression, leading to dopamine accumulation, which may subsequently enhance dopamine release.

To directly measure the alterations in dopamine release, we used a well-established dopamine sensor, rDA3m [21, 22], to monitor dopamine dynamics in brain nuclei downstream of midbrain dopaminergic neurons. Dopamine neurons in the SNc project to the dorsal striatum, forming the nigrostriatal pathway, which controls movements and motivated behaviors. Using fiber photometry, we monitored dopamine dynamics in the dorsal striatum of control and Bmal1-cKO mice in OFT (Fig. 3D). During the resting state of OFT, dopamine fluctuations, measured as variations (1.12 ± 0.43 versus 3.34 ± 0.79 s.d., p = 0.034, unpaired Student’s t test) in rDA3m fluorescence (ΔF/F0) [23], were markedly increased in cKO mice (Fig. 3E and F). Indicating hyperactive dopamine dynamics due to selective ablation of Bmal1 in dopaminergic neurons. Additionally, by analyzing the positive standard deviation (s.d.) of ΔF/F0 as a threshold to quantify dopamine fluctuation bouts, we found that cKO mice exhibited larger peak amplitudes (1.97 ± 0.75 versus 5.15 ± 1.10%, p = 0.038, unpaired Student’s t test) and longer durations of dopamine release events (0.37 ± 0.09 versus 0.86 ± 0.11 s, p = 0.008, unpaired Student’s t test) compared to controls (Fig. 3G and H).

Together, these findings highlight the critical role of Bmal1 in maintaining dopaminergic homeostasis and regulating dopamine release dynamics, suggesting potential implications for neuropsychiatric disorders associated with dysregulated dopamine signaling.

Fig. 3

Dopamine homeostasis and release dynamics in Bmal1-cKO mice. (A) Representative immunofluorescence images of MaoA expression in TH-positive neurons in the VTA of Ctrl and cKO mice. Scale bar: 10 μm. (B) The quantification of fluorescence intensity of MaoA was analyzed by ImageJ. *P < 0.05, unpaired Student’s t test. n = 5 per genotype. (C) ELISA test for the contents of dopamine in the downstream nuclei (nucleus accumbens) of dopaminergic neurons in Ctrl (n = 5) and cKO (n = 7) mice; *P < 0.05, unpaired Student’s t test. (D) Schematic diagram of the experimental setup. The fiber photometric canula was inserted into the right dorsal striatum. (E) Example traces of rDA3m fluorescence. Dotted lines indicate the positive standard deviation value (s.d., threshold for high-level dopamine release bouts) of ΔF/F0. (F-H) Quantification of fluorescence variation measured as (F) the s.d., (G) the peak amplitude, and (H) the peak duration of ΔF/F0 of rDA3m fluorescence. Each data point represents one 30-second trace analyzed. *P < 0.05, unpaired Student’s t test. Data are presented as mean ± SEM

Enhanced neuronal excitability of striatal MSNs in Bmal1 conditional knockout mice

Considering enhanced dopamine release is likely to increase dopamine signaling, which would modulate the neuronal activities of dorsal striatum cells, we performed the whole-cell patch-clamp recordings to evaluate the excitability of medium spiny neurons (MSNs) in the dorsal striatum. Indeed, MSNs from cKO mice exhibited significantly greater excitability compared to those from control mice, as evidenced by higher action potential (AP) frequency in response to a series of current pulses injections (Fig. 4A and B), decreased rheobase current (376.67 ± 34.8 versus 280.45 ± 20.06 pA, p = 0.014, unpaired Student’s t test; Fig. 4C), more depolarized resting membrane potential (RMP) (-70.55 ± 1.72 versus − 64.98 ± 1.34 mV, p = 0.015, unpaired Student’s t test; Fig. 4D), and increased input resistance (Rin) of neurons (39.28 ± 5.19 versus 63.07 ± 4.2 MΩ, p = 0.001, unpaired Student’s t test; Fig. 4E). In addition, we performed whole-cell patch-clamp recordings to evaluate the excitability of pyramidal cells in the PFC, and no significant differences in neuron excitability were observed between two groups (Supplementary Fig. 3A-E). Collectively, our data indicate that the nigrostriatal pathway is hyperactive in cKO mice, suggesting that dysregulated dopamine release and signaling contribute to altered neuronal function in dorsal striatum.

Fig. 4

Striatal MSNs exhibit increased neuronal excitability in Bmal1-cKO mice. (A) Representative APs of dorsal striatum neurons induced by current injections of 350 pA (black) or -50 pA (grey) in Ctrl and cKO mice. The recording was performed at ZT09-12 timepoint. (B) The AP frequencies in response to a series of current pulses injections in dorsal striatum neurons from two groups. n = 15 neurons from 5 Ctrl mice versus 23 neurons from 6 cKO mice; **P < 0.01, *P < 0.05, repeated-measure ANOVA test. (C) The rheobase of dorsal striatum neurons from two groups. *P < 0.05, unpaired Student’s t test. (D) The resting membrane potential (RMP) of dorsal striatum neurons from two groups. *P < 0.05, unpaired Student’s t test. (E) The input resistance (Rin) of dorsal striatum neurons from two groups. **P < 0.01, unpaired Student’s t test. Data are presented as mean ± SEM

Pharmacological rescue of hyperactivity behavior via modulation of dopamine signaling

To further investigate the long-term hyperactive behavioral phenotype of Bmal1-cKO mice and test the rescue effects of pharmacological interventions on dopamine signaling, we conducted a 30-minute OFT on both control and cKO mice. During this test, we recorded the distance traveled by the mice every 5 min. The results revealed that cKO mice traveled significantly greater distance in the OFT during each 5-minute interval compared to control mice (Fig. 5A).

Next, we challenged mice with amphetamine in a long-term OFT to explore the functional consequences of altered dopamine homeostasis. Amphetamine is commonly used as a treatment for ADHD [24] and is also a compound that elevates extracellular dopamine level [25]. Mice were acutely treated with intraperitoneal (i.p.) injections of either amphetamine (0.2 mg/kg) or saline 30 min after the onset of OFT. Before amphetamine administration, cKO mice exhibited significantly greater locomotor activity compared to control mice, consistent with our previous findings. However, following amphetamine injections, cKO mice traveled significantly shorter distances than control mice (Fig. 5B), indicating an alleviation of hyperactivity by amphetamine. The reduced responsiveness to amphetamine in cKO mice, coupled with their baseline hyperactivity, supports the notion of elevated dopamine levels under normal conditions, indicating lower resilience in behavior response to dopamine dynamic when Bmal1 is lost in dopamine neurons.

Finally, we attempted to rescue the hyperactivity by reducing hyperactive dopamine signaling in cKO mice. We treated mice with the dopamine D1 receptor antagonist SCH23390 (15 µg/kg, i.p., a relatively low dose, which is expected to have minimal impact on behavior in wildtype mice) [26] and assessed their behavior 30 min later in the OFT. The results shown that SCH23390 normalized the hyperactivity of cKO mice in OFT (Saline: 4391.05 ± 270.63 versus 5496.51 ± 386.02 cm, p = 0.040; SCH: 4237.56 ± 163.49 versus 3829.13 ± 376.79 cm, p = 0.320, two-way ANOVA test; Fig. 5C and D). However, SCH23390 cannot rescue the abnormal behavior of cKO mice in Y-maze (Saline: 77.68 ± 2.97 versus 61.52 ± 2.98 alternation%, p = 0.040; SCH: 72.72 ± 3.83 versus 64.22 ± 4.51 alternation%, p = 0.530, two-way ANOVA test) and TST (Saline: 155.4 ± 7.64 versus 89.6 ± 10.96 s, p = 0.002; SCH: 150 ± 13.47 versus 91.4 ± 8.51 s, p = 0.005, two-way ANOVA test; Fig. 5E and F). These data indicate that enhanced dopamine signaling, particularly through D1 receptor-mediated pathways, play a critical role in the hyperactivity phenotype observed in Bmal1-cKO mice.

Fig. 5

Amphetamine and D1 receptor antagonist attenuate the hyperactivity behavior in cKO mice. (A) cKO mice traveled significantly longer distances than Ctrl mice during a 30-minute OFT; ***P < 0.001, **P < 0.01, *P < 0.05, repeated-measure ANOVA test. n = 16 mice per genotype. (B) After 30 min of basal activity, amphetamine (0.2 mg/kg) or saline was injected i.p. (indicated by the arrow) and locomotor activity was monitored for 60 min. cKO mice travelled significantly shorter distances than Ctrl mice after amphetamine injection; **P < 0.01, *P < 0.05, repeated-measure ANOVA test. Saline: n = 5 Ctrl versus 6 cKO mice; Amphetamine: n = 10 Ctrl versus 11 cKO mice. (C) Representative movement traces of animals in the OFT after saline or SCH23390 treatment. (D) Saline or dopamine D1 receptor antagonist SCH23390 (15 µg/kg) were injected i.p., and behaviors were measured 30 min post-injection. SCH23390 normalized the hyperactivity in the OFT. Saline, n = 6 Ctrl versus 7 cKO mice; SCH, n = 9 Ctrl versus 8 cKO mice. *P < 0.05, two-way ANOVA test. Data are presented as mean ± SEM. (E) SCH23390 cannot normalized the reduced alteration in the Y-maze. n = 5 mice per genotype and treatment. *P < 0.05, two-way ANOVA test. (F) SCH23390 cannot normalized the anti-despair behavior in TST. n = 5 mice per genotype and treatment. **P < 0.01, two-way ANOVA test. Data are presented as mean ± SEM