0021-972X/00/$03.00/0 The Journal of Clinical Endocrinology & Metabolism Copyright © 2000 by The Endocrine Society
Vol. 85, No. 6 Printed in U.S.A.
Physiological Relationships of Uncoupling Protein-2 Gene Expression in Human Adipose Tissue in Vivo* JONATHAN H. PINKNEY, OLIVIER BOSS, GEORGE A. BRAY, KAREN BULMER, SIMON W. COPPACK, AND VIDYA MOHAMED-ALI Pennington Biomedical Research Center, Louisiana State University (J.H.P., G.A.B.), Baton Rouge, Louisiana 70808-4124; Department of Medicine, University of Bristol, Bristol Royal Infirmary (J.H.P.), Bristol, United Kingdom BS2 8HW; Faculty of Medicine, Medical Biochemistry, University of Geneva (O.B.), 1211 Geneva 4, Switzerland; and Department of Medicine, University College London Medical School, Whittington Hospital (K.B., S.W.C., V.M.-A.), London, United Kingdom N19 3UA ABSTRACT The physiological significance of changes in uncoupling protein-2 (UCP-2) gene expression is controversial. In this study we investigated the biochemical and functional correlates of UCP-2 gene expression in sc abdominal adipose tissue in humans in vivo. UCP-2 messenger ribonucleic acid expression was quantified by nuclease protection in adipose tissue from lean and obese humans in both the fasting and postprandial states. Plasma fatty acids, insulin, and leptin were all determined in paired samples from the superficial epigastric vein and radial artery, and local production rates were calculated from 133Xe washout. In the fasting state UCP-2 expression correlated inversely with body mass index (r ⫽ ⫺0.45; P ⫽ 0.026), percent body fat (r ⫽ ⫺0.41; P ⫽ 0.05), plasma insulin (r ⫽ ⫺0.47; P ⫽
NCOUPLING PROTEIN-2 (UCP-2) is a recently characterized member of the mitochondrial carrier protein family (1, 2). UCP-2 is widely expressed in human tissues, including white adipose tissue (WAT). However, most of the literature has focused on factors that alter levels of UCP-2 gene expression, so the physiological role of UCP-2 remains controversial. In studies of rodents, UCP-2 messenger ribonucleic acid (mRNA) expression was increased in WAT of ob/ob and db/db mice (2), but was reduced (3) or unchanged (4) in fa/fa rats. In three studies it was found that ␤3-agonists had no effect on UCP-2 gene expression in WAT (1, 5, 6), although others (3, 7) demonstrated a ␤3-mediated induction of UCP-2 gene expression in the WAT of mice and rats. UCP-2 gene expression in WAT also correlated with plasma insulin levels (5), and leptin was found to induce UCP-2 expression in pancreatic islet cells (8). Furthermore, the central administration of leptin exhibited a biphasic effect on adipose tissue UCP-2 gene expression dependent on the age of the animals (9). Peroxisome proliferator-activated receptor agonists and fatty acids have also been observed to increase UCP-2 gene expression in adipose cells (10). Finally, a high Received June 11, 1999. Revision received February 9, 2000. Accepted February 27, 2000. Address all correspondence and requests for reprints to: Dr. Jonathan H. Pinkney, University of Bristol, Department of Medicine, Research Centre for Neuroendocrinology, Bristol Royal Infirmary, Marlborough Street, Bristol BS2 8HW, United Kingdom. E-mail: [email protected]
* This work was supported by a Wellcome Trust Travelling Fellowship (to J.P.), the British Heart Foundation (to K.B.), and the Wellcome Trust.
0.02), epigastric venous fatty acids (r ⫽ ⫺0.45; P ⫽ 0.04), and leptin (r ⫽ ⫺0.50; P ⫽ 0.018). UCP-2 expression remained inversely related with plasma leptin after controlling for percent body (r ⫽ ⫺0.45; P ⫽ 0.038). At 2 or 4 h postprandially, there were no significant relationships between UCP-2 expression and biochemical parameters. In conclusion, 1) UCP-2 messenger ribonucleic acid expression in sc adipose tissue is inversely related to adiposity and independently linked to local plasma leptin levels; and 2) UCP-2 expression is not acutely regulated by food intake, insulin, or fatty acids. Reduced UCP-2 expression may be a maladaptive response to sustained energy surplus and could contribute to the pathogenesis and maintenance of obesity. (J Clin Endocrinol Metab 85: 2312–2317, 2000)
fat diet has been shown to increase UCP-2 mRNA expression in WAT in some mouse strains (1, 11, 12), but not in others (12). Together, these studies suggest that in rodents UCP-2 in WAT may be up-regulated, through multiple mechanisms, in the presence of energy surplus and/or excess fatty acid flux. The metabolic significance of this phenomenon remains unclear. In contrast, less information is available regarding UCP-2 in human adipose tissue. Studies in humans have suggested a correlation between UCP-2 gene expression in WAT and body mass index (BMI), and increasing levels of expression with fasting (13). However, these investigators found no effect on UCP-2 expression with less severe, but more prolonged caloric restriction (14). In contrast, reduced UCP-2 gene expression was observed in ip, but not extraperitoneal, adipose tissue of obese humans, with no change in expression with weight loss (15). Thus, the animal and human studies do not yet afford a clear view of the role of UCP-2 in human adipose tissue. In particular, doubts have been cast about UCP-2 as a prime regulator of thermogenesis because of its induction by starvation, its inconsistent response to adrenergic agonists, and the belief that WAT is not a major thermogenic tissue in humans. This idea would be in keeping with the hypothesis that the skeletal muscle homolog UCP-3 (16, 17) could play a role in lipid metabolism (18 –21). The present study was set up to examine quantitatively levels of UCP-2 gene expression in human adipose tissue and to relate this measure to major physiological variables, including local concentrations in adipose tissue of free fatty acids (FFA) and leptin, levels of obesity, and nutritional
UCP-2 EXPRESSION IN HUMAN ADIPOSE TISSUE
status, as indicated by responses to food intake. To facilitate these analyses, adipose tissue leptin and FFA production were measured directly in vivo using the arterio-venous balance technique, with adjustment for blood flow measured from 133Xe washout, and we established a sensitive nuclease protection assay for UCP-2 to examine, quantitatively, levels of gene expression in abdominal sc adipose tissue. Subjects and Methods Subjects The studies were approved by the institutional ethical committee, and subjects gave informed written consent. Human volunteers (15 women and 10 men) with a wide range of levels of adiposity were recruited [median BMI, 32.9 kg/m2 (range, 18.0 –77.1); percent body fat, 41.9% (range, 27.4 –50.8%)] from an obesity clinic and hospital staff. The mean ⫾ sd age was 40.5 ⫾ 13.8 yr. Subjects were all nondiabetic and euthyroid, and obese subjects had simple obesity.
Study design Subjects were instructed by a nutritionist on how to maintain a high carbohydrate diet for 2 days before the day of study and completed 3-day food diaries, which were analyzed using Dietplan 5 (Forestfield Software Ltd., West Sussex, UK). Subjects were studied in the postabsorptive state (at least 15 h of fasting) and for 5 h after a high carbohydrate meal (energy content adjusted to represent 50% of the basal metabolic rate, 70% energy from carbohydrate, 20% from fat, and 10% from protein). Blood was sampled at baseline (in duplicate) after a 15-h fast and at 2 and 4 h after the meal.
Anthropometric determinations Subjects were weighed, height was recorded, and BMI was calculated (kilograms per m2). Body composition was measured by electrical bioimpedance (Bodystat, Douglas, Isle of Man).
Adipose tissue biopsy An open biopsy of sc anterior abdominal adipose tissue was performed, with the removal of 2–5 g adipose tissue. Biopsies were carried out in the fasting state and in a subset at 2 h (n ⫽ 16) and 4 h postprandially (n ⫽ 11).
Blood sampling Cannulas were inserted, using local anesthesia, into a radial artery and a superficial epigastric vein draining the sc abdominal adipose tissue. All lines were kept patent by a slow infusion of isotonic saline. Blood samples were taken simultaneously from the two sites. Previous work has shown that venous blood from superficial epigastric veins approximates the effluent from an adipose tissue bed, and arteriovenous differences across the abdominal adipose tissue yield results in good agreement with microdialysis studies (22).
Blood flow measurements and assays Adipose tissue blood flow was determined with 133Xe washout (23) during the baseline sampling and after food intake. A tissue/blood coefficient of 10 was assumed for all subjects, as previously advocated (23, 24). Plasma flow was calculated from blood flow and hematocrit. The local leptin production by adipose tissue was calculated as the product of the arterio-venous difference and local plasma flow (25). Leptin was measured in plasma with an in-house RIA, using an antihuman leptin rabbit polyclonal antibody and 125I-labeled leptin. This assay shows good correlation with the Linco leptin RIA (r ⫽ 0.92; Linco Research, Inc., St. Charles, MO). The assay was sensitive to 0.1 ng/mL, with inter- and intraassay coefficients of variation less than 10% (26). FFA were measured by an enzymatic colorimetric assay (Roche Molecular Biochemicals, Lewes, UK). Insulin concentrations were determined with a specific immunoenzymometric assay (DAKO Corp., Cambs, UK),
with a detection limit of 3 pmol/L. Plasma glucose was measured by the glucose oxidase method (Beckman Coulter, Inc., Brea, CA).
UCP-2 mRNA expression Total RNA was extracted from adipose tissue with Tri-Reagent (Sigma, Poole, UK). The UCP-2 riboprobe was made from human UCP-2 complete coding sequence (GenBank U82819) (16) cloned in pBluescript SK⫺. Purified plasmid DNA was linearized with the enzyme RsaI (New England Biolabs, Inc., Beverley, MA). A high specific activity riboprobe was synthesized by in vitro transcription with T3 polymerase (Maxiscript-T3, Ambion, Inc., Austin, TX) and randomly labeled by incorporation of [␣-32P]UTP (New England Nuclear Corp., Boston, MA) in the absence of limiting nucleotide. A 300-nucleotide UCP-2 riboprobe was produced, protecting a 270-nucleotide fragment of UCP-2 mRNA. For the internal control, a ␤-actin riboprobe protecting a 125-nucleotide sequence of ␤-actin mRNA was transcribed from p-TRI-Actin-125human (Ambion, Inc.) with T3 polymerase, and labeled by random incorporation of [␣-32P]UTP. Riboprobes were purified by spin column (CLONTECH Laboratories, Inc., Palo Alto, CA), and isolated by 10% urea/formaldehyde gel electrophoresis. RNA Century Marker Template (Ambion, Inc.) was used as the RNA marker. Reagents for the nuclease protection assay (RPA-III kit) were obtained from Ambion, Inc. Hybridization reactions were set up with 10 g RNA/reaction and incubated with both riboprobes for 24 h. Single stranded RNA was digested with ribonuclease A/ribonuclease T1 (1:100). Hybridized fragments were resolved by urea-formaldehyde gel electrophoresis, with appropriate yeast RNA controls. Gels were exposed in a PhosphorImager cassette (Molecular Dynamics, Inc., Sunnyvale, CA) and quantified with ImageQuant. No cross-reactivity was apparent between the UCP-2 riboprobe and additional homologous mRNAs, suggesting very low levels of expression of UCP-1 and UCP-3 messages. Thus, in pilot studies UCP-2 and ␤-actin were readily detectable in as little as 1 g total adipose tissue RNA, whereas, using similar in vitro transcription and nuclease protection assay protocols for human UCP-1 and UCP-3 (GenBank U84763) (16), neither of these messages was detectable in 20 g sc adipose tissue RNA.
Statistical analyses Statistical analyses were performed with the program SPSS, Inc. (Statistical Package for the Social Sciences), version 6.0. UCP-2 gene expression was expressed as a ratio to ␤-actin expression. Data were analyzed using parametric methods. Cross-sectional between-group comparisons were made with unpaired t test, and differences between variables at baseline and during a high carbohydrate diet were made using paired t test. Relationships between continuous variables were determined with linear regression analysis. Two-tailed P values were determined, and statistical significance was defined as P ⬍ 0.05.
Figure 1 is a representative gel showing expression of UCP-2, but neither UCP-1 nor UCP-3, in human abdominal sc white adipose tissue. UCP-2 was detected in all samples measured, whereas UCP-1 and UCP-3 were not detected. The faint bands noted around 230 nucleotides in the UCP-1 assay were present in human and yeast control lanes alike, excluding the possibility of significant binding to human UCP-2. However, the faint additional band around 180 nucleotides in the UCP-2 lane could represent a signal from a homologous sequence. The absence of significant UCP-3 expression reduces the likelihood that this signal represents a UCP-3 fragment bound by the UCP-2 riboprobe. It seems probable that this band represents binding to either homologous sequences of UCP-1 or other homologous RNAs. Fasting state
All subjects were studied in the fasting state. The expected positive correlations were observed for both arterial and
PINKNEY ET AL.
FIG. 1. Nuclease protection assays for human UCP-1, -2, and -3 mRNA in abdominal sc adipose tissue. Urea/formaldehyde gel. Orientation right to left: lanes 1– 6, Plus RNA marker. Gel shows a 270-nucleotide fragment of UCP2 mRNA in human RNA sample (lane 3). In contrast, UCP-1 was not detected (lane 1). Undigested UCP-1 probe is seen at 400 nucleotides in lanes 1 and 2, and nonspecific binding of UCP-1 probe is also present at 230 nucleotides in both sample and control. UCP-3 was not detected in sample or control (lanes 5 and 6). Twenty micrograms of total RNA were loaded in each reaction. Yeast RNA was used as a control to demonstrate nonspecific binding, and control lanes show residual unhybridized UCP probes after ribonuclease A/ribonuclease T1 (1:100) digestion. ␤-Actin was used as internal control mRNA.
venous insulin and FFA with BMI (r ⫽ 0.44 – 0.66; 0.05 ⬍ P ⬍ 0.001) and for percent body fat (r ⫽ 0.40 – 0.55; 0.06 ⬍ P ⬍ 0.007). However, the UCP-2/␤-actin ratio correlated inversely with both indexes of adiposity, BMI (r ⫽ ⫺0.45; P ⫽ 0.026) and percent body fat (r ⫽ ⫺0.41; P ⫽ 0.05; Fig. 2, A and B). UCP-2/␤-actin gene expression ratios of biopsies obtained in the fasting state were then examined in relation to biochemical variables in both radial arterial and epigastric venous plasma taken concurrently with the biopsy. The UCP-2/␤-actin ratio correlated inversely with arterial plasma insulin (r ⫽ ⫺0.47; P ⫽ 0.02) and epigastric venous FFA levels (r ⫽ ⫺0.45; P ⫽ 0.04; Fig. 3, A and B). Both arterial and epigastric venous plasma leptin concentrations correlated inversely with the UCP-2/␤-actin ratio (r ⫽ ⫺0.51; P ⫽ 0.014 and r ⫽ ⫺0.50; P ⫽ 0.018; Fig. 3, C and D). In view of the inverse correlations between the UCP-2/␤-actin ratio and measures of adiposity as well as biochemical measures, this analysis was repeated with biochemical indexes expressed as a ratio to percent body fat to determine whether the relationships were independent of body fat. The UCP-2/␤-actin ratio no longer correlated with arterial or venous insulin or FFA (r ⫽ ⫺0.002– 0.29; 0.23 ⬍ P ⬍ 0.99), but was still significantly related to arterial and venous leptin concentrations
JCE & M • 2000 Vol 85 • No 6
FIG. 2. Correlations between UCP2 expression in adipose tissue and indexes of obesity. A, Correlation between UCP2/␤-actin ratio and BMI (kilograms per m2; r ⫽ ⫺0.45; P ⫽ 0.02). B, Correlation between UCP2/␤-actin ratio and percent body fat (r ⫽ ⫺0.41; P ⫽ 0.05).
(r ⫽ ⫺0.44 and ⫺0.45; both P ⫽ 0.04). However, neither the arterio-venous difference in leptin concentrations nor the adipose tissue leptin production rate was significantly related to the UCP-2/␤-actin ratio (r ⫽ ⫺0.31; P ⫽ 0.16 and r ⫽ ⫺0.21; P ⫽ 0.37). Arterial triglycerides (r ⫽ ⫺0.39; P ⫽ 0.07) also tended to correlate inversely with the UCP-2/␤-actin ratio, although this correlation did not reach statistical significance. Postprandial study
In the postprandial study both FFAs and insulin concentrations changed significantly, with a reduction in epigastric venous FFA (P ⫽ 0.0007) and an increase in arterial insulin concentrations (P ⫽ 0.0001). However, no significant changes in gene expression were observed among these three time points. No significant relationship was observed between plasma FFA and the UCP-2/␤-actin ratio in the postprandial studies, either singly or combining the 2 and 4 h study groups.
UCP-2 EXPRESSION IN HUMAN ADIPOSE TISSUE
FIG. 3. Correlations between UCP2 expression in adipose tissue and levels of insulin, free fatty acids, and leptin. This figure shows correlations between the UCP2/␤-actin ratio and log fasting arterial insulin (A; r ⫽ ⫺0.47; P ⫽ 0.02), fasting epigastric vein FFA (B; r ⫽ ⫺0.44; P ⫽ 0.03), fasting arterial leptin (C; r ⫽ ⫺0.51; P ⫽ 0.01), and fasting epigastric vein leptin (D; r ⫽ ⫺0.50; P ⫽ 0.02).
The present study is the first to report on the relationship between UCP-2 gene expression in human sc adipose tissue and local concentrations of putative regulators, such as fatty acids and leptin, sampled directly from the same adipose tissue depot in vivo. By bringing together molecular and physiological studies, the present data provide new insights into the potential physiological significance of UCP-2 in humans. These studies show that UCP-2 gene expression in human sc adipose tissue is inversely related to both adiposity and its biochemical correlates. In the present studies UCP-2 gene expression was measured by nuclease protection assay. This approach was adopted by Oberkofler et al. (15), although these investigators observed reduced UCP-2 expression in visceral adipose tissue, but not sc abdominal adipose tissue of obese patients compared to lean subjects. Using this approach together with adipose tissue obtained at open biopsy may offer more accurate measurement of gene expression than PCR-based methods on smaller samples, where amplification introduces
further error. In agreement with the results of previous studies, neither UCP-1 nor UCP-3 (16) was expressed to a significant degree in human sc adipose tissue, but the present studies demonstrate this point well in the same assay. The principal conclusion of the present study was that UCP-2 gene expression was reduced with increasing levels of adiposity. This conclusion was supported by concordant biochemical data, including inverse relationships with circulating plasma insulin, leptin, and fatty acid concentrations. In the fasting state epigastric venous FFA concentrations were inversely related to levels of UCP-2 gene expression. The use of epigastric vein sampling, as opposed to measurements of mixed venous blood from the forearm, is a more sensitive way to examine concentrations of substrates and hormones emitted from same local WAT depot. The increase in epigastric venous FFA during fasting is a convincing index of lipolysis, implying mobilization and release of adipocyte lipid stores. The elevated venous FFA levels make it unlikely that intracellular FFA concentrations are lower in obesity. Therefore, in obese subjects and in the fasting state, we an-
JCE & M • 2000 Vol 85 • No 6
PINKNEY ET AL.
ticipated that raised intracellular FFA might lead to increased UCP-2 gene expression, as suggested by Aubert et al. (10). However, to the contrary, UCP-2 gene expression was lower in more obese subjects and was not further increased in the fasting state. In the fed state, no differences were noted in UCP-2 gene expression, suggesting that individual meals do not affect UCP-2 gene expression in human adipose tissue. Regarding insulin, measurements were made on arterial plasma, in view of the previous demonstration by our group that adipose tissue contributes significantly to the extraction of insulin from plasma (27). There are relatively few data regarding the effects of insulin on the expression of any member of the UCP family. Although insulin has been shown to increase UCP-1 protein (28), euglycemic clamps affected neither UCP-2 mRNA expression in human sc adipose tissue and skeletal muscle nor UCP-3 mRNA expression in skeletal muscle (13, 18). In our study, in which obesity was linked to hyperinsulinemia, we cannot distinguish independent effects of adiposity and insulin on UCP-2 expression. Both arterial and epigastric venous leptin concentrations correlated inversely with UCP-2 expression in the fasting state, after adjusting for adiposity, suggesting an independent relationship between local leptin concentrations and UCP-2 gene expression. In previous studies in rodents leptin was suggested to increase UCP-3 mRNA expression in skeletal muscle (7, 29 –31) and increase UCP-2 mRNA expression in WAT (29, 31), whereas some researchers reported a lack of effect of leptin on UCP-2 mRNA levels in certain WAT depots (31, 32). However, UCP regulation increasingly appears to be dependent on the UCP subtype, the tissue in question, the location of the WAT depot, and the time course studied (9, 31, 33, 34). In the postprandial studies no significant relationship was observed between UCP-2 expression and the metabolic variables addressed. This may in part be on account of the smaller size of the postprandial study groups together with the postprandial time points being relatively early to observe significant effects on levels of gene expression. These data suggest that there is no acute effect of food intake on UCP-2 gene expression in WAT, and therefore that changes in UCP-2 gene expression are likely to occur in relation to more long-term adaptive changes in adipose tissue. From the present work and previous literature on UCP-2 mRNA expression, it is difficult to draw firm conclusions about the functional significance of observed differences in levels of UCP-2 mRNA expression. The present results could be consistent with an impairment of thermogenesis in the sc adipose tissue of obese human subjects. However, it is unclear whether this stands in a primary or a secondary relationship to obesity. Reduced UCP-2 expression also might be predicted to enhance the mitochondrial electrochemical gradient and to promote adipocyte ATP synthesis, which might inappropriately facilitate biosynthetic processes, including fatty acid reesterification as well as de novo lipogenesis. In conclusion, this study, by bringing together in vivo physiological techniques with molecular methods in humans, affords new insights into physiological relationships of gene expression that may not necessarily be apparent from work in animals or in vitro. UCP-2 gene expression in sc abdominal adipose tissue is inversely related to adiposity and its bio-
chemical correlates. Furthermore, it is inversely related to plasma leptin levels independent of body fat, supporting an inverse relationship between leptin and UCP-2 gene expression in vivo. Reduced UCP-2 expression in obese individuals may be a maladaptive phenomenon and could offer a target of therapeutic potential in the treatment of obesity. Acknowledgments We are grateful to Prof. J.-P. Giacobino and Dr. P. Muzzin for their generous assistance with this project and comments on the manuscript.
References 1. Fleury C, Neverova M, Collins S, et al. 1997 Uncoupling protein-2: a novel gene linked to obesity and hyperinsulinaemia. Nat Genet. 15:269 –272. 2. Gimeno RE, Dembski M, Weng X, et al. 1997 Cloning and characterisation of an uncoupling protein homolog. A potential molecular mediator of human thermogenesis. Diabetes. 46:900 –906. 3. Emilsson V, Summers RJ, Hamilton S, Liu Y-L, Cawthorne MA. 1998 The effects of the ␤3-adrenoceptor agonist BRL 35135 on UCP isoform mRNA expression. Biochem Biophys Res Commun. 252:450 – 454. 4. Hidaka H, Kakuma T, Yoshimatsu H, Yasunaga S, Kurokawa M, Sakata T. 1998 Molecular cloning of rat uncoupling protein 2 cDNA and its expression in genetically obese Zucker (fa/fa) rats. Biochim Biophys Acta 1389:178 –186. 5. Savontaus E, Rouru J, Boss O, Huuoponen R, Koulu M. 1998 Differential regulation of uncoupling protein by chronic treatments with ␤3-adrenergic agonist BRL 35135 and metformin in obese fa/fa Zucker rats. Biochem Biophys Res Commun. 246:899 –904. 6. Yoshitomi H, Yamazaki K, Abe S, Tanaka I. 1998 Differential regulation of mouse uncoupling proteins among brown adipose tissue, white adipose tissue, and skeletal muscle in chronic beta 3 adrenergic receptor agonist treatment. Biochem Biophys Res Commun. 253:85–91. 7. Gong D-W, He Y, Karas M, Reitman M. 1997 Uncoupling protein-3 is a mediator of thermogenesis regulated by thyroid hormone and ␤3-adrenergic agonists and leptin. J Biol Chem. 272:24129 –24132. 8. Zhou Y-T, Shimabukuro M, Koyama K, et al. 1997 Induction by leptin of uncoupling protein-2 and enzymes of fatty acid oxidation. Proc Natl Acad Sci USA. 94:6386 – 6390. 9. Qian H, Hausman GJ, Compton MM, Azain MJ, Hartzell DL, Baile CA. 1998 Leptin regulation of peroxisome proliferator-activated receptor-␥, tumor necrosis factor, and uncoupling protein-2 expression in adipose tissues. Biochem Biophys Res Commun. 246:660 – 667. 10. Aubert J, Champigny O, Saint-Marc P. 1998 Up-regulation of UCP-2 gene expression by PPAR agonists in preadipose and adipose cells. Biochem Biophys Res Commun. 238:606 – 611. 11. Matsuda J, Hosoda K, Itoh J, et al. 1997 Cloning of rat uncouplin protein-3 and uncoupling protein-2 cDNAs: their gene expression in rats fed high fat diet. FEBS Lett. 418:200 –204. 12. Surwit RS, Wang S, Petro AE, et al. 1998 Diet-induced changes in uncoupling proteins in obesity-prone and obesity-resistant strains of mice. Proc Natl Acad Sci USA. 95:4061– 4065. 13. Millet L, Vidal H, Andreelli F, et al. 1997 Increased uncoupling protein-2 and -3 mRNA expression during fasting in obese and lean humans. J Clin Invest. 100:2665–2670. 14. Barbe P, Millet L, Larrouy D, et al. 1998 Uncoupling protein-2 messenger ribonucleic acid expression durng very-low calorie diet in obese premenopausal women. J Clin Endocrinol Metab. 83:2450 –2453. 15. Oberkofler H, Liu YM, Esterbauer H, Hell E, Krempler F, Patsch W. 1998 Uncoupling protein-2 gene: reduced mRNA expression in intraperitoneal adipose tissue of obese humans. Diabetologia. 41:940 –946. 16. Boss O, Samec S, Paoloni-Giacobino A, et al. 1997 Uncoupling protein-3:a new member of the mitochondrial carrier family with tissue-specific expression. FEBS Lett. 408:39 – 42. 17. Vidal-Puig A, Solanes G, Grujic D, Flier JS, Lowell BB. 1997 UCP3:An uncoupling protein homologue expressed preferentially and abundantly in skeletal muscle and brown adipose tissue. Biochem Biophys Res Commun. 235:79 – 82. 18. Boss O, Bobbioni-Harsch E, Assimacoupoulos-Jeanet F, et al. 1998 Uncoupling protein-3 expression in skeletal muscle and free fatty acids in obesity. Lancet. 351:1933. 19. Weigle DS, Selfridge LE, Schwartz MW, et al. 1998 elevated free fatty acids induce uncoupling protein-3 expression in muscle. Diabetes. 47:298 –302. 20. Samec S, Seydoux J, Dulloo AG. 1998 Role of UCP homologues in skeletal muscles and brown adipose tissue: mediators of thermogenesis or regulators of lipids as fuel substrate? FASEB J. 12:715–724. 21. Samec S, Seydoux J, Dulloo AG. 1998 Interorgan signaling between adipose tissue metabolism and skeletal muscle uncoupling protein homologs: is there a role for circulating free fatty acids? Diabetes. 47:1693–1698.
UCP-2 EXPRESSION IN HUMAN ADIPOSE TISSUE 22. Frayn KN, Coppack SW, Humphries SM, Whyte P. 1989 Metabolic characteristics of human adipsoe tissue in vivo. Clin Sci. 76:509 –516. 23. Bulow J, Jelnes R, Astrup A, Madsen J, Vilmann P. 1987 Tissue/blood partition coefficients for xenon in various adipose tissue depots in man. Scand J Clin Lab Invest. 47:1–3. 24. Summers LKM, Samra JS, Humphreys SM, Morris RJ. 1996 Subcutaneous abdominal adipose tissue blood flow: variation within and between subjects and relationship to obesity. Clin Sci. 91:679 – 683. 25. Klein S, Coppack SW, Mohamed-Ali V, Landt M. 1996 Adipose tissue leptin production and plasma leptin kinetics in humans. Diabetes. 45:984 –987. 26. Pinkney JH, Coppack SW, Mohamed-Ali V. 1998 Effect of isoprenaline on plasma leptin and lipolysis in humans in vivo. Clin Endocrinol (Oxf). 48:407– 412. 27. Bulmer K, Clarke DK, Katz J, et al. 1998 Insulin clearance by adipose tissue in vivo [Abstract]. Diabetic Med 15(Suppl 1):41. 28. Geloen A, Trayhurn P. 1990 Regulation of the level of uncoupling protein in brown adipose tissue by insulin. Am J Physiol 258:R418 –R424.
29. Cusin I, Zakrzewska KE, Boss O, et al. 1998 Chronic central leptin infusion enhances insulin-stimulated glucose metabolism and favors the expression of uncoupling proteins. Diabetes. 47:1014 –1019. 30. Liu Q, Bai C, Chen F, et al. 1998 Uncoupling protein-3: a muscle-specific gene upregulated by leptin on ob/ob mice. Gene. 207:1–7. 31. Scarpace PJ, Nicolson M, Matheny M. 1998 UCP2, UCP3 and leptin gene expression: modulation by food restriction and leptin. J Endocrinol. 159:349 –357. 32. Surwit R, Petro AE, Parekh P, Collins S. 1997 Low plasma leptin in response to dietary fat in diabetes- and obesity-prone mice. Diabetes. 46:1516 –1520. 33. Matsuda J, Hosoda K, Itoh H, et al. 1998 Increased adipose expression of the uncoupling protein-3 gene by thiazolidinediones in Wistar fatty rats and in cultured adipocytes. Diabetes. 47:1809 –1814. 34. Lin B, Coughlin S, Pilch PF. 1998 Bidirectional regulation of uncoupling protein-3 and GLUT-4 mRNA in skeletal muscle by cold. Am J Physiol 275: E386 –E391.