Pharmacological and Genetic Modulation of REV-ERB Activity and Expression Affects Orexigenic Gene Expression

The nuclear receptors REV-ERBα and REV-ERBβ are transcription factors that play pivotal roles in the regulation of the circadian rhythm and various metabolic processes. The circadian rhythm is an endogenous mechanism, which generates entrainable biological changes that follow a 24-hour period. It regulates a number of physiological processes, including sleep/wakeful cycles and feeding behaviors. We recently demonstrated that REV-ERB-specific small molecules affect sleep and anxiety. The orexinergic system also plays a significant role in mammalian physiology and behavior, including the regulation of sleep and food intake. Importantly, orexin genes are expressed in a circadian manner. Given these overlaps in function and circadian expression, we wanted to determine whether the REV-ERBs might regulate orexin. We found that acute in vivo modulation of REV-ERB activity, with the REV-ERB-specific synthetic ligand SR9009, affects the circadian expression of orexinergic genes in mice. Long term dosing with SR9009 also suppresses orexinergic gene expression in mice. Finally, REV-ERBβ-deficient mice present with increased orexinergic transcripts. These data suggest that the REV-ERBs may be involved in the repression of orexinergic gene expression.


Introduction
The circadian rhythm is an autonomous, 24-hour, self-sustained oscillation regulating the physiology and behavior of organisms ranging from bacteria to humans [1][2][3][4][5][6][7]. The master clock is located in the suprachiasmatic nucleus (SCN) in the hypothalamus and is maintained by a negative feedback loop in which a CLOCK-BMAL1 complex activates E-box containing genes, including Cryptochrome (Cry1 and Cry2), and Period (Per1, Per2, and Per3). In turn, the CRY-PER complex inhibits the activating action of the CLOCK-BMAL1 heterodimers. The REV-ERBs have been demonstrated to be an essential part of the accessory loop by inhibiting the transcription of the core cock genes BMAL1, NPAS2, and CLOCK. [8][9][10][11][12][13]. The molecular oscillations that occur in the brain also occur in peripheral organs and are involved in the regulation of metabolic functions [12,[14][15][16].
(method of euthanasia). All procedures were approved and conducted in accordance to the Scripps Florida Institutional Animal Care and Use Committee.

General Mouse Studies
For acute injections, SR9009 was injected one time at either ZT0 (lights on) or ZT6, corresponding to the middle of the lights on period. ZT0 injections-groups of animals (n = 6) were sacrificed every six hours at ZT0, ZT6, ZT12 and ZT18. ZT6 injections-groups of animals (n = 6) were sacrificed one hour later at ZT7. For chronic studies, mice were administered SR9009 (100mg/kg, twice a day) for 10 consecutive days. Injections occurred at ZT0 (lights on) and ZT12 (lights off) in order to try and maintain exposure of SR9009 over a 24-hour period and be consistent with previously published dosing regimens of chronic SR9009 administration [17]. Animals (N = 5) were sacrificed at ZT6 after last injection. For REV-ERBβ KO mice, groups of animals (n = 5) were sacrificed at ZT6.

Statistical analysis
All data are expressed as mean ±S.E.M. All statistical analysis was performed using GraphPad Prism6 software. A two-way analysis of variance (ANOVA) was performed to determine significant differences between Time of day x treatment. One-way ANOVA was used to analyze intra-gene differences across several time points. All other analysis was performed using a Student's t-test. (N is indicated in the figure legends). Significance was assessed as follows: Ã p<0.05, ÃÃ p< 0.01, ÃÃÃ p < 0.001.

Results
In order to determine whether modulation of REV-ERB activity affected expression of PPO (Hcrt), OX 1 R (Hcrtr1), and OX 2 R (Hcrtr2), we administered SR9009 to mice at ZT0 and collected brain tissue every six hours to monitor gene expression changes ( Fig 1A).  Administration of the REV-ERB agonist, SR9009, at ZT0 affects expression of orexinergic genes in the brain over a 24-hour period. A. Mice were injected at ZT0 with the REV-ERB agonist SR9009 [100mg/kg, i.p.] and tissue was collected every six hours for a twenty-four hour period. The brain was dissected into hypothalamus or reticular formation and processed for mRNA levels at the different time points, ZT0, ZT6, ZT12, and ZT18. B. Mice administered with SR9009 or vehicle control were assessed for expression of PPO (Hcrt), Bmal1 (Arntl), and REV-ERBα (Nr1d1) in the hypothalamus and OX 1 R (Hcrtr1) and OX 2 R (Hcrtr2) in the brainstem. N = 6. One way analysis was used to determine intra-gene variation across time. Two-way ANOVA was used to assess differences between groups (time of day x treatment). *P<0.05, **P<0.01, and **P<0.001, n.s. = not significant.
doi:10.1371/journal.pone.0151014.g001 that, consistent with published data, the mRNA expression of Hcrt fluctuates over a 24-hour period, peaking at ZT6 [40]. Surprisingly, SR9009 suppressed this peak in Hcrt expression at ZT6 relative to vehicle controls. Furthermore, SR9009 also affected the expression of Hcrtr1 transcripts, inhibiting the initial upregulation of its expression at ZT6 relative to vehicle control. Similar effects were observed with SR9009 on Hcrtr2, although to a lesser degree. REV-ERBα contains a REV-ERB response element (RevRE) in its promoter region and has been shown to modulate its own expression. We also characterized the 24-hour expression pattern of the Bmal1 (Arntl) and REV-ERBα (Nr1d1) genes after ZT0 injections with SR9009 and observed repression of Arntl at ZT6 and little to no change in the expression of Nr1d1 relative to vehicle control, which is consistent with our previously published data (Fig 1B) [17]. Thus, modulation of REV-ERB activity using a synthetic REV-ERB agonist affects expression of orexinergic genes over a 24-hour period.
We recently demonstrated that acute injections at ZT6, which corresponds to peak REV-ERB mRNA expression, results in increased wakefulness and locomotion as well as reduced rapid-eye movement (REM) and slow-wave sleep (SWS) [51]. Since orexin is implicated in wake and alertness maintenance [33-36, 38, 52-58], we evaluated the effects of acute injections of SR9009 at ZT6 on orexinergic transcripts, as a possible mechanistic pathway mediating REV-ERB agonist wake-inducing effect. Mice were injected at ZT6 and sacrificed at ZT7, the time point where SR9009 has maximal wake-inducing effects, to collect brain tissue (Fig 2A). SR9009 suppressed Hcrt and Bmal1 (Arntl) in the hypothalamus (Fig 2B and 2C) and Hcrtr1 and Hcrtr2 transcript levels in the brainstem of mice (Fig 2D and 2E). Arntl was used as a positive control (Fig 2B).
To determine how chronic administration of REV-ERB ligands affects expression of Hcrtr1 and Hcrtr2, we administered SR9009 to mice for 10 days [100mg/kg, i.p., twice a day] after which time mice were sacrificed at ZT6 to collect brains for mRNA analysis. We assessed the expression of orexinergic genes in the hypothalamus, and the brainstem. In the hypothalamus, SR9009 inhibited Hcrt expression, with suppression of Arntl used as a control. (Fig 3A and 3B). Similarly, in the brainstem, mRNA expression of Hcrtr1 and Hcrtr2 were also reduced ( Fig 3C  and 3D respectively). These results indicate that the REV-ERBs may be repressing orexinergic transcription in these brain areas.
To determine how loss of REV-ERB expression affects orexin gene expression, we used fullbody REV-ERBβ deficient mice to assess Hcrt and Arntl levels in the hypothalamus, and Hcrtr1 and Hcrtr2 levels in the brainstem ( [51] and upublished data). Consistent with REV-ERBβ being a transcriptional repressor, we observed increased expression of Hcrt and Arntl in the hypothalamus (Fig 4A and 4B) and of Hcrtr1 and Hcrtr2 in the brainstem (Fig 4C  and 4D). These data suggest that REV-ERBβ may act as a transcriptional repressor of Hcrt, Hcrtr1, and Hcrtr2.
Our results show that modulation of REV-ERB activity affects transcription of the orexinergic genes [PPO (Hcrt), OX 1 R (Hcrtr1), and OX 2 R (Hcrtr2)]. Acute pharmacological manipulation of REV-ERB activity resulted in alterations of orexinergic genes over a 24-hour period after single injections at ZT0, relative to vehicle control. Similar results were observed with chronic pharmacological manipulation of REV-ERB activity (Fig 3). Delayed/repressed transcription of the orexinergic genes suggests that they may be REV-ERB target genes (Fig 1). Alternatively, orexin is also regulated by peripheral nutrient signaling, directly affecting the hypothalamus [52,58,[62][63][64][65] as well as the nucleus of the solitary tract (NTS) [67]. These signaling macronutrients and peptides may be activated by REV-ERB agonist administration, thus indirectly affecting orexinergic expression. Finally, modulation of orexigeneric gene expression may be due to post-translational effects at the genes, effects incited by nutrient signaling. Future studies examining this phenomenon are warranted in order to definitively determine whether this is the case.
Consistent with the REV-ERBs actively repressing orexinergic genes, mice deficient in REV-ERBβ demonstrate de-repression of orexinergic transcripts at ZT6, which would appear REV-ERB agonist SR9009 administration at ZT6 causes decreased transcript expression of orexinergic genes in the brain. A. Hypothalamic sections were analyzed for transcript levels of PPO (Hcrt) at ZT7. B. Hypothalamic sections were analyzed for transcript levels of Bmal1 (Arntl) at ZT7. C. Reticular formation sections were analyzed for transcript levels of OX 1 R (Hcrtr1) at ZT7. D. Reticular formation sections were analyzed for transcript levels of OX 2 R (Hcrtr2) at ZT7. N = 6. Student's t-test was used to assess differences between groups. *P<0.05, **P<0.01, and ***P<0.001. to further corroborate direct regulation of Hcrt, Hcrtr1, and Hcrtr2 by the REV-ERBs. However, recently published ChIP-seq data generated by the Lazar laboratory determining REV-ERBα binding sites in the brain do not support this as no REV-ERBα binding to any regions in or surrounding orexigenic genes was observed [68]. However, in this data set, REV-ERBα does bind in the promoter region of the Nur77 gene (Nr4a1), along with other transcription factors that have been demonstrated to regulate orexin expression, which may account for the changes in orexinergic transcription observed in our studies [69]. REV-ERBα and REV-ERBβ are thought to bind to the same DNA response element and regulate similar genes [18,70,71]. In fact, REV-ERBβ is thought to be a redundant protein to REV-ERBβ [18,71]. Therefore, the ChIP-seq data would suggest that the effects observed with SR9009 treatment and in the absence of REV-ERBβ are indirect effects. However, ChIP-seq studies determining the REV-ERBß cistrome in the brain, coupled with RNA-seq studies are warranted to definitively determine any direct/indirect effects of the REV-ERBs on orexin.
Orexin signaling via OX 1 R in the striatum and nucleus accumbens is associated with reward feeding and addictive behavior toward nicotine and other drugs [72][73][74]. In line with this, our lab recently demonstrated that administration of SR9009 injection decreased the addictive phenotypic properties of cocaine [19]. Suppression of reward behavior by the REV-ERBs may occur in part via orexinergic signaling. Future mechanistic studies may aid in describing the method of action of the REV-ERBs in reward. Therefore, pharmacological modulation of the REV-ERBs may be a viable therapeutic option to treat addiction, anxiety, and depression via regulation, at least in part, of the orexinergic pathway at the transcriptional level.

Conclusions
Modulation of REV-ERB activity and expression leads to changes in expression of the orexinergic genes Hcrt, Hcrtr1, and Hcrtr2. Our laboratory recently published REV-ERB ligands effects on sleep, anxiety, and metabolism. Given the overlap in REV-ERB pathways with the orexinergic system, our data suggests there may be an interplay between the orexin and REV-ERB signaling pathways. However, further studies exploring this overlap are warranted. Collectively, our data indicate that REV-ERB ligands may be a means to regulate orexin expression and could be a potential therapeutic avenue for disorders associated with aberrant orexin signaling. and OX 2 R (Hcrtr2) transcript levels in the brainstem were assessed using RT-PCR. Increased orexinergic and Bmal1 (Arntl) transcripts were observed at ZT6 in REV-ERBβ-deficient versus wild-type mice. N = 8. Student's t-test was used to assess differences between groups. *P<0.05, **P<0.01, and ***P<0.001. doi:10.1371/journal.pone.0151014.g004

Author Contributions
Conceived and designed the experiments: AA LAS TPB. Performed the experiments: AA SB YW. Analyzed the data: AA LAS TPB YW. Contributed reagents/materials/analysis tools: TMK. Wrote the paper: AA LAS TPB.