Role of ATP Production and Uncoupling Protein-2 in the Insulin Secretory Defect Induced by Chronic Exposure to High Glucose or Free Fatty Acids and Effects of Peroxisome Proliferator-Activated Receptor-γ Inhibition
In rat pancreatic islets chronically exposed to high glucose or high free fatty acid (FFA) levels, glucose-induced insulin release and mitochondrial glucose oxidation are impaired. These abnormalities are associated with high basal ATP levels but a decreased glucose-induced ATP production (Δ of increment over baseline 0.7 ± 0.5 or 0.5 ± 0.3 pmol/islet in islets exposed to glucose or FFA vs. 12.0 ± 0.6 in control islets, n = 3; P < 0.01) and, as a consequence, with an altered ATP/ADP ratio. To investigate further the mechanism of the impaired ATP formation, we measured in rat pancreatic islets glucose-stimulated pyruvate dehydrogenase (PDH) activity, a key enzyme for pyruvate metabolism and for the subsequent glucose oxidation through the Krebs cycle, and also the uncoupling protein-2 (UCP-2) content by Western blot. In islets exposed to high glucose or FFA, glucose-stimulated PDH activity was impaired and UCP-2 was overexpressed. Because UCP-2 expression is modulated by a peroxisome proliferator- activated receptor (PPAR)-dependent pathway, we measured PPAR-γ contents by Western blot and the effects of a PPAR-γ antagonist. PPAR-γ levels were overexpressed in islets cultured with high FFA levels but unaffected in islets exposed to high glucose. In islets exposed to high FFA concentration, a PPAR-γ antagonist was able to prevent UCP-2 overexpression and to restore insulin secretion and the ATP/ADP ratio. These data indicate that in rat pancreatic islets chronically exposed to high glucose or FFA, glucose-induced impairment of insulin secretion is associated with (and might be due to) altered mitochondrial function, which results in impaired glucose oxidation, overexpression of the UCP-2 protein, and a consequent decrease of ATP production. This alteration in FFA cultured islets is mediated by the PPAR-γ pathway.
Patients with type 2 diabetes are characterized by a progressive decline of insulin secretion that becomes more severe with the increasing duration of the disease (1–4). In these patients, the mechanisms that cause the progressive β-cell failure are currently under investigation: the altered insulin secretory pattern depends, at least in part, on the negative influence of chronic high glucose (5–8) and/or high free fatty acid (FFA) (9–12) plasma concentrations (gluco- or lipotoxicity). These metabolites are believed to affect pancreatic β-cell function by chronic β-cell stimulation and consequent “desensitization” to glucose. However, the molecular mechanisms of glucose desensitization induced by hyperglycemia or hypernefemia are still unclear (13,14).
Rat pancreatic islets that are chronically exposed to high glucose or FFAs have an impaired glucose oxidation (15). Because glucose oxidation generates ATP and the rise of ATP and ATP/ADP ratio plays a central role in glucose-induced insulin release by causing K+-ATP channel closure, membrane depolarization, increased calcium influx, and insulin granule exocytosis, we first measured ATP and ADP levels in islets that were chronically exposed to high glucose or FFAs. In addition to mitochondrial glucose oxidation, ATP synthesis and ATP/ADP ratio are regulated by uncoupling protein-2 (UCP-2) expression. UCP-2 is a member of a family of proteins that are located in the mitochondrial inner membrane and that act as proton channels to uncouple mitochondrial oxidative phosphorylation. By this mechanism, energy is wasted through heat, and cellular ATP synthesis is decreased. UCP-2 overexpression in rat pancreatic islets has already been shown to inhibit glucose-stimulated insulin secretion by decreasing ATP formation (16,17). Therefore, to investigate the role of two mechanisms involved in ATP synthesis in pancreatic islets that are chronically exposed to high glucose and FFAs, we studied the activity of pyruvate dehydrogenase (PDH), a key enzyme for pyruvate metabolism, and the subsequent glucose oxidation through the Krebs cycle and UCP-2 levels.
To study further the mechanisms implicated in the transcriptional regulation of the UCP-2 gene in pancreatic islets cultured with high glucose or FFAs, we evaluated the role of peroxisome proliferator-activated receptor (PPAR)-γ (18). In several tissues, this transcription factor has been shown to mediate the response of UCP-2 to fatty acids (19–23), because it has been found that fatty acids act as natural ligands for PPAR-γ (24,25). Therefore, in pancreatic islets exposed to chronically high glucose and FFAs, we measured the PPAR-γ protein expression and analyzed whether a PPAR-γ antagonist was able to prevent the effects (26).
RESEARCH DESIGN AND METHODS
Crude collagenase was obtained from Boehringer Mannheim (Mannheim, Germany). Culture medium CMRL-1066, heat-inactivated FCS, glutamine, and gentamicin were obtained from Gibco (Glasgow, U.K.). Bisphenol A diglycidyl ether (BADGE) was obtained from Fluka (Buchs, Switzerland). All other chemicals were of analytical grade.
Islet preparation and culture conditions.
Pancreatic islets were isolated by the collagenase method from 200- to 250-g fed male Wistar rats that received an intraperitoneal injection of 0.2 ml of a 0.2% pilocarpine solution 2 h before being killed by decapitation. With this technique, 300–400 islets were isolated from each pancreas (27). Purified islets were first cultured overnight at 5.5 mmol/l glucose in CMRL-1066 medium and then either for 48 h at 5.5 or 16.7 mmol/l glucose or for 72 h in a culture medium containing 5.5 mmol/l glucose and 2% BSA (FFA-free) with or without 2 mmol/l long-chain fatty acids (oleate/ palmitate 2:1), as described by Lee et al. (28). In a set of experiments, long-chain fatty acid concentration was 0.4 mmol/l. In some experiments, islets were cultured in the presence or absence of 100 μmol/l BADGE, a synthetic antagonist for PPAR-γ.
At the end of the culture period, islets were washed twice in Krebs Ringer HEPES buffer (KRHB; containing 115 mmol/l NaCl, 5.4 mmol/l KCl, 2.38 mmol/l CaCl2, 0.8 mmol/l MgSO4, 1 mmol/l Na2HPO4, 10 mmol/l HEPES, and 0.5% BSA [pH 7.35]). Groups of five purified islets were then incubated with glucose 2.8 or 16.7 mmol/l (30 min, 37°C), and then insulin was measured in the medium by radioimmunoassay. Results are expressed as insulin released in the medium (pg · islet− 1 · 30 min−1) (29).
Glucose oxidation was determined by measuring the 14CO2 formation from [U-14C] glucose (New England Nuclear, Boston, MA) (30). At the end of the culture period, islets were washed twice in KRHB. Groups of 15 islets were incubated in a plastic cup in 100 μl of Krebs-Ringer bicarbonate buffer (118 mmol/l NaCl, 4.8 mmol/l KCl, 2.5 mmol/l CaCl2, 1.2 mmol/l MgSO4, 1.2 mmol/l KH2PO4, and 25 mmol/l NaHCO3) supplemented with 10 mmol/l HEPES (pH 7.4) and containing 3 μCi d-[U-14C]glucose (specific activity 302 mCi/mmol) plus nonradioactive glucose to a final concentration of either 2.8 or 16.7 mmol/l. The cups, suspended in standard 20-ml glass scintillation vials, were gassed with O2:CO2 (95:5) and capped airtight. The vials were then shaken continuously at 37°C for 120 min. The islet metabolism was stopped by injecting 100 μl of 0.05 mmol/l Antimycin A (dissolved in 70% ethanol) into the cup. This was immediately followed by an injection of 250 μl of Hyamine hydroxide (New England Nuclear) into the vials. 14CO2 was liberated from the incubation medium by a subsequent injection into the cup of 100 μl of 0.4 mmol/l Na2HPO4 solution adjusted to pH 6.0. After 2 h at room temperature (to allow liberated 14CO2 to be trapped by the Hyamine hydroxide), the cup was removed, 10 ml of a scintillation fluid was added to each flask, and the radioactivity was measured in a liquid scintillation counter. The 14CO2 recovery (88.8 ± 0.9%) was assessed as previously described (31).
Measurements of adenine nucleotides.
Adenine nucleotides were measured according to Detimary et al. (32). At the end of the culture period, islets were washed twice in KRHB. Groups of five islets were preincubated for 30 min at 37°C in 675 μl of KRHB with 2.8 or 16.7 mmol/l glucose. The incubation was stopped by the addition of 0.125 ml of TCA to a final concentration of 5%. The tubes were then mixed, placed on ice for 5 min, and centrifuged. For eliminating TCA, 400 μl of supernatant was mixed with 1.5 ml of diethylether, and the ether phase containing TCA was removed. This procedure was repeated three times to ensure complete elimination of TCA. The extracts were then diluted with 0.4 ml of a buffer containing 20 mmol/l HEPES, 3 mmol/l MgCl2, and KOH as required to adjust pH to 7.75 (assay buffer). The diluted extracts were frozen at −70°C until assayed.
After appropriate sample dilution, ATP and ADP were measured in triplicate by a luminometric method (33). For measuring total ATP+ADP, ADP was first converted into ATP by mixing 100 μl of the diluted extract with 300 μl of the assay buffer supplemented with 1.5 mmol/l phosphoenolpyruvate and 2.3 units/ml pyruvate kinase and incubating at room temperature for 15 min. Samples with known ADP concentrations but without ATP were run in parallel as controls for complete transformation. ATP was measured by the addition of 100 μl of an ATP monitoring reagent containing luciferase and luciferin (Sigma, St. Louis, MO). The emitted light was measured in a luminometer (Turner TD-20/20). For measuring only ATP, the same procedure was followed, except for the first incubation step, where pyruvate kinase was lacking. ADP levels were then calculated by subtracting ATP from the total ATP+ADP. Blanks and ATP standards were run through the entire procedure, including the extraction steps.
Preparation of islet homogenate extracts and radiochemical assay of PDH activity.
Islet PDH complex was extracted and measured as previously described (34). With this method, we measured active, not total, PDH activity. At the end of the culture period, groups of 200 islets were washed twice in KRHB and then incubated with 16.7 mmol/l glucose (3 h, 37°C). After this period, islets were washed twice in PBS and homogenized in 300 μl of homogenization buffer (50 mmol/l HEPES [pH 7.5], 0.2 mmol/l KCl, 3 mmol/l ethylenediaminetetraacetate, 5 mmol/l dithiothreitol, 0.1 mmol/ LNα-p-tosyl-l-lysine chloromethyl ketone, 0.1 mg/ml trypsin inhibitor from egg white, 200 KIE/ml aprotinin, 2% rat serum, and 0.25% [vol/ vol] Triton X-100). For extracting the PDH complex, islets were frozen-thawed three times and then passed 20 times through a 0.5-ml insulin syringe (Becton-Dickinson, Rutherford, NJ). Before measurement of PDH activity, islet extracts were left on ice for 30 min. Then, 50 μl of extracts and 100 μl of the enzyme reaction mixture (50 mmol/l HEPES [pH 7.5], 1 mmol/l MgCl2, 3 mmol/l NAD+, 0.4 mmol/l dithiothreitol, 0.1% vol/vol Triton X-100, 7.5 units/ml lipoamide dehydrogenase, EC 188.8.131.52, and 1.0 mmol/l pyruvate) were placed in 1.5-ml uncapped Eppendorf tubes inside a capped 20-ml scintillation vial containing 1.0 ml of KOH (1.2 mol/l). The reaction was started by adding 0.1 μCi of [1–14C] pyruvate (NEN, Boston, MA). After 20 min in a 37°C shaking water bath, the reaction was stopped by injecting into the Eppendorf tube 100 μl of 15% of trichloroacetate. After an additional 2-h incubation at 37°C (to allow the liberated CO2 to be absorbed by KOH), the Eppendorf tubes were removed, 5 ml of the scintillation fluid was added to each vial, and the radioactivity was measured. Blanks (lacking islets) containing 50 μl of homogenization buffer were run with each series, and the averaged blank count was subtracted from sample counting. One unit of PDH complex activity was defined as the production of 1 μmol CO2/min at 37°C. Activity was linear with time between 5 and 30 min and with the amount of islet extract from 5 to 20 islets.
UCP-2 protein measurement.
The UCP-2 protein levels were measured by Western blot analysis as previously described (35). Briefly, at the end of the culture period, groups of 200 islets were washed twice in PBS and homogenized by sonication in SDS-PAGE sample buffer. Islet proteins were denatured by boiling for 3 min, and protein content was measured by the BCA protein assay kit (Pierce, Rockford, IL), using BSA as standard. Equivalent amounts of proteins were used in the different experimental groups. Homogenates (25 μg of total cellular protein) were separated on 10% SDS-polyacrylamide gel (Mini-Protean; Bio-Rad) and electrophoretically transferred onto nitrocellulose membrane (Amersham Pharmacia Biotech, Buckinghamshire, U.K.). Blotting efficiency and the position of protein standards were assessed by Ponceau staining. Membranes were then washed at room temperature with PBS containing 1% nonfat dried milk and 0.2% Tween 20 (two changes within 30 min) and blocked with the same buffer containing 10% nonfat dried milk at room temperature for 1 h. After blocking, the membranes were washed two times, as described above, and incubated with a rabbit polyclonal anti-UCP-2 antibody (Alpha Diagnostic International, San Antonio, TX) at 1:4,000 dilution in blocking solution, 4°C, overnight. Membranes were then washed and blotted with a donkey anti-rabbit Ig peroxidase-linked whole antibody (Pierce) diluted 1:40,000 for 1 h at room temperature. Peroxidase activity was detected using a Super Signal-CL Kit (Pierce). As verified in each experiment by Ponceau red staining, the total amount of protein load was similar in each lane for the different culture conditions.
PPAR-γ protein measurement.
The protein levels were measured by Western blot analysis. As for UCP-2 protein measurement, islets were homogenized, separated on 10% SDS-polyacrylamide gel, and transferred electrophoretically onto nitrocellulose membranes. Membranes were then washed at room temperature with PBS containing 1% nonfat dried milk and 0.2% Tween 20 (two changes within 30 min) and blocked with the same buffer containing 10% nonfat dried milk at room temperature for 1 h. After blocking, the membranes were washed two times, as described above, and incubated with a rabbit polyclonal anti-PPAR-γ antibody (Affinity Bioreagents, Golden, CO) at 1:4,000 dilution in blocking solution, 4°C, overnight. Membranes were then washed and blotted with a donkey anti-rabbit Ig peroxidase-linked whole antibody (Pierce) diluted 1:4,000, 1 h at room temperature. Peroxidase activity was detected using an enhanced chemiluminescence detection reagent (Amersham Pharmacia Biotech) for 1 min at room temperature.
Data are mean ± SE. Statistical significance was assessed by the Student’s t test for unpaired comparison and two-tailed analysis.
In control rat islets, basal insulin release (in the presence of a nonstimulatory glucose concentration, 2.8 mmol/l) was 73 ± 15 pg · islet−1 · 30 min−1 (mean ± SE, n = 5) and significantly increased in response to 16.7 mmol/l glucose (879 ± 63 pg · islet− 1 · 30 min−1; n = 5; Fig. 1). In islets preexposed for 48 h to 16.7 mmol/l glucose, as expected, basal insulin secretion was significantly increased (461 ± 40 pg · islet− 1 · 30 min−1; n = 5; P < 0.01 vs. control) and glucose-stimulated insulin release was markedly reduced (567 ± 43 pg · islet−1 · 30 min−1; n = 5; P < 0.01; Fig. 1). Similar data were obtained in islets exposed to FFA (Fig. 1): basal insulin release was increased (292 ± 48 vs. 89 ± 22 pg · islet−1 · 30 min−1 in control islet; n = 5; P < 0.01) and glucose-stimulated insulin release was reduced (301 ± 60 vs. 717 ± 62 pg · islet−1 · 30 min−1 in control islet; n = 5; P < 0.01) in comparison with control islets. As requested by a reviewer, these data were also confirmed in islets cultured for 72 h with 0.4 mmol/l FFA: glucose-stimulated insulin release was reduced in comparison with control islets (389 ± 78 vs. 899 ± 109 pg · islet−1 · 30 min−1; n = 5; P < 0.01); basal insulin secretion was increased (153 ± 24 vs. 78 ± 18 pg · islet− 1 · 30 min−1 in control islet; n = 5; P < 0.05).
In response to the glucose concentration increase, glucose oxidation also increased in control islets. In contrast, in islets preexposed to high glucose, whereas glucose oxidation at 2.8 mmol/l glucose was similar to that of control islets (14.4 ± 0.5 vs. 14.8 ± 1.9 pmol · islet−1 · 120 min−1, respectively; n = 4), glucose oxidation was significantly reduced after stimulation with 16.7 mmol/l glucose (29.8 ± 3.6 vs. 47.5 ± 2.4 pmol · islet−1 · 120 min−1 in control islets; n = 4; P < 0.01; Fig. 2). Similar data were obtained in islets preexposed to 2 mmol/l FFA: glucose oxidation at 2.8 mmol/l was unchanged (11.5 ± 1.2 vs. 8.9 ± 1.4 pmol · islet−1 · 120 min−1 in control islet; n = 4), but the value measured at 16.7 mmol/l glucose was significantly reduced in respect to control islets (35.0 ± 2.2 vs. 53.7 ± 3.7 pmol · islet−1 · 120 min−1, respectively; n = 4; P < 0.01; Fig. 2).
ATP and ADP content.
In response to glucose concentration increase (16.7 mmol/l), in control islets, ATP levels increased, ADP levels decreased, and, as a consequence, the ATP/ADP ratio clearly increased. In islets preexposed for 48 h to 16.7 mmol/l glucose, the basal (2.8 mmol/l glucose) ATP level was higher in respect to control islets (35.5 ± 2.5 vs. 25.7 ± 1.6 pmol/islet, respectively; n = 3; P < 0.05). After stimulation with 16.7 mmol/l glucose, the ATP level did not further increase over baseline in high glucose-exposed islets (Δ of increment over baseline: 0.7 ± 0.5 vs. 12.0 ± 0.6 pmol/islet in control islets; n = 3; P < 0.01). Therefore, in islets incubated for 48 h with high glucose, the ATP/ADP ratio in response to glucose stimulation was clearly reduced (4.5 ± 0.9 vs. 10.1 ± 1.4; n = 3; P < 0.05; Table 1). Also, in islets cultured for 72 h with high FFA, the basal level of ATP was increased in respect to control islets cultured without FFA (38.5 ± 1.9 vs. 30.7 ± 1.7 pmol/islet, respectively; n = 3; P < 0.05), the glucose-induced ATP raise was impaired (Δ of increment: 0.5 ± 0.3 pmol/islet in FFA exposed islets vs. 12.1 ± 1.5 pmol/islet in control islets; n = 3; P < 0.01), and the ATP/ADP ratio in response to high glucose was reduced (5.8 ± 0.8 vs. 13.8 ± 1.5 pmol/islet; n = 3; P < 0.01; Table 2).
To study the mechanism responsible for the impaired glucose oxidation after exposure to high glucose or FFA, we measured PDH activity in the pancreatic islet homogenates. In islets exposed for 48 h to high glucose, the glucose-stimulated PDH activity was significantly decreased at ∼ 54% of control islets (1.28 ± 0.08 vs. 2.38 ± 0.16 μU/islet, respectively; n = 4; P < 0.01). Similar results were obtained after 72 h exposure to FFAs: the islet PDH activity in response to glucose stimulation was greatly reduced, at ∼45% of control islets (1.49 ± 0.22 vs. 3.29 ± 0.42 μU/islet, respectively; n = 4; P < 0.01).
To investigate whether metabolic pathways other than glucose oxidation might play a role in modulating mitochondrial energy metabolism, we measured the UCP-2 protein level in islets exposed to either high glucose (48 h, 16.7 mmol/l) or FFA (72 h, 2 mmol/l). The UCP-2 protein content was significantly increased in islets exposed to high glucose for 48 h with respect to control islets (191 ± 29%; n = 4; P < 0.05). Similar data were obtained in islets preexposed to 2 mmol/l FFAs: also in these islets, the UCP-2 expression was significantly increased with respect to control islets (197 ± 13%; n = 4; P < 0.01). Protein recovery was similar in islets exposed for 48 h to either 5.5 or 16.7 mmol/l glucose or for 72 h with or without 2 mmol/l FFAs (Fig. 3).
To investigate whether chronic exposure to high glucose or FFA can modulate PPAR-γ protein levels, known transcription factors of UCP-2, we measured PPAR-γ expression in islets exposed to either high glucose (48 h, 16.7 mmol/l) or FFA (72 h, 2 mmol/l). In islets preexposed to 2 mmol/l FFAs, the PPAR-γ expression was significantly increased with respect to control islets (138 ± 6%; n = 3; P < 0.05). In contrast, PPAR-γ protein content was unchanged in islets exposed to high glucose for 48 h with respect to control islets (n = 4; Fig. 4).
To study the role of PPAR-γ in modulating UCP-2 expression and, as consequence, in impaired glucose-stimulated insulin secretion induced by chronic culture with high glucose and FFA levels, we analyzed whether PPAR-γ inhibition by synthetic antagonist of PPAR-γ was able to prevent the deleterious effects of gluco- and lipotoxicity. BADGE is a ligand for PPAR-γ and can antagonize the ability of FFA agonist ligand to stimulate the transcriptional activity of PPAR-γ. When islets were cultured for 72 h with or without FFA in the presence of BADGE, an inhibition of FFA-induced UCP-2 overexpression was observed; in addition, BADGE presence in control islets induced a reduction of UCP-2 protein levels (−28 ± 7%; n = 3; P < 0.05; Fig. 5A).
In this condition, a reversal of insulin release pattern to controls was found. In fact, in islets cultured with FFA and BADGE for 72 h, glucose-stimulated insulin release was increased compared with islets cultured with FFA alone (510 ± 19 vs. 323 ± 18 pg · islet−1 · 30 min−1; n = 4; P < 0.05) and basal insulin release was reduced (38.5 ± 6.7 vs. 176 ± 10 pg · islet−1 · 30 min−1; n = 4; P < 0.05; Fig. 5B). Similar effects were obtained in islets cultured in the presence of 0.4 mmol/l FFA and BADGE: glucose-stimulated insulin secretion was increased compared with islets cultured with FFA alone (934 ± 150 vs. 389 ± 78 pg · islet−1 · 30 min−1; n = 5; P < 0.01), and basal insulin secretion was reduced (86 ± 13 vs. 153 ± 24 pg · islet−1 · 30 min−1; n = 5; P < 0.01). Accordingly, BADGE addition to culture medium containing FFAs induced an increase of ATP/ADP ratio (Table 3) and a recovery of PDH activity (2.64 ± 0.21 vs. 1.20 ± 0.13 μU/islet, respectively; n = 3; P < 0.01) with respect to FFA-exposed islets.
In contrast, in islets incubated for 48 h with 5.5 or 16.7 mmol/l glucose in the presence or absence BADGE, PDH activity was unchanged (data not shown) and glucose-induced UCP-2 overexpression was unaffected, but BADGE presence in control islets induced a reduction of UCP-2 protein levels (−32 ± 9%; n = 3; P < 0.05; Fig. 6).
Our data indicate that the blunted insulin release as a result of chronic exposure of pancreatic islets to either high glucose or FFA levels (gluco- and lipotoxicity) is associated with a decrease in ATP synthesis as a result of the inability of β-cells to increase the ATP levels in response to glucose stimulation. We investigated the mechanism of the reduced ATP production and observed that both a defect in substrate supply to the Krebs cycle and UCP-2 protein overexpression could contribute to this abnormality. For the first pathway, we studied PDH, a key enzyme for pyruvate metabolism and for the subsequent glucose oxidation through the Krebs cycle. In fact, pyruvate, the final product of the glycolytic pathway, enters mitochondrial metabolism by decarboxylation to acetyl-CoA (a reaction that is catalyzed by PDH, an enzymatic complex of the mitochondrial inner membrane).
In islets exposed to high glucose or FFA levels, we found a significant decrease of glucose-induced PDH activity. This finding suggests that the reduced glucose oxidation observed in islets chronically exposed to high glucose or FFA might be due, at least in part, to the impaired PDH activity. Because in several tissues PDH activity has been demonstrated to be inhibited by high mitochondrial ATP levels (36), it is possible that also in our model the increased ATP availability (as a result of the substrate excess in the environment) could exert an adverse effect on PDH activity. It is also possible that high glucose or FFAs could decrease PDH activity either by increasing the PDH kinase activity (37) or by affecting the PDH protein expression (38).
Indirect evidence suggests that PDH could play a role in pancreatic islet desensitization induced by hyperglycemia. In fact, a decreased PDH activity has already been observed in vivo, in GK rats, a genetic model of type 2 diabetes (34). In this article, we first demonstrate a direct effect of chronic hyperglycemia on PDH activity and the association with a reduced ATP synthesis and impaired insulin secretion. In addition, we demonstrate that both chronic hyperglycemia and chronically high FFA levels reduce PDH activity in pancreatic islets and that this alteration could be due to similar mechanism: an overload fueled by increasing basal ATP levels could blunt glucose-induced PDH activity and could be responsible, at least in part, for impaired insulin secretion. Zhou et al. (39) reported a decrease of PDH activity in pancreatic islets exposed to high levels of FFA but suggested that this defect was related to the glucose-FFA cycle (Randle cycle) (40).
The second pathway that can contribute to the impaired ATP formation that we have observed in our model involves UCP-2 and the respiratory chain function. In fact, after prolonged culture with high glucose and FFA, we observed a UCP-2 protein upregulation. This increase could be a protective response induced by high energetic fuels (glucose and FFA), resulting in basal ATP load. Because UCP-2 uncouples fuel oxidation from conversion of ADP to ATP, the increased UCP-2 expression may contribute to the altered glucose-induced ATP formation and the decreased insulin release that we observed at the end of chronic incubation with high glucose or FFA.
Recent data indicate that in tumoral β-cell line, chronic exposure to high FFA both reduced insulin secretion and increased UCP-2 levels by regulating glucose-induced ATP formation (41). In normal rat islets, overexpression of UCP-2 inhibits glucose-stimulated insulin secretion (16,17). In other reports, UCP-2 overexpression by reducing ATP levels but enhancing the ATP/ADP ratio restored insulin secretion in islets from ZDF rat.
Conflicting data have been reported in islets exposed to high glucose provided on UCP-2 expression: in islets from hyperglycemic 90% pancreatectomized rats, UCP-2 mRNA was increased (42); in contrast, in INS β-cell line, chronic exposure to high glucose reduced UCP-2 mRNA levels (43). Furthermore, we investigated the mechanism by which high glucose or FFA levels increase UCP-2 expression, and we observed that PPAR-γ were to some extent involved. PPAR-γ are nuclear transcription factors involved in several functions ongoing from cellular cycle control to the regulation of enzymes implicated in glucose and lipid metabolism (44,45).
In our study, we observed an increase of PPAR-γ protein levels in islets exposed to high FFAs and the reversal effects induced by a PPAR-γ antagonist. BADGE added to the culture medium was able to prevent UCP-2 protein increase and to restore both insulin secretion and the ATP/ADP ratio. In addition, in the presence of BADGE, we observed a decrease of UCP-2 expression; accordingly, in both control and FFA-treated islets, very high ATP levels further increased in response to glucose stimulation. These data could indicate that UCP-2 modulates the basal ATP content and that only when UCP-2 is upregulated, new ATP synthesis is impaired. In contrast, when UCP-2 was downregulated by PPAR-γ antagonist, the basal ATP levels were high and further increased in response to glucose stimulation, preserving glucose-induced insulin secretion (46). In addition, our data show that BADGE is able to reverse the decrease of PDH activity. This effect could be explained by a direct effect of PPAR-γ on PDH. Alternatively, BADGE could normalize lipid oxidation and then reduce acetyl-CoA levels and, by this mechanism, reverse PDH inhibition. This last view is supported by data obtained in ZDF rats, in which troglitazone, a PPAR-γ agonist, is able to increase FFA oxidation (47). These data support the hypothesis that the FFA effects on β-cell function are mediated by PPAR-γ transcription factors, whereas PPAR-γ could not be involved in the glucose-induced β-cell dysfunction. In fact, PPAR-γ levels were not altered by chronic hyperglycemia, and both the UCP-2 levels and PDH activity were not influenced by PPAR-γ antagonist. The role of PPAR-γ in pancreatic islets is still not clarified. Protective effects of PPAR-γ on β-cells were reported in a genetic model of lipotoxicity (47, 48). In contrast, after chronic hyperglycemia, an impaired islet function was associated with a PPAR-γ RNA increase (42, 49). In our report, hyperglycemia does not modify PPAR-γ protein levels, and this, with respect to previous reports, could be due to our relatively short exposure time or an impaired mRNA translation.
In conclusion, our data show that the chronic exposure to either high glucose or FFA levels impairs glucose-induced ATP formation trough at least two different mechanisms. On the one hand, it decreases PDH activity and glucose oxidation through the Krebs cycle. On the other hand, it increases UCP-2 expression. It is conceivable that when an increased fuel influx through the metabolic pathways overcomes the oxidative capacity of pancreatic β-cells, large amounts of high-energy substrates might load the cells and impair their function (48,50). In this condition, the β-cell, like many other cell types, protects itself from fuel overload by decreasing glucose oxidation and increasing energy disposal by increased UCP-2 protein expression. This compensatory mechanism, by causing a decrease of ATP production, a key regulator of insulin secretion, however, impairs glucose-induced insulin release and further contributes to the long-term deleterious effects of excess substrate availability.
Address correspondence and reprint requests to Francesco Purrello, Unità Operativa di Medicina Interna, Ospedale Cannizzaro, Via Messina 829, 95126 Catania, Italy. E-mail:.
Received for publication 21 February 2002 and accepted in revised form 3 May 2002.
BADGE, bisphenol A diglycidyl ether; FFA, free fatty acid; KRHB, Krebs Ringer HEPES buffer; PDH, pyruvate dehydrogenase; PPAR, peroxisome proliferator-activated receptor; UCP-2, uncoupling protein-2.