Nuclear Hormone Receptors for Heme: REV-ERBα and REV-ERBβ Are Ligand-Regulated Components of the Mammalian Clock

Abstract

The nuclear hormone receptors (NHRs), REV-ERBα and REV-ERBβ, regulate a number of physiological functions including the circadian rhythm, lipid metabolism, and cellular differentiation. These two receptors lack the activation function-2 region that is associated with the ability of NHRs to recruit coactivators and activate target gene transcription. These NHRs have been characterized as constitutive repressors of transcription due to their lack of an identified ligand and their strong ability to recruit the corepressor, nuclear receptor corepressor. Recently, the porphyrin heme was demonstrated to function as a ligand for both REV-ERBs. Heme binds directly to the ligand-binding domain and regulates the ability of these NHRs to recruit nuclear receptor corepressor to target gene promoters. This review focuses on the physiological roles that these two receptors play and the implications of heme functioning as their ligand. The prospect that these NHRs, now known to be regulated by small molecule ligands, may be targets for development of drugs for treatment of diseases associated with aberrant circadian rhythms including metabolic and psychiatric disorders as well as cancer is also addressed.

NUCLEAR HORMONE RECEPTORS (NHRs) are transcription factors that regulate a wide range of biological processes including carboyhydrate and lipid metabolism, cellular proliferation and differentiation, reproductive system development and function, and circadian rhythms. NHRs have a conserved domain structure with a variable amino-terminal domain (A/B region) and highly conserved DNA-binding domain (C region), a hinge region (D region), and a carboxy-terminal ligand-binding domain (E region, or LBD) (Fig. 1) (1). A small number of receptors also contain an F region that has unclear function. Forty-eight members of the NHR superfamily exist in humans, approximately half that number exist in Drosophila, and six times that number are found in the nematode. In humans, approximately half of the NHRs have identified ligands in which the receptors clearly function as ligand-activated transcription factors whereby, small hydrophobic molecules, such as steroid hormones, fatty acids, lipophilic vitamin derivatives, and dietary metabolites, act as ligands for the NHR proteins binding within the LBD. The remaining half of the human receptor superfamily and nearly all of the Drosophila and Caenorrhabitis elegans NHRs are orphan receptors, so called due to the lack of an identified ligand.

Fig. 1.

Comparison of the Orphan NHRs, REV-ERBα and REV-ERBβ, to the Drosophila melanogaster E-75

DBD, DNA-binding domain. Numbers indicate amino acids. Letters A–F refer to the domain structure common to most NHRs as described in the text. dE75, Drosophila melanogaster E-75.

NHRs bind directly to specific DNA sequences within the genome near their cognate target genes and regulate the rate of transcription in a ligand-dependent manner. Most bind to DNA as dimers: either homodimers (as is the case for steroid receptors) or heterodimers with another member of the NHR superfamily, the retinoid X receptor (2, 3). A limited number of NHRs have the ability to bind to DNA as monomers. Ligand binding (agonist) induces a conformational change within the LBD that allows for a unique surface to be created that is recognized by a class of proteins known as coactivators (4). Once the coactivator proteins are recruited to the NHR LBD, and hence the promoter region of the gene, the rate of mRNA synthesis is increased due to stabilization of the preinitiation complex and modification of the chromatin structure (5). Whereas most of the NHRs behave in this manner, others may exhibit constitutive activity due to some degree of constant binding to coactivator proteins or exhibit transcriptional silencing (active repression) of gene transcription due to constitutive binding of corepressor proteins that mediate recruitment of histone deacetylases (6).

The wide range of physiological and pathological process regulated by NHRs has made this superfamily of receptors a rich source of targets for drug development. The steroid receptors, including the mineralocorticoid, estrogen, androgen, glucocorticoid, and progesterone receptors, have been successful targets for a range of diseases including cardiovascular disorders, cancer, and inflammation. The thyroid hormone, peroxisome proliferator-activated, and retinoic acid and retinoid X receptors have been successful target for development of drugs for metabolic and skin disorders as well as cancer. In addition, peroxisome proliferator-activated receptor (PPAR)α and PPARγ are targets for the fibrates and thiazolidienedione classes of drugs used to treat dyslipidemia and type 2 diabetes mellitus, respectively. Recently deorphanized receptors such as the liver X receptors α and β [NR1H3 and NR1H2] and farnesoid X receptor (NR1H4) hold great promise for development of additional drugs targeting metabolic diseases (7, 8). Additional deorphanization efforts will likely yield an even greater number of targets for drug discovery as well as characterization of novel physiological regulatory processes.

REV-ERBα AND REV-ERBβ, UNUSUAL MEMBERS OF THE NHR SUPERFAMILY

REV-ERBα and REV-ERBβ are two unique orphan members of the NHR superfamily. REV-ERBα is encoded by the opposite DNA strand of the c-erbA oncogene, which is the thyroid hormone receptor-α (9, 10, 11). Both REV-ERBα and the closely related REV- ERBβ that was identified a few years later have an atypical LBD in which the carboxy-terminal activation function-2 is missing (Fig. 1) (12, 13, 14 , 107). This region of the LBD forms helix 12 and is required for recognition of coactivators required for transcriptional activation by NHRs (4). Consistent with the lack of the activation function-2 region, both of these NHRs have been demonstrated to be to be constitutive repressors of transcription due to their inability to bind to coactivators and their constant binding corepressors such as nuclear receptor corepressor (NCoR) (15). These two NHRs display an overlapping pattern of expression both spatially and temporally. Expression is exceptionally high in the liver, adipose tissue, skeletal muscle, and brain, and both exhibit circadian patterns of expression (16, 17, 18, 19). REV-ERBα and REV-ERBβ recognize two classes of DNA response elements and can bind to DNA as either monomers or homodimers, depending on the nature of the response element. REV-ERBs bind as a monomer to a (A/G)GGTCA half-site with a 5′ AT-rich extension or as a homodimer to a direct repeat 2 element (AGGTCA sequence with a 2-bp spacer) (20, 21). When bound to their response element [REV-ERB response element (RevRE)] within the promoter of a target gene, the REV-ERBs constitutively recruit corepressors to the promoter, resulting in repression of the target gene due to active histone deacetylation and condensation of the chromatin (15, 21). Another orphan NHR subclass known as the retinoic acid receptor-related orphan receptor [ROR (α,β, and γ); NR1F1, NR1F2, and NR1F3] also recognize at least a subset of RevREs and are often coexpressed in the same tissues as the REV-ERBs (22). In contrast to the REV-ERBs, the RORs are constitutive activators of transcription. Thus, the balance of expression of ROR (activator) vs. REV-ERB (repressor) has been suggested to be critical for dynamic regulation of target genes containing these elements. An interesting feedback loop also exists between these two NHRs. The Rev-erbα promoter contains a RevRE and is thus responsive to both autoregulation and regulation RORα (23, 24, 25). Additionally, expression of a dominant-negative form of REV-ERBβ in C2C12 myocytes results in an approximate 6-fold increase in RORα expression indicating that REV-ERB represses expression of RORα and thus completes the regulatory loop (26). Both REV-ERBs and RORα share a number of common target genes indicating that they likely coordinately regulate physiological processes.

REV-ERBs AND REGULATION OF METABOLISM

Loss of function experiments provide unequivocal data illustrating a role for REV-ERBs in lipid metabolism. Rev-erbα^−/− mice exhibit a dyslipidemic phenotype with elevated very-low-density lipoprotein (VLDL) triglyceride levels along with increased liver and serum apolipoprotein CIII (apoCIII) expression (Table 1) (27, 28). apoCIII is a component of both high-density lipoprotein and VLDL and plays a role in regulation of lipoprotein lipase activity and triglyceride levels. The staggerer (sg/sg) mouse, which harbors a RORα loss of function mutation, displays reduced apoCIII expression as well as lowered triglyceride levels (28). These results are consistent with the in vitro observations that apoCIII is a direct target gene of these NHRs and is repressed by REV-ERBα and activated by RORα (both in human and rodent cell lines) (27, 28). Apoliprotein AI, a component of high-density lipoprotein, is also regulated by both REV-ERBα and RORα; however, this appears to be species specific and does not occur in human cell lines (29). Recently, additional REV-ERBα target genes, including ElovI3 and plasminogen activator inhibitor type I (PAI-1), have been identified, further solidifying its role in lipid metabolism and atherosclerosis (30, 31). ElovI3 encodes a very long-chain fatty acid elongase found in the liver and adipose tissue whereas PAI-1 is a regulator of the fibrinolytic system and is believed to play a role in atherogenesis.

Table 1.

Metabolic Phenotypes of Mice Harboring Mutations in the Genes of the Circadian Oscillator

Mice	Phenotype	Mice	Phenotype
Rev-erbα−/−	↑Triglcyerides	Rorasg/sg	↑VLDL
	↑VLDL		↓apoAI
	↑apoCIII		↓HDL
	↓20% Body weight		Susceptible to atherosclerosis
Bmal1 −/−	↓Gluconeogenesis	Clockmut	Obesity
	↓Life span		↑Cholesterol
	Progeria-like phenotype		↑Glucose
			↑Triglycerides
	Impaired adipogenesis (in null embryonic fibroblasts)		↓Insulin

Additional loss of function experiments in which a dominant-negative form of REV-ERBβ (lacking its LBD and thus unable to repress transcription) was expressed in mouse myogeneic C2C12 cells provides additional insight into the role of REV-ERBs in lipid metabolism (26). This study first demonstrated that REV-ERBβ was expressed in proliferating myoblasts and decreased 2- to 3-fold as the cells exit the cell cycle and form differentiated multinucleated myotubes. Cells expressing the dominant-negative form of REV-ERBβ exhibit decreases in expression of critical genes involved in lipid metabolism including Fat/CD36, Fabp-3, Fabp-4, Ucp3, Srebp-1c, and Scd-1. Although it is not clear whether REV-ERBβ directly regulates these genes, it clearly demonstrates a role for REV-ERBβ regulating genes (at least indirectly) that are involved in fatty acid and lipid absorption, energy expenditure, and lipogenesis in muscle. Consistent with the observation that REV-ERBs inhibit myogenesis (32, 33), the expression myostatin (a growth factor that limits muscle tissue growth) was reduced approximately 24-fold in the C2C12 cells overexpressing the dominant-negative REV-ERBβ (26). Interestingly, mice deficient in myostatin have reduced accumulation of body fat, increased muscle mass, and increased insulin sensitivity (34), and Scd-1^−/− mice have increased insulin sensitivity and reduced VLDL and triglyceride levels and are resistant to diet-induced obesity (35, 36).

In contrast to the role of REV-ERBs inhibiting myogenesis, these NHRs are involved in induction of adipogenesis. Chawla and Lazar (37) first demonstrated that Rev-erbα was highly induced during adipogenesis. A PPAR response element has been found in the promoter of the Rev-erbα gene (38) and the role of REV-ERBα in adipogenesis may be linked to PPARγ, which is also induced during adipogenesis (39). This Rev-erbα PPAR response element is also responsive to PPARα, which is particularly interesting given the role of PPARα in regulation of Bmal1 expression and regulation of the circadian rhythm in the liver (40). PPARγ agonists increase Rev-erbα expression in adipose tissue both in vivo and in 3T3-L1 preadipocytes as do PPARα agonists in hepatocytes (38, 39). Overexpression of REV-ERBα in 3T3-L1 preadipocytes results in increased expression of markers of adipogenesis including aP2, PPARγ, and CCAAT enhancer binding protein-α along with small increase in lipid accumulation (39). Furthermore, overexpression of REV-ERBα in these cells synergized with the PPARγ ligand rosiglitazone to increase these markers of adipogenesis (39). Although the Rev-erbα^−/− mice do not exhibit an overt adipose or muscle phenotype (41), the mice do display alterations in lipid metabolism as discussed above and, in addition, exhibit alterations in myosin heavy chain isoform expression in skeletal muscle (42). Functional redundancy between REV-ERBα and REV-ERBβ may be one explanation for the limited phenotype. Development of Rev-erbβ null mice as well as double-knockout mice will be needed to address this issue in the future.

REV-ERBs AND THE MAMMALIAN CLOCK

Circadian rhythms play an essential role in aspects of physiology and behavior including the sleep-wake cycle, feeding, metabolism, body temperature, blood pressure, and renal function. The circadian rhythm is generated by feedback loops in gene expression in which heterodimers of brain and muscle Arnt-like protein 1 (BMAL1) and circadian locomoter output cycles kaput (CLOCK) induce the expression of the Cryptochrome (Cry1, Cry2) and Period (Per1, Per2, Per3) genes (Fig. 2). As CRY and PER reach critical levels, they are able to repress the stimulatory effect of the CLOCK/BMAL1 dimer on the expression of their own genes. REV-ERBα has been demonstrated to be the major regulator of the cyclic expression of Bmal1 (18, 43). Two ROR response elements (ROREs)/RevREs are located in the Bmal1 promoter, and Bmal1 transcription is repressed by REV-ERBα. REV-ERBα expression is, in turn, regulated by the BMAL1/CLOCK heterodimers via E-boxes found within the Rev-erbα promoter (44, 45). Rev-erb^−/− mice exhibit aberrant expression of Bmal1 and display alterations in the period and phase of their circadian behavior patterns (43). The RORs have also been demonstrated to play a key role in regulation of Bmal1 via the identical RORE/RevRE, and consistent with this observation the staggerer mice display alterations in the circadian oscillator (46). When this element responsive to both REV-ERBs and RORs is placed upstream of a reporter construct, this element alone is enough to drive the circadian expression of the reporter gene (47). This is consistent with the observation that REV-ERBα (the repressor) and RORα (the activator) are expressed in an oscillatory fashion 180° out of phase with one another (43, 46) (Fig. 2). Interestingly, the REV-ERB target gene, Bmal1, which is such a critical component of the circadian oscillator, is also essential for adipogenesis (48).

Fig. 2.

Mechanism of the Cellular Circadian Oscillator

The top portion of the figure illustrates the oscillations in Bmal1/Clock and Per/Cry and RORα/REV-ERBα during the 24-h cycle. The lower portion of the figure illustrates the molecular mechanism responsible for creation of the oscillation. The roles of NHRs, RORα, and REV-ERBα are highlighted in this figure.

The DROSOPHILIA MELANOGASTER NHR, E-75, HEME, AND REV-ERBs

A Drosophila ortholog of human REV-ERBα, E75, was recently demonstrated to contain a heme prosthetic group (49). The heme is apparently tightly bound, and removal of the heme is associated with loss of E75 stability. The oxidation state of heme within E75 determines whether this NHR can interact with its heterodimer partner, DHR3 and, interestingly, this dimer interaction is regulated by binding of either NO or CO to the heme sensor, indicating that E75 may function as a diatomic gas sensor (49). Coordination of heme was assessed by site-directed mutagenesis in this study (49) and well as in a more recent study (50). E75 functions in Drosophila development and a number of potential roles for heme were proposed for regulation of E75 activity including regulation of stability of the protein, a redox sensor, and a sensor of diatomic gases (49). The closest human NHR relatives of Drosophila E75 are REV-ERB α and REV-ERB β (Fig. 1). Alignment of the LBDs of the receptors reveals conservation of several residues that have been demonstrated to be important for coordination of heme (49, 50). Thus, it becomes conceivable that the REV-ERBs may also bind to heme.

REV-ERBs ARE HEME RECEPTORS

Recent work from our laboratory (in collaboration with Fraydoon Rastinejad) as well as that from the Lazar group demonstrated that heme is the ligand for REV-ERBα (51, 52) and REV-ERBβ (51). However, unlike the Drosophila E-75 in which heme is essential for the stability of the protein, the REV-ERBs appear to function as true receptors for heme and alter their function in response to changing intracellular heme concentrations. Both groups demonstrated that heme binds reversibly and specifically to the LBD of REV-ERB and modulates the ability of the receptor to recruit the corepressor NCoR (51, 52). Based on isothermal titration calorimetry, the affinity of heme for REV-ERBα and REV-ERBβ is approximately 2–4 μm (51). Heme binding is required for efficient NCoR recruitment, and mutations that eliminate the ability of the receptor to bind heme also eliminate its ability to recruit NCoR. Alterations in intracellular heme levels also modulate the expression of REV-ERB target genes such as Bmal1 and Elovl3 in a manner consistent with heme stabilizing the REV-ERB-NCoR interaction. Diatomic gasses such as nitric oxide did not appear to modulate REV-ERB activity. A model for ligand regulation of REV-ERB activity is shown in Fig. 3. Although the idea that heme is a ligand for a NHR may seem peculiar, a closer look reveals that this is not at all an unusual situation. In terms of molecular weight, although heme is at the high end of the range of sizes of NHR ligands, it is clearly not the largest (Fig. 4). T₃, a thyroid hormone receptor ligand, as well as ligands for pregnane X receptor such as rifampacin, are larger. Certainly the hydrophobic character of heme is consistent with the characteristics commonly associated with NHR ligands. A recent study reported the crystal structure of the REV-ERBβ LBD indicating that the ligand-binding pocket is too small to accommodate any ligands of the size characteristic of nuclear receptor ligands (53). The receptor used by these investigators lacked helices 1 and 2 as well as a long amino acid insert between helices 2 and 3 that is characteristic of REV-ERBs; thus, this structure may not represent the physiologically relevant structure.

Fig. 3.

Model Illustrating the Proposed Mechanism for Heme Modulation of REV-ERB Activity

REV-ERB functions as a receptor for heme. REV-ERB recognizes and binds specific DNA sequences known as RevREs in the promoter regions of its target genes. Heme binds directly to the LBD of REV-ERB and increases the affinity of the receptor for the corepressor NCoR. When NCoR is recruited to the promoter by REV-ERB, the gene is repressed due to the histone deacetylase activity that is associated with the NCoR/histone deacetylase complex. Modulation of REV-ERB target genes is associated with regulation of physiological functions such as lipid metabolism, circadian rhythm and cellular differentiation. DBD, DNA-binding domain.

Fig. 4.

Examples of Ligands for NHRs

Heme is a ligand for REV-ERBα and REV-ERBβ, T₃ is a ligand for the thyroid hormone receptor, 1,25-dihydroxyvitamin D₃ is a ligand for the vitamin D receptor, estradiol is a ligand for the estrogen receptor α and β, chenodeoxycholic acid is a ligand for farnesoid X receptor, and rifampicin is a ligand for pregnane X receptor. MW, Molecular weight.

The discovery of heme as a ligand for the REV-ERBs also has implications with respect to our understanding of how NHRs typically function. In the classical model for NHR action, the receptor is inactive, and upon ligand binding the receptor undergoes a conformational change allowing recruitment of coactivators, thus becoming an activator of transcription. The REV-ERBs were previously believed to lack ligands and function as constitutive repressors of gene transcription. However, these two studies have demonstrated a new model in which binding of a physiological ligand transforms an inactive receptor into one that efficiently recruits corepressor and is a repressor of transcription (51, 52). Given that there are other members of the NHR superfamily that are thought to be constitutive repressors of transcription, such as the chicken ovalbumin upstream promoter-transcription factors, the example of the REV-ERBs may not be unique.

REGULATION OF HEME SYNTHESIS

Like the REV-ERBs, the synthesis of heme has been directly correlated to the circadian rhythm via regulation of the expression of the rate-limiting enzyme in heme synthesis, aminolevulinic acid synthase (ALAS) (54, 55, 56) (Fig. 5). Approximately 15% of heme synthesis occurs in the liver where heme is incorporated into a number of metabolic enzymes (e.g. p450s), whereas most of the remaining 85% is synthesized in erythroid progenitors for incorporation primarily into hemoglobin (57). Heme synthesis occurs both in the cytosolic and mitochondrial compartments in both systems; however, liver vs. erythroid regulation of synthesis exhibits major differences. In the erythroid cells heme synthesis is regulated primarily by the availability of iron, whereas in the liver, heme biosynthetic enzymes are subject to rapid turnover and are able to respond to the changing metabolic environment (57). The rate-limiting step in heme synthesis is catalyzed by two distinct ALAS enzymes, ALAS1 and ALAS2. ALAS1 is ubiquitously expressed whereas ALAS2 is expressed only in erythroid progenitors. Distinctions between the regulation of ALAS1 and ALAS2 provide for the observed differences in regulation of heme synthesis in the liver vs. the erythroid cells. Interestingly, when Rev-erbβ was first described, it was isolated in chickens and was found to be highly expressed in the spleen, which is the major organ of erythropoesis in birds (14). Given the proximity of Rev-erbβ expression, heme synthesis, and erythropoesis in this tissue, it is important to investigate the potential role of Rev-erbβ in erythropoesis.

Fig. 5.

Schematic Illustration the Pathway Responsible for Biosynthesis of Heme, the Ligand for REV-ERBα and REV-ERBβ

CoA, Coenzyme A.

HEME AS A REV-ERB LIGAND REGULATING THE CIRCADIAN RHYTHM

Alas1 expression is regulated in a circadian fashion and is dependent on the PER1 and PER2 components of the clock (56). Additionally, components of the cellular circadian clock such as neuronal PAS domain protein 2 (NPAS2), a homolog of CLOCK expressed in the forebrain, which also functions as a heterodimer with BMAL1, and PER2 contain heme as a prosthetic group (54, 58). NPAS2 activity has been shown to be a gas-sensitive transcription factor (58), and the NPAS2/BMAL1, along with PER, is an essential regulator of Alas1 expression (54, 56), leading to the hypothesis that heme may play a role in circadian feedback loop whereby it regulates the activity of critical clock components and they, in turn, regulate heme synthesis. Consistent with this hypothesis was the observation that injection of heme into mice altered mPer1 and mPer2 expression and modified circadian locomotor behavior relative to injection of vehicle alone (54).

Both the NPAS2/BMAL1 and CLOCK/BMAL1 heterodimers are sensitive to the redox state of nicotinamide adenine dinucleotide and may utilize this mechanism to allow the entrainment of the circadian clock to the nutritional state (59). Feeding clearly has the ability to entrain the circadian clock, particularly in the liver where metabolism of carbohydrates and lipids are coordinated; interestingly, this occurs in the absence of input from the central circadian oscillator in the suprachiasmatic nucleus of the brain (60, 61, 62). A link between the circadian rhythm and glucose/lipid metabolism has been recognized for some time. Plasma glucose and triglyceride levels display a circadian pattern in addition to circulating levels of hormone involved in glucose and lipid metabolism including corticosterone and adiponectin (63). This link between the circadian rhythm and metabolic disorders is clearly demonstrated in mice deficient in either Bmal1 or harboring a mutant Clock gene (Table 1). These mice display abnormalities similar to metabolic syndrome including obesity, dyslipidemia, and impaired glucose metabolism (63, 64).

A heme-nutritional status relationship was revealed during the analysis methods of treatment for genetic diseases associated with accumulation of heme precursors, porphyrias. The effects of accumulation of heme precursors lead to mental disturbances and can be precipitated by fasting. Glucose infusion is the typical treatment of choice during a porphyric attack, but the exact mechanism of action was not understood until recently. Handschin et al. (65) found that the nuclear receptor coactivator, PPAR coactivator (PGC)-1α, directly regulates the expression of Alas1, and hence heme synthesis. The expression of PGC-1α is significantly up-regulated in response to fasting and mediates the transition toward oxidation of fatty acids and gluconeogenesis when plasma glucose is low (66). Glucose infusion results in decreased expression of PGC-1α and thus, a decrease in expression of Alas1 and decreased heme synthesis, which results in decreased accumulation of heme precursors in the porphyric individual. This analysis of the role of PGC-1α in regulation of heme synthesis provides significant insight into how heme is providing a signal as to the nutritional status of the organism. Because heme production is coupled to PGC-1α expression, there is a clear link between nutritional status and heme; therefore, heme may provide a key connection as a signaling molecule to indicate the nutritional status to the circadian oscillatory machinery and allow it to respond appropriately. Normal operation of the circadian oscillator is required for normal metabolic regulation, and dysregulation is associated with a metabolic disease phenotype. Additionally, methods to modify the activity of various components of the oscillator, such as the REV-ERBs, may be useful for treatment of metabolic diseases.

HEME AS A REV-ERB LIGAND REGULATING DIFFERENTIATION

Because the REV-ERBs play a role in both adipogenesis and myogenesis, this brings about the question of what role heme may be playing as a regulator of REV-ERB function in these processes. Heme is a porphyrin that is an essential prosthetic group for several proteins critical for skeletal muscle function including myoglobin and cytochrome oxidase. Heme serves not only a functional role as a prosthetic group within proteins but also as an independent regulatory role by which it increases myoglobin expression and mitochondrial biogenesis (67, 68, 69). Heme has also been shown to induce differentiation and maturation of rat primary skeletal muscle cells as well as increase their aerobic capacity (70, 71). Interestingly, the expression and activity of ALAS1 in muscle is increased by exercise (68, 69, 72, 73, 74). This suggests that one adaptive response of skeletal muscle in response to exercise will be an increase in intracellular heme levels, which correlates with increased differentiation/maturation and mitochondrial biogenesis (aerobic capacity). It is possible that this increase in heme synthesis may be mediated by the exercise-induced increase in PGC-1α expression (65, 75, 76). Thus, it is clear that intracellular heme levels vary under differing physiological conditions in muscle and may affect the activity of the REV-ERBs. Heme has also been demonstrated to play an important role in adipogenesis. If heme biosynthesis is blocked using pharmacological methods, 3T3-L1 cells fail to differentiate into adipocytes whereas addition of heme induces adipogenesis (77). The role that REV-ERB plays in mediating these effects of heme is unclear currently; however, evidence suggests that these recently deorphanized NHRs may be playing an important role.

LIGANDS FOR REV-ERBs: IMPLICATIONS FOR DISEASES ASSOCIATED WITH CIRCADIAN RHYTHM DYSFUNCTION

Aberrant circadian rhythms are associated with numerous disorders in the human. Sleep disorders including chronic insomnia are clearly associated with an abnormal circadian rhythm (78, 79). Mutations or polymorphisms in either key components of the mammalian clock mechanism or genes that directly regulate the stability of the components of the mechanisms affect the sleep-wake cycle (80, 81). Mood disorders in humans, such as bipolar disorder, major depressive disorder, and seasonal affective disorder, have been demonstrated to be associated with alterations in the circadian rhythm (82, 83). Mutations in the Clock gene in mice induce mania-like behavior (84), and polymorphisms in the human Clock and Bmal1 genes are associated with bipolar disorder (85, 86, 87, 88, 89). Mood-stabilizing drugs such as lithium and valproate have been demonstrated to have direct effects on circadian clock. Lithium, which lengthens the circadian period in numerous organisms, including humans (90), has been recently shown to affect the pattern of Bmal1 expression by modulation of REV-ERBα stability (91). Valproate alters the level of expression of several circadian genes in the brain (92) as does the selective serotonin reuptake inhibitor antidepressant fluoxitene (93).

In addition to clear linkage of the circadian clock to psychiatric disorders, the circadian rhythm plays an essential role in regulation of metabolism as previously discussed. However, in addition to the experimental genetic evidence linking the circadian rhythm to metabolic regulation, there are considerable epidemiological studies pointing to a clear role of the clock in metabolic disease. Several studies have shown that the incidence of cardiovascular disease and metabolic disturbances is higher in individuals who alter their normal sleep-wake pattern for shift work (94, 95, 96). Epidemiological as well as biochemical data have also shown a relationship between abnormal circadian rhythms and cancer (97, 98, 99). Two independent studies have shown that women shift workers, who are required to reverse their sleep-wake cycle, have as much as a 60% increase in the risk of developing breast cancer (100, 101). Consistent with the observation of a link between circadian regulation and cancer, Fu et al. (102) showed that mice deficient in Per2 expression were prone to development of cancer and suggested that PER2 is a tumor suppressor. Interestingly, Per genes have been shown to be deregulated in up to 95% of human breast cancers (103). Additionally, Matsuo et al. (104) demonstrated a direct regulatory link between the circadian rhythm and cell cycle where wee1, a critical cell cycle regulator, was directly regulated by components of the circadian oscillator. Previous studies have also shown that REV-ERBα regulates the expression of the c-myc protooncogene (105). Thus, there is clear evidence that alterations in the normal circadian rhythm are associated with human disease.

CONCLUSIONS

The discovery that the orphan NHRs, REV-ERBα and REV-ERBβ, are receptors for heme opens many new areas of investigation. From the physiological perspective, this observation has solidified a recently observed link demonstrating close association between regulation of the circadian rhythm and metabolic function. Additionally, previous studies demonstrating circadian oscillations in the expression of heme synthetic enzymes appear to bear additional significance now. We have recently observed that intracellular heme levels oscillate in a circadian fashion in synchronized NIH3T3 cells (106). The fact that the REV-ERBs are receptors for heme suggests that heme is likely to play an important role in regulation of functions that have been associated with REV-ERB including circadian function, glucose and lipid metabolism, adipogenesis, and myogenesis. Because these NHRs regulate such a critical physiological function, the circadian rhythm, the prospect that synthetic REV-ERB ligands can be designed treat disorders associated with aberrant circadian function, such as metabolic and psychiatric diseases and cancer, is certainly a possibility. It is also possible that short acting synthetic REV-ERB ligands may be useful for entrainment or resetting the circadian clock and thus be useful for sleep disorders and/or jet lag.

NURSA Molecule Pages:

• Coregulators: NCOR | PGC-1α;

• Nuclear Receptors: Eip75 | PPARα | PPARγ | REV-ERBα | REV-ERBβ | RORα.