Which CIDE are you on? Apoptosis and energy metabolism

Abstract

Around 1998, cell death-inducing DNA fragmentation factor-alpha (DFFA)-like effector (CIDE) proteins including CIDEA, CIDEB and CIDEC/fat specific protein 27 (Fsp27) were first identified by their sequence homology with the N-terminal domain of the DNA fragmentation factor (DFF). Indeed, in vitro analysis revealed that all three CIDE proteins are involved in apoptosis. However, recent gene-targeting studies have provided novel insights into the physiological function of CIDE proteins. Mice deficient in each CIDE protein exhibit lean phenotypes, a reduction of lipid droplet size in white adipose tissue and increased metabolic rate. Thus, all CIDE proteins play an important role in energy metabolism and lipid droplet formation. More recently, a glycoproteomics approach has shown that post-translational regulation of CIDE proteins via glycosylation modulates transforming growth factor (TGF)-beta 1-dependent apoptosis. Another recent study using mouse embryonic fibroblasts derived from CIDEA-deficient mice revealed that 5′AMP-activated protein kinase (AMPK) activity is regulated by CIDEA-mediated ubiquitin-dependent proteasomal degradation via a protein interaction with the AMPK beta subunit. Even after a decade of study, the physiological roles of CIDE proteins have still not been completely elucidated. This review aims to shed light on the novel functions of CIDE proteins and their physiological roles.

---

Tomo Yonezawa

Tomo Yonezawa received a BSc (Agr.), BA and PhD from Tohoku University, Japan. He was a Research Fellow (DC1) of the Japan Society for the Promotion of Science as a PhD student. Subsequently, he joined Genodive Pharma Inc. established by Prof. Hidetoshi Inoko and held a Postdoctoral Fellowship at the Institute of Medical Science, the University of Tokyo, Japan. Currently, he is a Postdoctoral Fellow at the Department of Molecular Life Science, Division of Basic Medical Science and Molecular Medicine Tokai University School of Medicine. His research interests are focused on elucidating the pathogenesis of and identifying drug candidates against metabolic disorders and autoimmune diseases using molecular biology, especially a proteomics approach.

Riho Kurata

Riho Kurata received a BSc and MSc from Tokai University, Kanagawa, Japan. She is a PhD student of the Department of Molecular Life Science, Tokai University School of Medicine, Kanagawa, Japan. Her research interests are focused on identifying novel druggable biomarkers such as SNPs and proteins in autoimmune diseases, particularly Behçet's Disease and validating their association with these diseases.

Minoru Kimura

Minoru Kimura received a BSc, MS and DSc from Kyoto University, Japan. He got a position of Assistant Professor at Keio University, and currently he is a Professor at Tokai University School of Medicine. He is a molecular biologist and especially interested in the regulation of gene expression and gene function in whole organisms. He is one of the pioneers of making genetically engineered mice in Japan. In the course of his study, he went to Max-Planck-Institute at Freiburg i. Br. in Germany and the Jackson Laboratory at Bar Harbor, Maine, USA to analyze gene expression in early mouse embryos as a collaborative research.

Hidetoshi Inoko

Hidetoshi Inoko received a BSc, BA and PhD from Kyoto University, Kyoto Japan. He held a postdoctoral position at the Department of Biochemistry, Virus Research Institute, Kyoto University. Currently, he is a Professor at the Department of Molecular Life Science, Division of Basic Medical Science and Molecular Medicine Tokai University School of Medicine. His research interests are centered on elucidating the genes and their function of disease genes using genome analyses.

---

Introduction

CIDE family proteins

Apoptosis is the execution of a cell suicide program in all animals.¹ Cells undergoing apoptosis exhibit various morphological changes including membrane blebbing, cytoplasmic and nuclear condensation, chromatin aggregation, development of apoptotic bodies and DNA fragmentation.¹ DNA fragmentation factor (DFF) plays a critical role in caspase-dependent apoptosis and is composed of two subunits; one of 40 kDa (DFF40; also known as caspase-activated DNase (CAD)) possessing the DNA fragmentation activity and one of 45 kDa (DFF45; also known as inhibitor of CAD (ICAD)), its inhibitor.² In the steady state, DFF45 inhibits the activity of DNA fragmentation by binding to DFF40.² Once death signaling is initiated, DFF45, but not DFF40, is cleaved by active caspase-3 into three proteolytic fragments, liberating DFF40 which then executes DNA fragmentation.² However, the DFF system has not yet been fully elucidated. Recently, new family members have been identified in the DFF system.³

The cell death-inducing DFF45-like effector (CIDE) family of proteins consists of three molecules; CIDEA, CIDEB and CIDEC/fat specific protein 27 (Fsp27).³ Human CIDEA and CIDEB were first identified in 1998 by sequence homology with the N-terminal domain of DFF.³ Before this report, in 1984, Fsp27 was cloned after identification as an upregulated gene by a differential screening approach during adipogenesis of murine TA1 preadipocytes.⁴ Subsequently, the human CIDEC gene (originally named CIDE-3) was cloned as a homolog of murine Fsp27 and shown to translate into two proteins, a long form (CIDE-3, six exons) and a short form (CIDE-3 alpha, five exons) (Fig. 1).⁵ As shown in Fig. 1, all three CIDE proteins possess an evolutionarily conserved CIDE-N domain (N-terminus) sharing homology to DFF and a CIDE-C domain (C-terminus) that is unique among CIDE proteins.⁶ The CIDE-N domain of CIDEA, CIDEB and CIDEC possesses homology with DFF45 of 39, 29 and 38%, respectively (Fig. 1). Furthermore, the CIDE-C domain of CIDEA or CIDEB possesses homology with that of CIDEC of 54 and 53%, respectively (Fig. 1). In fact, Inohara et al.³ have demonstrated that in various human cell lines, ectopic-expression of CIDEA or CIDEB induces apoptosis and DNA fragmentation. However, recent gene targeting studies have revealed that CIDE proteins are also involved in energy metabolism.^7–10 This review aims to shed light on the novel functions of CIDE proteins and their physiological roles, with particular emphasis on their involvement in apoptosis and energy metabolism.

Fig. 1 Schematic structure of the human CIDE protein family. The open and dark gray boxes indicate the CIDE-N and -C domains, respectively.^3,5,6

CIDE proteins and apoptosis

The three CIDE proteins possess different tissue distribution patterns. In the rodent, CIDEA is exclusively expressed in brown adipose tissue (BAT) converting excess energy into heat in adaptive thermogenesis,⁷ whilst in the human CIDEA is highly expressed in white adipose tissue (WAT), storing excess energy as triacylglycerol (TAG).^11,12 Human CIDEB is mainly expressed in liver and small intestine,⁵ while mouse CIDEB is expressed in liver and kidney.⁴ Mouse Fsp27 is highly expressed in BAT and WAT,^9,10 but human CIDEC is robustly expressed only in WAT.^13,14 It was originally believed that the amounts of BAT in humans gradually reduced during postnatal development and that it was completely lost by adulthood.¹⁵ However, several recent studies using a combination of positron emission tomography (PET) and computed tomography (CT) showed that adult humans possess substantial and functional BAT.^16–18 Additionally, human multipotent adipose-derived stem (hMADS) cells, which possess a normal karyotype and high self-renewal ability and are not tumorigenic, are established¹⁹ and possess the ability to differentiate into various lineages including adipocytes, myoblasts and osteoblasts.^19–22 These cells can be differentiated by chronic exposure to peroxisome proliferator-activated receptor (PPAR) gamma, but not PPAR alpha or delta, agonists, into functional BAT accompanied by remarkable elevation of CIDEA and UCP1.²³ It is likely that CIDEA is highly expressed and plays an important role in human BAT. Inohara et al.³ were the first to reveal that murine and human CIDE proteins share homology with the N-terminus of DFF45 and they demonstrated that ectopic expression of CIDEA or CIDEB induces DNA fragmentation and apoptosis in several mammalian cells (Fig. 2 and 3). This induction of apoptosis is inhibited by DFF45 via a protein interaction with each CIDE-N domain while the CIDE-C domain is both necessary and sufficient for CIDEA-induced apoptosis.³ The NMR solution structure of the human CIDE-N domain of CIDEB, DFF40 and DFF45 indicated that each CIDE-N domain binds an opposite topology like a yin/yang shape via a set of conserved EDG loop residues.²⁴ Chen et al.²⁵ have demonstrated by mutational analysis that human CIDEB-induced apoptosis requires mitochondrial localization and homophilic binding via 30 amino acid residues in the CIDE-C domain and also that CIDEB interacts with CIDEA via the CIDE-C domain (Fig. 3). Overexpression of both the two human CIDEC isoforms including CIDE-3 and CIDE-3 alpha also localize to mitochondria and induces DNA fragmentation and apoptosis in mammalian cells, but apoptosis-induced activity of CIDE-3 alpha was less than that of CIDE-3.⁵ Interestingly, human chromosome 3p25 region containing CIDEC was associated with loss of heterozygosity (LOH) and was deleted at a high rate in many cancers.⁵ Mouse Fsp27 possesses a similar ability to induce DNA fragmentation and apoptosis in mammalian cells^26,27 (Fig. 4). From a protein interaction assay using the yeast two hybrid method, mouse Fsp27 was shown to form a homodimer via the CIDE-N and/or CIDE-C domain and a heterodimer with CIDEA via the CIDE-C, but not the CIDE-N, domain.²⁴ Inohara et al.³ initially reported that CIDEA- or CIDEB-induced apoptosis is independent of the caspase pathway due to the failure of caspase inhibitors to block apoptosis. Further study indicated that CIDEB-induced apoptosis is dependent on caspase-3 activity and release of mitochondrial cytochrome c.²⁸ Similarly, mouse Fsp27-induced apoptosis occurred in the same manner as CIDEB.²⁷ It is striking that hepatitis C virus nonstructural protein 2 (NS2) binds the CIDE-C domain of CIDEB, protecting infected host hepatocytes from CIDE-B-induced apoptosis²⁸ (Fig. 3). CIDEB expression correlates with neuronal apoptosis after nerve injury.²⁹ In skeletal muscle after burn injury, apoptosis occurs accompanied by a remarkable enhancement of CIDEA expression.³⁰ Even after a decade of studies, the detailed mechanism of CIDE-induced apoptosis remains to be determined.

Fig. 2 Schematic diagram of the functions of CIDEA as described in this review. Bold arrows indicate activation. Barred lines indicate inhibition. AMPK: 5′-AMP-activated protein kinase; CIDEA: cell death-inducing DFF45-like effector A; DFF: DNA fragmentation factor; Expi: extracellular proteinase inhibitor; LD: lipid droplet; PPARs: peroxisome proliferator-activated receptors; PGC: PPAR gamma coactivator; RIP140: receptor interacting protein 140; UCP1: uncoupling protein 1. CIDEA localizes to the mitochondria, nucleus, ER and LD, inhibits the activity of UCP1 and lipolysis, and enhances apoptosis and LD formation. CIDEA is a short-lived protein regulated by a ubiquitin-dependent proteasomal degradation system. AMPK is regulated by CIDEA-mediated ubiquitin-dependent proteasomal degradation via the CIDE-C domain interacting with the AMPK beta subunit that controls its signaling pathway in the ER, resulting in suppression of beta-oxidation. PPARs synergize with PGC1s to enhance the transcription of CIDEA gene, whilst RIP140 suppresses this induction. Expi induces apoptosis and the up-regulation of CIDEA. TGF-beta 1-induced deglycosylated CIDEA preferentially localizes to the cytoplasm but not the nucleus, inhibiting TGF-beta 1-induced apoptosis.

Fig. 3 Schematic diagram of the functions of CIDEB as described in this review. Bold arrows indicate activation. Barred lines indicate inhibition. ACC: acetyl-CoA carboxylase; ApoB: apolipoprotein B; CIDEB: cell death-inducing DFF45-like effector B; DFF: DNA fragmentation factor; FAS: fatty acid synthase; HNF4α: hepatocyte nuclear factor 4 alpha; LD: lipid droplet; NS2: nonstructural protein 2 of hepatitis C virus; p-Akt: phosphorylated Akt; p-IRS1: phosphorylated insulin receptor substrate 1; PPARs: peroxisome proliferator-activated receptors; PGC: PPAR gamma coactivator; SCD1: stearoyl-CoA desaturase 1; SREBP-1c: sterol response element binding protein-1c; VLDL: very-low-density lipoprotein. CIDEB localizes to LDs and smooth ER, and enhances apoptosis. Hepatitis C virus NS2 binds the CIDE-C domain of CIDEB, protecting infected host hepatocytes from CIDEB-induced apoptosis. CIDEB binds apoB proteins and leads to maturation of VLDL packaging triacylglycerol. CIDEB-deficient mice were shown to enhance the phosphorylation of insulin signaling molecules such as IRS1 and AKT2, and to suppress SREBP-1c and its downstream target such as ACC, FAS and SCD, resulting in insulin sensitivity. HNF4α regulates the transcription of CIDEB gene.

Fig. 4 Schematic diagram of the functions of CIDEC as described in this review. Bold arrows indicate activation. Barred lines indicate inhibition. CIDE: cell death-inducing DFF45-like effector: COX: cyclooxygenase; DFF: DNA fragmentation factor; FoxC2: forkhead box protein C2; Glut4: glucose transporter type4; IRS1: insulin receptor substrate 1; LD: lipid droplet; NRF1: nuclear respiratory factor 1; PPARs: peroxisome proliferator-activated receptors; PGC: PPAR gamma coactivator; Rb: retinoblastoma protein; RIP140: receptor interacting protein 140; T3: triiodothyronine; UCP1: uncoupling protein 1; VLCAD very long chain acyl-CoA dehydrogenase. CIDEC localizes to the mitochondria, nucleus, ER and LD, and enhances apoptosis and LD formation. CIDEC is a short-lived protein regulated by a ubiquitin-dependent proteasomal degradation system. CIDEC-deficient mice were shown to increase the number and activity of mitochondria accompanied by up-regulation of mitochondria genes such as NRF1, VLCAD, CIDEA, UCP1, COX I and COX II, and IRS-1, AKT2, and Glut4 proteins, resulting in enhancement of energy expenditure and insulin sensitivity. PPARs, PGC1s, NRF1 and FoxC2 are positive regulators of the transcription of CIDEC gene, whilst RIP140, Rb and p107 are negative regulators. T3 action could functionally antagonize CIDEC.

Lipid droplet localization and function

Initially, ectopic expression of CIDEA, CIDEB or CIDEC proteins was found to co-localize with mitochondria specific marker (mitotracker), suggesting these localize to the mitochondria^5,7,25 (Fig. 2–4). This localization pattern is likely due to the later stage of apoptosis.²⁵ Recent studies have demonstrated that CIDE proteins localize to lipid droplets (LDs) and endoplasmic reticulum (ER) but not mitochondria^{9,10,27,31–36} (Fig. 2–4). Puri et al.^31,35 showed for the first time that mouse CIDEA and Fsp27 share four homology domains with perilipin, a major LD protein which protects LDs from cytosolic lipases, and that CIDEA localizes to the regions surrounding LDs in differentiated 3T3-L1 adipocytes. These domains include a short N-terminal sequence (I) with shared homology to adipophilin, which is also a major LD protein, a segment (II) with homology to a region of perilipin thought to shield LDs from lipases, and two regions (III and IV) thought to function in the targeting and binding of LDs (Fig. 5). CIDEA also localizes to ER and, to a smaller extent, Golgi apparatus³² (Fig. 2). Similarly, CIDEB localizes to LDs and ER, especially the smooth ER-enriched fraction in hepatocytes and is absent from the cytosolic and mitochondrial fractions³³ (Fig. 3). The C-terminal 166–195 amino acid residues of CIDEB are a prerequisite for ER and LD localization.³³ Initially, proteomic analysis of LDs in 3T3-L1 adipocytes demonstrated that Fsp27 was isolated from the fat cake fraction, not excluding the possibility of contamination by other cellular fractions.³⁴ In agreement with this first screening report,⁴ several recent studies have demonstrated that Fsp27 is increased in a PPAR gamma-dependent manner during adipocyte differentiation.^{26,27,31,35,36} Overexpression of Fsp27 induced the accumulation of LDs and enhancement of LD size not only in adipocytes but also in many non-lipogenic cells,^27,35 whilst knockdown of Fsp27 decreased LD size and increased LD number in 3T3-L1 adipocytes^10,36 (Fig. 4). Although the localization of CIDEC/Fsp27 proteins in the ER still has not been determined, LDs are budded from the smooth ER and composed of proteins and polar lipid.^37–39 It is likely that CIDEC/Fsp27 also exists in the ER because of the important role of LD formation. Overexpression of CIDEA as well as Fsp27 also increased the accumulation and size of LDs in 3T3-L1 adipocytes and COS cells³¹ (Fig. 2). Interestingly, Liu et al.²⁷ have demonstrated that oleate-induced LD formation canceled Fsp27-induced apoptosis in adipocytes and cancer cells (Fig. 4). This mechanism is still not defined but may be due mainly to a change in the localization of Fsp27 in mitochondria to the LD surface because CIDEB-induced apoptosis requires mitochondrial localization.²⁵ Chen et al. also examined the protective effects of Fsp27 as well as perilipin on LD breakdown by cytosolic lipases such as adipose triacylglycerol lipase (ATGL).²⁵ Fsp27 was unable to protect the LDs from breakdown by ATGL.²⁵ However, other cytosolic lipases including hormone sensitive lipase (HSL), ATGL and the calcium-independent phospholipase A₂ (iPLA₂) family consisting of iPLA₂-epsilon, -zeta and -eta have also been identified.^40,41 Detailed investigation of the protective effects of CIDEs against these cytosolic lipases has not yet been delivered. Collectively, all three CIDE proteins induce apoptosis and exist in not only mitochondria but also ER and LDs, and cellular localization, particularly in mitochondria or LD, are important for functions of CIDE proteins involved in apoptosis or LD formation. It is now clear that CIDE proteins possess different important roles in living cells.

Fig. 5 Sequence homology between human CIDE proteins and perilipin. Four conserved regions (I–IV) were identified between CIDEA, Fsp27 and perilipin.³¹ I: adipophilin-like sequence; II: TAG shielding region that plays a critical role for protecting stored TAG from cytosolic lipases; III and IV: targeting and anchoring LD regions.

CIDE proteins and energy metabolism

Astonishingly, recent gene-targeting studies on CIDE proteins have provided unexpected insights into physiological functions of CIDE proteins in mammals.^7–10 In 2003, Professor Peng Li's group for the first time generated CIDEA-deficient mice and analyzed their phenotype.⁷ These mice were not observed to have any apparent differences in apoptosis, but did exhibit a remarkable lean phenotype accompanied by lower WAT mass but without a reduction in adipogenesis, plasma leptin and glucose levels and with an increase in core body temperature compared to wild type, resulting in enhancement of metabolism and insulin sensitivity and resistance to diet-induced obesity.⁷ They also demonstrated that CIDEA is highly expressed in BAT and that CIDEA-deficiency induces a reduced response to cold exposure.⁷ This molecular mechanism is explained, in part, by the fact that CIDEA localizes to the mitochondria and inhibits the activity of uncoupling protein 1 (UCP1), which plays a crucial role in adaptive thermogenesis, by directly binding to BAT⁷ (Fig. 2). This result established the new concept that CIDE proteins play an important role in energy metabolism.

Recently, another molecular mechanism of CIDEA was revealed when increased activity and protein levels of 5′AMP-activated protein kinase (AMPK) subunits consisting of alpha, beta and gamma were identified in CIDEA-deficient brown adipocytes and embryonic fibroblasts.³² These authors demonstrated, using CIDEA-deficient brown adipocytes, that AMPK is regulated by CIDEA-mediated ubiquitin-dependent proteasomal degradation via the CIDE-C domain interacting with the AMPK beta subunit that controls its signaling pathway in the ER³² (Fig. 2). The C-terminal 85 amino acid residues in the AMPK beta subunit (186–270) are required for binding CIDEA (amino acid residues 232–248 is the minimum sequence required for CIDEA binding) as well as the AMPK alpha and gamma subunits.²⁹ AMPK signaling plays a crucial role in energy homeostasis, being involved in beta-oxidation, glucose uptake and lipolysis.^42–44 Acetyl-CoA carboxylase (ACC) plays a pivotal role between de novo lipid synthesis and beta-oxidation, and its activity is negatively regulated by AMPK via phosphorylation of ACC.^42–44 In fact, the phosphorylation of ACC was augmented in CIDEA-deficient brown adipocytes, resulting in reduction of ACC activity.³² The molecular mechanism of CIDEA-mediated ubiquitination remains to be elucidated. Subsequently using generation of CIDEB-deficient mice, Li et al. provided further evidence that CIDE proteins play important roles in regulation of energy metabolism.⁸ CIDEB-deficient mice also possess a lean phenotype accompanied by lower plasma TAG, free fatty acids, insulin and leptin levels, increased plasma adiponectin and whole body metabolism, and drastic reduction of WAT mass.⁸ These mice were resistant to diet-induced obesity and hepatic steatosis.⁸ In liver but not other organs, these mice were shown to enhance the phosphorylation of insulin signaling molecules such as insulin receptor substrate (IRS) 1 and AKT2, resulting in augmentation of insulin sensitivity⁸ (Fig. 3). Additionally, in hepatocytes derived from these mice, sterol response element binding protein (SREBP)-1c, but not SREBP-1a and -2, a critical regulator of fatty acid synthesis, was significantly down-regulated, resulting in the reduction of its downstream targeting genes including ACC, fatty acid synthase (FAS) and stearoyl-CoA desaturase (SCD)⁸ (Fig. 3). How CIDEB regulates SREBP-1c still remains to be clarified.

This analysis also shed light on the underlying mechanism of lower plasma lipids in CIDEB-deficient mice.³³ In CIDEB-deficient liver, very-low-density lipoproteins (VLDLs) contain higher TAG compared to wild type, whilst secreted VLDLs are present in lower amounts and are immature and small.³³ CIDEB localizing to smooth ER and LDs binds apoB proteins and leads to maturation of VLDL packaging TAG via the C-terminal 118–165 amino acid residues of CIDEB³³ (Fig. 3). Adenovirus-mediated CIDEB expression showed that the CIDE-C domain is sufficient to rescue this phenotype, correcting VLDL impairment and lowering plasma TAG levels.³³

Independently, two groups reported the phenotypes of Fsp27-deficient mice.^9,10 Both groups showed that Fsp27-deficient mice have drastically lean phenotypes accompanied by smaller LDs in WAT, larger LDs in BAT, and lower plasma glucose and leptin levels, resulting in enhancement of insulin sensitivity and resistance to diet-induced obesity.^9,10 These two groups further demonstrated that the number and activity of mitochondria were elevated, accompanied by up-regulation of mitochondrial genes such as nuclear respiratory factor (NRF) 1, PPAR gamma coactivator (PGC) 1-alpha, very long chain fatty acyl-CoA dehydrogenase (VLCAD) and cyclooxygenase (COX) I and COX II in WAT but not BAT^9,10 (Fig. 4). Additionally, Peng Li's group further showed that in Fsp27-deficient WAT, negative regulators of BAT differentiation such as retinoblastoma (Rb), p107 and receptor interacting protein (RIP) 140 are reduced, while in contrast, positive regulators such as forkhead box (Fox) C2, PPARs and PGC-1alpha, and BAT functional genes such as uncoupling protein (UCP) 1, CIDEA and deionidase (Dio) 2 are increased¹⁰ (Fig. 4). They also demonstrated that Fsp27 deficiency ameliorates insulin resistance, hyperinsulinemia, hyperlipidemia and obesity in leptin-deficient (ob/ob) mice via increases in IRS-1, AKT2 and glucose transporter type 4 (Glut4), which play critical roles in insulin-dependent glucose uptake in WAT, although a single-deficiency of Fsp27 is sufficient for the induction¹⁰ (Fig. 4). Studies from the two groups showed that Fsp27-deficiency shifted the character of WAT into that of BAT although Fsp27-deficient mice are not resistant to cold exposure.^9,10 Peng Li's group further emphasized that Fsp27-deficient mouse embryonic fibroblasts (MEFs) preferentially differentiated into brown adipocytes in the presence of thyroid hormone triiodothyronine (T3) while wild-type MEFs only moderately differentiated,⁹ suggesting that the functions of Fsp27 are regulated by exogenous factors such as T3 (Fig. 4). A summary of the phenotypes of each CIDE-deficient mouse is shown in Table 1. These results strongly indicate that all three CIDE proteins play important roles in energy metabolism and lipid synthesis.

Table 1

Gene name	CIDEA	CIDEB	Fsp27
Blood parameter
Plasma triacylglycerol	No change	Decreased (high-fat diet)	Decreased^9 or no change^10
Plasma non-esterified fatty acids	Not determined	Decreased (normal diet)	Decreased^9 or no change^10
Plasma ketone body	Not determined	No change	Decreased^9
Plasma insulin	No change	Decreased	No change
Plasma glucose	Decreased (high-fat diet)	Decreased	No change^9 or decreased^10
Plasma adiponectin	Not determined	Increased (high-fat diet)	No change^10
Plasma leptin	Decreased	Decreased	Decreased
Metabolic condition
Glucose disposal	Increased	Increased	Increased
Insulin sensitivity	No change	Increased	Increased
Adipose index	Decreased	Decreased	Decreased
Food intake	No change	Increased	Increased^9 or no change^10
Body temperature	Increased	Not determined	Increased^9 or no change^10
Whole body metabolic rate	Increased	Increased	Increased

---

Association with disease

In a genetic association study of CIDE proteins, a single nucleotide polymorphism (SNP) of a G to T transversion in CIDEA exon 4, which is equivalent to a Val 115 Phe substitution, associates with body mass index in Swedish male and female obese patients.⁴⁵ In Japanese patients, the same SNP associates with fasting-blood glucose levels, abdominal obesity and metabolic syndrome.⁴⁶ This nonsynonymous substitution exists in the central region in CIDEA proteins and does not involve either the CIDE-N or the CIDE-C domain. Further studies may reveal whether this substitution produces any change in CIDEA function.

In 2009, a female patient with unexplained acanthosis nigricans and partial lipodystrophy (affecting the limb, femorogluteal and subcutaneous abdominal fat) was found.⁴⁷ She was found to be homozygous for a SNP involving a G to T transversion in CIDEC exon 6, which causes a Glu 186 X nonsense mutation in the CIDE-C domain, identified as an autosomal recessive form of familial partial lipodystrophy.⁴⁷ Coincident with the phenotype of the Fsp27-deficient mouse, this mutation reduces fat mass and induces the formation of white adipocytes with multilocular lipid droplets and an increased number of mitochondria in WAT.³⁸ From an in vitro study, the Glu 186 X mutation fails to localize LDs in COS and differentiated 3T3-L1 cells.⁴⁷ This patient exhibited insulin resistance, a tendency to ketosis and diabetes although CIDEC/Fsp27-deficient mice are protected against diet-induced obesity and insulin resistance.^9,10,47

Three PPARs belonging to a family within the lipid-activated receptor superfamily have been identified in mammals; PPAR alpha, beta/delta and gamma.^48–50 These genes are thought to play a critical role in energy homeostasis and lipid metabolism.^48–50 PPAR gamma is highly expressed in WAT^48–50 but is expressed at a lower level in liver at only 10–30% of the levels of WAT.⁵¹ PPAR gamma expression levels are remarkably elevated in the severe fatty liver of murine models of diabetes.^52,53 There is no direct evidence for elevation of PPAR gamma expression in human liver steatosis. Nonetheless, accumulating evidence suggests that PPAR gamma is a primary mediator of hepatic lipid accumulation in nonalcoholic hepatic steatosis.^54–56 Matsusue et al.⁵⁷ have generated liver specific PPAR gamma-deficient mice on an ob/ob genetic background using the albumin promoter and demonstrated that these mice show a significant decrease in hepatic TAG and ameliorate steatosis. They further performed subtraction cloning using the liver of these mice and identified three direct downstream target genes of liver PPAR gamma including SCD1, CIDEA and Fsp27⁵⁸ (Fig. 2 and 4). Adenovirus-mediated overexpression of Fsp27, but not CIDEA, increased hepatic TAG in vitro and vivo, whilst adenovirus-mediated knock-down of Fsp27 decreased hepatic TAG.⁵⁸ In another human study, in body mass index-matched obese humans, CIDEA and CIDEC expression in WAT were positively correlated with insulin sensitivity.³¹ Restricted caloric intake in obese patients induced down-regulation of CIDEC.¹³ CIDEA expression in WAT is inversely correlated with basal metabolic rate.¹⁴ These results suggest that the functions of CIDE proteins involved in energy metabolism may be conserved in the human as well as the mouse.

Transcriptional regulation

CIDE proteins are finely modulated by transcriptional regulation. By analyzing the human gene around the 1.5-kb CIDEA promoter region, Li et al.⁵⁹ identified a 125-bp minimal transcription module between −137 to −13 bp from its translation start site that contains two specific protein (sp) 1/sp3 binding sites. They demonstrated that, in fact, sp1 and sp3 regulate basal transcription of CIDEA via binding to specific sites. They have found that this minimal transcription module locates in a CpG island at position −285 to +39 and demonstrated that the methylation condition of this region is responsible for the specificity of CIDEA transcription in a tissue- and cell-dependent manner.⁵⁹ Strikingly, demethylation by 5-aza-dC, which is a DNA methyl-transferase inhibitor, induces ectopic-expression of CIDEA in CIDEA non-expressing cells including 293T and Hela cells.⁵⁹

In analysis of the human CIDEB promoter region, the same group have identified that CIDEB exists as two transcript variants, a long and a short form, which consist of 7 (additional 5′ untranslated exons 1 and 2) and 5 exons (3A, 4, 5, 6 and 7) respectively and are mediated by dual transcription modules such as an upstream promoter (Pu) and an internal promoter (Pi) in a tissue- and cell-specific manner.⁶⁰ Both transcripts possess the same open reading frame.⁶⁰ They identified a 284-bp minimal Pu between −3628 to −3478 bp from its translation start site that contains one sp1 binding site shared with the leukotriene B4 receptor promoter in the reverse strand and demonstrated that sp1 regulates basal transcription of the CIDEB long form variant by binding specific sites. They also showed that Pu is located in the CpG island as well as CIDEA⁵⁹ and is highly methylated in cells not expressing the CIDEB long form variant.⁶⁰ They also identified a 204-bp minimal Pi between −204 to −1 bp that contains one sp1/sp3 and two hepatocyte nuclear factor (HNF) 4 alpha binding sites and demonstrated that sp1 and sp3 regulate basal transcription of the CIDEB short form variant via binding to specific sites⁶⁰ (Fig. 3). Furthermore, HNF 4 alpha synergizes with sp1 to enhance transcription of the CIDEB short form variant in these cells which do not express the CIDEB short form.⁶⁰ These dual promoters regulate tissue- and cell-specific CIDEB transcription.

By analysis of the mouse gene around the 3.0-kb CIDEA promoter region, Viswakarma et al.⁶¹ identified 1.3 kb-minimal transcription modules and three PPAR response elements (PPREs) in its upstream region and demonstrated that agonist activated-PPAR alpha and gamma regulate a marked induction of CIDEA via binding of the most proximal PPRE but not another two PPRE in the liver (Fig. 2). Another group has identified estrogen-related receptor (ERR) alpha and NRF1 binding sites in the minimal transcription modules and demonstrated that PGC1-alpha induces CIDEA transcription via binding of ERR alpha and NRF1, although PGC1-beta also binds it in immortalized brown adipocytes⁶² (Fig. 2). They have also demonstrated that RIP140, which is a co-repressor of the nuclear receptor, represses PGC1-alpha-induced CIDEA induction via binding of PGC1-alpha and that this molecular interaction occurs in a differentiation-dependent manner in brown adipocytes⁶² (Fig. 2).

Kim et al.⁶³ analyzed the mouse 2.0-kb Fsp27 promoter region and identified four PPREs and demonstrated that three of these, except for the most distal one, binds PPAR gamma 2 in differentiated 3T3-L1 adipocytes. This induction of Fsp27 occurs in a differentiation-dependent manner in adipocytes.⁶³ Additionally, CCAAT/enhancer binding proteins (C/EBPs),⁶⁴ which play a crucial role in adipocyte differentiation, and retinoid X receptor (RXR),⁶⁵ which synergizes PPAR-induced transcription via its binding, binds the Fsp27 promoter regions and controls its transcription during differentiation of adipocytes (Fig. 4).

To summarize the above, transcription of CIDE proteins is regulated by specific transcription factors and/or methylation of promoter regions in a tissue- and cell-specific manner.

Post-translational regulation

Recent studies have shown that CIDE proteins are finely modulated by post-translational regulation. Chan et al.⁶⁶ have demonstrated that human CIDEA is a short-lived protein regulated by a ubiquitin-dependent proteasomal degradation system in several cells including primary cultured brown adipocytes (Fig. 2). Human CIDEA possesses ten lysine residues. Chan and colleagues generated two mutants of human CIDEA to create an N-terminal lysine-less CIDEA (N-5KA) by substituting five lysine residues in the N-terminal to alanines, as well as a C-terminal lysine-less CIDEA (C-5KA) and they demonstrated that the N-5KA mutant is more stable than the wild type or C-5KA.⁶⁶ They also demonstrated that an N-terminal lysine (K23) mutation is mainly involved in polyubiquitination signaling and protein instability.⁶⁶ In the same manner, Fsp27 is regulated by a ubiquitin-dependent proteasomal degradation system involving the C-terminal three lysine residues⁶⁷ (Fig. 4). A mutant in which these three lysine residues were substituted with alanines remarkably increased protein stability and enhanced TAG storage in differentiated 3T3-L1 adipocytes.⁶⁷ It was also demonstrated that a beta-adrenergic agonist rapidly enhances the stability of Fsp27 proteins and the localization of Fsp27 around LDs, suggesting that rapid regulation of Fsp27 proteins via the ubiquitin-dependent proteasomal degradation system may play a negative feedback role in beta-agonist-induced lipolysis.⁶⁷ It is likely that CIDEB is also regulated by the ubiquitin-dependent proteasomal degradation system.

Recently, a glycoproteomics approach has revealed that post-transcriptional regulation of CIDE proteins via glycosylation modulates transforming growth factor (TGF)-beta 1-dependent apoptosis in human breast cancer cells.⁶⁸ Iwahana et al.,⁶⁸ using Helix pomatia agglutinin (HPA) lectin have purified O-glycosylation proteins (serine or threonine residues) in MCF-7 cells treated with TGF-beta 1 and identified these purified proteins including CIDEA by 2-dimensional electrophoresis (2-DE) combined with matrix-assisted laser desorption/ionization-time of flight mass spectrometry (MALDI-TOF-MS).⁶⁸ TGF-beta 1 rapidly induced the disappearance of CIDEA in 2-DE of HPA-purified proteins although the total CIDEA protein level was not changed.⁶⁸ They have demonstrated that TGF-beta 1-induced deglycosylated CIDEA preferentially localizes to the cytoplasm but not the nucleus, inhibiting TGF-beta 1-induced apoptosis (Fig. 2). They also showed that inducible overexpression of CIDEA induces deglycosylation of CIDEA, rescuing the cells from apoptosis.⁶⁸ Qi et al.³² have also identified the localization of CIDEA in the nucleus. Nuclear CIDEA may play some important role in apoptosis and transcriptional conditions. Mammary gland involution is well known as a good model for physiological apoptosis.⁶⁹ The extracellular proteinase inhibitor (Expi) gene was first identified as an involution-induced gene using differential screening of a mouse mammary gland cDNA library.⁷⁰ Indeed, Expi induces apoptosis and the up-regulation of CIDEA in mouse mammary epithelial cells⁷¹ (Fig. 2). Moreover, CIDEA expression is significantly higher during lactation and involution of the mammary gland.⁷² These results suggest that CIDEA may play an important role in involution-dependent apoptosis via post-transcriptional regulation such as glycosylation in the mammary gland although mammary epithelial cells are known to accumulate LDs in the cytosol and to be highly active in lipid metabolism.

Taken together, these results show that CIDE proteins and their localization are regulated by the ubiquitin-dependent proteasomal degradation pathway or by glycosylation.

Conclusions

As we have outlined in this review, research studies in the past decade have established that CIDE proteins are the executors of apoptosis and/or regulators of energy metabolism. Gene targeting studies strongly indicate that the physiological actions of all three CIDE proteins play important roles in energy metabolism, insulin sensitivity and lipid storage. However, in addition, evidence that CIDE protein functions are involved in apoptosis has been accumulating, accompanied by evidence of novel post-translational modifications. It is likely that cellular localization of CIDE proteins is important for their activity in apoptosis or metabolism. However, the molecular mechanism of CIDEs in the mouse and human are somewhat distinct such as that Fsp27-, but not CIDEC-, deficiency ameliorates insulin sensitivity and diabetes in the mouse and that tissue-specificity and promoter regions of each are different. All CIDE proteins are highly active and exhibit different cellular localization in varying conditions and cells, resulting in modulation of cellular transcriptional and signaling conditions. However, the molecular mechanisms involved are still not fully understood. The functions of CIDE proteins may provide a link between energy metabolism and apoptosis. Schematic summaries of the molecular mechanism of action of CIDEA, CIDEB and CIDEC are shown in Fig. 2–4, respectively. Further studies on CIDE proteins should drive advances in therapeutic treatments of metabolic disorders such as obesity and diabetes and lead to a better understanding of the molecular mechanisms in apoptosis.

Acknowledgements

This work was supported by Tokai University School of Medicine Research Aid. The authors have no financial conflict of interest.

References

1. A. H. Wyllie, Curr. Opin. Genet. Dev., 1995, 1, 97–104.
2. X. Liu, H. Zou, C. Slaughter and X. Wang, Cell, 1997, 89, 175–184.
3. N. Inohara, T. Koseki, S. Chen, X. Wu and G. Núñez, EMBO J., 1998, 17, 2526–2533.
4. A. B. Chapman, D. M. Knight, B. S. Dieckmann and G. M. Ringold, J. Biol. Chem., 1984, 259, 15548–15555.
5. L. Liang, M. Zhao, Z. Xu, K. K. Yokoyama and T. Li, Biochem. J., 2003, 370, 195–203.
6. C. Wu, Y. Zhang, Z. Sun and P. Li, BMC Evol. Biol., 2008, 8, 159.
7. Z. Zhou, S. Yon Toh, Z. Chen, K. Guo, C. P. Ng, S. Ponniah, S. C. Lin, W. Hong and P. Li, Nat. Genet., 2003, 35, 49–56.
8. J. Z. Li, J. Ye, B. Xue, J. Qi, J. Zhang, Z. Zhou, Q. Li, Z. Wen and P. Li, Diabetes, 2007, 56, 2523–2532.
9. S. Y. Toh, J. Gong, G. Du, J. Z. Li, S. Yang, J. Ye, H. Yao, Y. Zhang, B. Xue, Q. Li, H. Yang, Z. Wen and P. Li, PLoS One, 2008, 3, e2890.
10. N. Nishino, Y. Tamori, S. Tateya, T. Kawaguchi, T. Shibakusa, W. Mizunoya, K. Inoue, R. Kitazawa, S. Kitazawa, Y. Matsuki, R. Hiramatsu, S. Masubuchi, A. Omachi, K. Kimura, M. Saito, T. Amo, S. Ohta, T. Yamaguchi, T. Osumi, J. Cheng, T. Fujimoto, H. Nakao, K. Nakao, A. Aiba, H. Okamura, T. Fushiki and M. Kasuga, J. Clin. Invest., 2008, 118, 2808–2821.
11. E. A. Nordström, M. Rydén, E. C. Backlund, I. Dahlman, M. Kaaman, L. Blomqvist, B. Cannon, J. Nedergaard and P. Arner, Diabetes, 2005, 54, 1726–1734.
12. A. I. Su, M. P. Cooke, K. A. Ching, Y. Hakak, J. R. Walker, T. Wiltshire, A. P. Orth, R. G. Vega, L. M. Sapinoso, A. Moqrich, A. Patapoutian, G. M. Hampton, P. G. Schultz and J. B. Hogenesch, Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 4465–4470.
13. B. Magnusson, A. Gummesson, C. A. Glad, J. H. Goedecke, M. Jernås, T. C. Lystig, B. Carlsson, B. Fagerberg, L. M. Carlsson and P. A. Svensson, Metab., Clin. Exp., 2008, 57, 1307–1313.
14. A. Gummesson, M. Jernås, P. A. Svensson, I. Larsson, C. A. Glad, E. Schéle, L. Gripeteg, K. Sjöholm, T. C. Lystig, L. Sjöström, B. Carlsson, B. Fagerberg and L. M. Carlsson, J. Clin. Endocrinol. Metab., 2007, 92, 4759–4765.
15. B. Cannon and J. Nedergaard, Physiol. Rev., 2004, 84, 277–359.
16. K. A. Virtanen, M. E. Lidell, J. Orava, M. Heglind, R. Westergren, T. Niemi, M. Taittonen, J. Laine, N. J. Savisto, S. Enerbäck and P. Nuutila, N. Engl. J. Med., 2009, 360, 1518–1525.
17. A. M. Cypess, S. Lehman, G. Williams, I. Tal, D. Rodman, A. B. Goldfine, F. C. Kuo, E. L. Palmer, Y. H. Tseng, A. Doria, G. M. Kolodny and C. R. Kahn, N. Engl. J. Med., 2009, 360, 1509–1517.
18. W. D. van Marken Lichtenbelt, J. W. Vanhommerig, N. M. Smulders, J. M. Drossaerts, G. J. Kemerink, N. D. Bouvy, P. Schrauwen and G. J. Teule, N. Engl. J. Med., 2009, 360, 1500–1508.
19. A. M. Rodriguez, C. Elabd, F. Delteil, J. Astier, C. Vernochet, P. Saint-Marc, J. Guesnet, A. Guezennec, E. Z. Amri, C. Dani and G. Ailhaud, Biochem. Biophys. Res. Commun., 2004, 315, 255–263.
20. A. M. Rodriguez, D. Pisani, C. A. Dechesne, C. Turc-Carel, J. Y. Kurzenne, B. Wdziekonski, A. Villageois, C. Bagnis, J. P. Breittmayer, H. Groux, G. Ailhaud and C. Dani, J. Exp. Med., 2005, 201, 1397–1405.
21. L. E. Zaragosi, G. Ailhaud and C. Dani, Stem Cells, 2006, 24, 2412–2419.
22. C. Elabd, C. Chiellini, A. Massoudi, O. Cochet, L. E. Zaragosi, C. Trojani, J. F. Michiels, P. Weiss, G. Carle, N. Rochet, C. A. Dechesne, G. Ailhaud, C. Dani and E. Z. Amri, Biochem. Biophys. Res. Commun., 2007, 361, 342–348.
23. C. Elabd, C. Chiellini, M. Carmona, J. Galitzky, O. Cochet, R. Petersen, L. Pénicaud, K. Kristiansen, A. Bouloumié, L. Casteilla, C. Dani, G. Ailhaud and E. Z. Amri, Stem Cells, 2009, 27, 2753–2760.
24. A. A. Lugovskoy, P. Zhou, J. J. Chou, J. S. McCarty, P. Li and G. Wagner, Cell, 1999, 99, 747–755.
25. Z. Chen, K. Guo, S. Y. Toh, Z. Zhou and P. Li, J. Biol. Chem., 2000, 275, 22619–22622.
26. J. Y. Kim, K. Liu, S. Zhou, K. Tillison, Y. Wu and C. M. Smas, Am. J. Physiol.: Endocrinol. Metab., 2008, 294, E654–667.
27. K. Liu, S. Zhou, J. Y. Kim, K. Tillison, D. Majors, D. Rearick, J. H. Lee, R. F. Fernandez-Boyanapalli, K. Barricklow, M. S. Houston and C. M. Smas, Am. J. Physiol.: Endocrinol. Metab., 2009, 297, E1395–1413.
28. L. Erdtmann, N. Franck, H. Lerat, J. Le Seyec, D. Gilot, I. Cannie, P. Gripon, U. Hibner and C. Guguen-Guillouzo, J. Biol. Chem., 2003, 278, 18256–18264.
29. J. M. Wang, B. C. Li, L. Chen, G. B. Wang, H. Sun, Z. Chen and L. C. Zhang, Mil. Med., 2006, 171, 793–799.
30. H. Duan, J. Chai, Z. Sheng, Y. Yao, H. Yin, L. Liang, C. Shen and J. Lin, Apoptosis, 2009, 14, 52–65.
31. V. Puri, S. Ranjit, S. Konda, S. M. Nicoloro, J. Straubhaar, A. Chawla, M. Chouinard, C. Lin, A. Burkart, S. Corvera, R. A. Perugini and M. P. Czech, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 7833–7838.
32. J. Qi, J. Gong, T. Zhao, J. Zhao, P. Lam, J. Ye, J. Z. Li, J. Wu, H. M. Zhou and P. Li, EMBO J., 2008, 27, 1537–1548.
33. J. Ye, J. Z. Li, Y. Liu, X. Li, T. Yang, X. Ma, Q. Li, Z. Yao and P. Li, Cell Metab., 2009, 9, 177–190.
34. D. L. Brasaemle, G. Dolios, L. Shapiro and R. Wang, J. Biol. Chem., 2004, 279, 46835–46842.
35. V. Puri, S. Konda, S. Ranjit, M. Aouadi, A. Chawla, M. Chouinard, A. Chakladar and M. P. Czech, J. Biol. Chem., 2007, 282, 34213–34218.
36. P. Keller, J. T. Petrie, P. De Rose, I. Gerin, W. S. Wright, S. H. Chiang, A. R. Nielsen, C. P. Fischer, B. K. Pedersen and O. A. MacDougald, J. Biol. Chem., 2008, 283, 14355–14365.
37. D. P. Dylewski, C. H. Dapper, H. M. Valivullah, J. T. Deeney and T. W. Keenan, Eur. J. Cell Biol., 1984, 35, 99–111.
38. J. T. Deeney, H. M. Valivullah, C. H. Dapper, D. P. Dylewski and T. W. Keenan, Eur. J. Cell Biol., 1985, 38, 16–26.
39. I. H. Mather and T. W. Keenan, J. Mammary Gland Biol. Neoplasia, 1998, 3, 259–273.
40. C. M. Jenkins, D. J. Mancuso, W. Yan, H. F. Sims, B. Gibson and R. W. Gross, J. Biol. Chem., 2004, 279, 48968–48975.
41. R. Zimmermann, J. G. Strauss, G. Haemmerle, G. Schoiswohl, R. Birner-Gruenberger, M. Riederer, A. Lass, G. Neuberger, F. Eisenhaber, A. Hermetter and R. Zechner, Science, 2004, 306, 1383–6.
42. B. B. Kahn, T. Alquier, D. Carling and D. G. Hardie, Cell Metab., 2005, 1, 15–25.
43. Y. Minokoshi, Y. B. Kim, O. D. Peroni, L. G. Fryer, C. Müller, D. Carling and B. B. Kahn, Nature, 2002, 415, 339–343.
44. E. Louden, M. M. Chi and K. H. Moley, Reproduction, 2008, 136, 335–344.
45. I. Dahlman, M. Kaaman, H. Jiao, J. Kere, M. Laakso and P. Arner, Diabetes, 2005, 54, 3032–3034.
46. L. Zhang, K. Miyaki, T. Nakayama and M. Muramatsu, Metab., Clin. Exp., 2008, 57, 502–505.
47. O. Rubio-Cabezas, V. Puri, I. Murano, V. Saudek, R. K. Semple, S. Dash, C. S. Hyden, W. Bottomley, C. Vigouroux, J. Magré, P. Raymond-Barker, P. R. Murgatroyd, A. Chawla, J. N. Skepper, V. K. Chatterjee, S. Suliman, A. M. Patch, A. K. Agarwal, A. Garg, I. Barroso, S. Cinti, M. P. Czech, J. Argente, S. O'Rahilly and D. B. Savage, LD Screening Consortium, EMBO Mol. Med., 2009, 1, 280–287.
48. B. M. Spiegelman and J. S. Flier, Cell, 1996, 87, 377–89.
49. K. Schoonjans, B. Staels and J. Auwerx, Biochim. Biophys. Acta., 1996, 1302, 93–109.
50. S. Green, Mutat. Res., Fundam. Mol. Mech. Mutagen., 1995, 333, 101–109.
51. P. Tontonoz, E. Hu, R. A. Graves, A. I. Budavari and B. M. Spiegelman, Genes Dev., 1994, 8, 1224–1234.
52. M. Bedoucha, E. Atzpodien and U. A. Boelsterli, J. Hepatol., 2001, 35, 17–23.
53. R. A. Memon, L. H. Tecott, K. Nonogaki, A. Beigneux, A. H. Moser, C. Grunfeld and K. R. Feingold, Endocrinology, 2000, 141, 4021–4031.
54. B. A. Neuschwander-Tetri, E. M. Brunt, K. R. Wehmeier, C. A. Sponseller, K. Hampton and B. R. Bacon, J. Hepatol., 2003, 38, 434–440.
55. A. J. Sanyal, P. S. Mofrad, M. J. Contos, C. Sargeant, V. A. Luketic, R. K. Sterling, R. T. Stravitz, M. L. Shiffman, J. Clore and A. S. Mills, Clin. Gastroenterol. Hepatol., 2004, 2, 1107–1115.
56. K. Promrat, G. Lutchman, G. I. Uwaifo, R. J. Freedman, A. Soza, T. Heller, E. Doo, M. Ghany, A. Premkumar, Y. Park, T. J. Liang, J. A. Yanovski, D. E. Kleiner and J. H. Hoofnagle, Hepatology, 2004, 39, 188–196.
57. K. Matsusue, M. Haluzik, G. Lambert, S. H. Yim, O. Gavrilova, J. M. Ward, B. Brewer Jr, M. L. Reitman and F. J. Gonzalez, J. Clin. Invest., 2003, 111, 737–747.
58. K. Matsusue, T. Kusakabe, T. Noguchi, S. Takiguchi, T. Suzuki, S. Yamano and F. J. Gonzalez, Cell Metab., 2008, 7, 302–311.
59. D. Li, L. Da, H. Tang, T. Li and M. Zhao, Nucleic Acids Res., 2008, 36, 330–341.
60. L. Da, D. Li, K. K. Yokoyama, T. Li and M. Zhao, Biochem. J., 2006, 393, 779–788.
61. N. Viswakarma, S. Yu, S. Naik, P. Kashireddy, K. Matsumoto, J. Sarkar, S. Surapureddi, Y. Jia, M. S. Rao and J. K. Reddy, J. Biol. Chem., 2007, 282, 18613–18624.
62. M. Hallberg, D. L. Morganstein, E. Kiskinis, K. Shah, A. Kralli, S. M. Dilworth, R. White, M. G. Parker and M. Christian, Mol. Cell. Biol., 2008, 28, 6785–6795.
63. Y. J. Kim, S. Y. Cho, C. H. Yun, Y. S. Moon, T. R. Lee and S. H. Kim, Biochem. Biophys. Res. Commun., 2008, 377, 297–302.
64. U. Danesch, W. Hoeck and G. M. Ringold, J. Biol. Chem., 1992, 267, 7185–7193.
65. R. Nielsen, T. A. Pedersen, D. Hagenbeek, P. Moulos, R. Siersbaek, E. Megens, S. Denissov, M. Børgesen, K. J. Francoijs, S. Mandrup and H. G. Stunnenberg, Genes Dev., 2008, 22, 2953–2967.
66. S. C. Chan, S. C. Lin and P. Li, Biochem. J., 2007, 408, 259–266.
67. Z. Nian, Z. Sun, L. Yu, S. Y. Toh, J. Sang and P. Li, J. Biol. Chem., 2010, 285, 9604–9615.
68. H. Iwahana, I. Yakymovych, A. Dubrovska, U. Hellman and S. Souchelnytskyi, Proteomics, 2006, 6, 6168–180.
69. C. A. Palmer, M. C. Neville, S. M. Anderson and J. L. McManaman, J. Mammary Gland Biol. Neoplasia, 2006, 11, 269–282.
70. B. Choi, K. Myung, M. Lee, H. Kim, D. Jeon, I. Hwang, Y. Choi, S. Paik and M. Baik, Mol. Cells, 1996, 6, 311–315.
71. D. J. Jung, J. J. Bong and M. Baik, Exp. Cell Res., 2004, 292, 115–122.
72. T. Yonezawa, S. Haga, Y. Kobayashi, K. Katoh and Y. Obara, Biochem. Biophys. Res. Commun., 2009, 384, 535–9.

---