Transcriptional Regulation of Adipocyte Hormone-Sensitive Lipase by Glucose

Transcriptional Regulation of Adipocyte Hormone-Sensitive Lipase by Glucose Fatima Smih,1 Philippe Rouet,1 Ste´phanie Lucas,1 Aline Mairal,1 Coralie Sengenes,1 Max Lafontan,1 Sophie Vaulont,2 Marta Casado,2 and Dominique Langin1

Hormone-sensitive lipase (HSL) catalyzes the rate- maximum lipolytic capacity of human subcutaneous adi- limiting step in the mobilization of fatty acids from pocytes stimulated by a ␤-adrenergic agonist (2). More- adipose tissue, thus determining the supply of energy over, targeted disruption of the HSL gene in the mouse substrates in the body. HSL mRNA was positively regu- results in blunted ␤-adrenergic agonist-induced lipolysis lated by glucose in human adipocytes. Pools of stably (3,4). Clinical studies also support a role for HSL as a transfected 3T3-F442A adipocytes were generated with limiting factor in adipose tissue lipolysis and show that, ؊ human adipocyte HSL promoter fragments from 2,400/ besides the short-term modulation of activity by phosphor- ؉38 to ؊31/؉38 bp linked to the luciferase gene. A glucose-responsive region was mapped within the prox- ylation, variations in HSL expression are associated with imal promoter (؊137 bp). Electromobility shift assays changes in lipolytic capacity. Indeed, obese patients and showed that upstream stimulatory factor (USF)-1 and normal-weight subjects with a family trait for obesity show USF2 and Sp1 and Sp3 bound to a consensus E-box and decreased maximal lipolytic effect of catecholamines and -two GC-boxes in the ؊137-bp region. Cotransfection of blunted HSL expression (5,6). Furthermore, genetic stud -the ؊137/؉38 construct with USF1 and USF2 expres- ies suggest that HSL participates in the polygenic back sion vectors produced enhanced luciferase activity. ground of obesity and type 2 diabetes (7,8). Moreover, HSL mRNA levels were decreased in USF1- Increased mobilization of fatty acids from adipose tissue and USF2-deficient mice. Site-directed mutagenesis of the HSL promoter showed that the GC-boxes, although stores in diabetic patients leads to high free fatty acid contributing to basal promoter activity, were dispens- concentrations that contribute to the development of able for glucose responsiveness. Mutation of the E-box ketoacidosis, a major acute complication of type 1 diabe- led to decreased promoter activity and suppression of tes (9). Several mechanisms may account for the enhanced the glucose response. Analogs and metabolites were lipolysis. Hypoinsulinemia and an increase in the concen- used to determine the signal metabolite of the glucose trations of stimulatory hormones stimulate HSL activity response. The signal is generated downstream of glu- via posttranslational modifications. Moreover, isolated fat cose-6-phosphate in the glycolytic pathway before the cells from streptozotocin-induced diabetic rats show in- triose phosphate step. Diabetes 51:293–300, 2002 creased maximal lipolytic response to agents acting at the postreceptor level (10) and increased HSL mRNA and protein levels (11). Insulin deficiency and hyperglycemia could both account for the upregulation of HSL expression ormone-sensitive lipase (HSL) is a key enzyme in diabetes. Insulin per se had no effect on HSL mRNA for the hydrolysis of adipose tissue triacylglyc- concentrations in 3T3-F442A murine adipocytes (12). erol into fatty acids that are the major source of body energy in the absence of dietary lipids. However, in this cellular model, glucose deprivation re- H sulted in a twofold decrease in HSL mRNA and total Catecholamines and insulin are important regulators of lipolysis in adipocytes through a modulation of intracellu- activity levels (13). The effect of glucose was reversible lar cAMP levels and reversible phosphorylation of HSL (1). and was not due to an impairment of the differentiation The lipase is thought to catalyze the rate-limiting step in program, since the mRNA levels of CCAAT/enhancer bind- ␣ ␣ cAMP-dependent lipolysis. In agreement, a strong linear ing protein- (C/EBP ), a transcription factor important correlation was found between HSL protein levels and the for maintenance of the terminally differentiated state, and adipocyte triglyceride content were not affected by glucose deprivation. Moreover, it was shown that prolonged From 1INSERM Unite´ 317, Institut Louis Bugnard, Centre Hospitalier Univer- exposure of isolated rat adipocytes to glucose in the sitaire de Rangueil, Universite´ Paul Sabatier, Toulouse, France; and the 2Institut Cochin de Geńe´tique Molećulaire, INSERM Unite´ 129, Paris, France. presence of insulin resulted in an increase of basal and Address correspondence and reprint requests to Philippe Rouet or Dominique stimulated lipolysis and a maintenance of HSL protein Langin, INSERM U317, Baˆtiment L3, CHU Rangueil, 31403 Toulouse Cedex 4, levels (14). France. E-mail: [email protected] or dominique.langin@toulouse .inserm.fr. The adipocyte form of HSL is encoded by nine exons F.S. and P.R. contributed equally to this work. M.C. is currently affiliated spanning 11 kb (15,16). The transcriptional start site has with the Instituto de Biomedicina de Valencia (CSIC), Jaume Roig 11, 46010 Ј Valencia, Spain. been mapped in a short 5 -noncoding exon located 1.5 kb Received for publication 4 June 2001 and accepted in revised form 5 upstream of the first coding exon. Transient transfections November 2001. of various portions of the 5Ј-flanking region linked to the DMEM, Dulbecco’s modified Eagle’s medium; EMSA, electromobility shift assay; FCS, fetal calf serum; HSL, hormone-sensitive lipase; USF, upstream luciferase gene into rat adipocytes showed the existence stimulatory factor. of an active promoter (17). However, the regulatory ele-

DIABETES, VOL. 51, FEBRUARY 2002 293 HORMONE-SENSITIVE LIPASE REGULATION BY GLUCOSE

ments within the human adipocyte HSL promoter have not Quantification of mRNA levels in human adipocytes. Adipose primary been characterized. Thus, we decided in this study to cell culture total RNA was extracted using the Qiagen RNeasy kit. Total RNA was stored at Ϫ80°C until analysis. Quantitation of HSL and cyclophilin mRNA characterize the cis-acting regions that are critical for levels were performed by reverse transcription–competitive PCR as described promoter activity and mediate the responsiveness to glu- previously (21). The reverse transcription step was performed on 100 ng of cose. total RNA with a specific antisense primer and Omniscript reverse transcrip- tase (Qiagen). cDNA samples were then amplified by PCR using sense and antisense primers in the presence of known amounts of a specific DNA RESEARCH DESIGN AND METHODS competitor. PCR products were separated by capillary electrophoresis and Plasmids. Constructs containing human adipocyte HSL 5Ј flanking regions quantified using the ABI PRISM 310 Genetic Analyzer system with the from Ϫ2,400 to Ϫ31 relative to the transcription start point and 38 bp of the Genescan program (PE Applied Biosystem). HSL mRNA levels were normal- 5Ј untranslated region in pGL3basic vector encoding a modified firefly ized with cyclophilin mRNA levels. luciferase (Promega, Madison, WI) have been described previously (17). Preparation of nuclear extracts and electromobility shift assay. 3T3- Mutant HSL promoter constructs were generated by synthetic nucleotide F442A preadipocyte and adipocyte nuclear extracts were prepared according assembly in pGL3basic using the KpnI site of the vector and the XbaI site of to standard procedures (18). Probe preparation and electromobility shift the HSL promoter. These constructs were made using standard procedures assay (EMSA) conditions were described previously (22). Briefly, 150,000 (18) and sequenced using a Perkin Elmer Dye terminator sequencing kit and cpm/lane (0.5 ng of end-labeled DNA) and 4 ␮g of nuclear extracts were an ABI 373 sequencer. Assembly of the double-stranded oligonucleotides led incubated for 15 min on ice. Upstream stimulatory factor (USF) and Sp1 to substitution of the characterized binding sites by a mutagenized box with competitor oligonucleotides were from Promega. The HNF1 oligonucleotide an XbaI site. pCR3-USF1 and -USF2a expression vectors, pCMV-Sp1, and used as nonspecific competitor has been described (22). Supershift assays pSVsport-ADD1 and pSVsport-ADD1403, were gifts from M. Raymondjean and using antibodies against USF1 and USF2 (gift of M. Raymondjean) and against M. Vasseur-Cognet (INSERM U 129, Paris), A.P. Butler (M.D. Anderson Cancer Sp1 (sc-59X; Santa Cruz) and Sp3 (sc-644X; Santa Cruz) were performed by Center, University of Texas, Houston, TX), and B. Spiegelman (Dana Farber incubating the nuclear extracts with the antibodies for 15 min on ice before Cancer Institute, Boston, MA), respectively. addition of the probe. After electrophoresis, the gel was fixed in 10% acetic Cell culture, transfection, and luciferase assay. In agreement with French acid/10% ethanol and dried. Gels were exposed and analyzed using a Molec- laws on biomedical research, subcutaneous abdominal adipose tissue was ular Dynamics SI445 Phosphorimager. obtained from female subjects undergoing plastic surgery. Human adipocytes Northern blot analysis of HSL mRNA in USF1- and USF2-deficient mice. in primary culture were differentiated as described by Hauner et al. (19), with Total RNA was isolated from perigonadal adipose tissue of fasted USF1- modifications. Briefly, all visible fibrous material and blood vessels were and USF2-deficient mice (23,24). Wild-type littermates were used as controls. discarded from adipose tissue samples. The remaining adipose tissue was cut Northern blot analysis was performed using mouse HSL (12) and ribosomal into small pieces and digested in PBS containing 0.15% collagenase and 2% 18S cDNA probes as described previously (25). Specific signals were quanti- BSA under gentle shaking at 120 cycles/min at 37°C. After centrifugation, the tated using a Phosphorimager (Molecular Dynamics). stromal cell fraction was resuspended and incubated in an erythrocyte-lysing Statistical analysis. The nonparametric Mann-Whitney U test for unpaired

buffer (155 mmol/l NH4Cl, 5.7 mmol/l K2HPO4, and 0.1 mmol/l EDTA). Cells values was used to compare HSL mRNA levels. ANOVA with least-square were filtered through a nylon mesh (150 ␮m). After additional centrifugation, difference post hoc test was used for comparisons of luciferase activities washing, and filtration (70 ␮m), stromal cells were suspended in Dulbecco’s (SPSS). P Ͻ 0.05 was the threshold of significance. modified Eagle’s medium (DMEM)/Ham’s F-12 medium supplemented with 10% fetal calf serum (FCS) and put into culture at 30,000 cells/cm2. After 16- to 20-h incubation, cells were refed with a chemically defined serum-free RESULTS medium consisting of DMEM/Ham’s F-12 supplemented with 15 mmol/l Glucose-mediated regulation of HSL mRNA in pri- ␮ ␮ NaHCO3, 15 mmol/l HEPES, 33 mol/l biotin, 17 mol/l panthotenate, 10 ␮g/ml human transferrin, and 50 ␮g/ml gentamicin. Insulin (66 nmol/l), 1 mary cultures of human adipocytes. To establish that nmol/l triiodothyronine, 100 nmol/l cortisol, and 1 ␮g/ml troglitazone were human HSL was regulated by glucose, a primary culture added for the first 3 days. Cells were cultured for 10 days at 37°Cina system was used to obtain human differentiated adipo- humidified atmosphere of 5% CO2. 3T3-F442A preadipocytes were cultured in cytes. Cells were treated for 48 h in a serum-free medium DMEM containing donor calf serum (10%), penicillin (200 units/ml), and containing 25 or 1 mmol/l glucose. HSL mRNA levels were streptomycin (50 mg/l) at 37°C in a humidified atmosphere of 7% CO2. For adipocyte differentiation, DMEM was supplemented with insulin (50 nmol/l) quantified using reverse transcription–competitive PCR. once cells reached confluence, and donor calf serum was replaced with 10% The low glucose concentration led to a decrease in the FCS. Medium was renewed every 2 days for 10 days. Differentiated human and ratio of HSL to cyclophilin mRNA levels from 0.049 Ϯ murine adipocytes were treated in insulin-free, serum-free DMEM containing 0.003 to 0.024 Ϯ 0.004 (P Ͻ 0.05, n ϭ 3). 25 mmol/l glucose or 1 mmol/l glucose supplemented with 0.1 mmol/l lactate and 1 mmol/l pyruvate to maintain energy balance of the cells. Use of 10 Effect of extracellular glucose concentration on hu- mmol/l instead of 0.1 mmol/l lactate did not modify the glucose response. man HSL promoter activity in murine 3T3-F442A Luciferase constructs (20 ␮g) were stably transfected into exponentially adipocytes. Because there is no established human pre- growing 3T3-F442A cells by cotransfection with 2 ␮g of G418-resistant adipocyte cell line available, we used the mouse 3T3- pMC1neo vector using CaPO -mediated transfection (18). After 2 weeks of 4 F442A preadipocyte cell line to analyze human HSL G418 (0.6 mg/ml) selection, an average of 100 colonies were recovered and pooled. This pool of clones showed adipocyte differentiation at a frequency promoter activity in response to glucose. Luciferase con- similar to that of the parental 3T3-F442A cell line. Primary rat white adipo- structs containing 2,400 to 31 bp of the 5Ј flanking region cytes were isolated by collagenase digestion of periepididymal fat pads from and 38 bp of the 5Ј untranslated region (17) were stably young (150–200 g) male Wistar rats. Luciferase construct (20 ␮g) and 5 ␮gof transfected into preadipocytes. After 10 days of adipose pCMVlacZ (gift of P. Furth, University of Maryland, Baltimore, MD) were introduced into rat primary adipocytes by electroporation using a Biorad differentiation, each pool of transformants was maintained apparatus and cultured as previously described (20). ␤-Galactosidase expres- for 24 h in DMEM containing 1 or 25 mmol/l glucose. In 25 sion was used to normalize transfection efficiency and was quantified using an mmol/l glucose medium, luciferase activity was high for o-nitrophenyl-␤-D-galactopyranoside colorimetric assay (18). HepG2 and COS7 the Ϫ2,400/ϩ38, Ϫ280/ϩ38, and Ϫ137/ϩ38 constructs. The cells were cultured in DMEM containing 10% FCS, 200 units/ml penicillin, and Ϫ86/ϩ38, Ϫ57/ϩ38, and Ϫ31/ϩ38 constructs showed very 50 mg/l streptomycin. The cells were transfected with Fugene-6 reagent Ϫ ϩ (Roche) according to the manufacturer’s protocol using a total of 2 ␮g plasmid weak or null promoter activity. For the 2,400/ 38, per 3.5-cm-diameter dish. In addition, 50 ng of the Renillia luciferase expres- Ϫ280/ϩ38, and Ϫ137/ϩ38 constructs, promoter activity sion vector (pRL-SV40 from Promega) was cotransfected to check for was higher in 25 mmol/l than in 1 mmol/l glucose medium transfection efficiency. Before harvesting, cell monolayers were rinsed with (Fig. 1A). These results indicated that the Ϫ137-bp DNA calcium- and magnesium-free PBS. Cells were then scraped in passive lysis buffer (Promega). Firefly and Renilia luciferase assays were performed as segment contained the regulatory elements mediating the recommended by the manufacturer’s protocol (Promega) in a Labsystem glucose response and that the Ϫ137/–86 region was crucial luminometer. for promoter activity in the chromatin context. A dose-

294 DIABETES, VOL. 51, FEBRUARY 2002 F. SMIH AND ASSOCIATES

FIG. 2. Analysis of the human adipocyte HSL promoter sequence. The adipocyte HSL promoter glucose-responsive region was analyzed with MatInspector and TF-Search computer programs. Sequences that share homology with known binding sites for nuclear factors are underlined. The HSL E-box and HSL 3؅ GC-box showed 100% identity to E-box and GC-box consensus sequences. A half E-box CACTGA is located 5 bp 3؅ to the HSL E-box. Bold lines indicate the sequences of the oligonucleotides used to perform EMSA.

response of glucose effect is shown in Fig. 1B. The maximal stimulatory effect was observed at 15 mmol/l of glucose. The addition of FCS and insulin in the culture medium containing 25 mmol/l glucose did not modify HSL promoter activity. EMSA of the potential binding sites in the ؊137-bp region. To determine more precisely the sequences involved in the glucose responsiveness, a computer-assisted analysis of the Ϫ137-bp region revealed several potential nuclear factor binding sites (Fig. 2). A consensus CACGTG E-box motif was found between Ϫ106 and Ϫ101, and two GC-boxes were located, between Ϫ30 and Ϫ43 and between Ϫ82 and Ϫ71. E-boxes are known to bind basic helix-loop-helix leucine zipper transcription factors such as USFs. The boxes are found in several glucose response elements of hepatic genes (26). GC-boxes are bound by the Sp family of zinc finger nuclear factors. Two GC-rich sequences in promoter II of acetyl-coA carboxylase have been shown to mediate glucose responsiveness (27). EMSA was performed with the set of oligonucleotides described in Fig. 2. EMSA of the HSL E-box performed with adipocyte nuclear extracts showed two bands (Fig. 3). Competition with unlabeled oligonuceotide showed that the DNA-protein interaction was specific for the two complexes. While the nature of the band of higher mobility (X in Fig. 3) is still elusive, the lower-mobility band clearly corresponded to binding of USF1 and USF2 dimers. In- deed, anti-USF1 and anti-USF2 antibodies induced a supershift demonstrating the presence of USF1/USF2 heterodimers as previously shown in rat liver (28). The residual binding observed in the presence of anti-USF1 and anti-USF2 antibodies corresponded to USF1 and USF2 homodimers, respectively. A similar binding pattern was resolved using preadipocyte nuclear extracts (data not shown). Some glucose-responsive genes harbor two E- boxes separated by 5 bp (29). Because a half E-box CACTGA is located 5 bp 3Ј from the HSL E-box (Fig. 2), we speculated that the sequence might be weakly bound by FIG. 1. Effect of glucose on adipocyte HSL promoter activity in 3T3-F442A adipocytes. A: Effect of glucose deprivation on luciferase USF factors. However, no specific binding was observed activity induced by various 5؅-flanking regions. Luciferase activity was using preadipocyte and adipocyte nuclear extracts or measured in stably transfected 3T3-F442A differentiated adipocytes. cellular extracts prepared from COS7 cells transfected Activities of different promoter regions (luciferase activity/␮g protein) are expressed as fold induction over the activity of each pool of with USF1 and USF2 expression vectors (data not shown). clones in 1 mmol/l glucose for 24 h. This profile represents mean ؎ SD EMSA of the HSL 5Ј GC-box (Fig. 4A) showed two bands of three experiments performed on pools of clones derived from that displayed a binding pattern characteristic of members independent transfections. Glucose induction ratio ranged from 2 to 5. B: Dose-response effect of glucose on luciferase activity driven by the of the Sp family of nuclear factors. The binding to the ؊137/؉38 construct. Means ؎ SD of three independent experiments. probe was specific, since competition was observed with *P < 0.05. the HSL 5Ј GC-box and the Sp1 control oligonucleotides but not with the unrelated HNF1 oligonucleotide. Antibody

DIABETES, VOL. 51, FEBRUARY 2002 295 HORMONE-SENSITIVE LIPASE REGULATION BY GLUCOSE

FIG. 3. Binding of adipose nuclear proteins to HSL E-box. Nuclear extracts were from differentiated 3T3-F442A adipocytes. –, nuclear extracts and probe only. Antibodies against USF1 and USF2 were added separately (Anti USF1, Anti USF2) or simultaneously (Anti ؅ USF1 ؉ 2). As a control, nonimmune serum (Non imm.) was also added. FIG. 4. Binding of adipose nuclear proteins to 5 HSL GC-box (A) and ؅ HSL E-box and HNF1 unlabeled probes were added at a 200-fold molar 3 HSL GC-box (B). Nuclear extracts were from differentiated 3T3- ؅ ؅ excess. Arrows indicate supershift induced by the antibodies, USF1/ F442A adipocytes. –, extracts and probe only. HSL 5 GC-box, HSL 3 USF2 heterodimers, an unidentiﬁed binding complex X, and the un- GC-box, sp1 control, and HNF1 unlabeled probes were added at a bound labeled probe. In the presence of the antibodies, USF2 or USF1 200-fold molar excess. Antibodies against Sp1 and Sp3 were added ؉ homodimers are observed as faint bands with mobilities comparable to separately (Anti Sp1, Anti Sp3) or simultaneously (Anti Sp1 3). USF1/USF2 heterodimers. Arrows indicate supershift induced by the antibodies, bands speciﬁc for Sp1 and Sp3, and the unbound labeled probe.

against Sp1 induced a weak supershift of the low-mobility EMSA (data not shown). Mutant constructs were tran- complex. Antibody against Sp3 was able to totally super- siently transfected into primary rat adipocytes maintained shift binding proteins in the high-mobility complex. Com- in serum-free DMEM containing 25 mmol/l glucose (Fig. bination of the two antibodies led to the disappearance of 5). Mutation of the HSL E-box induced a 50% decrease in the two bands. HSL 5Ј GC-box oligonucleotide did not luciferase activity. Mutation of the HSL 5Ј GC-box had a completely abolish Sp1 binding to the probe, but compe- strong negative effect, whereas mutation of the HSL 3Ј tition with the Sp1 control oligonucleotide led to a com- GC-box had no effect. Combined mutation of the two plete disappearance of the bands. This indicated that Sp1 GC-boxes led to lower luciferase activity than single and Sp3 had a lower affinity for the HSL 5Ј GC-box than for mutations. Finally, mutations of the three boxes totally the Sp1 control competitor. This observation is in accor- abolished luciferase activity. dance with previously published results showing a better Transactivation of the HSL promoter by transient affinity of Sp1 for GGGCGGG than for GGGTGGG motifs cotransfection of expression vectors for USF1, USF2, (30). The use of the probe corresponding to the HSL 3Ј and ADD1/SREBP1c. Because of the importance of the GC-box with the ideal consensus sequence GGGCGGG led HSL E-box in promoter activity and its in vitro binding to to the same pattern (Fig. 4B). Cross-competitions con- USF, we wished to determine the transactivation proper- firmed that HSL 3Ј GC-box was a better competitor of Sp1 ties of USF1 and USF2 on the HSL promoter. Cotransfec- and Sp3 binding than HSL 5Ј GC-box. tion of the Ϫ137/ϩ38 construct with USF1 and USF2 E- and GC-box mutagenesis analysis by transient expression vectors in HepG2 cells resulted in a marked transfection into adipocytes. To evaluate the functional increase in luciferase activity (Fig. 6). Similar results were importance of each of the boxes on HSL promoter activity obtained in COS7 cells (data not shown). USF1 and USF2 in adipocytes, we performed single and multiple site- overexpression showed reduced effect on the Ϫ86/ϩ38 directed mutageneses of the sites. Lack of binding of the construct, which does not contain the E-box. We also cognate factors to the mutated sites was controlled by studied the effect of ADD1/SREBP1c, a basic helix-loop-

296 DIABETES, VOL. 51, FEBRUARY 2002 F. SMIH AND ASSOCIATES

-FIG. 5. Transient transfection of mutated ؊137/؉38 HSL promoter luciferase gene constructs into primary rat adipocytes. Mutations in E-box and GC-boxes are indicated by crosses. Firefly luciferase activity FIG. 6. Transactivation of the adipocyte HSL promoter by USF1, USF2, was normalized by measuring ␤-galactosidase activity derived from and ADD1403. The ؊137/؉38 and ؊86/؉38 HSL promoter-luciferase cotransfection of pCMV-lacZ into rat primary adipocytes. Normalized gene constructs were cotransfected in HepG2 cells with pRL-SV40 and luciferase activity is expressed as percent activity from the wild-type expression vectors for USF1, USF2, and ADD1403 as indicated. The ؊137/؉38 construct. The data (means ؎ SD) shown are representative ؊86/؉38 construct was used as a control for specific transactivation of of two independent experiments performed in triplicate. *P < 0.05. the HSL E-box, since it does not contain an E-box. Firefly luciferase activities were normalized to Renilia luciferase activities. Luciferase activities are expressed as fold induction over the activity level obtained with the ؊137/؉38 and ؊86/؉38 constructs in the absence of helix leucine zipper transcription factor that has been nuclear factor expression vectors. Means ؎ SD of three independent proposed to play a role in the activation of hepatic genes experiments. *P < 0.05. by insulin and glucose (31,32). ADD1-403 is a mutant that lacks the membrane-attachment domain and, therefore, could be attributed to the E-box in the human adipocyte enters the nucleus directly and shows enhanced transcrip- HSL promoter. tional activity (33). ADD1/SREBP1c (data not shown) and Characterization of glucose metabolites involved in ADD1–403 expression had no effect on the Ϫ137/ϩ38 and the modulation of HSL promoter activity. To deter- Ϫ86/ϩ38 construct activity. This is consistent with the mine the metabolic pathway(s) involved in the glucose lack of binding of ADD1/SREBP1c observed on the double response, we tested several glucose analogs and metabo- E-box motif of the human uncoupling protein-2 promoter lites (Fig. 9). Compounds were added for 24 h after a 24-h (34). period in 1 mmol/l glucose medium. Supplementation of HSL gene expression in adipose tissue of USF1- and the medium with 25 mmol/l glucose restored luciferase USF2-deficient mice. To further assess the role of USF1 activity. Mannitol, which is not transported into cells, did and USF2 in adipocyte HSL gene expression, HSL mRNA not mimic glucose effect. Hence, the effect of glucose levels were determined by Northern blot analysis of adi- deprivation was not due to difference in osmotic pressure. pose tissue from USF1- and USF2-deficient mice (23,24). 3-O-methylglucose, which is transported inside the adipo- The amount of HSL mRNA was reduced to 49% in USF1- deficient mice and to 14% in USF2-deficient mice compared with wild-type mice (Fig. 7). Glucose response analysis of HSL promoter mutants stably transfected into 3T3-F442A adipocytes. To define the DNA elements involved in the glucose response, the mutant constructs were stably transfected into 3T3- F442A preadipocytes. Transfectant pools were subjected to glucose deprivation for 24 h after 10 days of adipocyte differentiation (Fig. 8). Mutations of the GC-boxes had no effect on glucose-induced luciferase activity. Mutation of the E-box resulted in a complete loss of the glucose response. In accordance with this observation, simulta- neous mutations of the GC-boxes and the E-box led to an absence of glucose response. Mutation of the half E-box FIG. 7. HSL mRNA level in adipose tissue of USF1- and USF2-deficient mice. Northern blot analysis was performed on total RNA using HSL CACTGA (Fig. 2) did not alter the glucose response (data and ribosomal 18S cDNA probes. Means ؎ SD of three mice in each not shown). Therefore, the glucose response element group. *P < 0.05.

DIABETES, VOL. 51, FEBRUARY 2002 297 HORMONE-SENSITIVE LIPASE REGULATION BY GLUCOSE

-FIG. 8. Glucose response of mutated ؊137/؉38 HSL promoter-lucif erase gene constructs in 3T3-F442A adipocytes. Luciferase activity was measured in stably transfected 3T3-F442A differentiated adipocytes. Cells were cultivated for 24 h in medium containing 1 mmol/l or 25 mmol/l glucose. Mutations of E-box and GC-boxes are indicated by crosses. Two to three pools of clones from independent transfections were tested and gave similar results. Promoter activity data (luciferase activity/␮g protein) are expressed as fold induction over the activity of -each pool of clones in 1 mmol/l glucose. Means ؎ SD of three indepen dent experiments. *P < 0.05. cyte but not phosphorylated, had no effect. A lack of effect was also observed for 2-deoxyglucose, which is transported and phosphorylated into 2-deoxyglucose-6-phosphate but not further metabolized. Therefore, glucose-6- phosphate is not the metabolite mediating the glucose response. Fructose and mannose, which enter glycolysis at the glucose-6-phosphate level, partially restored luciferase activity. Dihydroxyacetone, which is converted into dihydroxyacetone phosphate, had no effect. Similarly, xylitol, an intermediate of the nonoxidative branch of the pentose FIG. 9. Effect of glucose analogs and metabolites on the ؊137/؉38 HSL promoter activity. Several stable pools of clones of the ؊137/؉38 HSL phosphate pathway, and glucosamine, which enters the promoter-luciferase gene construct were tested and gave similar re- hexosamine pathway, could not substitute for glucose. sults. Cells were grown to day 10 of differentiation, and DMEM was replaced by medium containing 1 mmol/l glucose in the presence of Increased glucosamine and xylitol concentrations up to 20 lactate and pyruvate for 24 h. R., metabolite addition during 24 h after mmol/l did not modify the promoter activity (data not the 24-h period in 1 mmol/l glucose (called reversion). Promoter shown). activity data (luciferase activity/␮g protein) obtained after the treatment period are normalized to the activity in 1 mmol/l glucose. .Means ؎ SD of ﬁve independent experiments. *P < 0.05 DISCUSSION Glucose-mediated regulation of gene expression has been Interestingly, this glucose-responsive region contains two reported for a growing set of genes involved in metabolic functional GC-boxes binding the Sp family of nuclear and energetic pathways in the cell (35). We have recently factors and an E-box binding USF. Sp1 has been impli- shown that HSL, the gene encoding the enzyme catalyzing cated in the glucose response of the acetyl CoA carboxy- the breakdown of triacylglycerol into fatty acids in adipose lase and leptin genes in adipocytes (27,36). It was tissue, is positively regulated by glucose in a mouse proposed that glucose-induced Sp1 dephosphorylation en- adipocyte cell line (13). Here, we show that HSL mRNA is hanced Sp1 binding and transcriptional activity of the also responsive to glucose in primary culture of human acetyl CoA carboxylase promoter II (37). Glucose was also subcutaneous adipocytes. Because the molecular mecha- shown to promote glucosamine-mediated O-linked glyco- nisms underlying the glucose response are largely un- sylation of Sp1, which protects the transcription factor known in the adipocytes, we sought to determine the from proteolytic degradation (38). The glucosamine path- cis-acting elements in the human adipocyte HSL promoter way is involved in the regulation of the leptin gene (39). mediating upregulation by glucose. However, our data do not support a role for Sp1 or Sp3 in Determination of reporter gene activity in differentiated the glucose response of the HSL promoter. Indeed, indi- 3T3-F442A adipocytes showed that the proximal 137 bp of vidual and double mutations of the GC-boxes did not alter the HSL promoter contained a glucose-responsive region. the glucose response of several pools of stably transfected

298 DIABETES, VOL. 51, FEBRUARY 2002 F. SMIH AND ASSOCIATES

3T3-F442A adipocyte clones. Moreover, in our experimen- adipose tissue (39). The lack of glucosamine effect ex- tal conditions, no difference in Sp1 and Sp3 DNA binding cluded this pathway and its end product UDP-N-acetylglu- activity was observed with nuclear extracts prepared from cosamine in HSL transcriptional activation. Together, the either glucose-deprived or glucose-fed adipocytes (data data suggest that a metabolite generated during glycolysis not shown). However, we showed that the GC-boxes are between glucose-6-phosphate and triose phosphate could necessary for full HSL promoter activity in adipocytes. regulate the transcription of the human HSL gene. The A typical E-box motif, CACGTG, was located between fructose and mannose stimulatory effects are in agreement Ϫ106 and Ϫ101 in the HSL promoter. USF1/USF2 het- with this scenario, since they enter glycolysis at the erodimers from adipocyte nuclear extracts bound to the glucose-6-phosphate level. Hence, the metabolic steps E-box. Moreover, overexpression of the transcription fac- involved in the regulation of HSL promoter activity define tors increased the Ϫ137/ϩ38 construct–mediated lucif- a novel mechanism for the transcriptional effect of glu- erase activity, and mutation of the E-box lowered cose. promoter activity in rat adipocytes. The importance of In conclusion, we have shown that an E-box and two USF was also observed in vivo. Indeed, HSL gene expres- GC-boxes are critical cis-acting elements of the human sion was decreased in adipose tissue of USF1- and USF2- adipocyte HSL promoter. USF, through binding to the deficient mice compared with wild-type mice. Together, E-box, may participate in the promoter transactivation. these data demonstrate that USF1 and USF2 could partic- The E-box also mediates the glucose-mediated induction ipate in the transactivation of the HSL gene in adipose of HSL gene expression. Metabolism of glucose at a step tissue. Mutation of the HSL E-box abolished the glucose between glucose-6-phosphate and triose phosphates is response in stably transfected 3T3-F442A adipocytes. E- required. Further work is needed to identify the metabo- boxes have been identified as parts of glucose response lites and the transcription factor complex involved in the glucose response. elements for several hepatic genes (26). In USF1 and USF2 knockout mice (23–25), as well as in USF2-deficient hepa- tocytes in primary culture (S.V., unpublished results), ACKNOWLEDGMENTS endogenous USF was shown to be important for carbohy- This work was supported by grants from Fondation pour drate induction of hepatic genes. However, as previously la Recherche Medicale (to P.R.) and Produits Roche reported (26), these transcription factors cannot by them- France (to D.L.). F.S. was supported by Groupe Danone. selves explain the transcriptional regulation by glucose. We are indebted to J.P. Salier (INSERM U 519, Rouen, USF binding to the HSL E-box did not differ between France) and P.J. Romanienko (National Institutes of nuclear extracts from adipose cells grown in the presence Health, Bethesda, MD) for critical reading of the manu- or absence of glucose (data not shown). Our data show script. We also thank Corinne Lefort and Lydia Pinard for that the HSL E-box is an essential part of the glucose excellent technical assistance. response element. Whereas USF may be indirectly involved in the transcriptional activation by glucose, the REFERENCES data indicate that other factors are part of the glucose 1. Langin D, Holm C, Lafontan M: Adipocyte hormone-sensitive lipase: a sensor complex. A good candidate is the recently identi- major regulator of lipid metabolism. Proc Nutr Soc 55:93–109, 1996 fied carbohydrate response element binding protein, 2. Large V, Arner P, Reynisdottir S, Grober J, Van Harmelen V, Holm C, Langin D: Hormone-sensitive lipase expression and activity in relation to which is a member of the basic helix-loop-helix leucine lipolysis in human fat cells. J Lipid Res 39:1688–1695, 1998 zipper family of transcription factors expressed in liver 3. Osuga J, Ishibashi S, Oka T, Yagyu H, Tozawa R, Fujimoto A, Shionoiri F, (40). It is not known whether the factor is expressed in Yahagi N, Kraemer FB, Tsutsumi O, Yamada N: Targeted disruption of adipocytes and plays the same function as in liver. hormone-sensitive lipase results in male sterility and adipocyte hypertro- phy, but not in obesity. Proc Natl Acad SciUSA97:787–792, 2000 Several metabolic pathways have been proposed for 4. Wang SP, Laurin N, Himms-Hagen J, Rudnicki MA, Levy E, Robert MF, Pan glucose-mediated transcriptional activation (35). We L, Oligny L, Mitchell GA: The adipose tissue phenotype of hormone- therefore tested the effect of glucose analogs and metab- sensitive lipase deficiency in mice. Obes Res 9:119–128, 2001 olites on the activity of the Ϫ137/ϩ38 construct. In adi- 5. Hellstro¨ m L, Langin D, Reynisdottir S, Dauzats M, Arner P: Adipocyte pose tissue, hexokinase-dependent metabolism of glucose lipolysis in normal weight subjects with obesity among first-degree rela- tives. Diabetologia 39:921–928, 1996 into glucose-6-phosphate is required for the glucose-medi- 6. Large V, Reynisdottir S, Langin D, Fredby K, Klannemark M, Holm C, Arner ated induction of fatty acid synthase and acetyl CoA P: Decreased expression and function of adipocyte hormone-sensitive carboxylase gene expression (41). The lack of effect of lipase in subcutaneous fat cells of obese subjects. J Lipid Res 40:2059– 3-O-methyl glucose and 2-deoxyglucose on HSL promoter 2066, 1999 7. Magre´ J, Laurell H, Fizames C, Antoine PJ, Dib C, Vigouroux C, Bourut C, activity suggests that a metabolite downstream to glucose- Capeau J, Weissenbach J, Langin D: Human hormone-sensitive lipase: 6-phosphate in the glycolytic pathway was involved. How- genetic mapping, identification of a new dinucleotide repeat, and associ- ever, dihydroxyacetone, which has been shown to increase ation with obesity and NIDDM. Diabetes 47:284–286, 1998 hepatic glucagon receptor mRNA expression (42), did not 8. Klannemark M, Orho M, Langin D, Laurell H, Holm C, Reynisdottir S, Arner restore HSL promoter activity. It has been proposed that P, Groop L: The putative role of the hormone-sensitive lipase gene in the pathogenesis of type II diabetes mellitus and abdominal obesity. Diabeto- glucose acts in liver though the nonoxidative branch of the logia 41:1516–1522, 1998 pentose phosphate pathway to induce the L-pyruvate ki- 9. Foster DW, McGarry JD: The metabolic derangements and treatment of nase gene expression (43). Xylitol, which is converted into diabetic ketoacidosis. N Engl J Med 309:159–169, 1983 xylulose 5-phosphate, was unsuccessful in restoring HSL 10. Solomon SS, Heckemeyer CM, Barker JA, Duckworth WC: Hormonal control of lipolysis in perifused adipocytes from diabetic rats. Endocri- promoter activity. The hexosamine pathway may also nology 117:1350–1354, 1985 mediate the effect of glucose on gene expression, as 11. Sztalryd C, Kraemer FB: Regulation of hormone-sensitive lipase in strep- proposed for leptin expression in skeletal muscle and tozotocin-induced diabetic rats. Metabolism 44:1391–1396, 1995

DIABETES, VOL. 51, FEBRUARY 2002 299 HORMONE-SENSITIVE LIPASE REGULATION BY GLUCOSE

12. Pleé-Gauthier E, Grober J, Duplus E, Langin D, Forest C: Inhibition of 28. Viollet B, Lefrançois-Martinez A-M, Henrion A, Kahn A, Raymondjean M, hormone-sensitive lipase gene expression by cAMP and phorbol esters in Martinez A: Immunochemical characterization and transacting properties 3T3–F442A and BFC-1 adipocytes. Biochem J 318:1057–1063, 1996 of upstream stimulatory factor isoforms. J Biol Chem 271:1405–1415, 1996 13. Raclot T, Dauzats M, Langin D: Regulation of hormone-sensitive lipase 29. Shih HM, Liu Z, Towle HC: Two CACGTG motifs with proper spacing expression by glucose in 3T3–F442A adipocytes. Biochem Biophys Res dictate the carbohydrate regulation of hepatic gene transcription. J Biol Commun 245:510–513, 1998 Chem 270:21991–21997, 1995 14. Botion LM, Green A: Long-term regulation of lipolysis and hormone- 30. Letovsky J, Dynan W: Measurement of the binding of transcription factor sensitive lipase by insulin and glucose. Diabetes 48:1691–1697, 1999 Sp1 to a single GC-box recognition sequence. Nucl Acids Res 17:2639– 15. Langin D, Laurell H, Stenson Holst L, Belfrage P, Holm C: Gene organiza- 2653, 1989 tion and primary structure of human hormone-sensitive lipase: possible 31. Foretz M, Guichard C, Ferre P, Foufelle F: Sterol regulatory element significance of a sequence homology with a lipase of Moraxella TA144, an binding protein-1c is a major mediator of insulin action on the hepatic antarctic bacterium. Proc Natl Acad SciUSA90:4897–4901, 1993 expression of glucokinase and lipogenesis-related genes. Proc Natl Acad 16. Stenson Holst L, Langin D, Mulder H, Laurell H, Grober J, Bergh A, SciUSA96:12737–12742, 1999 Mohrenweiser HW, Edgren G, Holm C: Molecular cloning, genomic orga- 32. Foretz M, Pacot C, Dugail I, Lemarchand P, Guichard C, Le Lie`pvre X, nization, and expression of a testicular isoform of hormone-sensitive Berthelier-Lubrano C, Spiegelman B, Kim JB, Ferre´ P, Foufelle F: ADD1/ Genomics lipase. 35:441–447, 1996 SREBP-1c is required in the activation of hepatic lipogenic gene expres- 17. Grober J, Laurell H, Blaise R, Fabry B, Schaak S, Holm C, Langin D: sion by glucose. Mol Cell Biol 19:3760–3768, 1999 Characterization of the promoter of human adipocyte hormone-sensitive 33. Kim JB, Spiegelman BM: ADD1/SREBP1 promotes adipocyte differentia- lipase. Biochem J 328:453–461, 1997 tion and gene expression linked to fatty acid metabolism. Genes Dev 18. Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, 10:1096–1107, 1996 Stru¨ hl K: Current Protocols in Molecular Biology. New York, Wiley & 34. Medvedev AV, Snedden SK, Raimbault S, Ricquier D, Collins S: Transcrip- Sons, 2001 tional regulation of the mouse uncoupling protein-2 gene: double E-box 19. Hauner H, Entenmann G, Wabitsch M, Gaillard D, Ailhaud G, Negrel R, Pfeiffer EF: Promoting effect of glucocorticoids on the differentiation of motif is required for peroxisome proliferator-activated receptor-gamma- human adipocyte precursor cells cultured in a chemically defined medium. dependent activation. J Biol Chem 276:10817–10823, 2001 J Clin Invest 84:1663–1670, 1989 35. Girard J, Ferre´ P, Foufelle F: Mechanisms by which carbohydrates regulate 20. Rolland V, Dugail I, Le Liepvre X, Lavau M: Evidence of increased expression of genes for glycolytic and lipogenic enzymes. Annu Rev Nutr glyceraldehyde-3-phosphate dehydrogenase and fatty acid synthetase pro- 17:325–352, 1997 moter activities in transiently transfected adipocytes from genetically 36. Fukuda H, Iritani N: Transcriptional regulation of leptin gene promoter in obese rats. J Biol Chem 270:1102–1106, 1995 rat. FEBS Lett 455:165–169, 1999 21. Millet L, Vidal H, Andreelli F, Larrouy D, Riou JP, Ricquier D, Laville M, 37. Daniel S, Zhang S, DePaoli-Roach AA, Kim K-H: Dephosphorylation of Sp1 Langin D: Increased uncoupling protein-2 and -3 mRNA expression during by protein phosphatase 1 is involved in the glucose-mediated activation of fasting in obese and lean humans. J Clin Invest 100:2665–2670, 1997 the acetyl-CoA carboxylase gene. J Biol Chem 271:14692–14697, 1996 22. Rouet P, Raguenez G, Tronche F, Yaniv M, NGuyen C, Salier J: A potent 38. Han I, Kudlow JE: Reduced O glycosylation of Sp1 is associated with enhancer made of clustered liver-specific elements in the transcription increased proteasome susceptibility. Mol Cell Biol 17:2550–2558, 1997 control sequences of human alpha 1-microglobulin/bikunin gene. J Biol 39. Wang J, Liu R, Hawkins M, Barzilai N, Rossetti L: A nutrient-sensing Chem 267:20765–20773, 1992 pathway regulates leptin gene expression in muscle and fat. Nature 23. Vallet VS, Henrion AA, Bucchini D, Casado M, Raymondjean M, Kahn A, 393:684–688, 1998 Vaulont S: Glucose-dependent liver gene expression in upstream stimula- 40. Yamashita H, Takenoshita M, Sakurai M, Bruick RK, Henzel WJ, Shil- tory factor 2 –/– mice. J Biol Chem 272:21944–21949, 1997 linglaw W, Arnot D, Uyeda K: A glucose-responsive transcription factor 24. Vallet VS, Casado M, Henrion AA, Bucchini D, Raymondjean M, Kahn A, that regulates carbohydrate metabolism in the liver. Proc Natl Acad Sci Vaulont S: Differential roles of upstream stimulatory factors 1 and 2 in the USA98:9116–9121, 2001 transcriptional response of liver genes to glucose. J Biol Chem 273:20175– 41. Foufelle F, Gouhot B, Pegorier JP, Perdereau D, Girard J, Ferre P: Glucose 20179, 1998 stimulation of lipogenic enzyme gene expression in cultured white adipose 25. Casado M, Vallet V, Kahn A, Vaulont S: Essential role in vivo of upstream tissue: a role for glucose-6-phosphate. J Biol Chem 267:20543–20546, 1992 stimulatory factors for a normal dietary response of the fatty acid synthase 42. Burcelin R, Mrejen C, Decaux JF, Hauguel De Mouzon S, Girard J, Charron gene in the liver. J Biol Chem 274:2009–2013, 1999 MJ: In vivo and in vitro regulation of hepatic glucagon receptor mRNA 26. Vaulont S, Vasseur-Cognet M, Kahn A: Glucose regulation of gene tran- concentration by glucose metabolism. J Biol Chem 273:8088–8093, 1998 scription. J Biol Chem 275:31555–31558, 2000 43. Doiron B, Cuif M-H, Chen R, Kahn A: Transcriptional glucose signaling 27. Daniel S, Kim KH: Sp1 mediates glucose activation of the acetyl-CoA through the glucose response element is mediated by the pentose phos- carboxylase promoter. J Biol Chem 271:1385–1392, 1996 phate pathway. J Biol Chem 271:5321–5324, 1996

300 DIABETES, VOL. 51, FEBRUARY 2002