Pharmacological inhibition of REV-ERB α activity with SR8278 restored circadian mood-related behaviors in PD mouse models
To assess whether daily variations in mood-related behaviors were altered in the striatal 6-OHDA-lesioned PD mouse model, we observed the emotional behavioral phenotypes in vehicle (VEH)-injected or 6-OHDA-lesioned mice, at subjective dawn [circadian time (CT) 22 − 01] and dusk (CT10-13) at 5 weeks after the 6-OHDA lesion (Fig. 1a). To demonstrate that regulation of REV-ERB activity could restore changes in mood-related behaviors in 6-OHDA-lesioned mice, we observed the effect of the REV-ERB antagonist SR8278 on affective behaviors in 6-OHDA-lesioned mice by local administration of SR8278 at 3 h before behavioral tests.
In the despair-based FST and TST, VEH-injected control mice showed time-of-day variation in depression-like behaviors (Fig. 1b). However, 6-OHDA-lesioned mice exhibited a significant increase in immobility time only at CT22-01, with the disappearance of daily variations compared to the control groups. Interestingly, SR8278 microinjection into the VTA significantly suppressed the immobility times in the FST and TST, which were increased by a 6-OHDA lesion at CT22-01, showing an antidepressant effect in 6-OHDA-lesioned mice, although SR8278 did not affect the immobility time in VEH-treated mice (Fig. 1b). Unlike the effects of SR8278 at CT22-01, the SR8278 microinfusion significantly decreased the immobility time and exerted antidepressant effects in VEH-treated mice, but not in 6-OHDA-injected mice at CT10-13. SR8278 treatment rescued depression-like behaviors in 6-OHDA lesioned mice only at CT22-01, recovering the circadian pattern of FST and TST scores resembling those of VEH-injected control mice.
The EPM test revealed that 6-OHDA-injected mice showed more anxiety-like phenotypes only at CT22-01, with a shorter duration, leading to the elimination of circadian patterns of behavior shown in VEH-treated mice (Fig. 1c, upper panels). In the EPM test, the SR8278 microinfusion increased the duration and frequency in the open arms in 6-OHDA-lesioned mice only at CT22-01, exerting anxiolytic effects; however, this was not observed in VEH-treated mice. Anxiety-like behaviors were reduced in VEH-injected mice only at CT10-13 by SR8278 microinjection. SR8278 recovered the rhythm of anxiety-like behaviors in 6-OHDA-lesioned mice, similar to that observed in VEH-treated mice. There were no significant differences in the distance traveled and center crossings in the EPM with 6-OHDA lesion and SR8278 treatment, suggesting that anxiety responses in EPM are not attributable to changes in general activity (Fig. 1c, bottom panel). Our results revealed that SR8278 treatment recovered emotion-related disorders in a circadian time-dependent manner in 6-OHDA-lesioned mice, exerting antidepressant and anxiolytic effects, especially at dawn.
To confirm that 6-OHDA lesions induced PD-like motor deficits, we performed cylinder and rotarod tests (Supplementary Fig. 1a, b). The cylinder and rotarod tests revealed that 6-OHDA-injected mice exhibited significant motor asymmetry and decreased motor coordination performance regardless of the time of day. Furthermore, we assessed the circadian rhythmicity of active-rest cycles and body temperature in both LD and constant dark (DD) conditions (Supplementary Fig. 2a-c). The variation between light and dark periods in total activity and body temperature in the LD condition decreased at 2 weeks following 6-OHDA microinjection. Furthermore, 6-OHDA-lesioned mice showed a longer period, lower amplitude, and lower robustness of rhythm in locomotor activity and body temperature under DD conditions, exhibiting a disturbance of endogenous rhythm (Supplementary Fig. 2a, d-e).
SR8278 microinjection alters remaining DAergic neuron-specific transcription levels of Rev-erbα and Nurr1 in the VTA
To prove that 6-OHDA-lesioned mice show significant PD pathologies in DAergic neurons, we analyzed TH expression in both SNpc and VTA (Supplementary Fig. 3). Immunohistochemistry revealed that control mice exhibited a temporal variation with a greater number of TH + neurons at CT00 than at CT12 in both the SNpc and VTA (Supplementary Fig. 3b). The ipsilateral 6-OHDA-lesioned mice exhibited a significant decrease in the number of TH + neurons in the SNpc and VTA, regardless of the time of day, compared to the contralateral region or the VEH-injected control animals. Furthermore, 6-OHDA injection led to the disappearance of any temporal variation in the number of TH + neurons on the ipsilateral side. The degree of TH + neuronal loss on the ipsilateral side of the 6-OHDA-lesion mice was more severe in the SNpc than in the VTA, which is consistent with previous findings [16, 30]. Western blot analysis also supported these findings (Supplementary Fig. 3c, d). To assess whether 6-OHDA lesions altered the DAergic system and circadian clock, we assessed the circadian mRNA expression of DA-related genes and circadian clock genes in the midbrain of 6-OHDA-lesioned mice (Supplementary Fig. 4). TH and Bmal1 mRNA levels were elevated at dawn, which differed from the circadian oscillation of Per2 mRNA levels in the control. The expression levels of Nurr1 mRNA were constantly expressed, regardless of the time of day. TH and Nurr1 mRNA expression were reduced by 6-OHDA lesions (p < 0.0001 for TH, p < 0.0001 for Nurr1), whereas transcriptional levels of circadian clock genes, Bmal1 and Per2, were largely unaffected by 6-OHDA, showing their circadian patterns (p = 0.0002 for Bmal1, p = 0.0156 for Per2).
To study the mechanisms of action underlying SR8278-induced behavioral recovery in the 6-OHDA-lesioned mice, we measured the remaining DAergic-specific transcription levels of Rev-erbα and Nurr1, which are the upstream nuclear receptors of TH, in the VTA after SR8278 microinjection, as assessed by FISH analysis (Fig. 2a). Rev-erbα mRNA expression in the TH + neurons of the VTA was significantly affected by 6-OHDA or SR8278 treatment (p = 0.0172) and drug-by-time interaction (p = 0.0042) (Fig. 2a, b, left panel). In the control group, Rev-erbα mRNA levels were significantly elevated at CT12 compared to CT00. However, 6-OHDA lesions increased Rev-erbα transcript levels only at CT00 without altering the mRNA levels at CT12, resulting in the loss of daily variation in Rev-erbα transcript levels compared to VEH-treated mice. SR8278 treatment significantly reduced Rev-erbα mRNA expression in VTA DAergic neurons in 6-OHDA-lesioned mice at CT00, but not at CT12, thereby restoring rhythmic transcription of Rev-erbα. Rev-erbα mRNA expression was downregulated by SR8278 microinjection in the VTA DAergic neurons of VEH-treated mice at CT12, but not at CT00, showing time-dependent action of SR8278. Nurr1 transcription levels in the remaining DAergic-specific neurons of the VTA were also affected by drug-by-time interaction (p = 0.0115) (Fig. 2a, b). VEH-treated mice showed constant transcription levels of Nurr1, regardless of the time of day. 6-OHDA lesions slightly increased DAergic-specific Nurr1 mRNA levels at CT00 without an alternation at CT12, thereby disturbing the consistency of Nurr1 transcription. However, SR8278 treatment lowered Nurr1 mRNA levels in 6-OHDA-lesioned mice at CT00 and upregulated its level at CT12, showing time-dependent action of SR8278. These data suggest that the mRNA levels of the remaining DAergic-specific Rev-erbα and Nurr1 in the VTA were altered in a time-dependent manner by 6-OHDA lesions. Notably, SR8278 treatment restored the time-dependent transcription levels of Rev-erbα and Nurr1 in 6-OHDA-lesioned mice at dawn, consistent with the behavioral recovery (see Fig. 1).
SR8278 microinjection restores antagonistic crosstalk of REV-ERBα and NURR1 binding activity to TH promoter and TH protein levels in VTA at dawn
We performed a chromatin immunoprecipitation (ChIP) assay with VTA tissues to determine whether the competitive binding of REV-ERBα and NURR1 to R/N sites of the TH promoter would be altered by the 6-OHDA lesion and restored by SR8278 microinjection into the VTA. The REV-ERBα binding affinity to R/N sites was significantly affected by the time of day and drug injection of 6-OHDA and SR8278 (p = 0.0023) (Fig. 3a, left panel). In VEH-treated mice, REV-ERBα binding affinity was much higher at CT12 than at CT00, in accordance with Rev-erbα mRNA expression (Fig. 2b). In 6-OHDA-lesioned mice, the REV-ERBα binding activity to R/N sites was increased at CT00, but decreased at CT12, inducing a reversed pattern compared to the VEH-treated mice. Notably, SR8278 treatment rescued REV-ERBα binding affinity to R/N sites by reducing its affinity in 6-OHDA-lesioned mice only at CT00. On the other hand, SR8278 treatment tended to increase REV-ERBα binding affinity at CT12 in 6-OHDA-lesioned mice, showing the time-dependent action of SR8278. In VEH-treated mice, SR8278 local injection significantly reduced REV-ERBα binding activity to R/N sites only at CT12, but not at CT00. Because of the time-dependent effect of SR8278, it restored the rhythmic REV-ERBα binding affinity to R/N sites of the TH promoter in 6-OHDA-lesioned mice, similar to the control.
For NURR1 binding activity, injection of 6-OHDA and SR8278 altered relative binding to its cis-elements (R/N sites) in a time-dependent manner (p = 0.0284), which is the opposite pattern observed in the REV-ERBα ChIP assay (Fig. 3a, right panel). NURR1 binding activity was higher at CT00 than at CT12 in the VEH-treated mice. The NURR1 binding affinity to R/N sites was significantly diminished in 6-OHDA-lesioned mice at CT00, but not at CT12. Notably, SR8278 treatment rescued NURR1 binding affinity to R/N sites by increasing its binding affinity in 6-OHDA-lesioned mice at CT00, while NURR1 binding affinity to R/N sites was not altered at CT12. NURR1 binding affinity to R/N sites was decreased at CT00 by SR8278 treatment in the VEH-treated mice, with a tendency to increase its binding affinity at CT12. NURR1 binding activity patterns were opposite to REV-ERBα binding activity, exhibiting antagonistic crosstalk between REV-ERBα and NURR1.
Because REV-ERBα and NURR1 bind to the R/N sites as nuclear receptors regulating TH expression, we observed changes in TH protein levels in the VTA by western blot analysis. TH protein expression in the VTA was significantly affected by SR8278 microinjection at CT00 (p = 0.0468), whereas TH protein expression of VTA was affected by 6-OHDA lesions at CT12 (p = 0.0088), but not by SR8278 treatment (p = 0.7924), showing time-dependent action of SR8278 (Fig. 3b, c). Reduced TH protein level of VTA by 6-OHDA lesion was restored to the levels in VEH-treated mice only at CT00, while SR8278 did not affect TH expression at CT12 in 6-OHDA-lesioned mice. These data indicate that recovery of antagonistic crosstalk between REV-ERBα and NURR1 causes the elevation of TH expression in 6-OHDA-lesioned mice at dawn.
SR8278 treatment induces enrichments of REV-ERBα and NURR1 binding motifs
To further investigate the underlying mechanisms of SR8278 in TH expression, we performed an assay for transposase-accessible chromatin using high-throughput sequencing (ATAC-seq) and motif-based analysis for R/N sites. This genome-wide mapping of the chromatin accessibility technique is useful for detecting chromatin regions with increased accessibility, annotating the peak called regions, and comparing the enriched motifs in different experimental groups.
To include the effects of light cues and minimize the individual variations in the genomic architectures, we prepared the samples under the LD cycle, which is considered as zeitgeber (the external cue), and pooled VTA samples from five animals for each group (n = 5). We observed an increase in TH gene expression at CT00 compared to CT12 (Fig. 3b, c), which agrees with the greater number of peaks and the chromatin accessibility of the TH genomic regions at zeitgeber time (ZT) 00 in VEH-treated groups compared to that at ZT12 (Supplementary Fig. 5). We acquired consistent results on TH expression under circadian (without light cues) and zeitgeber conditions. This allowed us to presume that the peak calling of ATAC-seq is adequate for further analysis. For additional quality controls for ATAC-seq results, the annotated genomic locations of the called peaks did not show clear differences in fold changes of transcription start sites (TSS) and promoter regions among experimental groups, and we presented fragment distributions for each ATAC-seq sample (Fig. 4b, Supplementary Fig. 6).
To understand the epigenetic regulation of SR8278 in 6-OHDA-lesioned mice, we characterized the differential motif analysis between experimental groups using the ATAC-seq approach. For this, we used the known motif for R/N with REV-ERBα and NURR1 binding cis-acting sites in the TH promoter, as previously described (Fig. 4a). We analyzed the R/N motifs using HOMER and compared the motif densities within 1 kb from the peaks between the indicated experimental groups, which were presented at a specific time point, at ZT00 or ZT12 (Fig. 4c). The occurrence of R/N motifs was clearly altered in the 6-OHDA and/or SR8278 treated groups. 6-OHDA-lesioned mice showed a notable reduction in R/N motifs at ZT00 and ZT12 compared to the VEH-treated mice (Fig. 4c, first panel). SR8278 treatment in 6-OHDA-lesioned mice increased the number of R/N motifs only at ZT00 (Fig. 4c, third panel). Moreover, the increase in the enrichment of R/N motifs by SR8278 in 6-OHDA-lesioned mice was almost similar to that in VEH-treated mice only at ZT00. The SR8278 treatment did not seem to be effective at ZT12 in R/N motif enrichment (Fig. 4c, fourth panel). For the control analysis, we also performed differential motif analysis with E-box, the most documented cis-element in circadian biology (Supplementary Fig. 7a). Enrichment of E-box motifs remained unaffected by 6-OHDA or SR8278 treatments and did not show time-dependent differences at ZT00 and ZT12 (Supplementary Fig. 7b, c). Taken together, SR8278 not only changed the binding activities of REV-ERBα and NURR1 in 6-OHDA-lesioned mice but also altered the R/N motifs at specific time points, at ZT00 (Fig. 3a, 4c).