Hexosamines Regulate Leptin Production in 3T3-L1 Adipocytes ...

0013-7227/02/$03.00/0 Printed in U.S.A.

Endocrinology 143(1):99 –106 Copyright © 2002 by The Endocrine Society

Hexosamines Regulate Leptin Production in 3T3-L1 Adipocytes through Transcriptional Mechanisms PEILI ZHANG, ELLEN S. KLENK, MARC A. LAZZARO, LLOYD B. WILLIAMS, CONSIDINE

AND

ROBERT V.

Division of Endocrinology and Metabolism, Indiana University School of Medicine, Indianapolis, Indiana 46202 This study was undertaken to examine the regulation of leptin gene (LEP) transcription and leptin release by hexosamines in 3T3-L1 adipocytes. Glucosamine (1 mM), an intermediate in hexosamine biosynthesis, increased leptin release to 117.0 ⴞ 7.3% (P ⴝ 0.0430; n ⴝ 9) and 134.6 ⴞ 6.5% of the control value (P ⴝ 0.0367; n ⴝ 4) by 48 and 96 h, respectively. With 0.01 mM glucosamine, leptin release was increased to 120.0 ⴞ 3.0% of the control value (P ⴝ 0.0069; n ⴝ 4) by 96 h of treatment. Glucose at 5 and 20 mM stimulated leptin release to 759 ⴞ 227% and 1104 ⴞ 316% of the control value over the 96-h culture period. Inhibition of hexosamine biosynthesis with 6-diazo-5oxonorleucine (20 ␮M) reduced glucose-stimulated leptin release 13 ⴞ 2.3% and 29.9 ⴞ 6.6% at 24 and 96 h, respectively (n ⴝ 4; P < 0.05). A 24-h incubation in 5 mM glucose significantly increased (163.0 ⴞ 19.3%; n ⴝ 7) the activity of a human LEP

promoter electroporated into differentiated 3T3-L1 cells. Glucosamine (1 mM; 48 h) also increased LEP promoter activity 170.0 ⴞ 13.0% (n ⴝ 5). Mutation of the three Sp1 binding sites in the LEP construct significantly reduced promoter activity. However, glucose (5 mM; 24 h) and glucosamine (1 mM; 48 h) increased the activity of the mutated promoter to 165 ⴞ 40% (n ⴝ 8) and 143 ⴞ 13% of the control value (n ⴝ 8). Glucosamine significantly increased O-glycosylation of Sp1 by 16.1 ⴞ 4.5% (P ⴝ 0.0305; n ⴝ 3). These data demonstrate that glucose and hexosamines regulate leptin production through transcriptional mechanisms localized to the proximal portion of the LEP promoter. Hexosamine-mediated regulation of LEP gene expression does not depend on Sp1 binding to traditional sites on the promoter. (Endocrinology 143: 99 –106, 2002)

T

HE BIOSYNTHESIS of hexosamines has been hypothesized to be one measure of nutrient flux into the cell (1). This metabolic pathway for glucose utilization has been implicated in the development of insulin resistance in skeletal muscle and adipose tissue (2), and more recently in the regulation of leptin production in rodent adipocytes (3, 4). The end product of hexosamine biosynthesis, UDP-N-acetylglucosamine, is used in O-linked glycosylation reactions in which the N-acetylglucosamine moiety is covalently linked to serine or threonine residues of proteins. Several studies suggest that O-linked glycosylation/deglycosylation may provide a regulatory function analogous to protein phosphorylation/dephosphorylation (5). Many transcription factors have been demonstrated to contain O-linked N-acetylglucosamine, and it has been hypothesized that this modification may influence the transcriptional activity of these factors (6). O-Linked glycosylation of transcription factors that specifically act on the LEP gene promoter may link hexosamine biosynthesis to the regulation of leptin synthesis. We recently reported that the hexosamine biosynthetic pathway regulates leptin production in human adipose tissue (7). In this study we found a significant positive correlation between serum leptin and UDP-N-acetylglucosamine in sc adipose tissue. We also observed that exposure of human adipocytes to glucosamine, an intermediate in hexosamine biosynthesis, increased leptin release into the culture medium. In contrast, inhibition of the rate-limiting enzyme in hexosamine biosynthesis, glutamine:fructose

amidotransferase (GFAT), with 6-diazo-5-oxo-l-norleucine resulted in a significant attenuation of leptin production from cultured sc adipocytes. These findings strongly suggest that hexosamine biosynthesis regulates leptin production in human adipose tissue. The following studies were undertaken to more fully elucidate the mechanism through which hexosamines regulate leptin production. 3T3-L1 adipocytes were used to study changes in LEP gene promoter activity in response to alterations in hexosamine biosynthesis. The data suggest that metabolism of glucose to hexosamines regulates leptin production in 3T3-L1 cells through transcriptional mechanisms acting within the proximal promoter region. Materials and Methods Cell culture 3T3-L1 preadipocytes were differentiated to mature adipocytes by standard techniques (8) and were used between 7 and 14 d after the initiation of differentiation. Differentiated cells were maintained in DMEM, 10% FBS, and 10⫺6 m insulin until use. The specific medium used for each experiment is described in detail in Results. Medium was replaced with fresh medium every 24 h. DMEM (D5030), glucosamine (G4875), insulin (I5500), and 6-diazo-5-oxo-l-norleucine (D2141) were purchased from Sigma (St. Louis, MO). Sodium azide (BP922–500) was obtained from Fisher Scientific (Fairlawn, NJ), and BSA (no. 160069) was obtained from ICN Biomedicals, Inc. (Aurora, OH). FBS (no. 16000-036) was purchased from Life Technologies, Inc. (Grand Island, NY).

Leptin release For measurement of leptin released to the medium, cells were plated to confluence and differentiated in six-well Falcon culture dishes (Fisher Scientific). During each experiment cells were maintained in 1 ml culture medium. Leptin was measured in 200-␮l aliquots of culture medium in duplicate using a commercially available RIA kit (Linco Research, Inc.,

Abbreviations: DON, 6-Diazo-5-oxo-l-norleucine; GFAT, glutamine: fructose amidotransferase.

99

100

Endocrinology, January 2002, 143(1):99 –106

Zhang et al. • Hexosamine-Regulated Leptin Production

St. Charles, MO). The limit of detection of this assay is 0.5 ng/ml. The within- and between-assay coefficients of variation are 4.6% and 5.0% at 7.2 ng/ml.

Measurement of LEP mRNA LEP mRNA was determined by RT-PCR as previously described (9) using primers specific for the mouse LEP gene (upstream, 5⬘-atgtgctggagacccctgtg-3⬘; downstream, 5⬘-gggctaacatccaactgttg-3⬘). All comparisons between samples were made on the linear portion of the amplification curve (between cycles 20 –35), and no product was obtained in the absence of reverse transcriptase. The data are expressed as the ratio of LEP cDNA to actin cDNA. There was no difference in the amount of actin cDNA among the samples studied.

Transfection of differentiated 3T3-L1 cells Differentiated 3T3-L1 cells were transfected by electroporation as previously described (10). Briefly, differentiated cells were trypsinized on d 8 from 100-cm dishes, washed, and resuspended in Dulbecco’s PBS without Mg2⫹ or Ca2⫹ (Life Technologies, Inc.) at a concentration of approximately 1 ⫻ 107 cells/0.5 ml in the electroporation cuvette. The plasmid DNA to be transfected (300 ␮g) was added, and the cells were electroporated with a single pulse of 960 ␮F capacitance at 0.16 kV using a Gene Pulser (model 1652076, Bio-Rad Laboratories, Inc., Hercules, CA). The electroporated cells were then plated into six-well dishes in DMEM/ 10% FBS. The culture medium was changed to fresh DMEM/10% FBS 12 h after electroporation. Experiments were started 24 h after electroporation by changing to the appropriate culture medium. The transfection efficiency was approximately 50%.

Plasmids The LEP gene promoter construct was the gift of Dr. Da-Wei Gong, NIH (11). The proximal promoter sequence (⫺215 to ⫺1 bp) and the first exon (⫹1 to ⫹29 bp) were subcloned into the pGL3-basic vector containing the luciferase reporter gene. The plasmids were transformed into DH10B competent cells (Life Technologies, Inc.), which were grown under standard conditions. The plasmid DNA was isolated with Maxiprep kits (QIAGEN, Valencia, CA). Incorporation of point mutations to destroy Sp1-binding sites (Fig. 2) were made using the site-directed mutagenesis system (Promega Corp., Madison, WI) as previously described (12).

Western blotting Total cell lysates were prepared by solubilization of the 3T3-L1 cells in buffer containing 20 mm Tris-HCl (pH 7.5), 150 mm NaCl, 1 mm EDTA, 0.5% Triton X-100, 0.1% SDS, and protease inhibitor cocktail (no. 1687498, Roche Molecular Biochemicals, Mannheim, Germany). The protein content of the lysates was determined using the DC protein assay kit (Bio-Rad Laboratories, Inc.). Cell lysates (40 ␮g protein) were elec-

trophoresed through a 7.5% polyacrylamide gel and transferred to supported nitrocellulose (Life Technologies, Inc.). Blots were probed with primary antibodies to Sp1 (SC-59, Santa Cruz Biotechnology, Inc., Santa Cruz, CA) and O-glycosylated protein (RL2, Affinity BioReagents, Inc., Golden, CO) at a dilution of 1:666 for both. Secondary antibody was antirabbit IgG coupled to horseradish peroxidase (Amersham Pharmacia Biotech, Piscataway, NJ) used at a dilution of 1:1000. Protein bands were detected using the ECL chemiluminescence kit (Amersham Pharmacia Biotech, Piscataway, NJ) and Kodak X-OMAT film (Fisher Scientific). The films were scanned, and the band intensity was quantitated using Quantity One software (Bio-Rad Laboratories, Inc.).

Statistical methods All data in the text and tables are the mean ⫾ sem. Statistical comparisons were made using paired t test or, in the case of data not normally distributed, a one-sample t test of the percent change between treatment and control values. No adjustments for multiple comparisons were made. n denotes one comparison (control and treatment) on adipocytes in one experiment.

Results Glucosamine-stimulates leptin release from 3T3-L1 adipocytes

We have recently shown that treatment of human sc adipocytes with glucosamine increases leptin release. To investigate further the ability of hexosamines to regulate leptin production, we examined glucosamine-induced leptin release from differentiated 3T3-L1 adipocytes under conditions designed to maximize the response (glucose- and glutaminefree DMEM containing 1% BSA, 1 mm pyruvate, and 10⫺6 m insulin). In the absence of glucose 3T3-L1 adipocytes continuously released leptin into the medium throughout the 96-h treatment period. Within the first 24 h of treatment with 1 mm glucosamine, leptin released to the medium was reduced 13.6 ⫾ 5% compared with the control (Table 1). However, by 48 h of treatment, leptin released to the medium was significantly greater than that from untreated cells, an effect that was maintained until the end of the 96-h period. Glucosamine stimulates leptin production in a time- and dose-dependent manner. As shown in Table 1, leptin released into the medium by 48 h of culture with 0.1 mm glucosamine was significantly increased 14.7 ⫾ 4.7% over the control value. By 96 h in culture, 0.1 mm glucosamineinduced leptin release was similar to that induced by 1 mm

TABLE 1. Leptin production from 3T3-L1 adipocytes cultured in glucose or glucosamine Treatment

Glucosamine (1 mM; n ⫽ 9 at 24 and 48 h; n ⫽ 4 at 72 and 96 h) Glucosamine (0.1 mM; n ⫽ 4) Glucosamine (0.01 mM; n ⫽ 4) Glucose (5 mM; n ⫽ 7) Glucose (20 mM; n ⫽ 7) Glucose (5 mM) ⫹ DON (20 ␮M; n ⫽ 4)

Leptin released (% of control) 24 h

86.4 ⫾ 5.0 105.5 ⫾ 4.7 107.3 ⫾ 4.5 396.0 ⫾ 117.0c 285.5 ⫾ 80.9c 87.0 ⫾ 2.3e

48 h

72 h

116.9 ⫾ 7.3a

134.7 ⫾ 5.7a

114.7 ⫾ 4.7b 107.7 ⫾ 4.5 601.2 ⫾ 192.6c 650.7 ⫾ 195.3c 75.5 ⫾ 3.3e

125.4 ⫾ 5.8a 109.6 ⫾ 5.1 705.0 ⫾ 238.0c 917.0 ⫾ 287.6c,d 72.1 ⫾ 5.5e

96 h

134.6 ⫾ 6.5a 131.9 ⫾ 6.4a 120.0 ⫾ 3.0a 759.4 ⫾ 227.1c 1104.4 ⫾ 316.0c,d 70.1 ⫾ 6.6e

Differentiated adipocytes were cultured in DMEM and 1% BSA, 10⫺6 M insulin, and glucose or glucosamine as indicated. Differentiated adipocytes treated with DON were cultured in DMEM and 10% FBS, 10⫺6 M insulin, and glucose as indicated. Values represent the mean ⫾ SEM. a P ⬍ 0.05 compared with cells cultured without glucosamine. b P ⬍ 0.0518 compared with cells cultured without glucosamine. c P ⬍ 0.05 compared with cells cultured without glucose. d P ⬍ 0.05 compared with cells cultured with 5 mM glucose. e P ⬍ 0.05 compared with cells not treated with DON.


glucosamine. Glucosamine at a concentration of 0.01 mm had a small effect on leptin production, which became significant during the last 24 h of culture. Glucosamine-stimulated leptin release in 3T3-L1 cells represents only a part of that induced by glucose. As shown in Table 1, 5 mm glucose stimulated leptin production to 759 ⫾ 227% of the control value during 96 h in culture compared with a maximal increase of 134.6 ⫾ 6.5% of the control value induced by 1 mm glucosamine. Interestingly, increasing the glucose concentration to 20 mm did not increase leptin production over that achieved with 5 mm glucose during the first 48 h of culture. However, 20 mm glucose induced significantly more leptin production than 5 mm glucose at 72 and 96 h of culture. The effect of glucose and glucosamine on leptin production is additive at low glucose concentrations. Incubation of cells with 1 mm glucosamine in the presence of 1.25 mm glucose resulted in a significant increase in release (119 ⫾ 6% of glucose alone; n ⫽ 5, P ⫽ 0.0156) by 48 h in culture. This additive effect was maintained until 96 h of culture (118 ⫾ 5% of glucose alone; n ⫽ 5; P ⫽ 0.0170). Coincubation of glucosamine with higher concentrations of glucose (2.5 or 5 mm) resulted in no additional release of leptin over that achieved with the glucose alone, most likely due to competition between glucose and glucosamine for uptake into the cell. It has been reported that exposure to glucosamine can reduce the concentration of ATP in 3T3-L1 adipocytes (13). To determine whether glucosamine increased leptin production from 3T3-L1 adipocytes through a toxic effect to deplete intracellular ATP, we cultured the cells for 48 h in the presence of the mitochondrial toxin sodium azide. At a concentration of 7 mm, sodium azide inhibited leptin release within the first 24 h of treatment (2.29 ⫾ 0.77 vs. 0.20 ⫾ 0.07 ng/ml; n ⫽ 5; P ⫽ 0.0004). Leptin release in the first 24 h was also significantly inhibited in the presence of 1 mm azide (0.83 ⫾ 0.30 ng/ml; P ⫽ 0.0236). Inhibition of leptin release was maintained up to 48 h in the presence of 7 and 1 mm azide (3.58 ⫾ 1.32 vs. 0.15 ⫾ 0.05 and 0.84 ⫾ 0.26 ng/ml for control, 7 and 1 mm azide respectively; n ⫽ 4; P ⱕ 0.0191). Addition of 0.1 mm azide to the culture medium had no effect on leptin production. These observations demonstrate that a reduction in intracellular ATP is not a signal for increased leptin production.

Endocrinology, January 2002, 143(1):99 –106 101

maintained throughout the treatment period, reaching a maximum 29.9 ⫾ 6.6% reduction in release by 96 h. The reduction in leptin release in the presence of DON was associated with a decrease in LEP gene expression. By 48 h of DON treatment the amount of LEP mRNA was reduced 21.0 ⫾ 4.9% (12.56 ⫾ 0.37 vs. 9.88 ⫾ 0.57 relative units for control and DON treatment, respectively; n ⫽ 6; P ⫽ 0.0100). This reduction in LEP mRNA is comparable to the reduction in leptin release to the medium observed at 48 h. Hexosamines regulate the LEP gene promoter

To understand the mechanism through which hexosamines regulate leptin production the proximal portion of the human LEP gene promoter (⫺215 to ⫹29 bp) was transfected by electroporation into differentiated 3T3-L1 adipocytes, and the cells were cultured for 24 and 48 h. In the absence of glucose the luciferase activity of the LEP promoter construct was 91 times greater than that from the plasmid without the promoter sequence (18 ⫾ 9 light units for three observations in glucose-free medium). Addition of 5 mm glucose to the culture medium for 24 h significantly increased LEP promoter activity 63.0 ⫾ 19.3% over that in the absence of glucose (Fig. 1). In agreement with the leptin release data in Table 1, increasing the glucose concentration to 20 mm had no additional stimulatory effect on LEP promoter activity measured at 24 h (Fig. 1). Addition of 20 ␮m DON to cells cultured in 5 mm glucose resulted in a significant reduction in LEP promoter activity. As observed for glucose, glucosamine also increased LEP promoter activity. Glucosamine had no effect on LEP promoter activity in the first 24 h of treatment (125 ⫾ 20% control activity; n ⫽ 7). However, by 48 h glucosamine significantly increased LEP promoter activity to 170 ⫾ 13% of the control value (Fig. 3). These observations demonstrate that glucose regulates the transcription of leptin through actions localized within the proximal sequence of the LEP promoter and suggest a role for hexosamine biosynthesis in this regulation.

6-Diazo-5-oxo-L-norleucine (DON) attenuates leptin release from 3T3-L1 adipocytes

Culture of differentiated 3T3-L1 cells with glucosamine demonstrates that an increase in hexosamine biosynthesis stimulates leptin production. It is therefore reasonable that a reduction in hexosamine biosynthesis should decrease leptin release. DON is a competitive inhibitor of GFAT that reduces the flux of glucose through the hexosamine biosynthetic pathway. 3T3-L1 adipocytes were cultured in DMEM containing glucose (5 mm), glutamine (4 mm), and 10% FBS. As shown in Table 1, DON (20 ␮m) induced a small, but significant, 13.0 ⫾ 2.3% reduction in leptin release by 24 h of treatment. Significant inhibition of leptin release was then

FIG. 1. Glucose stimulates and DON inhibits LEP promoter activity. A human LEP gene promoter construct (⫺215 to ⫹29 bp) was electroporated into differentiated 3T3-L1 adipocytes. Cells were cultured in medium containing 5 or 20 mM glucose with or without 20 ␮M DON for 24 h. Values are the mean ⫾ SEM for seven paired experiments with 5 mM glucose and DON and six experiments with 20 mM glucose. *, P ⬍ 0.05 compared with activity in absence of glucose.

102


Sp1 is one of several transcription factors that have been observed to be O-glycosylated, and it has been hypothesized that this modification alters the transcriptional activity of this protein (6). Within the proximal promoter sequence of the LEP gene are three potential GC boxes (⫺128 to ⫺123, ⫺104 to ⫺92, and ⫺22 to ⫺16 bp) at which Sp1 could bind and influence LEP gene transcription (Fig. 2A). To determine whether LEP gene transcription is regulated by an effect of hexosamines on Sp1 protein binding to the promoter, the three GC boxes were sequentially disrupted by the introduction of clustered point mutations. We refer to the promoter with one mutated GC box (⫺104 to ⫺92) as Sp1-1X. The promoter with two mutated GC boxes (⫺104 to ⫺92 and ⫺128 to ⫺123) is called Sp1-2X, and the promoter with all

FIG. 2. Mutation of Sp1-binding sites in the LEP promoter and effect of glucose. A, Proximal region of the human LEP promoter with Sp1-binding sites (GC boxes) and C/EBP sites illustrated. Point mutations introduced to inactivate Sp1 binding in each GC box are noted as well as the name given to the mutated promoter construct. B, The three mutated LEP promoter constructs were electroporated into 3T3-L1 adipocytes, and the cells were cultured in medium with or without 5 mM glucose for 24 h. Values are the mean ⫾ SEM for eight, seven, and eight independent paired experiments for the Sp1-1X, Sp1-2X, and Sp1-3X promoters, respectively. *, P ⬍ 0.05 compared with activity in absence of glucose.


three GC boxes mutated is Sp1-3X. Mutation of the first GC box at ⫺104 to ⫺92 bp (Sp1-1X) resulted in a significant 6-fold reduction in activity at 24 h compared with the activity of the nonmutated promoter in glucose free medium (1621 ⫾ 531 vs. 283 ⫾ 92 relative light units; P ⫽ 0.0198). The activity of Sp1-2⫻ and Sp1-3X in glucose-free medium was also significantly lower (13- and 8-fold, respectively) than that of the nonmutated promoter (P ⫽ 0.0159). Despite the significant reduction in activity, the mutated promoters were, on the average, 11 times more active than reporter plasmid lacking an insert (18 ⫾ 9 relative light units). As shown in Fig. 2B, culture with 5 mm glucose for 24 h increased the activity of the Sp1-1X promoter to 139 ⫾ 7% of the control value (P ⫽ 0.0195). As observed for the nonmutated LEP promoter, increasing the glucose concentration to 20 mm had no additional stimulatory effect on the activity of Sp1-1X in the first 24 h (data not shown). In contrast to the LEP promoter, a 24-h exposure to 20 ␮m DON did not reduce the activity of Sp1-1X compared with that in 5 mm glucose alone (365 ⫾ 108 vs. 392 ⫾ 125 relative light units for glucose plus DON and glucose alone, respectively). As shown in Fig. 2B, a 24-h glucose treatment resulted in a significant increase in the activity of the Sp1-2X promoter (157 ⫾ 7% of control; P ⫽ 0.0526) as well as Sp1-3X (165 ⫾ 40% of control; P ⫽ 0.0122). As observed for the LEP promoter, 1 mm glucosamine did not significantly increase the activity of the Sp1-1X or Sp1-3X promoters in the first 24 h of treatment. The activity of Sp1-2X was significantly increased by a 24-h treatment with glucosamine (127 ⫾ 38 vs. 167 ⫾ 45 light units for control and treated, respectively; P ⫽ 0.0192). However, as shown in Fig. 3, culture of transfected cells with 1 mm glucosamine for 48 h significantly increased the activity of Sp1-1X (157 ⫾ 11% of control), Sp1-2X (147 ⫾ 15% of control), and Sp1-3X (143 ⫾ 13% of control).

FIG. 3. Glucosamine increases activity of the intact and mutated LEP promoter constructs. The intact LEP promoter and mutated promoter constructs described in Fig. 2 were electroporated into 3T3-L1 adipocytes, and the cells were cultured in glucose-free medium with or without 1 mM glucosamine for 48 h. Values are the mean ⫾ SEM for five, eight, seven, and eight independent paired experiments for the LEP, Sp1-1X, Sp1-2X, and Sp1-3X promoters, respectively. *, P ⬍ 0.05 compared with activity in absence of glucosamine.


Sp1 is O-glycosylated in 3T3-L1 cells

Sp1 has been shown to be glycosylated in NRK cells treated with glucosamine (14). To demonstrate that hexosamine biosynthesis results in glycosylation of transcription factors in differentiated 3T3-L1 cells, Western blots for Sp1 were performed on total cell lysates of differentiated cells treated with glucosamine. As illustrated in the representative blot in Fig. 4, Sp1 was detected as a single band of approximately 100 kDa. The identity of Sp1 was confirmed in a subsequent experiment by immunoprecipitation with Sp1specific antibody (data not shown). There was no difference in the amount of Sp1 protein between untreated and glucosamine-treated cells (5922 ⫾ 1672 vs. 6041 ⫾ 1795 density units for control and glucosamine treated, respectively; n ⫽ 3). However, culture of differentiated cells with 1 mm glucosamine significantly increased the amount of O-linked glycosylation 16.1 ⫾ 4.5% (Fig. 4, lower panel). These observations demonstrate that the O-glycosylation of Sp1 in 3T3-L1 adipocytes can be altered by changes in hexosamine biosynthesis.

FIG. 4. Glucosamine increases O-linked glycosylation of Sp1. 3T3-L1 adipocytes were treated with 1 mM glucosamine for 72 h. Total cell lysates were blotted with antibodies specific for Sp1 and O-glycosylation. Upper panel, Representative blot illustrating the detection of Sp1 as a single band of 100 kDa (left side) and the increase in Oglycosylation with glucosamine treatment (right side). Lower panel, Quantitation of band density demonstrating a significant increase in O-glycosylation of Sp1 with glucosamine. Values are the mean ⫾ SEM for three experiments. *, P ⫽ 0.0305 compared with untreated cells. C, Control; G, glucosamine-treated.


Discussion

Recent evidence obtained in both animal models (3, 4) and cultured human adipocytes (7) suggests that the biosynthesis of hexosamines plays an important regulatory role in leptin production. The findings of the current study, that glucosamine increases in a time- and dose-dependent manner and DON inhibits leptin production in 3T3-L1 adipocytes, provide additional support for this hypothesis. Further, we document that glucose and hexosamines regulate leptin production through transcriptional mechanisms localized to the proximal promoter region. Hexosamine-mediated regulation of transcription involves Sp1, as illustrated by the reduction in promoter activity after mutation of the GC boxes in the proximal promoter region. However, glucose/hexosamines can also influence LEP promoter activity through mechanisms that do not involve binding of Sp1 to the proximal promoter. In the current study 3T3-L1 adipocytes were cultured in various concentrations of glucosamine for up to 96 h. A significant effect of glucosamine to increase leptin release (117 ⫾ 7% of the control value) to the medium was detected by 48 h. Glucosamine-induced leptin release was further increased to 135 ⫾ 7% of the control value by 96 h of treatment. The effect of glucosamine to increase leptin release was both time and dose dependent, with significant stimulation achieved by 48 h of treatment with 0.1 mm, but not with 0.01 mm, glucosamine. However, a 96-h exposure to 0.01 mm glucosamine did result in a significant increase in leptin release. These data demonstrate that the synthesis of UDPN-acetylglucosamine from glucosamine stimulates leptin release. To maximize the effect of glucosamine to directly increase intracellular UDP-N-acetylglucosamine synthesis and to minimize other effects of glucose metabolism in these studies, 3T3-L1 cells were cultured with glucosamine in the absence of glucose (1 mm pyruvate was present in the medium). Glucosamine enters the cell through the GLUT4 transporter at one quarter the rate of glucose (15); therefore, it competes poorly for uptake into the cell. However, glucosamine significantly increased leptin release in the presence of a low (1.25 mm) concentration of glucose in the culture medium. This observation supports our findings of glucosamine-stimulated leptin production obtained under the nonphysiological condition of complete glucose deprivation. The stimulation of leptin release by glucosamine represents only a portion of that induced by 5 mm glucose. This is probably due to the utilization of glucose in additional metabolic pathways that influence leptin production, such as synthesis of ATP, in which glucosamine is not a suitable substitute. As discussed in greater detail below, the glucosamine-induced increase in LEP promoter activity was comparable to that induced by glucose, but required a 24-h longer incubation to be detectable. This observation further suggests that glucose metabolism regulates leptin production through mechanisms other than transcription, such as translation or movement of secretory vesicles. Of interest was the observation that increasing the medium glucose concentration from 5 to 20 mm resulted in greater leptin production only during the second 48 h in culture (i.e. from 48 –96 h).

104


These observations are consistent with those of Mueller et al. (16) in that leptin release from cultured rat adipocytes is more dependent on glucose than insulin in the medium. Muller et al. also observed that the effect of glucose concentration on leptin release was more apparent during longer culture periods. One interpretation for the delayed leptin response to 20 mm glucose may be that de novo lipid synthesis occurred to increase LEP gene transcription later in the culture period. Future experiments will be needed to address this possibility. Within the first 24 h of treatment with 1 mm glucosamine, leptin released to the medium was reduced 13.6 ⫾ 5.0% compared with the control in glucose-free medium. However, in the presence of a low (1.25 mm) glucose concentration, 1 mm glucosamine had no effect on leptin release (109 ⫾ 10% of release in the presence of 1.25 mm glucose alone; P ⫽ NS). Lower concentrations of glucosamine in glucose-free medium did not reduce leptin production at any time during the culture period. The early reduction in leptin release induced by 1 mm glucosamine was probably due to an effect of glucosamine to reduce intracellular ATP levels. Hresko et al. (13) have shown that a 2.5-h exposure to 1 mm glucosamine in glucose-free medium reduced intracellular ATP concentrations approximately 35% in 3T3-L1 cells. In our studies exposure of cells to 7 and 1 mm sodium azide resulted in almost complete inhibition of leptin production within the first 24 h of treatment, an effect that was maintained to 48 h. This experiment thus demonstrates that ATP depletion can inhibit leptin production and, in light of the observations of Hresko et al. (13), probably explains the early moderate reduction in leptin production in 3T3-L1 cells exposed to glucosamine in the absence of glucose. However, our finding that leptin production then increased to levels significantly greater than the control level during the next 24 h argues against glucosamine-induced reductions in ATP stimulating leptin production. Rather, our data are consistent with the idea that a glucosamine-mediated increment in UDP-Nacetylglucosamine results in increased leptin production. It is important to note that the culture medium used in these experiments contained pyruvate (1 mm), which would be used as an alternative, albeit less efficient, energy source for the generation of ATP in the absence of extracellular glucose. DON is a competitive inhibitor of GFAT, the rate-limiting enzyme in UDP-N-acetylglucosamine synthesis. At a concentration of 20 ␮m, DON significantly inhibited leptin release from 3T3-L1 cells within the first 24 h of treatment, and this effect was maintained for the remainder of the culture period. These data are consistent with our previous observation that DON can inhibit leptin release from human adipocytes and establish that the effect is maintained over a longer culture period. Also, as previously observed, the reduction in leptin release with DON treatment was associated with a decrease in LEP mRNA content, suggesting that hexosamines regulate either LEP gene transcription or mRNA stability. The findings presented above demonstrate that hexosamine biosynthesis in 3T3-L1 cells regulates leptin production in a manner consistent with that in human sc adipocytes. These cells are therefore a reasonable model system to study the regulation of human LEP promoter activity by hexosamines. To accomplish this goal, 3T3-L1 cells were


transfected with a luciferase reporter construct containing the proximal 215 bp (11) of the human LEP gene promoter. Although transfection of adipocytes is in general difficult, an electroporation technique for 3T3-L1 cells has been successfully developed that results in the incorporation of plasmid DNA into 50 –70% of differentiated cells (10). Using this technique, we examined the ability of glucose and glucosamine to regulate the activity of the LEP gene promoter. In agreement with previous reports, the proximal promoter region of the LEP gene was sufficient for basal transcriptional activity (11, 17, 18). Addition of 5 mm glucose to the culture medium significantly increased LEP promoter activity 63.0% over that in the absence of glucose. Increasing glucose further to 20 mm did not result in any additional stimulation of the promoter, in agreement with our observations of leptin production from nontransfected cells cultured under similar conditions. These findings thus suggest that a glucoseresponsive element is located within the proximal region of the human LEP promoter, in agreement with previous observations for rat (19) and mouse (20) genes. Two observations support the hypothesis that the promoter response to glucose may be at least in part mediated through hexosamine biosynthesis. The first is that glucosamine significantly increases the activity of the LEP promoter to about the same extent (70% over that in the absence of glucose) as that induced by 5 mm glucose, although this effect takes 24 h longer to achieve. The second observation is that inhibition of GFAT activity with DON attenuates the ability of glucose to increase LEP promoter activity. Taken together, these two findings support a role for UDP-N-acetylglucosamine synthesis in the regulation of the LEP promoter by glucose. Further support for the hypothesis is provided by the recent observation that insulin-stimulated glucose metabolism, but not insulin per se, mediates the effects of insulin to increase the activity of a mouse LEP promoter construct transfected into 3T3-L1 cells (20). Within the proximal region of the LEP gene promoter are three Sp1 consensus sequences (GC box) and a C/EBP motif (11). Mason et al. (12) demonstrated that the GC box located at ⫺104 to ⫺92 bp is an important determinant of murine LEP promoter activity. It has also been suggested that O-glycosylation of Sp1 regulates its transcriptional activity (14). To test whether hexosamines regulate the LEP promoter through Sp1 acting at this site, point mutations demonstrated by gel shift analysis to prevent Sp1 binding (12) were introduced within this consensus sequence. In agreement with the observations by Mason et al. (12), mutation of the GC box at ⫺104 to ⫺92 bp significantly reduced basal promoter activity 6-fold (determined in the absence of glucose); demonstrating that this Sp1-binding site is an important determinant of LEP promoter activity in humans as well as rodents. However, the addition of 5 mm glucose to the culture medium for 24 h induced a 40 ⫾ 7% increase in Sp1-1X promoter activity over that in the absence of glucose. Further, a 48-h incubation with glucosamine significantly increased the activity of the mutated Sp1-1X promoter by 57 ⫾ 11%. In both experiments the percent increase in activity of the mutated promoter was similar to that observed for the nonmutated promoter. These observations therefore suggest that hexosamines regulate LEP promoter activity at sites in addition to the GC box at


⫺104 to ⫺92 bp. There are two other GC boxes in the LEP promoter, one at ⫺17 to ⫺22 bp and the other at ⫺123 to ⫺128 bp. To determine whether hexosamines were stimulating LEP promoter activity through these other GC boxes, we generated promoter constructs with either two or all three of the GC boxes mutated. The creation of these additional mutations did not further reduce the basal activity of the promoter compared with that of Sp1-1X. Despite the introduction of the additional mutations, both glucose and glucosamine increased the activity of these promoters to levels not different from that observed for Sp1-1X. These observations therefore suggest that hexosamines can regulate the activity of the LEP promoter through actions that do not require these Sp1-binding sites. One possibility is that hexosamines may be acting on the mutated promoters through C/EBP transcription factors, important determinants of gene expression in adipocytes (21) that have been demonstrated to be responsive to nutritional manipulation (22). Sp1 has been demonstrated to alter the amount of C/EBP␣ in 3T3-L1 adipocytes through actions on the promoter of this transcription factor (23). Future experiments will be needed to determine whether hexosamines regulate LEP promoter activity through C/EBP␣ or other, as yet, unidentified transcription factors. Interestingly, treatment with DON did not block the glucose-induced increase in activity of the Sp1-1X promoter as was observed with the nonmutated promoter. One interpretation for this finding is that glucose is activating the mutated promoter through mechanisms that do not involve hexosamine synthesis. This cannot be the complete explanation given the observation that glucosamine also activates the mutated promoter. Alternatively, inhibition of GFAT for 24 h may not be sufficient to completely eliminate all intracellular hexosamine, and less hexosamine may be needed to activate the mutated promoter at sites other than the GC box at ⫺104 to ⫺92 bp. Future experiments will be needed to more fully examine this possibility. In support of the hypothesis that hexosamines regulate LEP gene transcription through effects on transcription factors was the finding of a modest increase in O-linked glycosylation of Sp1 in cells treated with glucosamine. Although the mechanism through which O-glycosylation alters Sp1-mediated transcriptional events is still not completely understood, it has been suggested that a reduction in Oglycosylation of Sp1 results in increased degradation of the transcription factor by the proteosome, an effect that can be abrogated by incubation with glucosamine (14). In our studies we did not observe any difference in immunodetectable Sp1 protein in 3T3-L1 adipocytes treated with or without glucosamine. An additional possible effect of O-glycosylation of Sp1 is that it regulates the interactions of Sp1 with other transcription factors and the DNA (24). Future experiments will be necessary to determine the exact mechanism(s) through which O-glycosylation regulates Sp1 protein function in adipocytes and whether this is the primary mechanism through which hexosamines regulate LEP gene transcription. Several limitations in the current study deserve mention. The first is the use of glucosamine and DON to manipulate hexosamine levels. As discussed above, glucosamine may


reduce the intracellular ATP content of the adipocyte. DON, as an inhibitor of glutamine transamidases (25), could have altered the activity of enzymes other than GFAT to influence leptin production. However, the results of pharmacological manipulation of hexosamine levels in the current study are in agreement with findings in transgenic animals overexpressing GFAT (4). One other concern is the variability in promoter activity from experiment to experiment, which may be due to a combination of transfection efficiency and the age of the cells transfected. To minimize variability, paired control treatments (i.e. no glucose condition) were performed in each experiment. Finally, although the activity of the LEP promoter in 3T3-L1 cells appears to be less than that in studies using transfected rat adipocytes (12), a lower promoter activity would be in agreement with observations that 3T3-L1 cells synthesize less leptin than primary adipocytes (26). In summary, the current study demonstrates that leptin release from 3T3-L1 adipocytes can be regulated by hexosamine biosynthesis as previously observed for human sc adipocytes. Further, glucose/hexosamines regulate leptin production through transcriptional mechanisms within the proximal promoter region. Although glucosamine-stimulated hexosamine biosynthesis can induce increases in LEP promoter activity comparable to that produced by glucose, actual leptin secretion is greater in the presence of glucose than glucosamine. These findings suggest that glucose regulates leptin production through mechanisms in addition to hexosamine biosynthesis. Sp1 is important in the regulation of the LEP promoter, but hexosamines may also regulate the promoter through other means. Hexosamines act as an intracellular signal linking glucose metabolism in adipocytes to leptin release. Acknowledgments The authors thank Dr. Dai-Wei Gong at the NIH for the generous gift of the ob promoter construct, and Dr. Jeffrey S. Elmendorf, Indiana University School of Medicine, for assistance with the electroporation technique. Received October 6, 2000. Accepted September 10, 2001. Address all correspondence and requests for reprints to: Robert V. Considine, Ph.D., Indiana University School of Medicine, 541 North Clinical Drive, Clinical Building 455, Indianapolis, Indiana 46202-5111. E-mail: [email protected]. This work was supported in part by grants from the NIH (DK-51140) and the American Diabetes Association. Portions of this work were presented in preliminary form at the 60th Scientific Sessions of the American Diabetes Association, San Antonio, Texas, June 2000.

References 1. Marshall S, Garvey WT, Traxinger RR 1991 New insights into the metabolic regulation of insulin action and insulin resistance: role of glucose and amino acids. FASEB J 5:3031–3036 2. McClain DA, Crook ED 1996 Hexosamines and insulin resistance. Diabetes 45:1003–1009 3. Wang J, Liu R, Hawkins M, Barzilai N, Rossetti L 1998 A nutrient-sensing pathway regulates leptin gene expression in muscle and fat. Nature 393: 684 – 688 4. McClain DA, Alexander T, Cooksey RC, Considine RV 2000 Hexosamines stimulate leptin production in transgenic mice. Endocrinology 141:1999 –2002 5. Wells L, Vosseller K, Hart GW 2001 Glycosylation of nucleocytoplasmic proteins: signal transduction and O-GlcNAc. Science 291:2376 –2378 6. Hayes BK, Hart GW 1998 Protein O-GlcNAcylation: potential mechanisms for the regulation of protein function. Adv Exp Med Biol 435:85–94

106 Endocrinology, January 2002, 143(1):99 –106

7. Considine RV, Cooksey RC, Williams LB, Fawcett RL, Zhang P, Ambrosius WT, Whitfield RM, Jones RM, Inman M, Huse J, McClain DA 2000 Hexosamines regulate leptin production in human subcutaneous adipocytes. J Clin Endocrinol Metab 85:3551–3556 8. Student AK, Hsu RY, Lane MD 1980 Induction of fatty acid synthetase synthesis in differentiating 3T3–L1 preadipocytes. J Biol Chem 255:4745– 4750 9. Considine RV, Sinha MK, Heiman ML, Kriaucinas A, Stephens TW, Nyce MR, Ohannesian JP, Marco CC, McKee LM, Bauer TL, Caro JF 1996 Serum immunoreactive-leptin concentrations in normal-weight and obese humans. N Engl J Med 334:292–295 10. Elmendorf JS, Boeglin DJ, Pessin JE 1999 Temporal separation of insulinstimulated GLUT4/IRAP vesicle plasma membrane docking and fusion in 3T3L1 adipocytes. J Biol Chem 274:37357–37361 11. Gong DW, Bi S, Pratley RE, Weintraub BD 1996 Genomic structure and promoter analysis of the human obese gene. J Biol Chem 271:3971–3974 12. Mason MM, He Y, Chen H, Quon MJ, Reitman M 1998 Regulation of leptin promoter function by Sp1, C/EBP, and a novel factor. Endocrinology 139: 1013–1022 13. Hresko RC, Heimberg H, Chi MM-Y, Mueckler M 1998 Glucosamineinduced insulin resistance in 3T3-L1 adipocytes is caused by depletion of intracellular ATP. J Biol Chem 273:20658 –20668 14. Han I, Kudlow JE 1997 Reduced O glycosylation of Sp1 is associated with increased proteasome susceptibility. Mol Cell Biol 17:2550 –2558 15. Marshall S, Bacote V, Traxinger RR 1991 Discovery of a metabolic pathway mediating glucose-induced desensitization of the glucose transport system. J Biol Chem 266:4706 – 4712 16. Mueller W, Gregoire FM, Stanhope K, Mobbs CV, Mizuno TM, Warden C,


17.

18. 19. 20.

21. 22.

23. 24. 25. 26.

Stern JS, Havel PJ 1998 Evidence that glucose metabolism regulates leptin secretion from cultured rat adipocytes. Endocrinology 139:551–558 Miller SG, De Vos P, Guerre-Millo M, Wong K, Hermann T, Staels B, Briggs MR, Auwerx J 1996 The adipocyte specific transcription factor C/EBP␣ modulates human ob gene expression. Proc Natl Acad Sci USA 93:5507–5511 He Y, Chen J. Quon MJ, Reitman M 1995 The mouse obese gene. J Biol Chem 270:28887–28891 Fukuda H, Iritani N 1999 Transcriptional regulation of leptin gene promoter in rat. FEBS Lett 455:165–169 Moreno-Aliaga M, Stanhope KL, Havel PJ 2001 Transcriptional regulation of the leptin promoter by insulin-stimulated glucose metabolism in 3T3-L1 adipocytes. Biochem Biophys Res Commun 283:544 –548 Mandrup S, Lane MD 1997 Regulating adipogenesis. J Biol Chem 272:5367– 5370 Wang Y, Lee-Kwon W, Martindale JL, Adams L, Heller P, Egan JM, Bernier M 1999 Modulation of CCAAT/enhancer-binding protein-␣ gene expression by metabolic signals in rodent adipocytes. Endocrinology 140:2938 –2947 Tang QQ, Jiang MS, Lane MD 1999 Repressive effect of Sp1 on the C/EBP␣ gene promoter: role in adipocyte differentiation. Mol Cell Biol 19:4855– 4865 Roos MD, Su K, Baker JR, Kudlow JE 1997 O Glycosylation of an Sp1-derived peptide blocks known Sp1 protein interactions. Mol Cell Biol 17:6472– 6480 Ahluwalia GS, Grem JL, Hao Z, Cooney DA 1990 Metabolism and action of amino acid analog anti-cancer agents. Pharmacol Ther 46:243–271 MacDougald O, Hwang C-S, Fan H, Lane MD 1995 Regulated expression of the obese gene product (leptin) in white adipose tissue and 3T3-L1 adipocytes. Proc Natl Acad Sci USA 92:9034 –9037