| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLE |
Garvan Institute of Medical Research, St. Vincent’s Hospital, Sydney, New South Wales 2010, Australia
Address all correspondence and requests for reprints to: Dr. Ji-Ming Ye, Diabetes and Obesity Research Program, Garvan Institute of Medical Research, 384 Victoria Street, Darlinghurst, Sydney, New South Wales 2010, Australia. E-mail: j.ye{at}garvan.org.au.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
agonists, such as the antidiabetes drugs thiazolidinediones (TZDs), sensitize insulin action and lower circulating lipids. The close association of PPAR-induced lipid lowering with enhancement of insulin action and the fact that PPAR is highly expressed in adipose tissues, compared with muscle or liver, have led to a hypothesis that PPAR agonists improve muscle insulin action by sequestering lipids in adipocytes, a mechanism that ultimately reduces lipid accumulation in muscle. Accordingly, the enhanced insulin action in muscle, and perhaps in liver, after TZD treatment is generally considered secondary to a systemic lipid-lowering effect via a principal action in adipose tissue (1, 2).
Some evidence suggests that PPAR agonists may directly enhance muscle insulin action. In mice with lipodystrophy in which adipose tissue is essentially absent, troglitazone still improved insulin sensitivity (3). Recently we have found that at doses with equivalent efficacy in lowering basal muscle and plasma lipid levels, the PPAR agonist Pioglitazone (Pio) is almost twice as effective in ameliorating muscle insulin resistance in high-fat-fed rats as the PPAR
agonist WY14643, again suggesting the involvement of factors other than just reduction of lipid availability (4). These functional observations are consistent with evidence that muscle PPAR protein levels may be much higher than previously thought (5). There is a positive correlation between the mRNA expression levels of PPAR and lipid metabolism genes (e.g. lipoprotein lipase, muscle carnitine palmitoyltransferase-1) in human skeletal muscle (6). In Zucker diabetic rats, chronic treatment with a PPAR agonist has been shown to alter gene expression in muscle as well as in liver (7). More recently, studies in isolated skeletal muscle have shown that TZDs have direct effects on muscle metabolism independent of PPAR-mediated gene expression (8). However, most studies in vivo have been unable to dissociate the insulin-sensitizing action of a PPAR agonist from its effect of lowering circulating lipids, although there is one report of PPAR-mediated lipid lowering without improvement of insulin sensitivity in mice devoid of fat tissues (9).
Lipid plus heparin infusion has been shown to induce insulin resistance with similar metabolic characteristics to chronic high-fat feeding (10, 11, 12). A recent study using this model showed that fatty acid-induced insulin resistance was prevented by troglitazone independent of the lowering of circulating fatty acids during maximal insulin stimulation (13). However, lipid levels in muscle and liver, which may be critical to insulin action in these tissues, were not examined. The first aim of this study was to investigate whether preconditioning with the PPAR agonist Pio can protect normal rats against insulin resistance induced by lipid infusion. The second aim was to use this model to investigate the relationship between the lipid-lowering effects of TZDs and their insulin-sensitizing action.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Animal pretreatment and cannulation
Male Wistar rats supplied from the Animal Resources Centre (Perth, Australia) were conditioned at 22 ± 0.5 C with a 12-h day/12-h night cycle (lights on 0600 h) for 1 wk and were fed a standard chow diet (5% calories as fat) ad libitum. Rats were randomly assigned to receive vehicle (control) (3 g jelly/rat) or Pio (3 mg·kg-1·d-1 in jelly) for 2 wk between 1500 and 1700 h. A week before study, the left carotid artery and right jugular vein were cannulated under ketamine/xylazine (90 mg/10 mg per kilogram, ip) anesthesia. The cannulae were exteriorized in the back of the neck. Rats were handled daily to minimize stress. Body weight and food intake were recorded daily, and only those rats with fully recovered body weight were used for the study.
Basal lipid metabolism
The experiment was performed in conscious rats pretreated with vehicle and Pioglitazone (n 
7 rats/group) after 5–7 h of fasting. Briefly, the jugular cannula was connected to a sampling line between 0900 and 1000 h. Rats were allowed to rest 30–40 min, and two blood samples (20 min apart) were taken. Plasma was snap frozen in liquid nitrogen for subsequent determination of circulating metabolites. Rats were then killed with an overdose of Nembutal and tissues were carefully dissected, cleaned of blood, and immediately freeze clamped for subsequent analysis. Visceral (epididymal, retroperitoneal, and mesenteric depots) and sc (trunk depots) fat was carefully dissected and weighed. Plasma levels of triglycerides, free fatty acids (FFAs), glycerol and leptin were determined to evaluate the effects of Pio on lipid metabolism at the whole-body level. Tissue triglycerides and long-chain acyl coenzyme A (LCACoA) were measured.
Hyperinsulinemic-euglycemic clamp
The hyperinsulinemic-euglycemic clamp with lipid infusion was performed using a protocol previously established in this laboratory (11). Briefly, experiments were performed in conscious rats fasted 5–7 h with free access to water. They were connected to an infusion apparatus (via the carotid line) and a blood-sampling syringe (via the jugular line) between 0900 and 1000 h. Two blood samples (at -20 and 0 min) were taken for the measurement of basal metabolic parameters after 30–40 min of rest. Intralipid emulsion (-Lip, 1.3 ml/h triglyceride, Travenol, Sydney, Australia) plus heparin (40 U/h) was infused to elevate circulating FFAs for 6 h. The hyperinsulinemic clamp [0.25 U·kg-1·h-1 of human insulin (Novo Nordisk A/S, Sydney, Australia)] was concomitantly performed, and glucose infusion rate (GIR) was adjusted to maintain euglycemia. Clamp studies were first carried out in three groups (>6 rats/group) as: 1) vehicle-pretreated rats infused with glycerol (-Gly, 0.4 ml/h)/heparin (40 U/h) (Con-Gly) as a control for infused lipid (12); 2) vehicle-pretreated rats infused with lipid (Con-Lip); and 3) Pio-treated rats infused with lipid (Pio-Lip). Blood samples (0.4 ml each) were taken at 60, 120, 300, and 360 min to measure plasma metabolic parameters, and red blood cells were returned to the rat in 0.25 ml sterile normal saline. Based on results of plasma lipid levels in the Pio-Lip group, another group of Pio-treated rats were infused at an increased rate of lipid (2.0 ml/h)/heparin (60 U/h) (Pio-Isolip) to match the plasma lipid levels to the Con-Lip group. During the clamp, [3H]-3-glucose (10 µCi/h) was simultaneously infused to determine glucose disappearance rate (Rd) and hepatic glucose output rate (HGO) (11, 14). Rats were killed with an overdose of Nembutal after 6 h of clamp with tissues collected as described above. Glucose incorporation into glycogen (15, 16) and lipids was calculated from the area under plasma [3H]-3-glucose curve (17) and tissue [3H] counts in glycogen or extracted lipids (18).
Metabolite measurements
Plasma glucose was determined using a glucose analyzer (YSI, Inc. 2300, Yellow Springs, OH). Plasma FFAs were determined spectrophotometrically using an acyl-CoA oxidase-based colorimetric kit (NEFA-C, WAKO Pure Chemical Industries, Osaka, Japan). Plasma triglyceride concentrations were measured using enzymatic colorimetric methods (Triglyceride INT, procedure 336 and GPO Trinder, Sigma, St. Louis, MO). Plasma leptin and adiponectin were determined by RIA using commercial kits (Linco Research, Inc., St. Charles, MO). Insulin concentration was determined by RIA using a rat insulin kit (Linco Research, Inc.) that measures both rat and human insulin. To differentiate infused human insulin from endogenously secreted rat insulin, a human insulin-specific kit (Linco Research, Inc.), which does not cross with rat insulin, was also used to examine plasma insulin concentrations during the hyperinsulinemic clamp. An estimate of endogenous insulin concentrations during the clamp was obtained from differences between the two insulin assays. Tissue triglycerides were extracted (19) and measured by a Peridochrom Triglyceride GPO-PAP kit (Boehringer Mannheim, Indianapolis, IN). Glycogen content was determined as previously described (20). Tissue LCACoA concentrations were determined by a fluorescence spectrophotometer based on a method described previously (21).
Measurement of protein kinase B content and activity
Muscle and liver samples collected after 6 h of hyperinsulinemic clamp were homogenized with 20 vol of 50 mmol/liter HEPES, pH 7.5; 150 mmol/liter NaCl; 1 mmol/liter MgCl2; 1 mmol/liter CaCl2; 2 mmol/liter Na3VO4; 10 mmol/liter Na pyrophosphate; 10 mmol/liter NaF; 2 mmol/liter EDTA; 1% Nonidet P-40; and 10% glycerol including 2 µg/ml aprotinin, 5 µg/ml leupeptin, and 34 µg/ml phenylmethylsulfonyl fluoride. The samples were solubilized on ice for 1 h, and then insoluble material was removed by centrifugation for 15 min at 13,000 g. The supernatants were subjected to 10% SDS-PAGE and transferred to polyvinylidene fluoride membranes for 1 h at 90 V, which were blocked overnight with 2% milk powder in Tris-buffered saline containing 0.025% Tween 20. The membranes were probed with anti-phospho protein kinase B (PKB)-serine 473 and -threonine 308 antibodies conjugated with donkey antirabbit horseradish peroxidase. The detected bands were visualized by enhanced chemiluminescence. The membranes were stripped and reprobed with anti-PKB to quantify total PKB protein content. The quantitation was performed against a protein reference made from the same type of rat muscle. Details are as described in previous reports (22).
Statistical analyses
All results are presented as means ± SE. A repeated-measure ANOVA was used to assess results measured at consecutive multiple time points. A two-way design was used to incorporate effects of different experimental groups followed by a post hoc protected least significant difference (PLSD) test to compare two individual groups. Other comparisons were made using a one-way ANOVA followed by a PLSD test to compare two individual groups. The Macintosh Statview SE + Graphic program (Abacus Concepts-Brain Power, Inc., Cary, NC) was used to perform the statistics.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). However, evidence for a 45% increase in adiposity by Pio was found in sc fat, whereas visceral fat was not altered. Pio treatment reduced plasma levels of triglycerides, FFAs, leptin, and insulin by 23%, 21%, 44%, and 29%, respectively, but did not alter glucose concentration. There was a slight (17%) increase in muscle triglyceride content, whereas no differences were found in glycogen and other lipid metabolites in muscle or liver among the groups.
|
) and triglyceride (Fig. 1B) levels were elevated by approximately 11- and 2.3-fold, respectively, above the values of Con-Gly rats (P < 0.01). Pio treatment (Pio-Lip) substantially reduced the elevation of plasma levels of FFAs (by 
30%) and triglycerides (by 36%) (P < 0.05 vs. Con-Lip, repeated ANOVA) during Intralipid/heparin infusion. Equalization of circulating FFA levels in Pio-treated rats (Pio-Isolip) to those of the Con-Lip group required a 50% increased lipid infusion rate (Fig. 1A). Although euglycemia was maintained in all groups (Fig. 1C), plasma insulin levels (Fig. 1D) were elevated to a greater degree in all lipid-infused groups, compared with the Con-Gly group. Compared with the Con-Lip group, plasma insulin levels in Pio-Lip rats tended to be lower (P = 0.07), but the values of Pio-Isolip rats were similar to the Con-Lip group. These differences resulted from differences in endogenous insulin concentrations (Table 2).
|
|
shows the time course of GIRs required to maintain euglycemia during the hyperinsulinemic clamp. In the Con-Lip group GIRs were inhibited after 120 min of lipid elevation (P < 0.01 vs. Con-Gly, repeated ANOVA) with an overall 46% decrease for the last 3 h (Table 2
). Pio treatment (Pio-Lip) preserved GIRs at normal levels (P > 0.05 vs. Con-Gly) with enhanced Rd (by 26%) and improved suppressibility of HGO (by 46%, both P < 0.01 vs. Con-Lip). In Pio-Isolip rats, the overall GIR was more suppressed than the Pio-Lip group (P < 0.01 vs. Pio-Lip) but remained 30% higher than the Con-Lip group (P < 0.05) along with sustained suppression of HGO. However, the Rd of the Pio-Lip group was reduced to a similar level to the Con-Lip group.
|
). There was no inhibition of glycogen synthesis in Pio-Lip rats. However, in Pio-Isolip rats it was inhibited to the same extent as in Con-Lip rats. Consistent with glycogen synthesis were glycogen content determined after the clamp (82 ± 2, 68 ± 5, 85 ± 4, 68 ± 3 µmol/g in Con-Gly, Con-Lip, Pio-Lip, and Pio-Isolip groups, respectively) with significant improvement in Pio-Lip (P < 0.01 vs. Con-Lip) but not in Pio-Isolip. In contrast, lipid synthesis from glucose was increased in both Pio-Lip (by 400%) and Pio-Isolip (by 100%), compared with Con-Lip (Fig. 3B). Compared with Con-Gly rats, LCACoAs were increased to various degrees in all lipid infusion groups, but differences among Con-Lip, Pio-Lip, and Pio-Isolip groups were not statistically significant (P > 0.05, Fig. 3C). Increases in muscle triglyceride content after 6 h of clamp were not significantly different among lipid-infused groups (Fig. 3D).
|
) and lipogenesis from glucose by 80% (Fig. 4B) during insulin stimulation. In Pio-Lip rats, insulin-stimulated glycogen synthesis and lipogenesis were increased by 37% and 58%, respectively (both P < 0.01 vs. Con-Lip). The improvement of both glycogen synthesis and lipogenesis from glucose remained intact in Pio-Isolip rats. Differences in glycogen content were not statistically significant among the four groups after the clamp (80 ± 10, 77 ± 10, 97 ± 11, and 93 ± 7 µmol/g for Con-Gly, Con-Lip, Pio-Lip, and Pio-Isolip, respectively).
|
). This increase was largely prevented in the Pio-Lip group (99 ± 4 nmol/g) and remained at a lesser level even in Pio-Isolip rats (96 ± 9 nmol/g). Compared with the Con-Gly group (9.5 ± 0.6 µmol/g), there was substantial accumulation of triglyceride content in liver in all lipid infusion groups after 6 h of clamp (P < 0.01), as shown in Fig. 4D. The accumulation of triglyceride was similar between the Con-Lip and Pio-Lip groups after 6 h of the clamp (Fig. 4D). In the Pio-Isolip group, triglyceride content (31.3 ± 1.6 µmol/g) was further increased after 6 h of clamp (P < 0.01 vs. both Con-Lip and Pio-Lip). The ratio of triglyceride/LCACoAs was similar in both untreated groups (170 ± 18 in Con-Gly vs. 162 ± 9 in Con-Lip) and increased in Pio-treated groups (Pio-Lip: 230 ± 25 and Pio-Isolip: 337 ± 30, both P < 0.05 vs. Con-Lip), suggesting a different partitioning of lipids in the liver related to Pio treatment. LCACoA content was inversely correlated with HGO among all lipid-infused groups (r2 = 0.428, P = 0.0013, data not shown).
Tissue PKB phosphorylation
In the basal state, there was no difference in muscle PKB content (4.3 ± 1.2 vs. 4.9 ± 0.5) or phosphorylation at the site of serine 473 (0.7 ± 0.3 vs. 0.9 ± 0.2 P-ser473/total) between control and Pio-treated rats. After a 6-h hyperinsulinemic clamp (Fig. 5, A and B), muscle PKB phosphorylation was increased 2.7-fold at serine 473 above the basal value (P < 0.01) in Con-Gly rats. Lipid infusion blunted insulin-stimulated PKB phosphorylation by 48% (1.7 ± 0.1 P-ser473/total), and this inhibition was completely prevented in Pio-treated rats. There was no significant difference in the protection of PKB phosphorylation between Pio-Lip and Pio-Isolip groups (4.1 ± 0.6 vs. 4.3 ± 0.6 P-ser473/total). Changes in threonine 308 phosphorylation followed a similar pattern.
|
Plasma levels of adiponectin
Recent studies have implicated adiponectin (also known as Acrp30 or AdipoQ), an important adipokine, in PPAR-mediated alterations in glucose and lipid metabolism (23). Thus, we determined its plasma levels under our experimental conditions. In the basal state, plasma adiponectin was almost 2-fold higher in Pio-treated rats than controls (Fig. 6A). The elevated levels of adiponectin in Pio-treated rats remained at similar high levels after 6 h of clamp (Fig. 6B). Its levels were negatively correlated with HGO and liver LCACoAs (Fig. 7).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
50%), presumably because of adipose tissue (24). However, when effects of elevated FFA clearance are avoided by increasing exogenous lipid administration to match FFA elevation, there is still substantial protection against FFA-induced insulin resistance (GIR 30% higher than untreated rats). This is mainly related to effects in the liver in which insulin maintains its ability to suppress hepatic glucose production, an effect highly negatively correlated to levels of liver lipid availability, as indicated by LCACoA content.
PPAR agonists are known to ameliorate existing insulin resistance that is associated with chronic lipid accumulation. It is generally believed that the lowering of circulating lipids mediated by PPAR activation reduces lipid availability to tissues and thus ameliorates insulin resistance (1, 2, 25). This is further supported by evidence of an attenuated accumulation of muscle LCACoAs (the intracellular active form of fatty acids) by Pio in high-fat-fed rats (4). However, it is not clear, quantitatively, how important this lowering of lipids is to PPAR-mediated insulin sensitization. Recently we have shown that Pio increases muscle insulin sensitivity to a much greater extent than does the PPAR agonist WY14643, even though both reduce systemic FFAs and muscle LCACoAs to a similar extent. This suggests a possible additional insulin-sensitizing effect induced by the PPAR agonist independent of the lowering of lipids (4). The present study provides further evidence for this interpretation by showing the protection of insulin action in muscle and liver by Pio, even though there was no reduction in lipids in the basal state or any lessening of triglyceride accumulation in liver and muscle.
In the Pio-Lip group, both plasma levels of triglycerides and FFAs were markedly lower than the Pio-Gly group. Although tissue insulin action is ultimately affected by lipid concentrations at the tissue level, reduced circulating lipid levels may result in less exposure of muscle and liver to lipid metabolites that impair their response to insulin stimulation. In an attempt to address this issue, we matched the plasma lipid levels of Pio-treated rats to the Con-Lip group. The requirement of a 50% greater lipid infusion to achieve this indicates substantially enhanced lipid clearance from the circulation in Pio-treated animals. This may be associated with the observed increase in subcutaneous fat mass that may have resulted from the known effect of TZDs to promote adipocyte proliferation (1, 26). However, even under the conditions of increased systemic lipid loading, the whole-body insulin sensitivity was still markedly protected in Pio-treated rats, suggesting a component of insulin sensitization in vivo by Pio is unrelated to modification of circulating FFA levels.
It is of interest to note marked differences in clamp plasma insulin levels between Con-Gly and all lipid infusion groups. This is consistent with the action of FFAs to stimulate insulin secretion acutely in vivo (27). Comparisons of endogenous and infused exogenous insulin concentrations confirmed that higher plasma levels of insulin were due to increased endogenous insulin, whereas exogenous insulin was the same among all groups. Thus, results in the present study suggest that lowering of circulating lipids may be a mechanism of the normalization of hyperinsulinemia in insulin-resistant states and diabetes following treatment with PPAR agonist drugs. However, because of the lower plasma insulin level in the Con-Gly group, compared with all lipid-infused groups, it was not possible to interpret our data as showing complete prevention of FFA-induced insulin resistance by Pio. Nonetheless, it is valid to conclude that there are substantial protective effects of Pio against FFA-induced insulin resistance because the observed enhancement of clamp GIR in the Pio-Lip group was achieved at a relatively lower (not higher) plasma insulin level, compared with the Con-Lip group. In addition, the protection of insulin sensitivity in the Pio-Isolip group occurred at the same plasma insulin concentration as the Con-Lip group.
We are aware of only one study using a similar protocol to that used in our study. Recently, Hevener et al. (13) showed that troglitazone treatment prevented FFA-induced insulin resistance via opposing an impairment of insulin-mediated Rd, which occurred in untreated rats. Although we also found protection of insulin sensitivity at the whole-body level after TZD treatment, there are several major differences between our study and Hevener et al. (13). First, whereas Hevener et al. found that protective effects were independent of systemic FFA levels, we found that the majority of the whole-body protective effect (all but 30%) could be accounted by matching FFA elevation. Second, in our study an enhanced insulin-mediated suppression of HGO appeared to be largely responsible for the component of protected insulin sensitivity that was independent of systemic FFA concentrations. Because Hevener et al. used maximal insulin levels in their study, effects on liver insulin sensitivity may not have been apparent. Third, Hevener et al. reported that the muscle FFA transporter/translocase FAT/CD36 was reduced by FFA elevation in untreated rats, but not in TZD-pretreated rats, and suggested that this may contribute to the protective effects in muscle. Although this is possible, it could also have opposite effects, and it could be argued that an increased ability of muscle to take up FFAs might be expected to exacerbate lipid accumulation and insulin resistance. However, the consequences of the different muscle FAT/CD36 levels on muscle lipids were not investigated by them. In our study, muscle triglyceride and LCACoAs during lipid infusion were not reduced by TZD pretreatment, and the protective effects on insulin-mediated Rd and glycogen synthesis seemed largely associated with plasma FFA levels.
These differences may be related to the duration of lipid infusion and the insulin infusion rates. In the previous study (13), maximal insulin stimulation was used and the intralipid infusion plus hyperinsulinemic-euglycemic clamp lasted for only 2 h. In comparison, our experiment was maintained for 6 h because FFA-induced insulin resistance was apparent only after 2 h of Intralipid infusion at a physiological range of hyperinsulinemia during the euglycemic clamp. The delayed onset of lipid-induced insulin resistance in this study is consistent with previous reports in humans (10) and rats (11, 28).
In muscle, excess lipid metabolites may impair insulin action by different mechanisms including interaction with the insulin signaling cascade (29) and activation of enzymes regulating glucose metabolism (30). In the current study, we measured insulin-stimulated PKB phosphorylation as a downstream index of the status of insulin signaling transduction. As recently reported (31), lipid infusion inhibited PKB phosphorylation mediated by insulin. This inhibition was prevented by the prior Pio treatment, which is consistent with the improvement of insulin-mediated lipogenesis in this study and a previous report that a TZD protects insulin-stimulated insulin receptor substrate-1 (IRS-1) phosphorylation and the associated phosphatidylinositol 3-kinase (PI3K) activity during lipid infusion (13). We elected to measure PKB phosphorylation in our 6-h study because its increase by insulin is less transient than upstream components such as IRS-1 phosphorylation and associated PI3K activity (32). Our findings are consistent with a previous report that a TZD preserves muscle insulin-stimulated IRS-1 phosphorylation and PI3K activity during lipid infusion (13). Further studies would be required to clarify effects of Pio on insulin-signaling cascade steps using a short period of insulin infusion as commonly employed (12, 13, 31). However, our results show that the effect of Pio to protect against muscle insulin resistance was largely associated with the lowering of plasma FFA concentrations. How circulating FFA levels could influence muscle insulin action without a change in lipid accumulation is not obvious and will involve further study. This seems to differ from effects of TZDs in the high-fat-fed rats in which enhanced muscle insulin action is associated with reductions in muscle lipid accumulation (4, 25). It is possible in the Intralipid-infused rat that increased muscle fat oxidation because of higher FFA levels may oppose, via the glucose-FFA cycle, any TZD-related effects to enhance insulin sensitivity.
In terms of the insulin-sensitizing action of Pio in the liver, several lines of evidence suggest that modulation of intracellular lipids may be involved. The present study clearly showed that liver LCACoA content was markedly reduced to a similar degree in association with improvement of insulin’s suppression of hepatic glucose production in both Pio-Lip and Pio-Isolip groups. Furthermore, there was a strong negative correlation between LCACoA content and insulin’s action on hepatic glucose production across all groups. LCACoAs are the metabolically active form of intracellular FFAs. Although exact mechanisms are unknown, there is a strong likelihood that they can inhibit insulin action in muscle and liver (30). Thus, the decrease in liver LCACoAs concentrations may play a role in protecting the action of insulin on HGO suppression and glycogen and lipid synthesis in liver. This seems to be consistent with the fact that PPAR agonists down-regulate the expression of hepatic gluconeogenic enzymes in insulin-resistant obese rodents (7, 33) but not in normal animals (33).
Because the reduction in liver LCACoAs occurred independently of plasma FFA levels, it appears unlikely that this decrease was caused by a lesser lipid influx to liver. PPAR agonist treatment has been reported to have no effect on FFA transport proteins (7) or cause a slight increase in FAT/CD36 (34) in the liver. Our results show that liver triglyceride content was markedly increased in Pio-Lip rats despite decreases in LCACoA levels in this tissue. These data strongly suggest that reduced accumulation of hepatic LCACoAs during lipid infusion in Pio-treated rat resulted from increased intracellular partitioning toward storage. The difference in LCACoA and triglyceride levels between muscle and liver may reflect the metabolic nature of these two tissues. Although both muscle and liver are important tissues of FFA oxidation, liver (but not muscle) is also a major site of lipogenesis. Although PPAR agonist treatment causes increases in FFA oxidative enzyme expression in both tissues, a number of lipogenic enzymes are coordinately up-regulated only in liver, including fatty acid synthase and stearyl-CoA desaturase acyl-CoA synthetase (7). Therefore, it is possible that the reduced LCACoA/triglyceride ratio in liver, compared with muscle of TZD-treated animals, may reflect increased liver lipogenesis. In addition, TZD treatment has been shown to increase the expression of PPAR in liver, which may contribute to PPAR-mediated responses (35).
PPAR agonists can alter adipokines such as leptin (4, 36), TNF (37), and adiponectin (23, 38), which may affect insulin sensitivity (39). Although leptin may improve insulin action, its plasma level was reduced after Pio treatment. In contrast, plasma adiponectin concentrations were 2-fold higher in Pio-treated rats to the level similar to that shown to enhance insulin-mediated suppression of HGO in primary hepatocytes (23) and in vivo (40). Further support for a role of adiponectin in liver comes from the observation that administration of adiponectin promotes liver triglyceride storage (23), a result consistent with the present finding that the ratio of liver triglyceride content to LCACoA content was increased in Pio-treated rats after the hyperinsulinemic clamp. However, further studies are required to investigate whether adiponectin sensitizes hepatic insulin action directly or indirectly via reducing LCACoAs.
In conclusion, pretreatment of normal rats with a TZD PPAR agonist Pio substantially protects against insulin resistance that otherwise would be induced by acute lipid oversupply. The principal protective effect in muscle is associated with a substantial increase in systemic lipid clearance. In addition, Pio protects liver insulin sensitivity independent of circulating FFA level but associated with a reduction in liver LCACoAs and an elevation of plasma adiponectin levels. Overall, the present study may have important clinical significance for the use of TZDs to prevent the development of lipid-related insulin resistance.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: Con-Gly, Vehicle-pretreated rats infused with glycerol; Con-Lip, vehicle-pretreated rats infused with lipid; FFA, free fatty acid; GIR, glucose infusion rate; HGO, hepatic glucose output rate; IRS-1, insulin receptor substrate-1; LCACoA, long-chain acyl coenzyme A; Pio, Pioglitazone; Pio-Isolip, Pioglitazone-treated rats infused with lipid; Pio-Lip, Pioglitazone-treated rats infused with lipid; PKB, protein kinase B; PLSD, protected least significant difference test; PPAR, peroxisome proliferator-activated receptor; Rd, glucose disappearance rate; TZD, thiazolidinedione.
Received April 8, 2002.
Accepted for publication August 6, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
: adipogenic regulator and thiazolidinedione receptor. Diabetes 47:507–514[Abstract]
activation lowers muscle lipids and improves insulin sensitivity in high fat-fed rats: comparison with PPAR- activation. Diabetes 50:411–417 expression in human skeletal muscle. J Clin Endocrinol Metab 85:4293–4297 activation has coordinate effects on gene expression in multiple insulin-sensitive tissues. Endocrinology 142:1269–1277-mediated changes in gene expression. Diabetes 50:2309–2315
and alterations in the insulin signaling cascade. Diabetes 48:1270–1274[Abstract]
/ agonist, improves insulin sensitivity and increases fatty acid uptake into white adipose tissue of high-fat-fed rats. Diabetes 50:A121–A121
/
but not on glycogen synthase kinase-3. J Biol Chem 277:32915–32922 in liver. Mol Cell Biol Res Commun 2:202–208[CrossRef][Medline]
. J Clin Invest 98:1004–1009[Medline]
-induced insulin resistance by a mechanism independent of adipogenic activity of peroxisome proliferator-activated receptor-. Diabetes 50:1083–1092 ligands increase expression and plasma concentrations of adiponectin, an adipose-derived protein. Diabetes 50:2094–2099This article has been cited by other articles:
 |
|
 |
|
  G. Frangioudakis and G. J Cooney Acute elevation of circulating fatty acids impairs downstream insulin signalling in rat skeletal muscle in vivo independent of effects on stress signalling J. Endocrinol., May 1, 2008; 197(2): 277 - 285. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. G. Haider, F. Mittermayer, G. Schaller, M. Artwohl, S. M. Baumgartner-Parzer, G. Prager, M. Roden, and M. Wolzt Free fatty acids normalize a rosiglitazone-induced visfatin release Am J Physiol Endocrinol Metab, November 1, 2006; 291(5): E885 - E890. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Dhindsa, D. Tripathy, N. Sanalkumar, S. Ravishankar, H. Ghanim, A. Aljada, and P. Dandona Free Fatty Acid-Induced Insulin Resistance in the Obese Is Not Prevented by Rosiglitazone Treatment J. Clin. Endocrinol. Metab., September 1, 2005; 90(9): 5058 - 5063. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. E. Cleasby, N. Dzamko, B. D. Hegarty, G. J. Cooney, E. W. Kraegen, and J.-M. Ye Metformin Prevents the Development of Acute Lipid-Induced Insulin Resistance in the Rat Through Altered Hepatic Signaling Mechanisms Diabetes, December 1, 2004; 53(12): 3258 - 3266. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Tonelli, W. Li, P. Kishore, U. B. Pajvani, E. Kwon, C. Weaver, P. E. Scherer, and M. Hawkins Mechanisms of Early Insulin-Sensitizing Effects of Thiazolidinediones in Type 2 Diabetes Diabetes, June 1, 2004; 53(6): 1621 - 1629. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Sutinen, K. Kannisto, E. Korsheninnikova, R. M. Fisher, E. Ehrenborg, T. Nyman, A. Virkamaki, T. Funahashi, Y. Matsuzawa, H. Vidal, et al. Effects of rosiglitazone on gene expression in subcutaneous adipose tissue in highly active antiretroviral therapy-associated lipodystrophy Am J Physiol Endocrinol Metab, June 1, 2004; 286(6): E941 - E949. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. R. Colca, W. G. McDonald, D. J. Waldon, J. W. Leone, J. M. Lull, C. A. Bannow, E. T. Lund, and W. R. Mathews Identification of a novel mitochondrial protein ("mitoNEET") cross-linked specifically by a thiazolidinedione photoprobe Am J Physiol Endocrinol Metab, February 1, 2004; 286(2): E252 - E260. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology |
): control rats infused with glycerol; Con-Lip (
): control rats infused with lipid (1.3 ml/h) plus heparin (40 U/h); Pio-Lip (
): Pio-pretreated rats infused with lipid (1.3 ml/h) plus heparin (40 U/h); Pio-Isolip (
): Pio-treated rats infused with increased lipid (2.0 ml/h) plus heparin (60 U/h). Statistical results are described in the text using a two-way repeated ANOVA (n = 7–8/group).
