The insulin receptor - CiteSeerX

Endocrine-Related Cancer (2008) 15 325–335

The insulin receptor: a new anticancer target for peroxisome proliferator-activated receptor-g (PPARg) and thiazolidinedionePPARg agonists V Costa, D Foti, F Paonessa, E Chiefari, L Palaia, G Brunetti1, E Gulletta, A Fusco2 and A Brunetti Dipartimento di Medicina Sperimentale e Clinica ‘G. Salvatore’, Universita` di Catanzaro ‘Magna Græcia’, 88100 Catanzaro, Italy 1 Dipartimento di Scienze Biomolecolari e Biotecnologie, Universita` di Milano, 20133 Milano, Italy 2 Dipartimento di Biologia e Patologia Cellulare e Molecolare c/o Istituto di Endocrinologia ed Oncologia Sperimentale del Consiglio Nazionale delle Ricerche, Universita` di Napoli Federico II, e NOGEC-CEINGE, Biotecnologie Avanzate, 80131 Napoli, Italy (Correspondence should be addressed to A Brunetti; Email: [email protected]) V Costa and D Foti contributed equally to this work

Abstract The peroxisome proliferator-activated receptor-g (PPARg) is a member of the nuclear hormone receptor superfamily. Ligand activation of PPARg is associated with differentiation and inhibition of proliferation in the normal and malignant cells. Herein, we studied the effects of PPARg and the PPARg agonists thiazolidinediones (TZDs) on the insulin receptor (IR), a cell membrane tyrosine kinase receptor protein, whose role is of paramount importance in mediating the metabolic and growth-promoting effects of the peptide hormone insulin. Overexpression of the PPARg1 in human hepatocellular (HepG2) cells was associated with decreased IR gene transcription and protein expression levels, and these reductions were more evident in the presence of TZDs. Since no PPARg response elements were identified on the IR promoter, we postulated that PPARg adversely affects the IR gene transcription by perturbing the assembly and stability of the transcriptionally active multiprotein-DNA complex identified previously, which includes the high-mobility group A1 protein, the ubiquitously expressed transcription factor (Sp1), the CAAT enhancer-binding protein (C/EBPb), and, in some cell lines, the developmentally regulated activator protein-2 (AP-2) transcription factor. Using glutathione S-transferase pull-down assays combined with electrophoretic mobility shift assay and chromatin immunoprecipitation, we demonstrated that by interacting with Sp1, C/EBPb, and AP-2, PPARg can prevent Sp1/AP-2 protein–protein association and inhibit binding of Sp1 and C/EBPb to DNA, thus reducing IR gene transcription. Our results demonstrate that IR is a new target gene of PPARg, and support a potential use of TZDs as anti-proliferative agents in selected neoplastic tissues overexpressing IRs. Endocrine-Related Cancer (2008) 15 325–335

Introduction The peroxisome proliferator-activated receptor-g (PPARg) is a member of a subfamily of nuclear receptors involved in the control of several aspects of lipid metabolism, including fatty acid transport, uptake by the cells, intracellular binding and activation, as

well as catabolism and storage (Desvergne & Wahli 1999, Kliewer et al. 2001, Berger & Moller 2002). It is highly expressed in adipose cells and its induction precedes the activation of most adipose-specific genes (Rosen et al. 1999, Keller et al. 1993a,b). Although PPARg is mainly expressed in the adipose tissue, its expression also occurs in many normal cells

Endocrine-Related Cancer (2008) 15 325–335 1351–0088/08/015–325 q 2008 Society for Endocrinology Printed in Great Britain

DOI: 10.1677/ERC-07-0226 Online version via http://www.endocrinology-journals.org

V Costa, D Foti et al.: Effects of PPARg and TZD-PPARg on IR expression (Vidal-Puig et al. 1997), and in several cancer cell lines and tissues, including breast, prostate, colon, lung, and B-lineage cells (Padilla et al. 2000, Allred & Kilgore 2005, Nunez et al. 2005). Besides a fundamental role in adipogenesis, PPARg plays a critical role in placental, cardiac, and embryonic development (Barak et al. 1999, Maloney & Rees 2005) and influences the production of cytokines, growth factor release, and cell cycle progression leading to differentiation-inducing, anti-inflammatory, and anti-proliferative effects (Kersten et al. 2000, Tan et al. 2005, Chou et al. 2007). The PPARg generally exerts its biological function as heterodimer with the retinoid X receptor, which binds to a specific cis-acting sequence on DNA, a peroxisome proliferator response element, initiating gene transcription (Desvergne & Wahli 1999, Tan et al. 2005). The PPARg activators identified so far include some inflammatory activators, like a prostaglandin J2 metabolite (15-deoxy-D12,14-PGJ2), and oxidized LDL particles (9-and 13-HODE; Desvergne & Wahli 1999, Kersten et al. 2000), and the synthetic thiazolidinedione (TZD) compounds (Lehmann et al. 1995). The TZDs are a new class of insulin-sensitizing drugs widely used in the treatment of human type 2 diabetes mellitus, including rosiglitazone, pioglitazone, and ciglitazone (Lehmann et al. 1995, Saltiel & Olefsky 1996). However, despite the positive role of TZDs in the control of glucose homeostasis, increasing evidences support a potential use of these drugs as antiproliferative agents. The peptide hormone insulin is a major regulator of glucose homeostasis and cell growth. The first step in insulin action is the binding of the hormone to the insulin receptor (IR), a phylogenetically ancient receptor protein embedded in the plasma membrane of insulin target cells. The IR belongs to the tyrosine kinase growth factors receptor superfamily and consists of two identical extracellular a-subunits that bind insulin, and two transmembrane b-subunits with intrinsic tyrosine kinase activity (Ullrich et al. 1985, Goldfine 1987, White & Kahn 1994). When insulin binds to the IR, the receptor is first activated by tyrosine autophosphorylation and then the IR tyrosine kinase phosphorylates various effector molecules, like the insulin receptor substrate-1 (IRS-1), leading to hormone action (Ullrich et al. 1985, Goldfine 1987, White & Kahn 1994). In target cells, the IR has been shown to be under the regulation of hormones, metabolites, and differentiation (Brunetti et al. 1989, Mamula et al. 1990). To understand the molecular mechanisms controlling IR gene regulation, many groups including our own group have previously identified and characterized the 5 0 -flanking region of 326

the IR gene (Araki et al. 1987, Seino et al. 1989, Brunetti et al. 1993). Later, we have provided evidence that transcriptional activation of the human IR promoter requires the assembly of a transcriptionally active multiprotein-DNA complex which includes, in addition to the architectural factor high-mobility group A1 (HMGA1), the transcription factors Sp1 and CAAT enhancer-binding protein (C/EBPb). Functional integrity of this nucleoprotein complex is required for full transactivation of the IR gene by Sp1 and C/EBPb in cells readily expressing IRs (Brunetti et al. 2001, Foti et al. 2003, 2005). Recently, we published data showing that the nuclear transcription factor activator protein-2 (AP-2), by interacting with Sp1 and HMGA1, may stabilize this multiprotein complex at the level of the IR promoter, leading to IR overexpression in breast cancer cells (Paonessa et al. 2006). In this paper, we demonstrate for the first time that PPARg and its ligands inhibit IR gene transcription. We show that PPARg physically interacts with Sp1, AP-2, and C/EBPb in vitro, and this protein–protein interaction significantly reduces the transactivation of the IR gene by Sp1, C/EBPb, and AP-2 in vivo, in the absence of a consensus sequence for PPARg on the IR promoter.

Materials and methods Cells and protein extracts

HepG2 human hepatoma cells, 3T3-L1 mouse fibroblasts, and MCF-7 human breast cancer cells (American Type Culture Collection, Manassas, VA, USA) were maintained in DMEM (GIBCO-BRL) supplemented with 10% fetal bovine serum. The 3T3-L1 fibroblasts were differentiated into 3T3-L1 adipocytes as described elsewhere (Foti et al. 2003). Nuclear and cytoplasmic extracts were prepared from cultured cells as previously described (Brunetti et al. 2001). For each nuclear extract, an equal number of nuclei were homogenized, and the final protein concentration in the extracts was determined by the modified Bradford method (Bio-Rad Laboratories). RT-PCR, immunoprecipitation, and western blot analysis

Total cellular RNA was extracted from cells using the RNAqueous-4PCR kit and subjected to DNase treatment (Ambion, Austin, TX, USA). The cDNA was synthesized from total RNA with the RETROscript first strand synthesis kit (Ambion) and used for PCR amplification. The PCR products were electrophoretically resolved on 2% agarose gel, visualized by www.endocrinology-journals.org

Endocrine-Related Cancer (2008) 15 325–335 ethidium bromide staining, and quantified by densitometry. The values obtained were then normalized to those of the tubulin gene. Western blot analyses of IR, PPARg, HMGA1, Sp1, C/EBPb, and AP-2 were done on total cellular lysates or nuclear extracts as previously described (Brunetti et al. 2001, Foti et al. 2003, 2005). For immunoprecipitation studies with IR antibody, aliquotes of HepG2 cytoplasmic extracts were incubated for 12 h with rotation at 4 8C with 10 ml IR antibody-coupled protein A beads. Covalent coupling of antibody was performed as previously described (Paonessa et al. 2006). Beads were recovered by gentle centrifugation and washed thrice with 500 ml NETN wash buffer (0.1% NP40, 150 mM NaCl, 1 mM EDTA, 50 mM Tris–HCl (pH 8.0)) for 5 min. Protein was removed from the beads by boiling in sample buffer for 5 min and analyzed by SDS-PAGE and immunoblotting. Antibodies used for these studies were as follows: anti-HMGA1 (Brunetti et al. 2001), anti-PPARg1 (H-100), anti-Sp1 (PEP 2), anti-C/EBPb (C-19), anti-AP-2a (H-79), and anti-IRb (C-19; Santa Cruz Biotechnology, Santa Cruz, CA, USA). GST pull-down assay

The 35S-labeled HA-tagged HMGA1, Sp1, C/EBPb, and AP-2 (Foti et al. 2003, Paonessa et al. 2006) were synthesized in vitro using the TNT-T7 quick-coupled transcription/translation system (Promega). The glutathione S-transferase (GST) fusion protein expression vectors pGEX4T-3-human PPARg2 (kindly provided by H Nishizawa and T Funahashi, University of Osaka; Nishizawa et al. 2002) and the pGEX-2TK control vector (Amersham Pharmacia Biotech) were transformed into the BL21(D3) strain of Escherichia coli (Stratagene, La Jolla, CA, USA), expanded in suspension culture, and induced for 2 h with 0.5 mM isopropyl-D-thiogalactopyranoside (Sigma). Bacteria were pelleted, sonicated in ice-cold PBS lysis buffer containing 1% NP-40, 10% glycerol, 1 mM dithiothreitol (DTT), 1 mM phenylmethylsulphonyl fluoride (PMSF), 10 mg pepstatin A, leupeptin, and aprotinin/ml, and 0.4 mg lysozyme/ml and centrifuged. The resultant supernatant was then added to 300 ml glutathioneagarose beads, mixed on a rotating wheel at 4 8C for 2 h and centrifuged. The bound GST-fused proteins in the pellet were washed five times with lysis buffer and resuspended in 300 ml binding buffer (50 mM NaCl, 20 mM Tris–HCl (pH 8.0), 0.05% NP-40, 0.25% BSA, 1 mM PMSF, 1 mM DTT). The bound protein was quantitated with the Coomassie protein assay reagent (Pierce Co., Rockford, IL, USA), and 0.5 mg of each GST-protein bound to glutathione-agarose beads was www.endocrinology-journals.org

incubated with 7 ml in vitro-translated 35S-labeled protein in 150 ml binding buffer at 4 8C for 2 h. Reactions were terminated by centrifugation, the precipitate was washed three times with proteinbinding buffer and subjected to a 10% PAGE (SDS-10% PAGE), and proteins were visualized by autoradiography. Plasmids, transfections, and electrophoretic mobility shift assay (EMSA)

Recombinant chloramphenicol acetyltransferase (CAT) plasmids (pIR-CAT, pCAT-C2 and pCAT-E3) containing different fragments of the human IR gene promoter have been described previously (Brunetti et al. 1993, Foti et al. 2003, Paonessa et al. 2006). The human PPARg1 expression vector was obtained using standard techniques after PCR amplification and cloning of PPARg full-length cDNA into the expression vector pcDNA3. The pC-hu/AP-2a expression vector was kindly provided by K Fujimori (Osaka Bioscience Institute). The IR-CAT reporter plasmids, together with effector vectors for PPARg and/or AP-2, in the absence or presence of the PPARg agonist rosiglitazone (a kind gift of Glaxo Smith Kline) or ciglitazone (Vinci-Biochem Alexis, Vinci, Italy) were transiently transfected into cultured cells by the calcium phosphate precipitation method, and CAT activity was assayed 48 h later, as previously described (Brunetti et al. 1993). The pSV-b-galactosidase control vector served as an internal control of transfection efficiency, together with measurements of protein expression levels (Brunetti et al. 2001). For EMSA, the C2 sequence (300 bp) of the human IR gene promoter was generated by PCR amplification, using the recombinant plasmid pCAT-C2 (Brunetti et al. 1993). The 32P-labeled C2 was used in gel shift assays as previously described (Brunetti et al. 2001). Chromatin immunoprecipitation (ChIP)

ChIP assay was done as described previously (Foti et al. 2005), using HepG2 cells transfected with the AP-2a expression vector and untransfected MCF-7 cells. Formaldehyde-fixed DNA–protein complex was immunoprecipitated with anti-HMGA1, anti-Sp1, antiC/EBPb, or anti-AP-2a antibody. Primers for the IR sequence (Paonessa et al. 2006) were used for PCR amplification of immunoprecipitated DNA (30 cycles) using PCR ready-to-go beads (Amersham Pharmacia Biotech). The PCR products were electrophoretically resolved on 1.5% agarose gel and visualized by ethidium bromide staining. 327

V Costa, D Foti et al.: Effects of PPARg and TZD-PPARg on IR expression Statistical analysis

All experiments were performed at least three times. The mean and S.E.M. were calculated for each set of results and the significance of the difference between values was assessed by the Student’s t-test. For all analyses, P!0.05 was considered significant.

Results PPARg and TZDs reduce IR protein and mRNA expression in HepG2 cells

To see whether PPARg had any effect on IR expression, we first measured the IR protein content in HepG2 cells, before and after forced expression of exogenous PPARg1. The HepG2 cells were considered ideally suited for studying the effects of PPARg on IR expression since IRs are relatively abundant in this cell line, while endogenous PPARg is barely detectable. Using immunoprecipitation and western blot analyses, we showed that IR protein content was considerably reduced in cells overexpressing PPARg (w50% less than control cells), and this reduction was more pronounced in PPARg1-overexpressing cells exposed to 10K5 M TZD (rosiglitazone; Fig. 1). It is notable that rosiglitazone alone also reduced IR protein content, and it is likely that this occurred, at least in part, by stimulating endogenous PPARg (Tontonoz et al. 1994). As measured by RT-PCR, the abundance of IR mRNA transcripts in HepG2 cells closely

Figure 1 Inhibition of IR protein and mRNA levels by PPARg and TZDs. (A) Total cellular extracts from untransfected (control) or HepG2 cells transfected with PPARg expression vector, in the absence or presence of TZD (rosiglitazone, 10K7 M), were resolved by SDS-PAGE, and the IR protein content was measured by immunoprecipitation and the western blot (IP/WB) with an anti-IRb antibody. Tubulin was measured as an internal control. (B) In parallel experiments, IR and tubulin mRNA levels were quantified by RT-PCR.

328

correlated with IR protein content in every condition analyzed, indicating that changes at the protein level reflected differences in IR mRNA (Fig. 1). PPARg and TZDs reduce IR gene transcription in cultured cells

In order to reveal whether PPARg and its agonists inhibited the transcriptional activation of IR gene, we performed reporter gene analysis using the pIR-CAT recombinant vector that contains the entire 5 0 -flanking region of the human IR gene (Foti et al. 2003). Forced expression of PPARg1 in HepG2 cells transfected with pIR-CAT determined a 40–50% reduction in CAT activity that was similar to the reduction observed in cells exposed to rosiglitazone alone, and resulted more marked in PPARg1-overexpressing cells simultaneously treated with TZD rosiglitazone (Fig. 2A). The possibility that PPARg could theoretically decrease IR gene transcription by reducing the intracellular expression of HMGA1, Sp1 or C/EBPb was excluded by western blot analysis of nuclear extracts showing that no changes were observed in the endogenous levels of these proteins under these experimental conditions (Fig. 2A). The hypothesis that PPARg and its agonists could negatively interfere with the crosstalk between these various transcription factors in the context of the IR gene was substantiated further by transiently cotransfecting the AP-2-null cell line HepG2 with the pIR-CAT promoter, together with an expression vector encoding the wild-type AP-2a. As shown in Fig. 2A, overexpression of AP-2 in HepG2 cells efficiently induced IR gene transcription. However, this effect was totally abolished by PPARg and TZDs that reverted gene transcription to levels similar to those observed in the absence of AP-2 (Fig. 2A). These findings were confirmed in PPARg overexpressing MCF-7 cells, a cell line naturally expressing only detectable levels of PPARg (Tontonoz et al. 1994), and 3T3-L1 adipocytes producing relatively high amounts of endogenous PPARg (Mueller et al. 1998). As shown in Fig. 2B and C, exposure of both cell lines to either rosiglitazone and/or ciglitazone reduced IR gene transcription in a dose-dependent manner. We have previously shown that transcriptional activation of the human IR gene by HMGA1, Sp1, and C/EBPb requires the assembly and cooperation among these various nuclear factors at the levels of two AT-rich sequences of the IR gene promoter, C2 and E3 (K674 to K874 and K1662 to K1818 bp upstream of the IR ATG codon respectively), which have a significant ability to drive transcription when introduced into mammalian cells (Brunetti et al. 1993, 2001, Foti et al. 2003). As demonstrated before, the www.endocrinology-journals.org


Figure 2 Functional significance of PPARg and TZDs for IR gene transcription. (A) HepG2, (B) MCF-7, and (C) 3T3-L1 cells were transfected with a CAT reporter plasmid containing the IR promoter element (pCAT-IR, 2 mg), in the absence or presence of PPARg (2 mg) and/or AP-2 (1 mg) expression vector, with or without treatment with rosiglitazone or ciglitazone at the indicated molar concentrations. For each cell line, data represent the meansGS.E.M. for three separate experiments; values are expressed relative to the CAT activity obtained in transfections with the pCAT-IR alone that is assigned an arbitrary value of 1. *P!0.05 versus cells transfected with the pCAT-IR alone; #P!0.01 versus cells transfected with the pCAT-IR plus AP-2 effector vector. In some experiments (D), HepG2 cells were transfected as above, using the CAT reporter vector containing the proximal C2 (pCAT-C2) or distal E3 (pCAT-E3) element of the IR gene promoter. *P!0.05 versus cells transfected with the pCAT-C2 alone. White bars, mock (no DNA); black bars, pCAT-basic (vector without an insert). Shown are representative western blots of PPARg, Sp1, C/EBPb, and HMGA1 from nuclear extracts of transfected HepG2 cells.

lack of any of these factors, or a scarcity of one or more, may severely perturb IR gene transcription in cells and tissues readily expressing IR (Brunetti et al. 2001, Foti et al. 2003, 2005). When the recombinant plasmid pCAT-C2 was transfected into HepG2 cells (and other cell lines), PPARg, and TZDs, either alone or in combination, inhibited pCAT-C2 activity to the same extent than they did in HepG2 cells transfected with the full-length pCAT-IR promoter (Fig. 2D), indicating that interference of PPARg and agonists with IR gene-transcription machinery occurs at the level of the proximal promoter region, C2. This conclusion was supported by the observation that in experiments using the pCAT-E3 reporter vector, containing the more distant E3 sequence of the IR www.endocrinology-journals.org

gene, no effects on CAT activity were observed in cells exposed to PPARg and/or TZDs (Fig. 2D). PPARg physically interacts with Sp1, C/EBPb, and AP-2, in vitro, in the absence of DNA

In our studies, no peroxisome proliferator response elements (PPRE) have been identified on the promoter region of the IR gene. A similar observation has been previously provided by us for the transcription factor AP-2, for which DNA-binding activity was undetectable with the IR gene promoter, and transactivation of the IR gene by AP-2 occurred indirectly through physical and functional cooperation with HMGA1 and Sp1 (Paonessa et al. 2006). Therefore, in an attempt to identify the biochemical mechanisms underlying the negative effects 329

V Costa, D Foti et al.: Effects of PPARg and TZD-PPARg on IR expression of PPARg on IR gene transcription, we performed experiments designed to investigate whether PPARg could affect this promoter function by physically interacting with HMGA1, Sp1, C/EBPb or AP-2, thus preventing the interaction between these factors in the preinitiation complex at the IR promoter. To this end, we performed GST pull-down assays, in which in vitro-translated 35S-labeled proteins were analyzed for their ability to be specifically retained by a GST-PPARg affinity resin. As shown in Fig. 3, under our experimental conditions, no interaction was observed between 35 S-labeled HA-tagged HMGA1 and GST-PPARg, excluding a physical contact between these two proteins in vitro. Conversely, the 35S-labeled Sp1 was retained by GST-PPARg but not GST alone, indicating that Sp1 physically interacts with PPARg. Similarly, the 35 S-labeled AP-2 and to a lesser extent labeled C/EBPb were also specifically bound by GST-PPARg, but not by GST alone (Fig. 3), indicating that PPARg directly interacts with these two proteins as well. A bona fide interaction between these factors and GST-PPARg was proved by the use of high concentrations of ethidium bromide (not shown), which has been shown to disrupt DNA-dependent protein–protein contact (Foti et al. 2003).

Physical interaction between PPARg, Sp1, AP-2, and C/EBPb has been characterized further in vitro by EMSA, using the previously described radiolabeled fragment C2 of the human IR promoter gene as probe (Brunetti et al. 1993). As shown in Fig. 4, binding of Sp1 and HMGA1 was easily detected in nuclear extracts

from untreated HepG2 cells, in which only trace amounts of PPARg are normally expressed. Binding of Sp1 was reduced in nuclear extracts from HepG2 cells overexpressing PPARg1, and was further decreased after treatment of these cells with TZD rosiglitazone. Under these conditions, no changes were observed with HMGA1-DNA binding. When EMSA was performed by nuclear extracts from HepG2 cells overexpressing AP-2, a co-complex containing Sp1, HMGA1, and AP-2 was also detected, in addition to Sp1 and HMGA1 complexes (Fig. 4). The reduction in Sp1DNA binding, in cells overexpressing PPARg1, paralleled the reduction in the co-complex-DNA binding formation and these interactions were further reduced by the addition of rosiglitazone. Similar results were obtained in the MCF-7 cells transfected with PPARg expression vector, either in the absence or in the presence of TZD. Like Sp1 and AP-2, binding of endogenous C/EBPb to DNA was reduced in nuclear extracts from cells overexpressing PPARg1 and was further decreased by treatment of cells with rosiglitazone (Fig. 4). The presence of HMGA1 in the more rapidly migrating complex was verified by its supershifting in the presence of the anti-HMGA1 antibody, whereas the presence of Sp1, AP-2, and C/EBPb in the slowly migrating complexes was verified by their supershifting in the presence of Sp1-, AP-2-, and C/EBPb-specific antibodies respectively. Since no effect on the electrophoretic migration of DNA–protein complexes was observed in the presence of controlunrelated antibodies (data not shown), the reduction of protein–DNA binding in supershift assays suggests that these factors are simultaneously binding the same labeled probe. The above findings were substantiated in vivo by ChIP analyses. As shown in Fig. 5, binding of Sp1, AP-2, and C/EBPb to the C2 promoter region of

Figure 3 Physical association between PPARg and IR-DNA binding proteins. SDS-PAGE of 35S-labeled HA-tagged HMGA1, Sp1, AP-2, and C/EBPb bound to GST-PPARg resin. In the first lane of each panel, 5% of the labeled protein was added directly onto the gel without binding to and elution from GST protein resin (input). Specifically bound proteins were visualized by autoradiography.

Figure 4 Nuclear protein–DNA interactions within the C2 region of the IR promoter. EMSAs of radiolabeled fragment C2 (0.2 ng) with 0.5 mg nuclear protein from (A) wild-type HepG2 cells or (B) AP-2-overexpressing HepG2 cells, and (C) wild-type MCF-7 cells, in the presence of 0.2 mg poly (dI-dC) as the competitor DNA. In supershift assays, nuclear protein was preincubated with 1 mg polyclonal antibody (Ab) to HMGA1, Sp1, AP-2, or C/EBPb before addition of the probe (none: probe alone).

Binding of Sp1, AP-2, and C/EBPb to the IR promoter is reduced by PPARg

330

www.endocrinology-journals.org


Figure 5 ChIP assay. ChIP of the IR promoter gene in AP-2aproducing HepG2 cells and MCF-7 cells both induced to overexpress PPARg, either in the absence or in the presence of TZD (rosiglitazone). ChIP was done using antibodies (Ab) against either HMGA1, Sp1, AP-2, or C/EBPb.

the IR gene was considerably attenuated in the HepG2 and MCF-7 cells overexpressing PPARg, either in the absence or presence of the TZD rosiglitazone. Thus, these data suggest that overexpression and/or activation of PPARg adversely affects IR gene transcription in the absence of PPRE on the IR gene promoter. We propose that, by causing a displacement of Sp1, AP-2, and C/EBPb, PPARg may play significant molecular roles in the transcriptional activities of these factors in the context of the IR gene, both in physiology and pathology (Fig. 6).

Discussion The TZDs are insulin-sensitizing drugs that improve insulin sensitivity in insulin-resistant states such as type 2 diabetes mellitus and obesity. These drugs are high-affinity ligands for the nuclear receptor PPARg, which regulate transcription of target genes involved in the homeostasis of nutrients. In addition to the effects on lipid and glucose metabolism, many evidences have shown that PPARg and its ligands play an important role in modulating many processes, including cell proliferation and differentiation, as well as inflammation, angiogenesis, and immune system function. The IR is critical in the insulin-mediated effects on cell metabolism and cell growth. Various studies have shown that IRs are increased in most human breast cancers, and both ligand-dependent malignant www.endocrinology-journals.org

transformation and increased cell growth occur in cultured breast cells overexpressing the IR (Osborne et al. 1978, Milazzo et al. 1992, Paonessa et al. 2006). Overexpression of functional IRs has also been involved in thyroid carcinogenesis (Farid et al. 1994). The IR can exert its oncogenic potential in malignant cells via abnormal stimulation of multiple cellular signaling cascades, enhancing growth factordependent proliferation and/or by directly affecting cell metabolism. Both breast and thyroid neoplastic cells express PPARg, and PPARg agonists have been shown to inhibit proliferation in these and other cell systems (Tontonoz et al. 1994, Martelli et al. 2002, Grommes et al. 2004). In this light, we undertook to investigate whether IR expression could be affected by PPARg. Our study shows that IR gene transcription and receptor protein content were reduced in cells with forced PPARg1 overexpression, or TZD-induced PPARg activation. Although these results apparently run contrary to what might have been predicted based on the known insulin-sensitizing effects of TZD, seeming to exclude the possibility that TZDs may act as insulin sensitizers through the IR, they are compatible with the pleiotropic effects of PPARg. In this regard, the IR may be considered a new target gene that accounts for the anti-mitogenic response to PPARg and its agonists, and this is the first description of a tyrosine kinase receptor involved in PPARginduced anti-proliferative mechanisms. To unravel the molecular basis underlying the decrease in IR gene expression produced by PPARg, we performed protein–protein and DNA–protein interaction studies together with ChIP analysis, combined with transient transcription assays in the living cells expressing variable amounts of the IRs. Since no PPRE have been detected within the IR gene promoter and no reduction in the expression of HMGA1, Sp1, or C/EBPb was observed in cells after PPARg or TZD stimulation, we propose that PPARg acts as a negative regulator of IR transcription by adversely affecting binding of Sp1, C/EBPb, or AP-2 to the IR gene. In other PPARg-responsive genes that carry a PPRE consensus in their sequence, a functional cooperation between Sp1 and PPARg has been described (Krey et al. 1995). On the other hand, functional cooperation between C/EBPb (and other members of the C/EBP family of proteins) and PPARg has been reported in the nutrient signaling system during fetal development (Maloney & Rees 2005), regulation of vascular inflammation (Takata et al. 2002), and inhibition of adhesive interaction between multiple myeloma and bone marrow stromal cells (Wang et al. 2007). Herein, we show that PPARg physically interacts with 331

V Costa, D Foti et al.: Effects of PPARg and TZD-PPARg on IR expression

Figure 6 Proposed model for the action of PPARg in the context of the IR gene promoter (C2 region). (A and B) By binding to sequence-specific sites located in the C2 regulatory region of the IR gene, HMGA1, Sp1, and C/EBPb exert a positive control over the rates of transcription. The PPARg can displace Sp1 and C/EBPb from their binding sites, decreasing the rate of transcription. (C and D) AP-2 overexpression accounts for increased IR expression in both breast cancer cell lines and breast cancer tissues, in which HMGA1 acts by facilitating and stabilizing interactions between AP-2 and Sp1 in the preinitiation complex at the IR promoter (Paonessa et al. 2006). By binding to AP-2 and Sp1, PPARg may attenuate the stimulatory role of AP-2 in IR gene expression in breast carcinogenesis.

Sp1, C/EBPb, and AP-2 reducing the IR gene transcription in the absence of PPARg DNA-binding sites on the IR gene. We suggest that, in the absence of PPRE in the context of the IR promoter, this nuclear receptor may produce its adverse effects on IR gene transcription by interacting physically with these factors, thus reducing their availability to the basic transcription machinery of the IR gene. With a similar mechanism involving Sp1/PPARg protein interaction, PPARg has been shown to exert an anti-proliferative role by suppressing transcription of the thromboxane receptor, the cyclin-dependent kinase inhibitor p21, and the fibronectin genes (Sugawara et al. 2002, Hong et al. 2004, Han et al. 2005). The molecular mechanism herein described is therefore in agreement with the increasing repertoire of ‘non-canonical’ PPARg target genes that now encompasses nonPPREs containing genes (Tan et al. 2005). Many potential target genes of PPARg have been already reported, including bcl-2, b-catenin, and the PTEN tumor suppressor gene (Elstner et al. 1998, Patel et al. 2001, Mulholland et al. 2005). However, how PPARg and its agonists may induce their antiproliferative effects is not fully understood yet. Recently, non-genomic crosstalks between PPARg and cytoplasmic proteins, like extracellular signalregulated kinase (ERK) 1/2 and MAPK kinases, have been reported in cancer cells and functional importance has been given to the subcellular localization of PPARg 332

(Burgermeister & Seger 2007, Papageorgiou et al. 2007). However, the last three decades of medical research examining the molecular pathogenesis of cancers have provided compelling evidence for the universal disruption of the cell cycle in human tumors, and recent studies have demonstrated a critical interface between hormonal signaling and the cell cycle (Hilakivi-Clarke et al. 2004). In this context, mitogens like insulin, via the IR, may promote the progression through the G1 phase by inducing competence of the cyclin D1/cyclin-dependant kinase 4 (CDK4) complex. It has been previously demonstrated that the PPARg agonists inhibit cyclin D1 (Wang et al. 2001). Our findings support the conclusion that PPARg and TZDs may interfere with the hormonal control of the cell cycle, at least in part, through the inhibition of the IR. Over the last decade, PPARg has emerged as an important drug target in type 2 diabetes mellitus (Savkur & Miller 2006), and TZDs are widely used for treatment of diabetic patients. However, conflicting results on the pro-carcinogenic and anti-tumorigenic effects of TZDs in humans with diabetes can be found in the literature. For instance, whereas a populationbased report showed that TZDs were associated with reduced risk of lung cancer in patients with diabetes (Govindarajan et al. 2007), a possible association between cancer and the use of TZD has been reported later in type 2 diabetic patients (Ramos-Nino et al. 2007). Taken together, our results consistently support www.endocrinology-journals.org

Endocrine-Related Cancer (2008) 15 325–335 the conclusion that IR gene may be considered a new anticancer target for PPARg, providing further evidence for the use of TZDs as anti-proliferative agents in selected tumors overexpressing the IR.

Acknowledgements The authors would like to thank K S Chang, K Fujimori, T Funahashi, H Nishizawa and M A Lazar for providing reagents. They would also like to thank A Malta, G Ceravolo, T Rossano and G Grandinetti for secretarial help. Grant support: Ministero dell’Universita` e della Ricerca, Italy protocol 2004062059-002 and Telethon, Italy grant GGP04245 (A Brunetti). The authors declare that there is no conflict of interest that would prejudice the impartiality of this scientific work.

References Allred CD & Kilgore MW 2005 Selective activation of PPARg in breast, colon, and lung cancer cell lines. Molecular and Cellular Endocrinology 235 21–29. Araki E, Shimada F, Uzawa H, Mori M & Ebina Y 1987 Characterization of the promoter region of the human insulin receptor gene. Journal of Biological Chemistry 262 16186–16191. Barak Y, Nelson MC, Ong ES, Jones YZ, Ruiz-Lozano P, Chien KR, Koder A & Evans RM 1999 PPARg is required for placental, cardiac, and adipose tissue development. Molecular Cell 4 585–595. Berger J & Moller DE 2002 The mechanisms of action of PPARs. Annual Review of Medicine 53 409–435. Brunetti A, Maddux BA, Wong KY & Goldfine ID 1989 Muscle cell differentiation is associated with increased insulin receptor biosynthesis and messenger RNA levels. Journal of Clinical Investigation 83 192–198. Brunetti A, Foti D & Goldfine ID 1993 Identification of inique nuclear regulatory proteins for the insulin receptor gene, which appear during myocyte and adipocyte differentiation. Journal of Clinical Investigation 92 1288–1295. Brunetti A, Manfioletti G, Chiefari E, Goldfine ID & Foti D 2001 Transcriptional regulation of human insulin receptor gene by the high-mobility-group protein HMGI-Y. FASEB Journal 15 492–500. Burgermeister E & Seger R 2007 MAPK kinases as nucleocytoplasmic shuttles for PPARgamma. Cell Cycle 6 1539–1548. Chou FS, Wng PS, Kulp S & Pinzone JJ 2007 Effects of thiazolidinediones on differentiation, proliferation, and apoptosis. Molecular Cancer Research 5 523–530. Desvergne B & Wahli W 1999 Peroxisome proliferatoractivated receptors: nuclear control of metabolism. Endocrine Reviews 20 649–688.


Elstner E, Mu¨ller C, Koshizuka K, Williamson EA, Park D, Asou H, Shintaku P, Said JW, Heber D & Koeffler HP 1998 Ligands for peroxisome proliferator-activated receptor gamma and retinoic acid inhibit growth and induce apoptosis of human breast cancer cells in vitro and in BNX mice. PNAS 95 8806–8811. Farid NR, Shi Y & Zou M 1994 Molecular basis of thyroid cancer. Endocrine Reviews 15 202–232. Foti D, Iuliano R, Chiefari E & Brunetti A 2003 A nucleoprotein complex containing Sp1, C/EBPb, and HMGI-Y controls human insulin receptor gene transcription. Molecular and Cellular Biology 23 2720–2732. Foti D, Chiefari E, Fedele M, Iuliano R, Brunetti L, Paonessa F, Manfioletti G, Barbetti F, Brunetti A, Croce CM et al. 2005 Lack of the architectural factor HMGA1 causes severe insulin resistance and diabetes in humans and mice. Nature Medicine 11 765–773. Goldfine ID 1987 The insulin receptor: molecular biology and transmembrane signalling. Endocrine Reviews 8 235–255. Govindarajan R, Ratnasinghe L, Simmons DL, Siegel ER, Midathada MV, Kim L, Kim PJ, Owens RJ & Lang NP 2007 Thiazolidinediones and the risk of lung, prostate, and colon cancer in patients with diabetes. Journal of Clinical Oncology 25 1476–1481. Grommes C, Landreth GE & Heneka MT 2004 Antineoplastic effects of peroxisome proliferator-activated receptor agonists. Lancet Oncology 5 419–429. Han SW, Ritzenthaler JD, Rivera HN & Roman J 2005 Peroxisome proliferator-activated receptor-gamma ligands suppress fibronectin gene expression in human lung carcinoma cells: involvement of both CRE and Sp1. American Journal of Physiology. Lung Cellular and Molecular Physiology 289 L419–L428. Hilakivi-Clarke L, Wang C, Kalil M, Riggins R & Pestell RG 2004 Nutritional modulation of the cell cycle and breast cancer. Endocrine-Related Cancer 11 603–622. Hong J, Samudio I, Liu S, Abdelrahim M & Safe S 2004 Peroxisome proliferator-activated receptor g-dependent activation of p21 in Panc-28 pancreatic cancer cells involves Sp1 and Sp4 proteins. Endocrinology 145 5774–5785. Keller H, Dreyer C, Medin J, Mahfoudi A, Ozato K & Wahli W 1993a Fatty acids and retinoids control lipid metabolism through the activation of peroxisome proliferator-activated receptors heterodimers. PNAS 90 2160–2164. Keller H, Mahfoudi A, Dreyer C, Hihi AK, Medin J, Ozato K & Wahli W 1993b Peroxisome proliferator-activated receptors and lipid metabolism. Annals of the New York Academy of Sciences 684 157–173. Kersten S, Desvergne B & Wahli W 2000 Roles of PPARs in health and disease. Nature 405 421–424.

333

V Costa, D Foti et al.: Effects of PPARg and TZD-PPARg on IR expression Kliewer SA, Xu HE, Lambert MH & Willson TM 2001 Peroxisome proliferator-activated receptors: from genes to physiology. Recent Progress in Hormone Research 56 239–263. Krey G, Mahfoudi A & Wahli W 1995 Functional interactions of peroxisome proliferator-activated receptor, retinoid-X receptor, and Sp1 in the transcriptional regulation of the acyl-coenzyme-A oxidase promoter. Molecular Endocrinology 9 219–231. Lehmann JM, Moore LB, Smith-Oliver TA, Wilkinson WO, Willson TM & Kliewer SA 1995 An antidiabetic thiazolidinedione is a high affinity ligand for peroxisome proliferator-activated receptorg (PPARg). Journal of Biological Chemistry 270 12953–12956. Maloney CA & Rees WD 2005 Gene–nutrient interactions during fetal development. Reproduction 130 401–410. Mamula PW, McDonald AR, Brunetti A, Okabayashi Y, Wong KY, Maddux BA, Logsdon C & Goldfine ID 1990 Regulating insulin-receptor-gene expression by differentiation and hormones. Diabetes Care 13 288–301. Martelli ML, Iuliano R, Le Pera I, Sama` I, Monaco C, Cammarota S, Kroll T, Chiariotti L, Santoro M & Fusco A 2002 Inhibitory effects of peroxisome proliferatoractivated receptor gamma on thyroid carcinoma cell growth. Journal of Clinical Endocrinology and Metabolism 87 4728–4735. Milazzo G, Giorgino F, Damante G, Sung C, Stampfer MR, Vigneri R, Goldfine ID & Belfiore A 1992 Insulin receptor expression and function in human breast cancer cell lines. Cancer Research 52 3924–3930. Mueller E, Sarraf P, Tontonoz P, Evans RM, Martin KJ, Zhang M, Fletcher C, Singer S & Spiegelman BM 1998 Terminal differentiation of human breast cancer through PPARg. Molecular Cell 1 465–470. Mulholland DJ, Dedhar S, Coetzee GA & Nelson CC 2005 Interaction of nuclear receptors with the Wnt/betacatenin/Tcf signaling axis: Wnt you like to know? Endocrine Reviews 26 898–915. Nishizawa H, Yamagata K, Shimomura I, Takahashi M, Kuriyama H, Kishida K, Hotta K, Nagaretani H, Maeda N, Matsuda M et al. 2002 Small heterodimer partner, an orphan nuclear receptor, augments peroxisome proliferator-activated receptor gamma transactivation. Journal of Biological Chemistry 277 1586–1592. Nunez NP, Liu H & Meadows GG 2005 PPAR-gamma ligands and amino acid deprivation promote apoptosis of melanoma, prostate, and breast cancer cells. Cancer Letters 236 133–141. Osborne CK, Monaco ME, Lippman ME & Kahn CR 1978 Correlation among insulin binding, degradation, and biological activity in human breast cancer cells in longterm tissue culture. Cancer Research 38 94–102. Padilla J, Kaur K, Harris SG & Phipps RP 2000 PPARgamma-mediated regulation of normal and malignant B lineage cells. Annals of the New York Academy of Sciences 905 97–109. Paonessa F, Foti D, Costa V, Chiefari E, Brunetti G, Leone F, Luciano F, Wu F, Lee AS, Gulletta E et al. 2006 Activator

334

protein-2 overexpression accounts for increased insulin receptor expression in human breast cancer. Cancer Research 66 5085–5093. Papageorgiou E, Pitulis N, Msaouel P, Lembessis P & Koutsilieris H 2007 The non-genomic crosstalk between PPARg ligands and ERK 1/2 in cancer cell lines. Expert Opinion on Therapeutic Targets 11 1071–1085. Patel L, Pass I, Coxon P, Downes CP, Smith SA & Macphee CH 2001 Tumor suppressor and anti-inflammatory actions of PPARgamma agonista are mediated via upregulation of PTEN. Current Biology 11 764–768. Ramos-Nino ME, MacLean CD & Littenberg B 2007 Association between cancer prevalence and use of thiazolidinediones: results from the Vermont Diabetes Information System. BMC Medicine 5 17–24. Rosen ED, Sarraf P, Troy AE, Bradwin G, Moore K, Milstone DS, Spiegelman BM & Mortensen RM 1999 PPARg is required for the differentiation of adipose tissue in vivo and in vitro. Molecular Cell 4 611–617. Saltiel AR & Olefsky JM 1996 Thiazolidinediones in the treatment of insulin resistance and tipe II diabetes. Diabetes 45 1661–1669. Savkur RS & Miller AR 2006 Investigational PPAR-g agonists for the treatment of Type 2 diabetes. Expert Opinion on Investigational Drugs 15 763–778. Seino S, Seino M, Nishi S & Bell GI 1989 Structure of the human insulin receptor gene and characterization of its promoter. PNAS 86 114–118. Sugawara A, Uruno A, Kudo M, Ikeda Y, Sato K, Taniyama Y, Ito S & Takeuchi K 2002 Transcription suppression of thromboxane receptor gene by peroxisome proliferatoractivated receptor,-g via an interaction with Sp1 in vascular smooth muscle cells. Journal of Biological Chemistry 277 9676–9683. Takata Y, Kitami Y, Yang Z-H, Nakamura M, Okura T & Hiwada H 2002 Vascular inflammation is negatively autoregulated by interaction between CCAAT/enhancerbinding protein d and peroxisome proliferator-activated receptor. Circulation Research 91 427–433. Tan NS, Michalik L, Desvergne B & Wahli W 2005 Multiple expression control mechanisms of peroxisome proliferator-activated receptors and their target genes. Journal of Steroid Biochemistry and Molecular Biology 93 99–105. Tontonoz PM, Hu E & Spiegelman BN 1994 Stimulation of adipogenesis in fibroblasts by PPAR gamma 2, a lipidactivated transcription factor. Cell 79 1147–1156. Ullrich A, Bell JR, Chen EY, Herrera R, Petruzzelli LM, Dull TJ, Gray A, Coussens L, Liao YC, Tsubokawa M et al. 1985 Human insulin receptor and its relationship to the tyrosine kinase family of oncogenes. Nature 313 756–761. Vidal-Puig AJ, Considine RV, Jimenezlinan M, Werman A, Pories WJ, Caro JF & Flier JS 1997 Peroxisome proliferator-activated receptor gene expression in human tissues. Journal of Clinical Investigation 99 2416–2422.


Endocrine-Related Cancer (2008) 15 325–335 Wang C, Fu M, D’Amico M, Albanese C, Zhou JN, Brownlee M, Lisanti MP, Chatterjee VK, Lazar MA & Pestell RG 2001 Inhibition of cellular proliferation through IkB kinase-indipendent and peroxisome proliferator-activated receptor g-dependent repression of cyclin D1. Molecular and Cellular Biology 21 3057–3070.


Wang LH, Yang XY, Zhang X & Farrar WL 2007 Inhibition of adhesive interaction between multiple myeloma and bone marrow stromal cells by PPARg crostalk with NF-kB and C/EBPb. Blood 110 4373–4384. White MF & Kahn CR 1994 The insulin signaling system. Journal of Biological Chemistry 269 1–4.

335