Transcriptional coregulator RIP140: an essential regulator of physiology
- Institute of Reproductive and Developmental Biology, Faculty of Medicine, Imperial College London, London, UK
- Correspondence should be addressed to J Nautiyal; Email: j.nautiyal{at}imperial.ac.uk
Abstract
Transcriptional coregulators drive gene regulatory decisions in the transcriptional space. Although transcription factors including all nuclear receptors provide a docking platform for coregulators to bind, these proteins bring enzymatic capabilities to the gene regulatory sites. RIP140 is a transcriptional coregulator essential for several physiological processes, and aberrations in its function may lead to diseased states. Unlike several other coregulators that are known either for their coactivating or corepressing roles, in gene regulation, RIP140 is capable of acting both as a coactivator and a corepressor. The role of RIP140 in female reproductive axis and recent findings of its role in carcinogenesis and adipose biology have been summarised.
Introduction
We have come a long way in the understanding of transcriptional coregulators from when they were first identified as transcription factor-binding proteins (Glass et al. 1990, Bastian & Nordeen 1991, Chatterjee et al. 1991, Wieland et al. 1991, Cavailles et al. 1994, Halachmi et al. 1994, Onate et al. 1995) to now when they are recognised as key players in influencing all physiological processes (Lonard & O’Malley 2007, O’Malley et al. 2008, York & O’Malley 2010, Nautiyal et al. 2013a, Dasgupta et al. 2014, Stashi et al. 2014). These proteins are essential for gene regulation, by regulating functions of transcription factors including all members of the nuclear receptor superfamily in the genome. Coregulators (a class of molecules that include both transcriptional coactivators and corepressors) function by bringing together multi-partner transcriptional machinery by acting as scaffolds and in several cases also possess enzymatic activities to activate (as coactivators) or inhibit (as corepressors) gene expression. In fact, recent findings suggest that these proteins perform functions beyond nucleus in the cytoplasm to regulate several biological functions. Coregulators are secondary transcription factors that are brought to the site of transcription by transcription factors but do not bind to the DNA directly. However, these are the factors that control gene expression by bringing enzymatic potentials to the gene regulatory sites. Numerous studies using transgenic technologies in mice and cell-based assays have revealed the powerful potential these proteins possess in regulating normal physiology and driving diseased states. Today 350+ proteins are known as coregulators and the number keeps increasing (Millard et al. 2013, Stashi et al. 2014). This article elaborates on the biological functions of RIP140, a coregulator that performs several important physiological functions. RIP140 functions in female reproductive tissues, adipose, and current knowledge on its role in carcinogenesis has been discussed.
Receptor-interacting protein140 (RIP140): a switch coregulator
Coregulators are either coactivators (examples, members of steroid receptor coactivator (SRC) and peroxisome proliferator-activated receptors gamma (PPARγ) coactivator (PGC-1) families) or corepressors (examples, nuclear receptor corepressors-1 (NCoR1) and silencing mediator for retinoid and thyroid hormone receptors (SMRT)). RIP140, however, is a uniquely reported coregulator that functions both as a coactivator and a corepressor (Fig. 1). It was first identified in human breast cancer cells in a screen to identify interacting partners of the hormone-binding domain of oestrogen receptor α (ER) (Cavailles et al. 1994, 1995). At this stage, only the basal transcriptional machinery, composed of the TATA box-binding protein (TBP) and TBP-associated factor (TAF), hTAFII30, was known to bind ER (Brou et al. 1993, Jacq et al. 1994), and this binding was not affected by treatment of the cells with oestrogen or anti-oestrogens. Thus, there was a gap in the understanding of binding partners of ER whose binding was selectively regulated or inhibited by the ligand oestrogen (E2) and antagonists (Tamoxifen) to effect ER-mediated transcription (Cavailles et al. 1995). As we know it today, RIP140 is a coregulator to most nuclear receptors and several other transcription factors (Nautiyal et al. 2013a). This elaborate interactome is due to 9 LXXLL motifs and a 10th LXXML motif mapped to the C-terminus on RIP140 protein. The LXXML motif is required for interaction with retinoic acid receptor (RAR) and retinoic X receptor (RXR) (Christian et al. 2006) (Fig. 1). Coregulator interactions with nuclear receptors and transcription factors are brought about by LXXLL motifs. This signature sequence (also called NR box), first identified by Heery and coworkers, is a short alpha helical sequence motif (where L is leucine and X is any amino acid) (Heery et al. 1997, Torchia et al. 1997). LXXLL motifs are common to several transcriptional coregulators. However, there is variation in the number of LXXLL sequences on different coregulators, which determines their respective interactomes. For example, SRC family of proteins have 3 LXXLL motifs, whereas RIP140 has 9 LXXLL motifs (Leo & Chen 2000, Christian et al. 2006, Dasgupta & O’Malley 2014).
RIP140 a switch coregulator. Cartoon of RIP140 protein showing different known domains (RD1–4) and interaction motifs (LXXLL and LXXML) (Christian et al. 2006) with transcription factors and enzymes, which modify histones, ie, histone deacetylase (HDAC) and CREB-binding protein (CBP). RIP140 can function both as a coactivator (ON switch) as well as a corepressor (OFF switch). Key processes regulated by RIP140 are highlighted in green (activation) and red (repression) boxes, respectively.
Initial studies indicated a coactivator role for RIP140 (Ikonen et al. 1997, Joyeux et al. 1997). However soon, duality in RIP140 function in switching from a corepressor to a coactivator was revealed on the rat prolactin promoter. This was driven by changes in the interaction dynamics of pituitary-specific transcription factor (Pit1) with ER and thyroid hormone receptors (Chuang et al. 1997) as a result of a truncation in Pit-1. Thus, RIP140 could act both as a coactivator and corepressor based on interaction dynamics with other factors. Several studies in physiological models and cells have testified to this remarkable ability of RIP140 to act as a coactivator (Zschiedrich et al. 2008, Rosell et al. 2011, Nautiyal et al. 2013a,b) in some situations and as a corepressor in others (Christian et al. 2005, Powelka et al. 2006, Debevec et al. 2007, Seth et al. 2007). Although mechanisms describing this dual role are not well understood, the best hypothesis so far is that, this switch is determined by the cellular environment and is promoter sequence specific.
Wei and coworkers presented the first biochemical evidence that RIP140 possesses a Trichostatin A-sensitive transrepressive activity and could directly recruit histone deacetylases (HDAC) 1 and 3 both in vivo and in vitro. The HDAC-interacting domain on RIP140 was mapped to its N terminus between amino acids 78 and 303 (Wei et al. 2000). HDACs are enzymes responsible for the removal of the acetyl group on the lysines and restoration of the DNA–histone attractive forces, leading to a closed chromatin confirmation. 4 classes of HDACs, representing about 18 enzymes are known. Of these, HDACs belonging to classes I, II and IV share a related catalytic mechanism involving a zinc metal ion and do not require any cofactor. Members of class III HDACs, which include sirtuins 1–7, employ a NAD+ dependent mechanism (Dawson & Kouzarides 2012). HDACs are frequently recruited to nuclear receptors by direct binding or brought to the site of transcription by other primary coregulators such as RIP140 (Heinzel et al. 1997, Nagy et al. 1997, Christian et al. 2006, Kato et al. 2011).
Using deletion mutants, Christian and coworkers identified 4 distinct autonomous repression domains (RD) RD1–RD4 in RIP140 (Christian et al. 2004). These RDs provide platforms for different corepressor complexes. Both RD1 and RD2 act by recruiting HDACs (Wei et al. 2000). Recruitment of HDACs to RD2 is mediated by corepressor, C-terminal-binding protein (CtBP). As repression is only relieved partially by HDAC inhibitor Trichostatin A, it is believed that CtBP acts by HDAC-dependent as well as -independent mechanisms. Four CtBP-binding motifs have been identified in RIP140 protein. Castet and coworkers identified the PIDLS and PINLS motifs, which are CtBP-binding motifs on RIP140 and demonstrated that RIP140 also interacts with HDAC 5. This interaction was mapped to the N terminus between residues 27 and 199 on RIP140 (Castet et al. 2004). As CtBP is a functional dehydrogenase that is sensitive to NADH:NAD redox ratio, an increase in this ratio may promote repression mediated by RIP140 (Christian et al. 2006). Thus, repression brought about by RIP140 is likely to be influenced by the redox status of the cells. Interactions between RIP140 and CtBP are regulated by post-translational modifications on RIP140. For example, acetylation through histone acetyl transferases (HATs) and E1A-binding protein p300 (p300)/cAMP response element-binding protein (CBP) of a N terminal lysine residue on RIP140 led to the disruption of the RIP140–CtBP complex and thus a loss of repressive potential of RIP140 (Vo et al. 2001).
Gupta and coworkers identified 11 phosphorylation sites on RIP140 using mass spectrophotometry. By studying constitutively phosphorylated and dephosphorylated mutants of RIP140 at different sites, this study reported that the strength of the repressive potential of RIP140 through HDACs is determined by its phosphorylation status. For example, a constitutive dephosphorylation at Thr202 and Thr207 on RIP140, significantly impaired its repressive activity and HDAC recruitment, whereas the opposite effects were seen with a constitutive phosphorylation of these residues (Gupta et al. 2005). Further threonine phosphorylation of RIP140 by Erk2, which promoted p300 recruitment for lysine acetylation on RIP140, enhanced its repressive activity. This was shown to influence fat accumulation in differentiated adipocytes (Ho et al. 2008). These studies indicate how post-translational modifications on RIP140 regulate its potential to act as a corepressor. Apart from the interactions with HDACs, several studies also suggested that the repressive mechanism of action of RIP140 is through competition with coactivators like SRC1 (Treuter et al. 1998) and p300/CBP-associated protein (PCAF) (Chen et al. 2004, Gupta et al. 2007). Corepressors NCOR1 and SMRT-1 also act by recruiting HDAC proteins and have specialised domains called the SANT domains through which they modulate the activities of HDACs and help in recognising histones in context with other proteins. The binding sites of several nuclear receptors and transcription factors on NCOR1 and SMRT have been mapped extensively (Wong et al. 2014).
With the background knowledge of corepressive functions of RIP140, the first mechanistic explanation to its coactivating function came from Stefan Herzig’s laboratory. Although an intrinsic activation function within RIP140 has not been found so far, this study reported that RIP140 activates the expression of several inflammatory cytokines in macrophages by coactivating the NFκB subunit RelA. This is achieved as a result of interaction of RIP140 with the C-terminus, of CBP through its RD1 domain (Zschiedrich et al. 2008). RIP140 was further shown to coactivate CREB-mediated transcription on the amphiregulin (Areg) promoter in the ovarian granulosa cells, and recent mapping of the RIP140-binding sites in breast cancer cells have shown that CBP and p300 co-occupy several RIP140-binding sites (Nautiyal et al. 2010, Rosell et al. 2014). Two distinct C-terminal activation domains (AD1/2) are present in coactivators of the SRC family. AD1 recruits CBP/p300 and pCAF and AD2 recruits arginine methyl transferase (CARM1) and protein arginine N methyl transferase (PRMT1) (Rollins et al. 2015).
Biological functions of RIP140
Knowledge of the biological functions of RIP140 came from studies on genetically modified mouse models of RIP140, i.e. the RIP140 knockout (RIPKO) (White et al. 2000) and constitutively overexpressing RIP140 transgenic (RIPTg) mice (Fritah et al. 2010). Two significant phenotypes came to light in RIPKO mice: (i) Female RIPKO mice were infertile because of an inability to ovulate (White et al. 2000) and (ii) both RIPKO male and female mice were significantly leaner than their WT counterparts (Leonardsson et al. 2004). Further studies helped in unravelling the regulatory role of RIP140 and its gene targets in ovary (Tullet et al. 2005, Nautiyal et al. 2010), uterus (Leonardsson et al. 2002), adipose tissue (Christian et al. 2005, Powelka et al. 2006, Hallberg et al. 2008, Park et al. 2009, Persaud et al. 2011), muscle (Seth et al. 2007, Fritah et al. 2012), liver (Herzog et al. 2007, Berriel Diaz et al. 2008), heart (Fritah et al. 2011, Chen et al. 2012), inflammatory cells (Zschiedrich et al. 2008, Ho et al. 2012), mammary gland (Nautiyal et al. 2013b), neural functions (Tsai et al. 2009, Duclot et al. 2012, Feng et al. 2014, 2015) and more recently in the pathophysiology of ageing (Yuan et al. 2012) and cancer (Docquier et al. 2013, Lapierre et al. 2014, 2015a, Rosell et al. 2014, Aziz et al. 2015, Gibson et al. 2016). The following sections describe RIP140 biology in hormone-responsive female reproductive tissues and recent findings of its role in carcinogenesis and metabolic functions in adipose tissue.
Female fertility and mammary gland development
RIPKO female mice show a delayed onset of sexual maturation, which is linked with ageing and to reduced levels of circulating insulin-like growth factor-1 (IGF1) (Yuan et al. 2012). Although these mice have a normal oestrus cycle (Nautiyal et al. 2013b), they are infertile (White et al. 2000). The infertility phenotype led to the examination of the hypothalamic–pituitary–gonadal axis (Leonardsson et al. 2002, Nautiyal et al. 2013b) and studies were carried out to examine the ovaries (White et al. 2000, Leonardsson et al. 2002, Tullet et al. 2005, Nautiyal et al. 2010), uterus (Leonardsson et al. 2002) and mammary gland (Nautiyal et al. 2013b). Cross-sections through ovaries of mature RIPKO mice revealed follicles at all developmental stages from primordial to the pre-ovulatory stage, indicating normal folliculogenesis. Differentiated corpora lutea were also observed in these mice pointing towards normal luteinisation. However, it was remarkable that the corpora lutea retained oocytes thus explaining the anovulatory phenotype (White et al. 2000). Further studies demonstrated defects in the transcriptional regulation and expression in RIPKO mice of the EGF-like growth factor family, which includes amphiregulin (AREG), epiregulin (EREG) and betacellulin (BTC) (Tullet et al. 2005, Nautiyal et al. 2010). In WT mice, the expression of these growth factors in the pre-ovulatory follicle is one of the earliest events after the luteinising hormone (LH) surge (Park et al. 2004). Expression of these factors initiates a cascade of events called cumulus expansion. Cumulus expansion is a biochemical sequence of processes after the LH surge, in the pre-ovulatory follicle where a series of downstream enzymes and factors from the cumulus granulosa cells get expressed, and their products constitute a secreted matrix on which the initially compactly packed cumulus granulosa cells loosen up and thus expand (Fig. 2). Cumulus expansion is essential for ovulation and several mouse models that are reported to be annovulatory show defects in this process (Richards 2007). Pre-ovulatory RIPKO follicles failed to show expression of EGF family members and thus a defective cumulus expansion (Tullet et al. 2005). Exposure of the RIPKO cumulus-oocyte complexes (COCs) to recombinant AREG both in vitro and in vivo rescued cumulus expansion and downstream expression of genes involved in the synthesis of the cumulus matrix. After the LH surge, the EGF-like growth factor family members are known to act in autocrine–paracrine manner and regulate their own expression as well as expression of several downstream enzymes that promote cumulus expansion (Richards 2007, Shimada et al. 2006, Richards & Pangas 2010). It was noteworthy, however, that exposure of RIPKO COCs to recombinant AREG in vitro rescued cumulus expansion and expression of several enzymes involved in the process but did not rescue Areg expression or that of Ereg and Btc indicating that RIP140 is essential for the expression of these genes of the EGF-like growth factor family (Nautiyal et al. 2010). Analysis of Areg promoter in granulosa cell lines demonstrated that RIP140 regulates gene expression of this growth factor by coactivating CREB and cJun transcription factors (Nautiyal et al. 2010). Ovarian and embryo transplant experiments in RIPKO mice resulted in only partial restoration of fertility (Leonardsson et al. 2002). About 50% of the embryos that were transplanted were resorbed at post coitum day 13.5, indicating an inadequate uterine environment to support developing embryos, and most of the pups were still born or died soon thereafter (Leonardsson et al. 2002), suggesting an inability to survive in the absence of maternal feeding. This indicated that although fertility could be partially restored by ovarian correction, the uterus and mammary gland in these mice required closer examination.
Biological processes regulated by RIP140. Ovary: Pre-ovulatory follicles with granulosa cells (pink) lining the follicle. The cumulus oocyte complex, which is the oocyte surrounded by cumulus granulosa cells (darker blue) is shown in an unexpanded state on left. After the LH surge, RIP140 activates amphiregulin (Areg) expression, which further activates downstream enzymes that synthesise the cumulus matrix. This results in cumulus expansion, i.e., cumulus granulosa cells spread on a self-synthesised matrix (lighter blue cells), shown on right (Nautiyal et al. 2010, Ochsner et al. 2003a,b). Mammary gland: Shown on the left is a mammary fat pad with rudimentary epithelium at the base representing a pre-pubertal mouse mammary gland. RIP140 activates mitogens and factors that are essential for mammary gland development, side branching and differentiation (Right). These factors include Areg, PGR, STAT5a and Ccnd1, which are all ER target genes that are coactivated by RIP140 in the mammary gland (Nautiyal et al. 2013a,b). Adipose tissue: (Extreme left) Fat storing WAT that is composed of adipocytes and M2 macrophages. WAT adipocytes expand in size on exposure to high-fat diet to meet the increasing demands of fat storage and also have an increased infiltration of macrophages, which are of the M1 (i.e., inflammatory type). Loss of RIP140 promotes the formation of BAT. RIP140 KO animals are lean and show a BAT adipose phenotype. BAT (extreme right) has smaller fat droplets and is abundant in mitochondria and generates heat. Loss of RIP140 from macrophages promotes the formation of Beige fat depots, which are adipocytes within the WAT that are intermediary between WAT and BAT. This is called ‘browning of fat,’ which is promoted by loss of RIP140 (Liu et al. 2015b).
RIPKO mice show clear defects in the mammary ductal elongation at puberty with adult mice developing a scanty epithelial network. Conversely, RIPTg mice show precocious growth with extensive cell division and side branching and develop ductal hyperplasia in the mammary gland (Nautiyal et al. 2013b). RIPTg mice have been reported to develop rapid, progressive postnatal cardiomyopathy. Examination of the heart tissue in these mice demonstrated mitochondrial degeneration and reduced expression of genes involved in mitochondrial activity, fatty acid metabolism and transport and impaired signalling of nuclear receptors, PPARα, PPARδ, oestrogen-related receptors ERRα and ERRγ (Fritah et al. 2010). As a result of cardiac defects, RIPTg mice die early and thus could not be used to study mammary tumourigenesis. Paracrine signalling and crosstalk between epithelium and stromal cells is essential for the normal development of the mammary gland (Sternlicht 2005). Tissue recombinant studies at pre-pubertal stages in the RIPKO and RIPTg mouse models helped in highlighting that RIP140 expression is essential in mammary epithelium as well as stroma for mammary gland development (Nautiyal et al. 2013b). Determining gene targets of RIP140 in isolated stromal and epithelial cell populations is warranted to obtain further mechanistic insights in stromal–epithelial crosstalk in mammary gland development. Global ChIP-seq analysis of ER binding sites in the mouse mammary tissue helped in finding several important regulators of mammary gland development that were ER targets and transcriptionally coactivated by RIP140. These genes included Areg, progesterone receptor (Pgr), cyclin D1 (ccnd1), c-Myc (Myc) and Signal Transducer and Activator of Transcription (Stat5a) (Nautiyal et al. 2013b). Analysis of mammary cell populations from RIPTg mice demonstrated that RIP140 overexpression leads to an increase in the basal mammary cell population. This is interesting as the basal cell compartment of the mammary gland is considered home for multi-potent stem cells. Within the luminal cell population, there was an increase in the progenitor cells over the differentiated ones indicating that RIP140 promotes proliferation and/or blocks differentiation (Nautiyal et al. 2013b). Thus, RIP140 levels determine cell fate in the mammary gland. Although work here was mainly done to look at ER target genes coregulated by RIP140, it is worth mentioning that its coregulatory potential with other transcription factors in the mammary gland remains to be explored.
Carcinogenesis
Insights obtained about the role of RIP140 in mammary gland development as a coactivator for ER warranted further mechanistic understanding of its involvement in ER-mediated gene transcription. Initial reports had implicated the involvement of RIP140 in breast cancer (Oh et al. 2006, Lee et al. 2007, Hannafon et al. 2011, Docquier et al. 2012). ChIP-seq analysis was performed to map RIP140 binding with ER in MCF7 breast cancer cell line. 82% of the RIP140-binding sites were co-occupied by ER and its interaction partners FOXA1, GATA3 and other coactivators p300, CBP and p160 family at H3K4me1-demarcated regions in the genome, H3K4me1 being an enhancer-specific epigenetic mark of ER binding (Joseph et al. 2010). Very similar to ER, only 5% of the RIP140 binding was seen at the promoters, whereas the rest was enriched in introns and distal intergenic regions (Fullwood et al. 2009, Rosell et al. 2014). siRNA knockdown experiments revealed that RIP140 is required for ER-mediated transcriptional complex formation and gene regulation, which stimulated breast cancer cell proliferation. Genes that changed as a result of RIP140 knockdown in the MCF7 cells were used to generate a gene expression classifier that could be successfully used to stratify patients with breast cancer, who received adjuvant tamoxifen treatment. Patients in the cluster representing poor outcome correlated with active RIP140 gene expression signature, whereas the cluster representing good outcome represented cases with less-active RIP140-responsive gene signature (Rosell et al. 2014). Aziz and coworkers demonstrated that RIP140-deficient mice were less susceptible to 7,12-dimethylbenz[a]anthracene (DMBA)-induced carcinogenesis compared to their WT counterparts. Although WT mice developed tumours in skin, mammary gland, liver and intestinal tract after DMBA exposure for 6 weeks, RIPKO mice did not develop any tumours, with the exception of smaller skin tumours. Analysis of several breast cancer cellular models showed that RIP140 knockdown induced apoptosis and inhibited cell growth (Aziz et al. 2015).
Although the role of RIP140 as a coactivator for ER is appreciated, Docquier and coworkers highlighted its role as a corepressor of ERβ and in antagonising E2-mediated cellular proliferation in ER-positive ovarian cancer cells, stably integrated with an oestrogen response element (ERE) luciferase reporter. With ERβ overexpression, there was increased recruitment of RIP140 on the ERE, and siRNA knockdown of RIP140 abolished the repressive effect exerted by ERβ in the regulation of E2-mediated activation of the ERE (Docquier et al. 2013). It has also been demonstrated in a study using ovarian cancer cells that RIP140 is involved in sensitisation of these cells to chemotherapeutic agents (Lapierre et al. 2015b). Recently, a tumour suppressor role for RIP140 has been suggested in colon cancer. Examination of RIPKO and RIPTg mice demonstrated that RIP140 inhibits the renewal of intestinal epithelium by regulating cell proliferation and apoptosis and thus influencing intestinal homeostasis. When mice were subjected to whole body irradiation, RIPKO mice demonstrated improved regenerative capacity in the intestine, whereas the overexpressing RIPTg mice had reduced renewal of the intestinal villi compared to WT counterparts (Lapierre et al. 2014). This was further confirmed in cellular models in mouse xenografts in which enhanced RIP140 expression reduced colon cancer cell proliferation. Clinically, RIP140 mRNA and protein levels were lower in human colon cancer compared to normal tissue, and low RIP140 expression correlated with poor prognosis. Mechanistically, RIP140 stimulates the transcription of multiprotein β catenin degradation complex subunit, adenomatous polyposis coli (APC), which inhibits β-Catenin activation in the intestinal epithelium (Lapierre et al. 2014). Recently, RIP140 has emerged as a prognostic marker in haematologic malignancies such as chronic lymphocytic leukaemia (CLL) (van’t Veer et al. 2006, Herold et al. 2011). High expression of RIP140 is a favourable prognostic marker in CLL. The exact mechanisms and pathways regulated by RIP140 in CLL are not known so far. It is speculated that it may well be through regulation of nuclear receptors, Wnt and NFκB signalling pathways (Lapierre et al. 2015a).
Two studies to identify mutations and drivers of endometrial cancer (EC) progression have recently identified several mutations in RIP140, suggesting its role in impaired ER signalling (Ferreira et al. 2014, Gibson et al. 2016). Microsatellite instability (MSI), which is a phenotype caused by mutations in the DNA mismatch repair (MMR) genes, is a common feature of EC (Cancer Genome Atlas Research Network et al. 2013). In a genome-wide search to identify genes effected by MSI, RIP140 was the highest mutated gene (34% of EC) (Ferreira et al. 2014). Exome sequencing of EC tumours in another study identified RIP140 mutation in a significant number of tumours (12% of EC). Although several missense mutations were identified, indel (single base pair insertion and deletion) mutations were highly recurrent on Asn516 and Lys728 on RIP140 (Gibson et al. 2016). Future studies should focus on the functional effects of RIP140 alterations in driving the development of EC.
Based on the examination of patient cohorts and work in mouse and cancer cellular models in the previously mentioned studies, it is clear that RIP140 plays a role in tumourigenesis. Just like in physiological conditions, the ability of RIP140 to act both as a coactivator and corepressor based on the cellular milieu and in a transcription factor-specific context is reflected in cancer as well. For example, RIP140 promotes cancer cell proliferation in breast cancer (Rosell et al. 2014), whereas it inhibits proliferation in colon cancer (Lapierre et al. 2014). Further functional and molecular studies in individual cancers will determine its use as a biomarker for patient stratification and help in determining individualised approaches for treatment.
Metabolic functions in adipose tissue
Adipose tissue is an endocrine organ that produces hormones such as E2, leptin, resistin, adiponectin and inflammatory cytokines and is a regulator of energy homeostasis (Stern et al. 2016). By regulating adipose composition in the body, RIP140 maintains the balance between energy storage and expenditure by mechanisms that are coming to light now (Kiskinis et al. 2014, Hu et al. 2015, Liu et al. 2015b). There are mainly two types of adipose tissues. (1) The fat-storing white adipose tissue (WAT), which expands under obese conditions and becomes inflammatory, leading to insulin resistance and (2) the fat-burning brown adipose tissue (BAT), which is rich in mitochondria and thus expends energy (Fig. 2). Originally, BAT was thought to exist only in neonates, but it has recently been shown to exist in adult humans (Cypess et al. 2009).
RIPKO mice are lean, with a 70% reduction in body fat and show resistance to high-fat diet-induced obesity. Although BAT tissue in these mice is histologically normal, the WAT tissue in RIPKO mice shows a significant reduction in the size of the adipocytes (Leonardsson et al. 2004). The WAT in RIPKO mice expresses genes that are more characteristic of brown fat, i.e., involved in increasing energy expenditure. These genes include uncoupling protein 1 (UCP1), cell death-inducing DNA fragmentation factor (CIDEA), a lipid droplet protein and carnitine palmitoyl transferase (CPT1), a mitochondrial fatty acid transporter (Leonardsson et al. 2004, Christian et al. 2005). Studies in genetically modified adipose cellular models demonstrated that RIP140 acts as a repressor of catabolic processes such as fatty acid oxidation, oxidative phosphorylation, glycolysis and tricarboxylic acid cycle (Powelka et al. 2006) and thus ablation of RIP140 leads to the upregulation of gene networks in these pathways and thus expenditure of energy. PGC1α a transcriptional coactivator first identified and cloned in BAT by the Speigelman group (Puigserver et al. 1998), modulates a number of genes involved in these metabolic pathways and shares several gene targets with RIP140, the two coregulators antagonising each other’s actions (Hallberg et al. 2008, Fernandez-Marcos & Auwerx 2011).
The existence of a third type of fat namely ‘the beige fat’ or BRITE has been recently appreciated in WAT. This fat has the potential to thrive within the WAT depots on exposure to cold and β2 adrenergic agonist treatment. This ‘browning of fat’ is considered beneficial as beige fat, like brown fat, is involved in thermogenesis although it is worth mentioning that the brown and beige adipocytes show distinct gene expression signatures (Walden et al. 2012) (Fig. 2). Coregulators PGC1α and PR domain-containing 16 (PRDM16) influence the expression of brown adipocyte genes within white adipocytes. Further investigation of the role of RIP140 in adipose cell lines indicated that RIP140 has little role in the regulation of brown fat gene expression. The way RIP140 acts is through preventing the development of BRITE depots within the white fat. This could either be by antagonising the coactivator potential of PGC1α on common genes or by directly regulating the expression of PGC1α (Kiskinis et al. 2014).
Obesity, insulin resistance and metabolic dysregulation are inflammatory conditions, and the role of macrophages is appreciated in this. Apart from being crucial effectors of the innate immune responses, macrophages play vital homeostatic roles, which are separate from their function as immune cells. For example, these phagocytic cells clear cellular debris generated during tissue remodelling and potentially harmful apoptotic products with precision (Mosser & Edwards 2008). Macrophages are known to have remarkable plasticity, and depending on the metabolic cues, they are capable of making phenotypic switches to different states. Broadly, macrophages can exist in two states, namely M1 and M2. These two states represent different gene expression and transcriptional profiles (Gautier et al. 2012), and their number and activation state can vary, based on the metabolic signals in the WAT tissue (Lumeng et al. 2007). For example, M2 maintains homeostasis and insulin sensitivity, whereas the M1 represents a state of chronic inflammation. In a high-fat diet-induced obesity model, the number of macrophages increases from 10 to 15% of the total stromal cells, mostly in the M2 state to 45–60% due to increased inflammation and recruitment of the M1 type (Odegaard & Chawla 2013). This leads to remodelling of the WAT tissue, which enlarges to meet the increased demands of energy storage and also secretes further harmful substances that aggravate the state of inflammation leading on to insulin resistance (Chawla et al. 2011, Wynn et al. 2013, Liu et al. 2014).
RIP140 has been reported to tip the balance of M1 vs M2 polarisation of the macrophage population (Liu et al. 2015b). High-fat diet increases RIP140 expression in macrophages. RIP140 knockdown specifically in macrophages in mice led to adipose tissue remodelling as white adipose tissue underwent browning and improved insulin sensitivity was observed in high-fat diet-fed conditions. In the absence of RIP140, the macrophage profile in these mice altered under high-fat diet conditions with increase in macrophages in M2 state and a decrease in M1 state, which improved insulin resistance. This study demonstrated that altering RIP140 expression in macrophages could be used as a new therapeutic strategy in inflammatory disorders and high-fat diet-induced insulin resistance (Liu et al. 2014). Exposure of preadipocytes to media obtained from cultures of macrophages, which are deficient in RIP140 led to differentiation into beige cells. Likewise injecting obese mice with M2 macrophages obtained from WAT of mice, which were deficient for RIP140 expression in macrophages, led to improvement of insulin resistance and browning of the white adipose tissue depots providing a proof of concept regarding the therapeutic potential in injecting engineered macrophages for RIP140 expression and correcting metabolic dysregulation (Liu et al. 2015a).
Concluding remarks
RIP140 is a coregulator that drives crucial physiological processes and plays a role in the development and progression of cancer, metabolic abnormalities and potentially even fertility-related pathologies. Reports of its coactivating and corepressing potential in different physiological or pathological settings are tantalising, and more mechanistic insights are needed to clarify this dual role of RIP140. The biomarker and therapeutic potential of RIP140 will be explored in future studies.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of this review.
Funding
This article did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector.
Acknowledgement
The author is grateful to Prof. Malcolm Parker, Prof. Simak Ali and Dr Mark Christian for critical reading of this manuscript.
- Received 26 December 2016
- Accepted 10 January 2017
- Made available online as an Accepted Preprint 10 January 2017
- © 2017 Society for Endocrinology
