From the
UMR-S 530 INSERM, Université Paris 5, Centre Universitaire, 45 rue des Saints-Pères, 75006 Paris, France,
|| Department of Cell Biology and Biochemistry, Texas Tech University Health Sciences Center, Lubbock, Texas 79430
Received for publication, July 12, 2002
, and in revised form, March 14, 2003.
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Obesity is a major risk factor for type 2 diabetes (5). The increased adipose tissue mass in obese people leads to a proportionally larger fatty acid release into the blood as compared with lean people. In one rodent model of obesity (Zucker rats), during development of insulin resistance, fasting plasma fatty acids increase before hyperglycemia arises, which is consistent with a cause and effect relationship (6). It is likely that one reason that a low-fat, low-calorie diet is often an effective treatment for type 2 diabetes is that it reduces circulating fatty acids.
Oral insulin-sensitizing drugs are an important adjunct to diet in the treatment of diabetes. Thiazolidinediones are a class of antidiabetic drugs that increase systemic insulin sensitivity in diabetic animal models and humans (1). Two such drugs, rosiglitazone and pioglitazone, are used in humans to treat type 2 diabetes. They are agonists for peroxisome proliferatoractivated receptor (PPAR
),1 a member of the nuclear hormone receptor family of transcription factors, which is expressed in fat cells (7). Although much is known about the molecular mechanisms of thiazolidinedione action, many questions remain about the physiological targets mediating better insulin sensitivity. Several lines of evidence have led to the idea that the increased insulin sensitivity of liver and skeletal muscle is secondary to the direct effect of these drugs on adipocytes (1).
A potential thiazolidinedione target in adipocytes concerns fatty acid release. Indeed, thiazolidinedione treatment in vivo is associated with an increase of fatty acid uptake into fat depots, as observed by metabolic studies (8), and a decrease in circulating fatty acid levels (8, 9). Interestingly, in Zucker rats, this PPAR agonist-induced decrease precedes the reduction in glucose levels. This suggests that decreases in fatty acid levels may be important for the insulin-sensitizing action of thiazolidinediones (10).
Plasma fatty acid levels represent a balance between their release from triacylglycerol stores in adipose tissue and their clearance into tissues that need energy. In humans, it is estimated that 30% or more of the fatty acids liberated by lipolysis are re-esterified into newly synthesized triacylglycerol. This apparent "futile cycle" of simultaneous lipolysis and re-esterification creates an important mechanism for energy homeostasis (11). Regulation of fatty acid re-esterification rate allows fat cells to adapt rapidly to changes in peripheral requirements for fatty acids. The re-esterification process requires the production of glycerol-3-phosphate as a substrate for fatty acid reesterification into triacylglycerol (12). Because glucose supply to the tissue is limited during lipolysis and because adipocytes have no significant glycerol kinase activity, under physiological conditions (13), glycerol-3-phosphate must be synthesized from non-carbohydrate precursors such as lactate, pyruvate, and amino acids. This pathway, glyceroneogenesis, was discovered more than 30 years ago (14, 15).
We have shown that expression of the gene encoding the key glyceroneogenic enzyme, phosphoenolpyruvate carboxykinase (PEPCK-C, EC 4.1.1.32
[EC]
; Refs. 12, 14, and 15), is strongly induced by thiazolidinediones via a mechanism involving PPAR binding and activation in adipose tissue (16). In the present work, we show that PEPCK-C is involved in the action of thiazolidinediones in lowering fatty acid release from adipose tissue. Glycerol kinase was recently demonstrated as another thiazolidinedione-induced enzyme involved in the same process (17). However, our analysis of the respective contributions of glycerol kinase and PEPCK-C suggests that in cultured adipocytes, glyceroneogenesis accounts for at least 75% of the thiazolidinedione effect.
![]() |
EXPERIMENTAL PROCEDURES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Cell CultureThe 3T3-F442A adipocyte cell line was differentiated as described previously (16) and then treated for 76 h with 1 µM rosiglitazone or 5 µM pioglitazone in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum. The medium was changed daily. The day of the experiment, the last thiazolidinedione treatment was performed in Dulbecco's modified Eagle's medium without glucose and serum but supplemented with 0.3% fatty acid-free bovine serum albumin for 3 h.
In the experiments carried out under lipolytic conditions, 1 µM isoproterenol was added in the same medium described above, supplemented with 25 mM unlabeled pyruvate and [14C1]pyruvate (0.5 µCi/ 60-mm plate), and 3-mercaptopicolinic acid where indicated. Fatty acid release and glyceroneogenesis (incorporation of pyruvate) were measured 30 min later by the same procedures used with fat explants.
In the case of the experiments described in Fig. 5, radiolabeled substrates (12 µCi) were added in Krebs-Ringer phosphate buffer with 0.3% fatty acid-free bovine serum albumin for 1 h before lipid extraction. Various glycerol and pyruvate concentrations and 80 µM oleic acid were used as indicated in the figure.
|
|
PEPCK-C was measured in adipose tissues that were homogenized in 50 mM triethanolamine buffer, pH 7.2, containing 0.25 M sucrose and submitted to a 100,000 x g centrifugation for 1 h to obtain the cytosol fraction. PEPCK-C enzymatic activity was determined spectrophotometrically in cytosol aliquots by the method of Chang and Lane (19), based on CO2 exchange between labeled KHCO3 and oxaloacetate in the presence of inosine triphosphate and Mn2+. Glycerol kinase (GyK) enzymatic activity assay was performed as described in Guan et al. (17), except that 0.5 mM unlabeled glycerol and 50 µM [3H]glycerol were used, and reactions were incubated at 37 °C for 30 min. Protein content was determined using the BCA protein assay kit from Sigma-Aldrich.
Student's t test was used for statistical comparisons between paired groups.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
We thus determined the effect of a 4-day treatment of rats with rosiglitazone on the subsequent in vitro release of glycerol and fatty acids from three sources of adipose tissue. Fatty acid release was measured with explants maintained in vitro for 2 h, with or without the glyceroneogenic precursor, pyruvate. Subcutaneous adipose tissue (SCAT) and two abdominal adipose tissues were chosen because of their known physiological differences (20). The two abdominal adipose tissue samples were from the fat pads adjacent to the kidneys (perirenal adipose tissue (PRAT)) and from the greater omentum (omental adipose tissue (OAT)).
Pyruvate had no effect on glycerol release (Fig. 1A), although it reduced fatty acid release about 2-fold in PRAT and 3-fold in OAT (Fig. 1B), a result consistent with an increased capacity for the adipocytes from abdominal adipose tissues to re-esterify fatty acids under these conditions. The complete lipolysis of 1 mol of triacylglycerol yields 1 mol of glycerol and 3 mol of fatty acid. Thus, if 100% of the fatty acids were released, then the fatty acid/glycerol ratio would be 3.0. However, fatty acid recycling causes this ratio to be less than 3.0 (Fig. 1C). The fatty acid/glycerol ratio in the absence of added pyruvate indicates that fatty acid re-esterification ranges from 25% in SCAT of control animals (because released fatty acids = 2.25/3.0, or 75%, and the remaining 25% are re-esterified) up to 40% in PRAT and OAT (Fig. 1C). Because pyruvate significantly decreased the fatty acid/glycerol ratio in PRAT, a stimulation of fatty acid recycling is probably occurring (Fig. 1C) in this tissue.
These data are easily explained by a pyruvate-dependent fatty acid recycling via glyceroneogenesis that contributes to a specific decrease in fatty acid release. Results also indicate that glyceroneogenesis is more active in adipose tissue samples taken from the abdomen than in those taken from the subcutaneous region.
Rosiglitazone caused a significant, 4164% decrease in fatty acid release from all three depots, in both the absence and presence of pyruvate (Fig. 1B). However, rosiglitazone caused only moderate, 728% decreases in glycerol release (Fig. 1A). Thus, rosiglitazone treatment reduced fatty acid output to a much greater extent than glycerol release from abdominal adipose tissues. Indeed, the fatty acid/glycerol ratio was decreased by 23-fold by rosiglitazone treatment of the rats, depending on the presence of pyruvate in PRAT and OAT (Fig. 1C). Such an observation suggests that the stimulation of fatty acid re-esterification is quantitatively more important than the inhibition of glycerol release for rosiglitazone to decrease fatty acid output from adipose tissue. This raises the possibility that thiazolidinediones block fatty acid release by stimulating glyceroneogenesis in a pyruvate-dependent manner.
Rosiglitazone Increases Fatty Acid Re-esterification via Glyceroneogenesis and PEPCK-CTo confirm that rosiglitazone actually increases glyceroneogenesis in adipose tissue, we carried out the same experiment as that shown in Fig. 1, but in the presence of [14C1]pyruvate. [14C1]Pyruvate was chosen because, in contrast to C2- or C3-labeled molecules, the C1 carbon of pyruvate is conserved in glycerol-3-phosphate synthesis and is thus a specific marker for glyceroneogenesis. Under this condition, the proportion of the released fatty acid molecules that are re-esterified in adipocytes can be detected because they are incorporated as [14C1]glycerol-3-phosphate in triacylglycerol. As shown in Fig. 2A, the incorporation of [14C1]pyruvate into triacylglycerol occurred in all three adipose tissue depots, although it was higher in visceral fat than in subcutaneous fat. Rosiglitazone increased [14C1]pyruvate incorporation 1.52-fold in visceral adipose tissues, whereas that in the subcutaneous depot was only slightly affected. Thus, rosiglitazone activates glyceroneogenesis in visceral adipose tissues with an inverse correlation to fatty acid output. The involvement of glyceroneogenesis in re-esterification implies that PEPCK-C specific activity is augmented. This was confirmed by direct measurement of PEPCK-C activity in Fig. 2B. Rosiglitazone treatment increased PEPCK-C specific activity 1.52-fold, and the magnitudes of the responses correlated with the depot-specific increase in [14C1]pyruvate incorporation (Fig. 2A) as well as with the decrease in fatty acid release (Fig. 1B). Hence, the rosiglitazone-induced increase in PEPCK-C mRNA that has been reported previously (16) results in the predicted increase of the enzyme itself.
|
Furthermore, we confirmed the functional involvement of PEPCK-C for fatty acid re-esterification by measuring the incorporation of [14C1]pyruvate into lipids in the presence of a specific inhibitor of PEPCK-C, 3-mercaptopicolinic acid (21). The addition of 150 µM 3-mercaptopicolinic acid in experiments carried out as described in Fig. 2A reduced the amount of labeled lipids in OAT (9.1 ± 1.3 versus 18.5 ± 2.6 dpm/mg in control rats and 13.8 ± 1.7 versus 32.2 ± 6.0 dpm/mg in rosiglitazone-treated rats) and in SCAT (5.0 ± 0.8 versus 10.0 ± 2.4 dpm/mg in control rats and 8.1 ± 0.6 versus 12.5 ± 0.8 dpm/mg in rosiglitazone-treated rats).
Rosiglitazone Specifically Decreases Fatty Acid Release in the Presence of a Lipolytic StimulusIt was originally demonstrated that glyceroneogenesis was stimulated by fasting, a physiological condition during which cAMP production and lipolysis are induced in adipose tissue (14). We thus asked whether rosiglitazone affected this pathway in the presence of a lipolytic stimulus. We used the -adrenergic agonist, isoproterenol, to induce cAMP in adipose tissues from rats that had been treated with rosiglitazone or vehicle, and we analyzed glycerol and fatty acid release. Results obtained with tissues cultured in the presence or absence of pyruvate are shown in Fig. 3 (compare AC with DF). As expected, isoproterenol stimulated glycerol and fatty acid release in all three adipose tissue depots whether cultured with pyruvate or not, although the magnitude of response was much higher in PRAT and OAT than in SCAT (Fig. 3, A, B, D, and E). Moreover, pyruvate did not modify glycerol release but strongly reduced fatty acid output with all treatment conditions. This is in agreement with a model whereby stimulation of fatty acid reesterification does not involve glycerol phosphorylation but requires glyceroneogenesis from pyruvate (Fig. 3, compare A and B with D and E). More importantly, rosiglitazone had either a weak effect or no effect on isoproterenol induction of glycerol release (Fig. 3, A and D), although it significantly reduced isoproterenol induction of fatty acid output in all three fat depots (Fig. 3, B and E). Hence, rosiglitazone action occurs whether or not cells are in a basal or lipolytic situation.
|
Pioglitazone and Rosiglitazone Activate PEPCK-C and Glyceroneogenesis and Decrease Fatty Acid Release from 3T3-F442A AdipocytesWe tested glyceroneogenesis-dependent fatty acid release in 3T3-F442A adipocytes because these cells have been established as a good model for adipose tissue metabolism. We analyzed the responses to rosiglitazone and a second thiazolidinedione, pioglitazone. We also used 3-mercaptopicolinic acid as the specific inhibitor of PEPCK-C (21) to determine the involvement of this enzyme in thiazolidinedione induction of glyceroneogenesis. Lipolysis was induced by isoproterenol for 30 min as described above, and then fatty acid release, glycerol release, and [14C1]pyruvate incorporation into lipids were measured in 3T3-F442A adipocytes that had been pretreated with 1 µM rosiglitazone or 5 µM pioglitazone for 3 days. Rosiglitazone and pioglitazone had similar effects (Fig. 4). Both glitazones significantly decreased fatty acid release (Fig. 4B) without affecting glycerol release (Fig. 4A). Moreover, both drugs increased glyceroneogenesis (Fig. 4D) and PEPCK specific activity 1.7-fold (Fig. 4C). In control cells, 3-mercaptopicolinate significantly enhanced fatty acid release (Fig. 4B) while inhibiting glyceroneogenesis as determined by the incorporation of [14C1]pyruvate into lipids (Fig. 4D). 3-Mercaptopicolinate not only abrogated the rosiglitazone and pioglitazone effects (Fig. 4D) but also increased fatty acid release, regardless of whether cells were thiazolidinedione-treated or not (Fig. 4B), without affecting glycerol release (Fig. 4A). These results indicate that PEPCK-C is involved in basal glyceroneogenesis and that an increase in its enzymatic activity is required for thiazolidinediones to stimulate glyceroneogenesis leading to an inhibition of fatty acid release.
|
Relative Contributions of PEPCK-C and GyK to Fatty Acid Re-esterification in Control and Rosiglitazone-treated AdipocytesThe results presented above suggest that thiazolidinediones increase fatty acid re-esterification by stimulating glyceroneogenesis. However, the thiazolidinedione induction of GyK, which was recently reported (17), suggested an alternative pathway that could provide glycerol-3-phosphate for fatty acid re-esterification. It was thus important to directly compare these two pathways in order to determine their relative contributions. The results shown in Fig. 5A confirm that rosiglitazone induces GyK 2.5-fold after a 72-h treatment in adipocytes. As an assessment of the relative contributions of the GyK and PEPCK-C pathways, the extent of lipid labeling was measured by adding either [3H]glycerol or [14C1]pyruvate to the culture media of 3T3-F442A adipocytes (Fig. 5, B and C). We reasoned that if rosiglitazone stimulated glyceroneogenesis and glycerol phosphorylation in proportion to the relative inductions of PEPCK-C and GyK, respectively, proportionate increases in triglyceride labeling should occur whether substrates are in limiting amounts or not. We thus used two different concentrations of glycerol (0.1 and 1 mM) and pyruvate (0.2 and 5 mM), the lowest of which was physiological, and the highest of which was saturating. We also reasoned that fatty acids originating from basal lipolysis could be rate-limiting at high glycerol-3-phosphate concentrations. We thus provided oleate (80 µM bound to 40 µM bovine serum albumin) when high substrate concentrations were used. Fig. 5, B and C, shows that rosiglitazone caused
1.42-fold increases in lipid labeling with both substrates. These increases were in agreement with the increases in the enzymatic activities described above (Figs. 4C and 5A). However, a maximum of 5.4 nmol of [3H]glycerol was incorporated into lipids per hour per mg of protein, whereas
3-fold more [14C1]pyruvate was incorporated after rosiglitazone treatment. Taken together, these results suggest that rosiglitazone stimulates fatty acid re-esterification by inducing two pathways: direct glycerol phosphorylation by GyK, and glyceroneogenesis via PEPCK-C. However, glyceroneogenesis accounts for at least 75% of the whole thiazolidinedione effect.
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The importance of glyceroneogenesis via adipocyte PEPCK in the regulation of lipid metabolism has very recently been documented in two transgenic mouse models (12). First, overexpression of PEPCK-C in adipose tissue resulted in obesity caused by activation of glyceroneogenesis and fatty acid reesterification (22). Unlike other animal models of obesity, these obese transgenic mice never develop insulin resistance, perhaps because their circulating fatty acid levels remain lower than those in non-obese control mice. Second, mice with a tissue-specific ablation of PEPCK-C gene expression in their adipose tissues showed phenotypes consistent with the loss of glyceroneogenesis (23). These adipose specific null mice were mildly lipodystrophic and appeared to be slightly insulin resistant, although not diabetic. Their fat tissue mass was reduced, whereas the release of fatty acids was increased and was not affected by pyruvate. Beyond our own data, such models emphasize the important physiological role of glyceroneogenesis in adipose tissue for controlling circulating fatty acid levels and perhaps preventing diabetes. Because elevated serum fatty acids can promote insulin resistance (4), taken together, these results suggest that glyceroneogenesis activation in adipose tissue, via its key enzyme (PEPCK-C), could help to abrogate diabetes by decreasing the systemic fatty acid supply.
Guan et al. (17) recently reported that thiazolidinediones induce GyK in adipocytes. They proposed this as the mechanism that activates the futile cycle in allowing fatty acid reesterification. They were apparently unaware of glyceroneogenesis and its contribution to the futile cycle. Guan et al. (17) showed that overexpression of GyK was less than half as efficient as rosiglitazone treatment in decreasing fatty acid release in 3T3-L1 adipocytes. Such a result is in accordance with our data linking glyceroneogenesis to the effect of thiazolidinediones on fatty acid release. Furthermore, glyceroneogenesis appears to be an important and immediate physiological pathway for regulation of fatty acid release, whereas GyK intervenes only as a response to thiazolidinediones because it has been detected in adipose tissue only at low levels in the absence of drug treatment. We recognize that glyceroneogenesis from pyruvate and direct phosphorylation of glycerol are not mutually exclusive pathways. Indeed, the present study suggests that their respective contributions to glycerol-3-phosphate synthesis for fatty acid re-esterification are in a ratio of 3:1 or more.
According to this revised model, thiazolidinediones would increase insulin sensitivity in tissues such as liver and skeletal muscle by decreasing the release of fatty acids from adipose tissue via the induction of glyceroneogenesis and GyK. In adipose tissue, the resulting increase of fatty acid re-esterification would be facilitated by the concomitant increase of proteins that allow their uptake and acyl-CoA activation, as expected from the reported increase in the corresponding mRNAs by PPAR agonists (10, 24). Furthermore, it was observed that a 6-month rosiglitazone monotherapy in patients with type 2 diabetes enhances the body weight, although moderately (25).
A number of lines of investigation have implicated PEPCK-C as one etiological factor in type 2 diabetes (12). The present study reinforces that notion by implicating PEPCK-C in an anti-diabetic action of thiazolidinediones. This is perhaps the reason why "fatless" mice (A-ZiP/F-1 mice, a lipodistrophic model) that develop type 2 diabetes are not responsive to thiazolidinedione treatment (26). In addition, PEPCK-C is known to lose its adaptive response to fasting with age in rats, whereas its expression is tightly regulated by insulin and counter-regulatory hormones in young animals (14). This could contribute to increased circulating fatty acid levels if such an age-related change occurs in humans. Finally, we found the regulation of glyceroneogenesis to occur mainly in visceral fat, the same fat depot that is specifically implicated in the progression of obesity to type 2 diabetes (27). This is important because the risk of insulin resistance can be predicted according to body fat distribution rather than generalized obesity in humans (27).
In summary, the major finding of this study is that stimulation of PEPCK-C by thiazolidinediones decreases fatty acid release from adipose tissues by increasing glyceroneogenesis. The recently developed adipose tissue PEPCK-C knockout model (23) should prove invaluable in testing the hypothesis that PEPCK-C is an important target for the anti-diabetic actions of thiazolidinediones. These mice will also be useful in testing whether PEPCK-C is involved in the etiology of type 2 diabetes.
![]() |
FOOTNOTES |
---|
Recipient of a fellowship from the French Ministère de l'Education Nationale et de la Recherche and a recipient of the 2002 Research Award from the French Nutrition Society.
¶ Supported by the Association Claude Bernard.
** Supported by National Institutes of Health Grant R01-GM39895.
A Centre National de la Recherche Scientifique researcher.
To whom correspondence should be addressed. Tel.: 33-142-863-861; Fax: 33-142-863-868; E-mail: benedicte.antoine{at}biomedicale.univ-paris5.fr.
1 The abbreviations used are: PPAR, peroxisome proliferator-activated receptor
; PEPCK-C, cytosolic phosphoenolpyruvate carboxykinase; GyK, glycerol kinase; SCAT, subcutaneous adipose tissue; PRAT, perirenal adipose tissue; OAT, omental adipose tissue.
![]() |
ACKNOWLEDGMENTS |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|