<<

International Journal of Obesity (2002) 26, 920–927 ß 2002 Nature Publishing Group All rights reserved 0307–0565/02 $25.00 www.nature.com/ijo PAPER Relationship between peroxisome proliferator- activated gamma and receptor alpha expression in obese human adipose tissue

A Redonnet1, S Bonilla1, C Noe¨l-Suberville1, V Pallet1, H Dabadie2, H Gin2 and P Higueret1*

1Laboratoire de Nutrition et Signalisation Cellulaire (USC-INRA), ISTAB, Universite´ Bordeaux I, Talence, France; and 2Service de Nutrition-Diabe´tologie, Hoˆpital du Haut Le´veˆque, Pessac, France

OBJECTIVE: To investigate in human adipose tissue a possible relationship between per oxisome proliferator-activated receptor gamma (PPARg) and retinoic acid receptor alpha (RARa) gene expression, two involved in the control of adipocyte differentiation. SUBJECTS: Ten lean control women (age 31 – 60 y, body mass index (BMI) 18 – 24.7 kg=m2) and an obese group of 15 women (age 27 – 62 y, BMI 30 – 57.5 kg=m2), of whom 10 subjects were in weight-gain phase and five were in weight-loss phase. MEASUREMENTS: We assessed the relative PPARg and RARa mRNA levels in subcutaneous abdominal adipose tissue using a real- time PCR method. RESULTS: PPARg mRNA level were significantly increased ( þ 91%; P < 0.01) in obese women compared to lean control women. In the obese group, we observed a PPARg mRNA level 42% lower in weight-loss obese than in weight-gain obese subjects. We obtained a positive correlation (r ¼ 0.56; P < 0.01) between PPARg mRNA level and the BMI of all subjects. Relative mRNA abundance level of RARa in subcutaneous adipose tissue of obese subjects is significantly lower than in control subjects ( 7 56%, P < 0.01), and a negative correlation was found between PPARg and RARa mRNA levels in subcutaneous adipose tissue of subjects study (r ¼ 7 0.75; P < 0.01). CONCLUSION: Our findings suggest that obesity is associated with an inverse relationship between PPARg and RARa expressions in human subcutaneous adipose tissue. Modulations of profile could be an important event in the body’s early adaptive mechanisms promoting adipose tissue plasticity and leading to the onset of obesity. International Journal of Obesity (2002) 26, 920 – 927. doi:10.1038/sj.ijo.0802025

Keywords: obesity; human adipose tissue; PPARg; RAR; gene expression

Introduction Among these factors, peroxisome proliferator-activated In humans, adipose tissue is dynamic with the lifelong receptor gamma (PPARg), a nuclear receptor which is highly potential for hyperplasia through preadipocyte replication expressed in adipose tissue, is known to be a dominant and differentiation, leading to the development of obesity. activator of fat cell differentiation. Furthermore, its expres- This process requires the continuous differentiation of new sion is induced very early in the adipocyte differentiation.4 adipocytes1 and involves a complex regulatory pathway con- Activated PPARg forms a heterodimer with retinoid X recep- trolled by the sequential and coordinate expression of spe- tors (RXRs) and binds to PPAR response elements occurring in cific regulatory genes and several factors.2,3 the promoter region of adipose specific genes involved in adipose differentiation, lipid storage and metabolism.4 White adipose tissue (WAT) plays a key role in the storage *Correspondence: P Higueret, Laboratoire de Nutrition et Signalisation and metabolism of lipids and is known as a potential target Cellulaire, ISTAB, Universite´ Bordeaux I, avenue des Faculte´s, 33405 organ for retinoic acid (RA) action. Indeed, recent studies have Talence cedex, France. indicated that adipose tissue may play an important action in E-mail: [email protected] 5 Received 11 September 2001; revised 11 February 2002; retinoid uptake, storage, mobilization and transport. More- accepted 14 February 2002 over, RA receptors, RAR and RXR, which are known to mediate PPARg and RARa expression in human adipose tissue A Redonnet et al 921 the RA action, are expressed in adipose tissue.6 It has been was washed in cold saline and frozen in liquid nitrogen established that activation of RA receptors, and in particular and stored at 7 80C until analysis. the RARa isoform, is required for the inhibition of adipocyte differentiation.7,8 Furthermore, recent reports have suggested that convergence of the RA and PPAR signaling pathways Total RNA preparation regulates the fate of the cultured preadipocytes in acquisition About 60 mg (wet weight) of frozen tissue samples were of the adipocyte phenotype.8,9 directly homogenized in 1 ml TRIzol reagent (Invitrogen, Given the key role that RAR and PPAR play as master France) and total RNA was extracted following the manufac- regulator genes for differentiation and pro- or antiprolifera- turer’s suggested protocol for small quantities of tissue. tion in many tissues, increased understanding of these cri- Purified RNA was quantitated and assessed for purity by UV tical factors and how they are regulated will provide insights spectrophotometry. Average yield of total RNA was into adipose tissue development. To illustrate this purpose, it 7.2Æ 0.7 mg=100 mg of adipose tissue and was not signifi- is suggested that modulation of adipose tissue cellularity by cantly different in tissue from control and obese subjects. food intake involves a potential dysregulation of the genetic development program. We obtained evidence in a previous study in which we examined the effect of exposure to an Reverse transcription obesity-inducing diet on the pattern of expression of the cDNA was synthesized with Superscript II reverse transcrip- RAR, RXR, TR (triiodothyronine receptor) and PPAR nuclear tase (Invitrogen, France) according to the protocol recom- receptors in rat WAT. Thus, data obtained for WAT of rats fed mended by the manufacturer with minor modifications. on a fat-rich high-energy diet showed that a significantly Briefly, 1 mg of total RNA was incubated at 70C for 5 min decreased expression of RARa and TR was concomitant with and then placed on ice before addition of RT reaction an increased expression of PPARg.10 Since few human studies reagents with a specific reverse primer (120 ng) in a final have been performed, the present work was undertaken to volume of 20 ml. The RT reaction was performed at 42C for obtain an insight into possible modulation of the expression 60 min. Parallel reactions for each RNA samples were run in of these different nuclear receptors in adipose tissue of obese the absence of RT to assess the degree of any contaminating subjects. genomic DNA.

Methods Analysis of gene expression using real-time PCR Subjects and tissues PCR was carried out using a LightCycler2 system (Roche This study included 25 unrelated white female subjects. We Diagnostics, Mannheim, Germany), which combines the studied 10 lean women (mean age 41.8Æ 3.4 y, body mass processes of amplification and detection (by fluorescence) index (BMI) 23.3Æ 2.8 kg=m2) and a group of 15 obese of a PCR product, enabling on-line and real-time detection. women (mean age 47.6Æ 2.3 y, BMI 40.9Æ 1.9 kg=m2). The For detection of target gene amplification products, Light- obese group was composed of 10 subjects in the weight-gain Cycler DNA Master SYBR Green I was used as described by phase (defined as weight gain  4% or more in body weight the manufacturer. PCR reactions were performed in micro- over the previous 6 months) and five subjects in the weight- capillary tubes in a final volume of 20 ml containing 1X LC- loss phase (defined as weight loss  4% or more in body DNA Master SYBR Green I mix, 4 mM of MgCl2, 0.5 mMof weight over the previous 6 months). These latter subjects each primer and 2 ml of cDNA. The amplification conditions received nutritional advice in a medical context and were were 95C for 8 min to activate the polymerase, followed by regularly followed. Dietary intake was assessed by food 45 cycles of denaturation at 95C for 6 s, annealing at 64C frequency. Among the 10 weight-gain obese patients, six for 6 s, and elongation at 72C for 10 s. After each elongation were impaired fasting glucose tolerant (IFGT). None of phase the fluorescence of SYBR Green I (a double-stranded them received insulin, thiazolidinedione or other hypogly- DNA binding dye) was measured and increasing amounts of cemic drugs. Moreover, no subject in the entire study was on PCR products can be monitored from cycle to cycle. The medications known to affect adipose tissue mass or lipopro- forward and reverse primer sequences are shown in Table 1. tein metabolism. All participants gave their written consent For each primer pair used, melting curve analysis showed a after being informed of the nature of the study. The experi- single melting peak after amplification, indicating specific mental protocol was approved by the Ethics Committee of product. the Bordeaux University Hospital and performed according Quantification data were analyzed using the LightCycler to French legislation (Huriet law). analysis software version 3.5 (Roche Diagnostics, Mannheim, After an overnight fast, blood was drawn and fasting Germany). In this analysis, the background fluorescence is serum insulin, glucose, triglycerides, cholesterol were mea- removed by setting a noise band. The log – linear portion of sured in blood samples. the standard amplification curve is identified and the cross- Abdominal subcutaneous adipose biopsies were taken ing point (Cp) is the intersection of the best-fit line through under local anesthesia by needle liposuction. The tissue the log – linear region and the noise band. The standard

International Journal of Obesity PPARg and RARa expression in human adipose tissue A Redonnet et al 922 2 Table 1 Primers used for LightCycler real-time PCR

PCR primer pair Reference Sequence Position Product length (bp)

0 0 F: 5 -GAGCTACGAGCTGCCTGACG-3 787 – 806 b-Actin 11 0 0 R: 5 -GTAGTTTCGTGGATGCCACAG-3 886 – 906 120 0 0 F: 5 -GGGTGGACATGCTGCAGGCGCC-3 1142 – 1163 RARa 12 0 0 R: 5 -CCTTGGCGCTGATGCTTCGCAG-3 1252 – 1273 132 0 0 F: 5 -CCGAAGCACTGTCCAGACCGAGAAC-3 334 – 358 TRb 13 0 0 R: 5 -TCAAAGACTTCCAAGAAGAGAGGC-3 399 – 423 90 0 0 F: 5 -CACAAGAACAGATCCAGTGGTTGCAG-3 339 – 364 PPARg 14 0 0 R: 5 -AATAATAAGGTGGAGATGCAGGCTCC-3 414 – 439 101

Sequences are shown for forward (F) and reverse (R) primers as well as size of amplicon.

curve is a plot of the Cp vs the amount of initial cDNA used for amplification. Standard curves were generated from 4- fold serial dilutions of target and housekeeping (b-actin) cDNA preparation (Figure 1). The relationship between the Cp and the initial amount of cDNA was found to be linear. The correlation coefficient (r) was 1 and PCR amplification efficiencies of the target and the housekeeping gene were similar and close to 100%. These standard curves are used to estimate the concentration of both the target and the b-actin gene in each sample. Then, the results were normalized by the ratio of the relative concentration of target to that of b- actin in the same sample.

Statistical analysis All results are presented as meanÆ s.e.m. Statistical signifi- cance was assessed using student’s t-test and correlation using Pearson’s coefficient. The level of significance was set at P < 0.05.

Results We analyzed dietary intake of weight gain (8456Æ 1150 kJ=day or 2021Æ 275 kcal=day) and weight-loss obese patients (6753Æ 1255 kJ=day or 1614Æ 300 kcal=day). Energy intake from diet was not statistically different between the two obese subgroups as compared to lean patients (7514Æ 255 kJ=day or 1796Æ 255 kcal=day). However, the declared energy intake ranged from 2326 to 24932 kJ=day (556 – 5959 kcal=day) in obese patients. Fat dietary energy did not differ between obese in weight-gain phase (40.8%), obese in weight-loss phase (39.1%) and lean patients (40.7%). The relevant clinical and metabolic characteristics of the patients are shown in Table 2. Fasting plasma insulin and triglycerides were higher in obese than lean control subjects. Figure 1 (A) Amplification curves of b-actin in 4-fold serial dilution of In obese subjects, the calculated insulin resistance index cDNA prepared from total RNA of adipose tissue. (B) Standard curve for (HOMA)15 values of IFGT subjects ranged from 4 to 6.67, b-actin. The figure shows the initial cDNA concentration plotted vs the crossing point for the detection of significant fluorescence. The slope was and ranged from 1.62 to 4.89 in non-IFGT subjects (P < 0.05). 7 3.20 and the regression coefficient was 1. There was a significant increase of plasma triglycerides in

International Journal of Obesity PPARg and RARa expression in human adipose tissue A Redonnet et al 923 IFGT obese subjects 2.4Æ 0.4 vs 1.3Æ 0.2 g=l in non-IFGT the weight-loss obese patients had a 42% lower PPARg level obese subjects (P < 0.05). compared with the weight-gain obese group (P < 0.05). We The relative adipose abundance of PPARg mRNA was obtained a positive correlation (r ¼ 0.51, P < 0.01) between significantly increase ( þ 91%, P < 0.01) in obese women PPARg mRNA level and the BMI in all subjects (Figure 3). compared to lean women. To determine the effect of obesity However, this positive correlation was due to the control status on expression of PPARg, RARa and TRb in subcuta- group. No correlation between the level of PPARg mRNA in neous abdominal adipose tissue, we examined separately the subcutaneous abdominal adipose tissue and BMI was found level of nuclear receptors mRNA in each subgroup of obese in weight-gain or weight-loss obese patients. It is important subjects. Figure 2 shows that PPARg level was 125% higher in to note that data obtained in IFTG obese patients were not the weight-gain subjects (P < 0.01) and only 30% higher in different from those obtained in non-IFTG obese patients. the weight-loss subjects relative to the lean controls. Thus Associated with body weight changes, a continuous decrease of plasma triglyceride concentrations both in weight-gain and weight — loss subjects and control subjects Table 2 Characteristics of study subjects was observed (1.9Æ 0.3, 1.4Æ 0.2 and 0.9Æ 0.1 g=l, respec- Obese tively). However, this reduction is significant only between weight-gain obese and control subjects (P < 0.05). In addi- Lean Total group Weight loss Weight gain g (n ¼ 10) (n ¼ 15) (n ¼ 5) (n ¼ 10) tion, we found a positive correlation between PPAR mRNA level and plasma triglycerides in all subjects (r ¼ 0.45, Age (y) 41.8Æ 3.4 47.6Æ 2.3 45.8Æ 1.7 48.5Æ 3.4 P < 0.05). BMI (kg=m2) 23.3Æ 2.8 40.9Æ 1.9* 39.2Æ 1.8* 41.8Æ 2.8* We also examined the RARa mRNA level in relation to the Fasting Glucose (mmol=l) 5.0Æ 0.2 5.5Æ 0.2 5.5Æ 0.3 5.5Æ 1 obesity status of subjects. Due to low and variable yield in Insulin (mU=l) 6.0Æ 0.1 16.3Æ 1.6* 16.6Æ 3* 16.2Æ 2* total RNA recovery in human adipose tissue, we could not Triglycerides (mmol=l) 0.9Æ 0.1 1.8Æ 0.2* 1.4Æ 0.2 1.9Æ 0.3* measure the mRNA level of RARa in all obese and control Total cholesterol (mmol=l) 4.6Æ 0.2 6.5Æ 0.7 7.7Æ 1.8 5.6Æ 0.8 subjects. Relative mRNA abundance level of RARa in sub- HOMA 1.3Æ 0.2 4.0Æ 0.5* 4.0Æ 0.8* 4.0Æ 0.9* cutaneous abdominal adipose tissue of obese subjects is Data are meanÆ s.e.m. Statistical significance, *P < 0.05 vs lean control. significantly lower than in control subjects ( 7 56%,

Figure 2 Levels of PPARg RARa and TRb mRNA in human subcutaneous adipose tissue. Values are meanÆ s.e.m. of weight-gain and weight-losing obese and are represented in percentage of variation compared to the levels of lean controls. *P < 0.05 and **P < 0.01 vs the control subjects; #P < 0.05 vs the weight-gain obese patients.

International Journal of Obesity PPARg and RARa expression in human adipose tissue A Redonnet et al 924 P < 0.01). Again, no IFGT-associated differences were found not reach significance. However, given the low number of in the patients. In details in Figure 2, in weight-gain obese RARa quantification in weight-loss obese group (n ¼ 3), these our results showed a clear decrease ( 7 60%, P < 0.01) of results have to be considered as partial data. RARa mRNA abundance level. Compared with weight-gain We extended our study to triiodothyronine nuclear recep- obese and control patients, weight-losing obese subjects tor (TR) expression because it has been shown that obese exhibited an intermediate value of RARa mRNA, but it did animals presented a tissue triiodothyronine resistance.16

Figure 3 Relationship between PPARg mRNA level and BMI of the study population. Lean control (squares), n ¼ 10; weight-loss obese (triangles), n ¼ 10, weight-gain obese (circle), n ¼ 5.

Figure 4 Relationship between PPARg and RARa mRNA levels in adipose tissue of the study population. Lean control (squares), n ¼ 7, and obese subjects (triangles and circles), n ¼ 12.

International Journal of Obesity PPARg and RARa expression in human adipose tissue A Redonnet et al 925 Moreover, various links between TR, RAR and PPAR signaling lean subjects.22 Our results show that increased PPARg mRNA pathways were described.17,18 Similar results were obtained level in subcutaneous adipose tissue is associated with obe- with the mRNA level of TRb in adipose tissue of the obese sity. Furthermore, it seems to be nutritionally regulated in (n ¼ 15) and lean control (n ¼ 10) subjects. However, the obese subjects as we have different results in obese patients decreased expression of TRb in all obese subjects was less losing or gaining weight. In our study, PPARg mRNA expres- marked ( 7 39%, P < 0.05) compared with control subjects sion was not correlated to BMI and triglycerides in all obese than decreased expression of RARa in the same individuals. subjects. In addition, we did not observe any correlation No difference was observed between the two subgroups of between fasting plasma insulinemia and PPARg mRNA level, obese (Figure 2). although we measured a wide range of fasting plasma insulin As shown in Figure 4, a significant negative correlation concentrations in obese subjects (7 – 28 mU=ml). Taken was found between the mRNA levels of PPARg and RARa in together, these observations argue for a relationship between adipose tissue of the subjects (r ¼ 7 0.75, P < 0.01). On the PPARg mRNA level in subcutaneous adipose tissue and obe- other hand, no correlation between mRNA levels of TRb and sity state independent of the degree of obesity or the blood PPARg was found in adipose tissue of the subjects (Figure 5). glucose regulation. Nevertheless, this finding must be taken with caution because we only determined the abundance of total PPARg Discussion mRNA without assessing the specific expression of gamma 1 It is important to human obesity that the regulatory system and 2 isoforms of PPAR in adipose tissue. Moreover, for the differentiation of adipocytes is well defined. It is differences in methodology for mRNA quantification added established that PPARg is a key that to a heterogeneous, but representative, population of obese trigger the entire of adipogenesis.3,19 Its role in the control used in the present study could explain the discrepancies in of body weight in animal models has been seen previously,4 results between the previous published studies and our own. and more recently suggested in humans.20 Thus it is crucial In obese patients, we reported for the first time a to identify the factors that are able to modulate PPARg mRNA decreased expression of RARa mRNA in adipose tissue com- expression at early stage of adipogenesis, and to determine pared with lean patients. Twenty years ago, RA was first their regulation in relation to obesity. identified as a potent inhibitor of adipocyte differentiation.23 Previous studies with conflicting findings concerning the After this it was shown that differentiation of preadipocyte association of PPARg with obesity have been reported. It has lines was prevented at an early stage by RA.24,25 To date, the been shown that only PPARg2 was increased and correlated mechanism for this effect of RA remains unclear, but this with BMI in subcutaneous adipose tissue of obese subjects,21 effect seems to be mediated by RARa.7 whereas other authors reported no alterations in PPARg1 or g2 The negative correlation between RARa and PPARg mRNA expression in adipose tissue of obese subjects compared to level found in obese and nonobese subjects added to the

Figure 5 Relationship between PPARg and TRb mRNA levels in adipose tissue of the study population. Lean control (squares), n ¼ 10, and obese subjects (triangles and circles), n ¼ 15.

International Journal of Obesity PPARg and RARa expression in human adipose tissue A Redonnet et al 926 same negative association found between the weight-gain ported by the Conseil Re´gional d’Aquitaine (Poˆle Aquitaine and the weight-loss subjects suggest that an adaptive Agro-Alimentaire and Nutrition), the French Ministe`re de response of the regulation of these two nuclear receptors is l’Enseignement Supe´rieur et de la Recherche, ‘Le Comite´ de related to the obesity state. This observation is consistent la Gironde de la Ligue Contre le Cancer’ and ‘l’Institut with results suggesting that a threshold level of RAR is National de la Recherche Agronomique’. required for retinoic acid inhibition of adipogenesis.8 In the current study, no relation was found between TR and PPARg or RAR mRNA level in adipose tissue, despite a lower expression of TR mRNA in obese subjects. This result References argue in favor of a preferential and dynamic cross-talk 1 Flier JS. The adipocyte: storage depot or node on the energy information superhighway. Cell 1995; 80:15– 18. between PPARg and retinoid signaling pathways associated 2 Fajas L, Fruchart JC, Auwerx J. Transcriptional control of adipo- with the regulation of adipose tissue plasticity. genesis. Curr Opin Cell Biol 1998; 10: 165 – 173. Obesity is often associated with a disturbance in energy 3 Spiegelman B, Flier J. Adipogenesis and obesity: rounding out the big picture. Cell 1996; 87: 1188 – 1193. balance in part due to an excessive food intake. In this study, 4 Brun RP, Kim JB, Hu E, Spiegelman BM. Peroxisome proliferator- we did not observe a significant difference in declared food activated receptor gamma and the control of adipogenesis. Curr intake between the three groups because the range of obese Opin Lipidol 1997; 8: 212 – 218. energy intake was too large. However, previous work in 5 Zovich DC, Orologa A, Okuno M, Kong LWY, Talmage DA, Piantedosi R, Goodman DS, Blaner WS. Differentiation-depen- rodents has reported that PPARg is regulated in vivo by diet dent expression of retinoid-binding in BFC-1 beta adipo- 26 rich in fatty acids. Similarly, in a previous study, we cytes. J Biol Chem 1992; 267: 13884 – 13889. observed an up-regulation of PPARg mRNA level in WAT of 6 Haq R, Chytil F. Expression of nuclear retinoic acid receptors in high-fat fed animals with a concomitant decreased level of rat adipose tissue. Biochem Biophys Res Commun 1991; 176: 1539 – 1544. a a RAR and TR mRNA level, whereas RXR mRNA level was 7 Kamei Y, Kawada T, Mizukami J, Sugimoto E. The prevention of 10 unchanged. These data suggest that nutrition can affect the adipose differentiation of 3T3-L1 cells caused by retinoic acid is retinoid and PPAR signaling pathways in WAT. Significance elicited through retinoid acid receptor alpha. Life Sci 1994; 55: as well as mechanisms of these phenomenons remain 307 – 312. 8 Xue JC, Schwarz EJ, Chawla A, Lazar MA. Distinct stages in unknown. However, it is possible that modulation in expres- adipogenesis revealed by retinoid inhibition of differentiation sion of RAR and PPAR (and maybe TR) occurring in these after induction of PPARg. Mol Cell Biol 1996; 16: 1567 – 1575. tissues could be an important event in the early body 9 Chawla A, Lazar MA. Peroxisome proliferator and retinoid signal- adaptive mechanisms leading to the hyperplasy and the ing pathways co-regulate preadipocyte phenotype and survival. Proc Natl Acad Sci USA 1994; 91: 1786 – 1790. hypertrophy of adipose tissue. 10 Redonnet A, Groubet R, Noe¨l-Suberville C, Bonilla S, Martinez A, More generally, excessive food intake (ie rich in lipids) Higueret P. Exposure to an obesity-inducing diet early affects the might induce an allostatic state because of disturbances in pattern of expression of peroxisome proliferator, retinoic acid, fatty acids and retinoid nuclear receptor signaling pathways and triiodothyronine nuclear receptors in the rat. Metabolism 2001; 50: 1161 – 1167. in different target tissues. For instance, RAR modulation will 11 Vandekerckhove J, Weber K. Mammalian cytoplasmic actins are be one event promoting the adipose tissue plasticity in the products of at least two genes and differ in primary structure response to a high energy intake, and leading to the devel- in at least 25 identified positions from skeletal muscle actins. Proc opment of obesity. On the other hand, such as abnormal Natl Acad Sci USA 1978; 75: 1103 – 1110. 12 Giguere V, Ong ES, Segui P, Evans RM. Identification of a receptor expression of the RAR could be related to the occurrence of for the retinoic acid. Nature 1987; 330: 624 – 629. certain types of cancer. Thus, in a colon carcinogenesis- 13 Weinberger C, Thompson CC, Ong ES, Lebo R, Gruol DJ, Evans induced animal model, we observed a decreased expression RM. The c-erb-A gene encodes a thyroid hormone. Nature 1986; of RARb mRNA in colon of rats fed a high-fat diet (unpub- 324: 641 – 646. 14 Mukherjee R, Jow L, Croston GE, Paterniti JR. Identification, lished observations). This information could be related to characterization, and tissue distribution of human peroxisome the epidemiological data showing that obese patients were at proliferator-activated receptor (PPAR) isoforms PPARgamma2 higher risk of colon cancer.27,28 versus PPARgamma1 and activation with retinoid X In conclusion, disturbances in the pattern of expression of receptor agonists and antagonists. J Biol Chem 1997; 272: 8071 – 8076. these nuclear receptors could favor the adverse health con- 15 Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, sequences associated with increasing body weight. A better Turner RC. Homeostasis model assessment : insulin resistance knowledge of the nuclear receptor signaling pathway and b-cell function from fasting plasma glucose and insulin interactions will provide new insights into adipose tissue concentrations in man. Diabetologia 1985; 28: 412 – 419. 16 Oh SS, Kaplan ML. Early treatment of obese (ob=ob) mice plasticity in the onset of obesity. with triiodothyronine increases oxidative metabolism in muscle but not in brown adipose tissue or liver. J Nutr 1995; 125: 112 – 124. 17 Lehmann JM, Zhang XK, Graupner G, Lee MO, Hermann T, Acknowledgements Hoffmann B, Pfahl M. Formation of homodimers leads to repression of T3 response: hormonal We are grateful to C Boue´ and N Combe from ITERG (Pessac, cross-talk by ligand-induced squelching. Mol Cell Biol 1993; France) for their assistance and advice. This work was sup- 13: 7698 – 7707.

International Journal of Obesity PPARg and RARa expression in human adipose tissue A Redonnet et al 927 18 Chu R, Madison LD, Lin Y, Kopp P, Rao MS, Jameson JL, Reddy JK. 23 Murray T, Russell TR. Inhibition of adipose conversion of murine Thyroid hormone (T3) inhibits ciprofibrate-induced transcription 3T3-L1 cells by retinoic acid. J Supramol Struc 1980; 23:255– 266. of genes encoding b-oxidation enzymes: cross-talk between per- 24 Sato M, Hiragun A, Mitsui H. Preadipocytes posses cellular reti- oxisome proliferator and T3 signaling pathways. Proc Natl Acad Sci noid binding and their differentiation is inhibited by USA 1995; 92: 11593 – 11597. retinoids. Biochem Biophys Res Commun 1980; 95: 1839 – 1845. 19 Tontonoz P, Hu E, Spiegelman BM. Stimulation of adipogenesis in 25 Stone RL, Bernlohr DA. The molecular basis for inhibition of fibroblasts by PPARg2, a lipid activated transcription factor. Cell adipose conversion of murine 3T3-L1 cells by retinoic acid. 1994; 79: 1147 – 1156. Differentiation 1990; 45: 119 – 127. 20 Ribot J, Rantala M, Kesa¨niemi YA, Palou A, Savolainen MJ. Weight 26 Vidal-Puig AJ, Jimenez-Lin˜an M, Lowell BB, Hamann A, Hu E, loss reduces expression of SREBP1c=ADD1 and PPARg2 in adipose Spiegelman BM, Flier JS, Moller DE. Regulation of PPARg gene tissue of obese women. Pflu¨gers Arch Eur J Physiol 2001; 441: 498 – expression by nutrition and obesity in rodents. J Clin Invest 1996; 505. 97: 2553 – 2561. 21 Vidal-Puig AJ, Considine RV, Jimenez-Lin˜an M, Werman A, Pories 27 Garfinkel L. Overweight and cancer. Ann Intern Med 1993; 103: WJ, Caro JF, Flier JS. Peroxisome proliferator-activated receptor 1034 – 1036. gene expression in human tissues. Effects of obesity, weight loss, 28 Giovannucci E, Colditz GA, Stampfer MJ, Willett WC. Physical and regulation by insulin and glucocorticoids. J Clin Invest 1997; activity, obesity, and risk of colo-rectal adenoma in women 99: 2416 – 2422. (United States). Cancer Causes Control 1996; 7:253– 263. 22 Rieusset J, Andrelli F, Auboeuf D, Roques M, Vallier P, Riou JP, Auwerx J, Laville M, Vidal H. Insulin acutely regulates the expression of the peroxisome proliferator-activated receptor-g in human adipocytes. Diabetes 1999; 48: 699 – 705.

International Journal of Obesity