Prefrontal cortex molecular clock modulates development of depression-like phenotype and rapid antidepressant response in mice

Abstract

Depression is associated with dysregulated circadian rhythms, but the role of intrinsic clocks in mood-controlling brain regions remains poorly understood. We found increased circadian negative loop and decreased positive clock regulators expression in the medial prefrontal cortex (mPFC) of a mouse model of depression, and a subsequent clock countermodulation by the rapid antidepressant ketamine. Selective Bmal1KO in CaMK2a excitatory neurons revealed that the functional mPFC clock is an essential factor for the development of a depression-like phenotype and ketamine effects. Per2 silencing in mPFC produced antidepressant-like effects, while REV-ERB agonism enhanced the depression-like phenotype and suppressed ketamine action. Pharmacological potentiation of clock positive modulator ROR elicited antidepressant-like effects, upregulating plasticity protein Homer1a, synaptic AMPA receptors expression and plasticity-related slow wave activity specifically in the mPFC. Our data demonstrate a critical role for mPFC molecular clock in regulating depression-like behavior and the therapeutic potential of clock pharmacological manipulations influencing glutamatergic-dependent plasticity.

Fig. 1. Repetitive swim stress alters circadian clock gene expression in the mouse mPFC.

a Schematic overview of the experimental design: CDM paradigm (5 days of 10 min swimming), tissue harvesting time points and gene expression analyses. b Relative mRNA expression of clock genes Per1, Per2, Cry1, Cry2, Bmal1, Npas2, Rorα, Rev-erbα, Rorβ, Rorγ normalized to s12, ApoE and GAPDH in mPFC samples harvested from naive (control) and CDM mice every 4 h at 12:12 h LD condition (n = 5 mice per group, two-way ANOVA: P values of the CDM effect is displayed inside the graph; Bonferroni post hoc test: *P < 0.05, **P < 0.01, ***P < 0.001 control vs. CDM). c Amplitude of rhythmic expression and acrophase (n = 30 mice: 5 ×6 ZT points; Extra sum of squares F test: **P < 0.01, ***P < 0.001). d Schema summarizing the effects of stress on clock gene expression in mPFC–CDM potentiates Per2 and Cry2 expression, while decreases Rors level. Data are presented as mean ± SEM. See also Supplementary Fig. 1 and Supplementary Data 1. Source data are provided as a Source Data file.

Fig. 2. Ketamine changes circadian clock gene expression in the mouse mPFC and opposes CDM effects.

a Schematic overview of the CDM paradigm, saline/ 3 mg/kg ketamine (KET) administration at ZT00 (start of the resting phase); tissue harvesting time points and gene expression analyses. b Relative mRNA expression of clock genes Per1, Per2, Cry1, Cry2, Bmal1, Rorα, Rev-erbα, Rorβ normalized to s12, ApoE, and GAPDH in mPFC samples from CDM mice injected at ZT00 with saline (CDM) or ketamine (KET) and harvested every 4 h from ZT06 (6 h after injection) till ZT06 (30 h post injection) (n = 5 mice per group, two-way ANOVA: P values of the CDM effect is displayed inside the graph, Bonferroni post hoc test: *P < 0.05, **P < 0.01, ***P < 0.001 CDM vs. KET). c Relative mRNA expression of clock genes Per1, Per2, Cry1, Cry2, Bmal1, Rorα, Rev-erbα, Rorβ normalized to s12, ApoE and GAPDH in mPFC samples from naive mice injected at ZT00 with saline (control) and ketamine (KET) and harvested at ZT06 (6 h after injection), ZT18 (18 h after injection) and ZT06 (30 h post injection) (n = 5 mice per group, two-way ANOVA with Bonferroni post hoc test: *P < 0.05, **P < 0.01, ***P < 0.001 control vs. KET). d Schema summarizing the effects of ketamine on clock gene expression in mPFC – ketamine downregulates Per and Cry clock suppressors and increases Rorα. Data are presented as mean ± SEM. See also Supplementary Fig. 2 and Supplementary Data 1. Source data are provided as a Source Data file.

Fig. 3. Targeted Bmal1 knockout in the excitatory CaMK2a neurons of mPFC affects depression-like behavior and Homer1a expression.

a Experimental strategy used for region- and cell type-specific virally induced deletion of Bmal1. b Time course of the AAV microinjections into mPFC, baseline behavioral assessment in IntelliCage followed by CDM paradigm, ketamine treatment and test phase. c Relative mRNA expression of Bmal1 in mPFC at ZT06 (n = 10 mice per group, two-tailed Student’s t test: ***P < 0.001). d Representative images showing viral EGFP and EGFP-Cre expression in mPFC (scale bar, 500 µm) (left) and Bmal1 labeling (red) in EGFP and EGFP-Cre expressing cells (scale bar, 20 µm) (right). e Representative western blot and quantitative data of BMAL1 protein expression of in mPFC at ZT06 (n = 4 mice per group) and ZT18 (n = 3 EGFP, n = 5 Bmal1KO mice), two-tailed Student’s t test: *P < 0.05, **P < 0.01). f Immobility time during the induction and test phase FST of Bmal1-floxed mice mPFC bilaterally injected with control CaMK2a-EGFP or CaMK2a-Cre AAVs. (n = 10 mice per group, repeated measures two-way ANOVA with Bonferroni post hoc test: *P < 0.05 and ^#P < 0.05, ^##P < 0.01, ^###P < 0.001 vs day 1). g Immobility time during baseline and test phase TST (n = 10 mice per group, repeated measures two-way ANOVA with Bonferroni post hoc test: *P < 0.05). h Sucrose preference assessed in IntelliCage during baseline and test phase (n = 5 mice, repeated measures two-way ANOVA with Bonferroni post hoc test: **P < 0.01 and ^#P < 0.05, ^##P < 0.01, ^###P < 0.001 vs baseline). i Relative mRNA expression of clock genes and Homer1a in mPFC at ZT06 of control (n = 4), CDM (n = 5) and CDM 24 h post ketamine mice (n = 5) (two-way ANOVA with Bonferroni post hoc test: *P < 0.05, **P < 0.01, ***P < 0.001). j Immobility time of FST and TST during the test phase of CaMK2a-EGFP/Bmal1KO mice 24 h post single 3 mg/kg ketamine i.p. injection (KET) (n = 10, two-way ANOVA with Tukey’s post hoc test: *P < 0.05). k Bmal1KO in mPFC blocks Homer1a modulation by stress and ketamine, and thus inhibits the development of depression-like behavior and the antidepressant response. Data are presented as mean ± SEM and the individual data points are depicted. See also Supplementary Fig. 3 and Supplementary Data 1. Source data are provided as a Source Data file.

Fig. 4. Specific knockdown of Per2 in mPFC elicits antidepressant effects.

a Schematic overview of the experimental design: CDM paradigm, followed by stereotactic microinjections of Accell siRNA and test phase. b Relative mRNA expression of Per2 in mPFC at ZT06 (n = 6 siCntr, n = 7 siPer2-injected mice, two-tailed Student’s t test: **P < 0.01). c Representative western blot (right) and quantitative data (left) of Per2 protein expression normalized to Histone 2B (H2B) in mPFC at ZT06 (n = 4 mice per group, two-tailed Student’s t test: *P < 0.05). d Immobility time during test phase TST of siCntr (n = 7) and siPer2 (n = 8) injected mice, two-tailed Student’s t test: *P = 0.0431). e Immobility time during test phase FST of siCntr (n = 7) and siPer2 (n = 8) injected mice, two-tailed Student’s t test: *P = 0.0231). f Relative mRNA expression of Homer1a at ZT06 in mPFC of siCntr (n = 6) and siPer2 (n = 7) injected mice (two-tailed Student’s t test: *P < 0.05). g Schema summarizing the effects of Per2 knockdown in mPFC on Homer1a expression and depression-like behavior. Data are presented as mean ± SEM and the individual data points are depicted. See also Supplementary Fig. 4 and Supplementary Data 1. Source data are provided as a Source Data file.

Fig. 5. REV-ERBα modulates depression-like phenotype and Homer1a expression via Bmal1.

a Experimental design schematic. b Immobility time during CDM induction and test phase FST of WT mice acutely i.p. injected with vehicle/Rev-ERB agonist SR10067 (30 mg/kg) after the swim on day 1 at ZT06 (n = 10, repeated measures two-way ANOVA with Bonferroni post hoc test: *P < 0.05, **P < 0.01, ***P < 0.001). c Immobility time during test phase TST (n = 10 mice, two-tailed Student’s t test: *P = 0.0267). d Sucrose preference assessed in IntelliCage (n = 10 mice, repeated measures ANOVA with Bonferroni post hoc test: **P < 0.01 and ^#P < 0.05, ^###P < 0.001 vs baseline). e Relative Bmal1 and Homer1a mRNA expression at ZT06 in mPFC of vehicle/SR10067 injected WT mice (n = 5, two-tailed Student’s t test: *P < 0.05, **P < 0.01). f Experimental strategy: 1 week after CDM, WT mice are vehicle/SR10067 and saline/ketamine injected 6 h prior to test. g Immobility time during test phase FST (n = 6 mice, two-way ANOVA with Bonferroni post hoc test: ^###P < 0.001 vs saline). h Immobility time during test phase TST (n = 6 mice, two-way ANOVA with Bonferroni post hoc test: ^###P < 0.001 vs saline). i Schema summarizing the effects of Rev-ERB activation on Bmal1/Homer1a expression. j Experimental design schematic. k Immobility time during induction and test phase of mPFC CaMK2a-Bmal1KO mice acutely injected with vehicle/SR10067 on day 1 at ZT06 (n = 7 mice, repeated measures ANOVA with Bonferroni post hoc test: ^#P < 0.05, ^##P < 0.01, ^###P < 0.001 vs day 1; *P < 0.05 day 5 vs test phase). l Immobility time during test phase TST (n = 7 mice, two-tailed Student’s t test). m Relative mRNA expression of Homer1a at ZT06 in mPFC of vehicle/SR10067 injected mPFC CaMK2a-Bmal1KO mice (n = 4 mice, two-tailed Student’s t test). n Immobility time during induction phase of WT (n = 8) and Rev-erbαKO mice (n = 6) (repeated measures two-way ANOVA with Bonferroni post hoc test: *P < 0.05, **P < 0.01, ***P < 0.001). o Immobility time during test phase FST and TST (n = 8 & 6 mice, two-tailed Student’s t test: ***P = 0.0005 and *P = 0.0119). p Relative mRNA expression of Bmal1 and Homer1a at ZT06 in mPFC of WT (n = 8) and Rev-erbαKO mice (n = 6) (two-tailed Student’s t test: *P < 0.05, ***P < 0.001). Data are presented as mean ± SEM and the individual data points are depicted. See also Supplementary Fig. 5 and Supplementary Data 1. Source data are provided as a Source Data file.

Fig. 6. Pharmacological activation of RORα/γ elicits rapid antidepressant effects dependent on Bmal1 and Homer1a in mPFC.

a Experimental design: baseline behavioral assessment in IntelliCage, followed by CDM paradigm and test phase 24 h after i.p. injection of vehicle/RORα/γ agonist SR1078 (10 mg/kg). b Sucrose preference assessed in IntelliCage at baseline and 24 h post vehicle/SR10067 i.p. injection (at ZT06) (n = 9 mice, repeated measures two-way ANOVA with Bonferroni post hoc test: **P < 0.01, ***P < 0.001). c Immobility time during test phase FST (n = 14 mice, two-tailed Student’s t test: ***P < 0.0001). d Immobility time during test phase TST (n = 14 mice, two-tailed Student’s t test: *P = 0.0003). e Relative mRNA expression of Bmal1 and Homer1a at ZT06 in mPFC 24 h post vehicle/SR1078 injection (n = 5 mice, two-tailed Student’s t test: *P < 0.05). f Experimental strategy: AAV-induced mPFC CaMK2a-Bmal1KO, followed after 3 weeks by CDM paradigm; acute i.p. injection of vehicle/SR1078 (10 mg/kg) followed 24 h later by test phase. g Immobility time of test phase FST (left) and TST (right) (n = 7 mice, two-tailed Student’s t test). h Relative mRNA expression of Homer1a at ZT06 in mPFC 24 h post vehicle/SR1078 injection of mPFC CaMK2a-Bmal1KO mice (n = 4 mice, two-tailed Student’s t test). i Experimental strategy of Homer1a KD: CDM protocol on WT mice before bilateral stereotactic injection of Accell siCntr/siHomer1a; followed by i.p. injection of vehicle/SR1078 (10 mg/kg) and test phase after 24 h. j Relative mRNA expression of Homer1a in mPFC of siCntr (n = 6) and siHomer1a (n = 7) injected mice (two-tailed Student’s t test: ***P < 0.001). k Immobility time during test phase FST (left) and TST (right) in siCntr/siHomer1a animals after administration of vehicle/SR1078 (n = 7 mice, two-way ANOVA with Bonferroni post hoc test: *P < 0.05). Data are presented as mean ± SEM and the individual data points are depicted. See also Supplementary Fig. 6 and Supplementary Data 1. Source data are provided as a Source Data file.

Fig. 7. Modulation of the circadian clockwork alters AMPA receptors synaptic expression and homeostatic plasticity in mPFC.

a, b Representative western blots and quantitative data of synaptic AMPA receptors subunits GluA1 and GluA2 levels normalized to Homer1b/c and PSD95 in mPFC of WT mice 24 h after SR10067 application (a), and 6 h and 24 h post SR1078 injection (b). (n = 5, a: two-tailed Student’s t test: *P = 0.0422 for GluA1, **P = 0.0069 for GluA2, b: one-way ANOVA with Bonferroni post hoc test: *P < 0.05, ***P < 0.001). c Experimental design: WT mice were subjected to the CDM protocol before undergoing implantation of ECoG, LFP and EMG recording electrodes. After recovery, mice were recorded in their home cage over 32 h from ZT0, with treatment (vehicle/ketamine/SR1078) administered at ZT06. d, e Delta power (0.5–4 Hz) of LFP signal recorded from the mPFC (d) and ECoG signal (e) recorded during slow wave sleep (SWS) across 12 h:12 h LD conditions (n = 14 mice CDM/vehicle; n = 6 SR1078; n = 4 KET; repeated measures mixed-effects model two-way ANOVA with Bonferroni post hoc test *P < 0.05, **P < 0.01 vs CDM/vehicle). Black arrow indicates time of treatment administration. f Schema summarizing the effects of RORα/γ activation on BMAL1, Homer1a and synaptic AMPAR expression in the mPFC and the relation to depression-like behavior. Data are presented as mean ± SEM and the individual data points are depicted. See also Supplementary Fig. 7 and Supplementary Data 1. Source data are provided as a Source Data file.

Fig. 8. Model of modulation of the mPFC molecular clock as a mechanism of mood-related changes.

Repetitive swim stress potentiates clock suppressor genes (Per2 and Cry2) and decreases positive regulators (RORs), while ketamine opposes CDM effects. Disrupted positive and negative elements affect Bmal1 expression and transcriptional activity, respectively, ultimately dysregulating Homer1a induction and modulating mechanisms of synaptic plasticity. Whereas, mPFC CaMK2a-Bmal1KO in mPFC excitatory neurons removes its regulation of Homer1a, inhibiting both pro-depressive and antidepressant responses. Experimental manipulation mimicking negative loop potentiation (Rev-ERB agonist SR10067) suppresses Bmal1, Homer1a and AMPAR expression, with a pro-depressive effect. In contrast, antidepressant interventions (ketamine, RORα agonist SR1078, siPer2 KD) inhibit negative and potentiate positive elements of the clock, increase Bmal1 expression and activity, ultimately driving induction of Homer1a. In turn, Homer1a induction is associated with increased markers of synaptic plasticity in the mPFC including increased AMPAR expression and elevated SWA, leading to positive effects on mood. Overall, this model describes specific changes to the local mPFC clock in response to stress and rapid antidepressant treatment, highlighting Bmal1 and Homer1a expression and subsequent plastic mechanisms as critical elements.