Circadian Rhythm: A Functional Connection Between SHP and DEC1 Transcription Factor

Circadian rhythms ensure physiological processes to be coordinated with daily changes of the surrounding. Circadian rhythm misalignment has been increasingly recognized to pose health risk for a wide range of diseases, particularly metabolic disorders. The liver maintains metabolic homeostasis and express many circadian genes, such as the genes encoding differentiated embryo chondrocyte-1 (DEC1) and small heterodimer partner (SHP). DEC1 is established to repress transcription through class B E-box elements, and SHP belongs to the superfamily of nuclear receptors and has multiple Ebox elements in its promoter. Importantly, DEC1 and SHP are expressed in an inverse oscillating manner. The present study was performed to test the hypothesis that the SHP gene is a target gene of DEC1. In cotransfection experiments, we have demonstrated that DEC1 repressed the SHP promoter and attenuated the transactivation of the classic circadian activator complex of Clock/Bmal1. Site-directed mutagenesis, electrophoretic mobility shift assay and chromatin immunoprecipitation established that the repression was achieved through the E-box in the proximal promoter. Overexpression of DEC1 led to decreased expression of SHP. In horse serum-shocked cells (induction of circadian rhythms), the widely used epileptic agent valproate inversely altered the expression of DEC1 and SHP. Both DEC1 and SHP are involved in energy balance and valproate is known to induce hepatic steatosis. Our findings collectively establish that DEC1 constitutes the negative loop of the SHP oscillating expression and that the DEC1-SHP pathway is intimately involved in energy homeostasis with profound pathophysiologic significance.

. Mammals have central and peripheral circadian clocks [8]. The central clock is located in the suprachiasmatic nucleus (SCN) of the hypothalamus and is entrained by such environmental cues as the light-dark cycle [9,10]. Through the retinohypothalamic tract, the SCN receives photic input signals and generates rhythms, which subsequently synchronize multiple peripheral clocks through neural and humoral signaling [9][10][11][12]. While each organ has its own clock, the liver clock is the most studied peripheral clocks, particularly related to food cues [13][14][15][16].
Circadian rhythms are controlled by a group of core clock genes such as Clock (circadian locomotor output cycles kaput) and Bmal1 (brain and muscle ARNT-like 1) [8,17]. Clock and Bmal1 form a heterodimer and transactivate through E-Box elements the Per (Period-1 and 2) and Cry genes (Cryptochrome-1 and 2). Per and Cry proteins in turn interact with Clock/Bmal1 and attenuate their own transactivation [5,8,17]. In addition to Clock/Bmal1, we and other investigators have demonstrated that DEC transcription factors, differentiated embryo chondrocyte-1 and 2, are strong E-box binding proteins [18][19][20]. However, binding to E-box by DEC transcription factors leads to efficacious repression of the target genes including Per1. DEC1 and DEC2 are both central and peripheral oscillators [15]. In particular, DEC1 has been recognized to play fundamental roles in orchestrating metabolism and resetting the liver clock. Indeed, DEC1 is one of the most sensitive genes in responding to food cues. A 30 min feeding significantly induces the hepatic expression of DEC1 [15].
In addition to DEC1, there are several other well-characterized hepatic circadian genes such as the gene encoding small heterodimer partner gene (SHP) [5,21,22]. SHP has been established as a major regulator of diverse metabolic pathways, particularly in bile acid synthesis, lipid metabolism, glucose homeostasis and liver fibrogenesis [23]. Structurally, SHP belongs to the superfamily of nuclear receptors [23,24].
However, it lacks the DNA binding domain, thus is referred to as an atypical nuclear receptor. SHP has been shown to interact with nuclear receptors and/or compete for co-factors (e.g., co-activators or co-repressor), delivering potent regulatory activities at the transcriptional level [23]. On the other hand, many nuclear receptors have been shown to support the induction of SHP in response to endobiotics such as bile acids.
In addition to nuclear receptors, the circadian complex of Clock/Bmal1 is a potent transactivator of the SHP gene [23]. The transactivation is achieved through the element CACGTG, a special type of E-box recognized by DEC transcription factors [23]. In mice, the expression of DEC1 and SHP is inversely oscillating [25,26].

Plasmid
Expression constructs of Bmal1 and Clock were gifts of Dr. Marina P Antoch of the Cleveland Clinic [27]. The Per1 promoter reporter (Per1-luc) was a gift of Dr. Joseph S. Takahashi of Northwestern University [28]. The SHP-2.2 Luc reporter was a gift of Dr. Hueng-Sik Choi of Keimyung University [29]. The 5' deletion mutants of the SHP reporter were prepared by inserting a Nhe I-Hind III fragment into the pGL3 basic vector. These fragments were generated by PCR with primers as described in Table   I. The SHP mutant reporter with a disrupted E-box (E1) was prepared with the same approach but the mutations were introduced in the forward primer (Table I). The DEC1 expression construct and its mutants (deletion or substitution) were described elsewhere [18,30]. All constructs were subjected to sequencing analysis.

Electrophoretic mobility shift assay (EMSA)
DEC1 stably transfected cells with a tetracycline-inducible construct [33] were cultured in the presence or absence of tetracycline (0.1 µg/mL) for 24. Cells were harvested and nuclear extracts were prepared with a nuclear extraction kit (Active Motif, Carlsbad, CA). The EMSA experiment was performed with Lightshift Chemiluminescent EMSA Kit (Thermo Fisher Scientific, Waltham, MA) as described previously [34]. The sense strand for SHP-187/162 (E1) was synthesized as labeled or non-labeled form (for competition). Nuclear protein (5 μg) was mixed with binding buffer and then incubated with a double-stranded biotinylated probe (0.1 pmol) on ice for 20 min. In competition assays, nuclear extracts were first incubated with an unlabeled probe at a 50x excess for 30 min before addition of the labeled probe. For antibody-disruption assay, the nuclear extracts were first incubated with anti-DEC1 antibody on ice for 20 min and then with the labeled probe. As a positive control, the EMSA experiment was performed with a DEC2 E-box containing probe [18].

Regulated expression of DEC1 and SHP by valproate in serum-shocked circadian induction
HepG2 cells were seeded in 6-or 24-well plates at a density of 6x10 5

Other analyses
The anti-DEC1 antibody against a peptide derived from the C-terminus was described elsewhere [33].

Repression of the SHP promoter by DEC1
DEC1 and SHP are established to play critical roles in a wide range of biological activities including metabolic homeostasis [23,35,36]. Both DEC1 and SHP are circadian genes and their expression is inversely rhythmed [25,26]. We have shown that DEC1 is a sequence-specific transcription factor that acts on Sp1 site as well as a specific type of E-box: CACAGT [30,37]. The SHP promoter and its immediate upstream sequence contain multiple E-boxes including a CACATG. We therefore hypothesized that DEC1 transcriptionally regulates the expression of SHP. To test this hypothesis, we first examined whether DEC1 represses the SHP promoter.
Specifically, cotransfection experiments were performed with an SHP promoter luciferase reporter in in 293T cells. For comparison, a Per1 reporter was included. We and other investigators have demonstrated that Per 1 is a circadian gene and negatively regulated by DEC1 [18,30].
As shown in Fig. 1, DEC1 repressed both SHP and Per1 reporters and the repression occurred in a dose-dependent manner. The repression was robust by as much as 90%.
Nevertheless, the SHP reporter was repressed to a greater extent than the Per1 reporter by 10-15% depending on the amount of DEC1 construct used for the transfection (Fig. 1A). We next tested whether DEC1 attenuates Clock/Bmal1transactivation of the SHP and Per1 promoters, as the Clock/Bmal1 heterodimer has been shown to regulate both DEC1 and SHP in a circadian manner [23,30]. Once again, cotransfection was performed. As shown in Fig. 1B, Clock/Bmal1 strongly transactivated both the SHP and Per1 reporters with the Per1 reporter being transactivated to a greater extent (7 versus 12 fold) (Fig. 1B). However, the transactivation of the SHP reporter was attenuated to a much greater extent than that of the Per1 reporter by DEC1. For example, DEC1 at 10 ng attenuated the Clock/Bmal1 transactivation of the SHP reporter by 97%. In contrast, the Clock/Bmal1 transactivation of the Per1 reporter was attenuated by 55% only with the same amount of DEC1. We next tested whether the repression of the SHP promoter requires DNA binding. As shown in Fig. 1C, no repression was detected with all constructs except DEC1 (wild-type) and DEC1P56A. We have previously shown that substitution of the residue proline-56 with an alanine remained the ability for DEC1 to bind to E-box and deliver repressive activity [18,30]. In contrast, substitution of the residue arginine-58 with a proline no longer bound to E-box element. These results conclude that DEC1 is a transcriptional repressor of SHP.

Repression of the SHP promoter by DEC1 through the E-box in the proximal promoter
The proximal promoter of SHP is an E-box rich region and has as many as 7 E-box elements [29]. However, these elements differ slightly with 2 of them being CACCTG, and 1 of the following: CACTTG, CATCTG, CAGCTG, CAGGTG and CACGTG. To specify whether one or more of these elements support DEC1 repression, deletion and site-directed mutants of the SHP reporter were prepared and tested for the responsiveness to DEC1. Once again, DEC1 stably transfected cells were cultured in the presence or absence of tetracycline and then transfected with a reporter. As shown in Fig. 2A, all deletion SHP reporters were repressed except SHP-116Luc, suggesting that the E-box (E1: CACGTG) in the SHP-191Luc reporter supported the repression.
To specify the role of this E-box in DEC1 repression, reporter SHP-191Luc was subjected to site-directed mutagenesis to selectively disrupt the E-box (CACGTG to AACGGG). As shown in Fig. 2A, disruption of this E-box completely attenuated DEC1mediated repression of the SHP promoter.
We next tested whether this E-box interacted directly with DEC1. The DEC1-stable line was cultured in the presence or absence of tetracycline, and nuclear extracts were prepared. Double-stranded oligonucleotides harboring this E-box were synthesized and biotinylated. The labeled probe was incubated with the nuclear extracts and To determine whether DEC1 occupies the SHP promoter region that harbors this Ebox (i.e., E1), ChIP experiment was performed. To gain specificity, primers were designed to amplify three fragments: the binding E-box (BE) fragment (E1 E-box); the other E-box (OE) fragment (other E-boxes but not binding) and the non E-box (NE) fragment (no E-box). As shown in Fig. 2C, chipped DNA showed the abundant presence of the BE-box fragment (labeled as lane 1) but not the other fragments. As expected, input DNA produced amplification of all three fragments (Right of Fig. 2C).
It should be noted that pre-immune IgG was used as a control but did not enrich any fragments. The ChIP experiment was performed with HepG2 cells transfected with To complement the transfection study, we next tested whether SHP and DEC1 are inversely regulated for their oscillating expression by valproate, a widely used antiepileptic that was established to down-regulate SHP [38]. Importantly, valproate is a steatotic agent and both DEC1 and SHP are metabolic regulators.
To mimic circadian rhythm, HepG2 cells were shocked by horse-serum and then treated with valproate. Cells were collected starting at 6 h after serum-shock and then at a 6 h interval. Total RNA was isolated and analyzed for the expression of SHP and DEC1. Both genes were expressed in a circadian manner and the patterns of the expression were inversed between these two genes (Fig. 3B).

DISCUSSION
Normal circadian rhythms ensure physiological processes to be coordinated with daily changes of the environment [1-3]. Circadian rhythm misalignment has been increasingly recognized to pose health risk for a wide range of diseases [4][5][6][7]. The DEC1 and SHP genes are members of the liver clock and have been shown to play critical roles in metabolic homeostasis [5,23,35,36].  1A and B). Consistent with the notion, the expression of DEC1 and SHP is inversely oscillating [25,26].
It remains to be determined whether the Per/Cry proteins negatively regulate the expression of SHP. Nonetheless, it is likely that DEC1 exerts a dominant repression, particularly on the oscillating expression of SHP. In this study, we have shown that DEC1 repressed both the SHP promoter and the Per1 promoter with the SHP promoter being repressed to a greater extent (Fig. 1A). Importantly, DEC1 at 10 ng attenuated the Clock/Bmal1 transactivation of the SHP promoter by as much as 97% but only 55% on the transactivation of the Per1 promoter (Fig. 1B). In both cases, DEC1 exerted repressive activity by binding to E-box CACGTG. The SHP E-box is flanked by GTGC (5') and GGGT (3'), respectively, whereas the Per1 by TAGC and ACAG, respectively.
It remains to be determined whether the differences in flanking sequences contribute to the differences in response to DEC1 repression. An early study demonstrated that the flanking sequences were important for interacting affinity with stra13, the mouse counterpart of DEC1 [26]. In this study, we have also shown that the antibody against DEC1 disrupted the shifted band with the SHP E-box (Fig. 2B). In contrast, we have shown that the same antibody caused a supershift of the DEC2 E-box and the Per1 Ebox [19,30].
The DEC1-SHP pathway likely plays critical roles in the synthesis of bile acids, particularly the circadian production of these endobiotics. While there are several pathways that have been shown to regulate the synthesis of bile acids, the pathway mediated by SHP and FXR (farnesoid X receptor) has been extensively studied for the bile acid-activated regulatory cascade [5,23]. This cascade is commonly referred as the bile acid negative feedback inhibition on the expression of the cytochrome P450 enzyme cholesterol 7α-hydroxylase (CYP7A1). CYP7A1 is the first and rate-limited enzyme in bile acid synthesis. Increased production of bile acids activates FXR, leading to the induction of SHP. Induction of SHP inactivates LRH-1 (liver receptor homolog-1) and HNF4α (hepatocyte nuclear factor 4α). In this study, we have shown that DEC1 downregulated SHP, thus counteracting the feedback inhibition.
Interestingly, the transcription factor DEC2 (functionally related to DEC1) reportedly repressed the expression of rat CYP7A1 through an E-box: CACATG [39]. This E-box is conserved in human and mouse based on a BLAST search. It remains whether DEC1 binds to this E-box and causes repression. Nevertheless, we have reported that DEC1 negatively regulated the expression of DEC2. It is likely that DEC1 de-represses CYP7A1 by downregulating SHP and DEC2.
The DEC1-SHP pathway likely serves as an important mechanism for lipid metabolism.
Although there are exceptions, SHP is generally considered to be lipogenic whereas DEC1 is anti-lipogenic. SHP reportedly augmented the transactivation by PPARγ (peroxisome proliferator-activated receptor-γ), leading to marked lipid accumulation in the liver [7]. Likewise, transgenic expression of SHP induced liver steatosis [41].
Consistent with these observations, SHP null mice were protected against diet-induced obesity. DEC1, on the other hand, has been shown to repress lipogenic genes such as fatty acid synthase and inhibit adipogenesis [42,43]. Overexpression of DEC1 by viral transduction alleviated fatty liver phenotypes accompanied by suppressed expression of the lipogenic gene Srebp-1c (sterol regulatory element-binding protein -1c) [35]. Interestingly, SHP null mice supported higher induction of Srebp-1c in response to cholic acid treatment, suggesting that SHP is a repressor of Srebp-1c [44].
It is not clear whether the observed repression has a broad implication.
The DEC1-SHP pathway may have profound significance in carbohydrate homeostasis. Patients with type 2 diabetes had higher frequency of loss-of-function SHP mutants than those without type 2 diabetes (61.5 versus 28.1%) [45]. In addition, SHP mutation carriers had significantly higher fasting plasma insulin levels than noncarriers [45]. In mice, knockout of SHP developed hepatic insulin resistance [46], and the antidiabetic metformin ameliorated cytokine-induced hepatic insulin resistance by inducing SHP [47]. These observations suggest that SHP positively regulates glucose homeostasis. In contrast, high glucose and high insulin significantly induced DEC1 [48,49]. The induction of DEC1 was inhibited by LY294002, a strong inhibitor of phosphoinositide 3-kinases. Importantly, DEC1 protein and the activity of AMPK (5' AMP-activated protein kinase) showed an inverse circadian rhythm, and knockdown of DEC1 expression increased AMPK activity [50]. AMPK is known to regulate glucose homeostasis and prevent insulin resistance [51].
In summary, SHP belongs to the superfamily of nuclear receptors and has been established to exert a wide range of biological activities, particularly related to metabolic