Increased activity of proinflammatory/stress pathways has been implicated in the pathogenesis of insulin resistance in obesity. However, the effects of obesity on the activity of these pathways in skeletal muscle, the major insulin-sensitive tissue by mass, are poorly understood. Furthermore, the mechanisms that activate proinflammatory/stress pathways in obesity are unknown. The present study addressed the effects of diet-induced obesity (DIO; 6 wk of high-fat feeding) and acute (6-h) hyperlipidemia (HL) in rats on activity of IKK/IκB/NF-κB c-Jun NH2-terminal kinase, and p38 MAPK in three skeletal muscles differing in fiber type [superficial vastus (Vas; fast twitch-glycolytic), soleus (Sol; slow twitch-oxidative), and gastrocnemius (Gas; mixed)]. DIO decreased the levels of the IκBα in Vas (24 ± 3%, P = 0.001, n = 8) but not in Sol or Gas compared with standard chow-fed controls. Similar to DIO, HL decreased IκBα levels in Vas (26 ± 5%, P = 0.006, n = 6) and in Gas (15 ± 4%, P = 0.01, n = 7) but not in Sol compared with saline-infused controls. Importantly, the fiber-type-dependent effects on IκBα levels could not be explained by differential accumulation of triglyceride in Sol and Vas. HL, but not DIO, decreased phospho-p38 MAPK levels in Vas (41 ± 7% P = 0.004, n = 6) but not in Sol or Gas. Finally, skeletal muscle c-Jun NH2-terminal kinase activity was unchanged by DIO or HL. We conclude that diet-induced obesity and acute HL reduce IκBα levels in rat skeletal muscle in a fiber-type-dependent manner.
- nuclear factor-κB inhibitor kinase
- nuclear factor-κB inhibitor
- nuclear factor-κB
recently, it has been proposed that elevated activity of proinflammatory/stress-signaling pathways plays an important role in the pathogenesis of insulin resistance in obesity, and perhaps type 2 diabetes. This hypothesis is supported by the observations that obesity and type 2 diabetes are states of chronic, low-grade inflammation (9, 12, 35), that increased expression of proinflammatory factors, such as TNF-α (13, 26, 27) and IL-6 (21), decreases insulin sensitivity, and by more recent studies demonstrating that decreased activity of proinflammatory/stress-signaling pathways improves, or prevents, the development of insulin resistance (1, 5, 14, 17, 20, 32, 38).
Cell responses to inflammatory and stress signals are mediated by a number of ubiquitously expressed signaling cascades. Of these, the nuclear factor-κB (NF-κB) inhibitor (IκB) kinase (IKK)/IκB/NF-κB and mitogen-activated protein kinase (MAPK) pathways, specifically c-Jun NH2-terminal kinase (JNK) and p38 MAPK, are best described. Altered activity of each of these pathways has been implicated in changing insulin action. Thus reduced IKK/IκB/NF-κB activity (1, 5, 17, 20, 32, 38) prevents the development of lipid-induced and/or obesity-induced insulin resistance in cells (32) and in rodents (1, 5, 20, 38) and improves insulin sensitivity in type 2 diabetes mellitus (17). Increased activity of IKK/IκB/NF-κB in mouse skeletal muscle induces dramatic tissue wastage, but insulin action appears to be unaltered (4), while increased IKK/IκB/NF-κB activity in liver (5) results in the development of insulin resistance. JNK-deficient mice are partially resistant to the development of insulin resistance resulting from a high-fat diet (14), whereas basal p38 MAPK phosphorylation is increased, and responsiveness to insulin is lost (22) in skeletal muscle in type 2 diabetes mellitus. Finally, studies in mouse skeletal muscle and L6 myotubes suggest that increased activity of p38 MAPK may be a negative regulator of glucose uptake (15).
Despite these observations, the effects of obesity on the activity of proinflammatory/stress pathways in skeletal muscle, the major insulin-sensitive tissue by mass in the body, remain unclear. Furthermore, the mechanisms of activation of proinflammatory/stress pathways in obesity are poorly understood. The present study was undertaken to address each of these issues. Specifically, we determined the effects of diet-induced obesity (DIO) and acute hyperlipidemia (HL) on levels of IκBα and p38 MAPK phosphorylation (commonly used indicators of activity of NF-κB and p38 MAPK, respectively) and JNK activity in three skeletal muscles differing in fiber-type composition and insulin sensitivity. The data demonstrate that DIO and acute HL decrease skeletal muscle IκBα, and that the effects of DIO and HL on IκBα levels are skeletal muscle fiber-type dependent. Furthermore, HL, but not DIO, decreases p38 MAPK activity in a fiber-type-dependent manner. However, DIO and HL have no effect on skeletal muscle JNK activity.
MATERIALS AND METHODS
Animal care and maintenance.
Male Wistar rats were purchased from Charles River (Madison, WI) at a weight of 175–275 g and maintained on a constant 12:12-h light-dark cycle, with free access to water. For obesity studies, rats (starting weight 175–200 g) were ad libitum fed either a standard rat chow (11% of calories from fat) or a high-fat diet (Harlan Teklad, Madison, WI, TD 96001, 45% of calories from fat) for 6 wk. For lipid infusion studies, all animals (starting weight 250–275 g) were ad libitum fed standard rat chow before the study. All experimental procedures were approved by the Institutional Animal Care and Use Committee of the University of Pittsburgh and were in accordance with the National Research Council’s Guide for the Care and Use of Laboratory Animals.
A 4- to 6-wk high-fat diet is sufficient to induce obesity and insulin resistance in rats (3, 16, 29, 33, 34). In the present study, 6-wk high-fat-fed animals and standard chow-fed control animals were fasted overnight (18 h) and anesthetized with pentobarbital sodium, and three skeletal muscles differing in fiber type [superficial vastus (Vas), fast-twitch glycolytic; soleus (Sol), slow-twitch oxidative; and gastrocnemius (Gas), mixed fast-twitch glycolytic and slow-twitch oxidative] were extracted and flash-frozen in liquid nitrogen. Tissues were stored at −70°C until analysis.
Lipid infusion studies.
Approximately 8 days before the study, catheters were inserted into the carotid artery and the contralateral jugular vein, as described previously (7, 28, 30). When animals had recovered to >90% of presurgery weight, they were fasted overnight. The following day, the indwelling cannulas were cleared with saline and extended with Silastic tubing, after which the animals were infused with Liposyn II/heparin (5 ml·kg−1·h−1 and 6 U/h, respectively, Abbott Laboratories, Chicago, IL) or saline (5 ml·kg−1·h−1) for 6 h. In preliminary studies, we confirmed that this intervention induced insulin resistance, as assessed by a hyperinsulinemic-euglycemic clamp (see results). In all other studies, after completion of the lipid or saline infusions, the animals were anesthetized with pentobarbital, and skeletal muscles were extracted and flash-frozen in liquid nitrogen. Tissues were stored at −70°C until analysis.
IκBα, phospho-p38, and p38 MAPK levels were measured by using a standard Western blotting procedure. Briefly, 50 mg of tissue were homogenized in lysis buffer, containing, as final concentrations, 20 mM Tris·HCl, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Tween-20, 2.5 mM pyrophosphate, 1 mM β-glycerophosphate, 1 mM sodium vanadate, and 1 μg/ml leupeptin. Equal amounts of protein were resolved by SDS-PAGE on a 10% Tris·HCl gel (Biorad, Hercules, CA) and then transferred to a polyvinylidene difluoride membrane (Immobilon, Bedford, MA). After blocking for 1 h in 5% nonfat dried milk in 1% Tris-buffered saline-Tween 20 at room temperature, membranes were incubated for 1 h with α-IκBα (1:1,000, Santa Cruz Biotechnology, Santa Cruz, CA), α-phospho-p38 MAPK (1:1,000, Cell Signaling, Beverly, MA) or α-p38 MAPK (1:1,000, Cell Signaling) at room temperature. After washing, the membrane was incubated with horseradish peroxidase-linked goat anti-rabbit IgG antibody for 1 h at room temperature. The membrane-bound antibodies were then detected by luminol chemiluminescence (Lumiglo, Cell Signaling Technology). The membrane was exposed to autoradiographic film, bands were scanned, and optical density was determined (NIH Image J, version 1.62). JNK activity was measured by an adaptation of the method of Hirosumi et al. (14). Briefly, 50 mg of tissue were homogenized in 0.5 ml of a lysis buffer containing, as final concentrations, 25 mM Tris·HCl, pH 7.4, 10 mM Na3VO4, 100 mM NaF, 50 mM Na4P2O7, 10 mM EGTA, 10 mM EDTA, 1% Igepal, 20 nM okadaic acid, 10 μg/ml leupeptin, 10 μg/ml aprotonin, and 2 mM PMSF. Five hundred micrograms of protein, 2.5 μg of α-JNK antibody (Santa Cruz Biotechnology), and 22 mg of protein A sepharose beads (Amersham Biosciences, Piscataway, NJ) were mixed and incubated overnight at 4°C. After centrifugation at 1,000 rpm at 4°C, the beads were washed three times in lysis buffer. The pellet was then resuspended in kinase buffer (25 mM HEPES, pH 7.4, 20 mM MgCl2, 20 mM β-glycerophosphate, 0.5 mM EGTA, 0.5 mM NaF, 0.5 mM Na3VO4, 1 mM PMSF, 1 mM dithiothreitol, 10 mM ATP). The reaction was incubated for 20 min at 30°C in the presence of 8 μg of c-jun fusion protein (Cell Signaling) and 1 μl (5 mCi/ml) of [γ-32P]ATP. The reaction was terminated by adding SDS and subsequent boiling of samples. Equal amounts of reaction product were resolved by SDS-PAGE and transferred to polyvinylidene difluoride membrane. The membrane was exposed to autoradiographic film and then scanned and quantified. Tissue triglycerides were measured, as described previously, with modifications. Briefly, ∼30 mg of tissue were crushed while under liquid nitrogen and homogenized in ice-cold (80%) methanol (MeOH)-H2O containing butylated-hydroxytoluene (5 μg/ml) as an antioxidant. The MeOH-H2O homogenate was mixed with chloroform (CHCl3) (2:1 vol/vol CHCl3-MeOH) to extract lipids, as described by Folch et al. (10). The extracted lipid sample was dried under vacuum, and lipids were resuspended in 500 μl of CHCl3. A 200 μl aliquot of extracted lipids was placed into a glass tube and allowed to air-dry overnight. The dried sample was dissolved in 60 μl of tert-butanol and 40 μl of Triton X-114-methanol (2:1) mixture. The triglycerides were then measured spectrophotometrically (Beckman DU 530) using the GPO-triglyceride kit (Sigma) with Lintrol lipids (Sigma) as standard (3).
Plasma triglycerides were measured by using the GPO-triglyceride kit (Sigma, Atlanta, GA) with Lintrol lipids as standards (Sigma, Atlanta, GA). Plasma free fatty acids (FFAs) were measured by using the FFA, half-micro test kit (Roche Diagnostics, Penzberg, Germany).
All results are expressed as means ± SE. Statistical significance was determined by unpaired t-test using the Systat statistical program (Evanston, IL). Significance was assumed at P < 0.05.
Skeletal muscle IκBα levels are decreased by DIO in a fiber-type-dependent manner.
In previous studies, our laboratory (3, 16) and others (29, 33, 34) have demonstrated in rats that a 4- to 6-wk high-fat diet is sufficient to induce obesity and insulin resistance. Thus we first determined the effects of DIO on levels of skeletal muscle IκBα, a commonly used indicator of activity of the IKK/IκB/NF-κB pathway (24). Body weight, adiposity, and plasma FFA and triglyceride concentrations were increased in DIO compared with standard chow-fed (Lean) controls (Table 1). In superficial Vas, a muscle composed predominantly of fast-twitch, glycolytic fibers, IκBα levels were decreased by 24 ± 3% (P = 0.001) in DIO compared with Lean (Fig. 1). Interestingly, these decreases did not occur in Sol, a slow-twitch, oxidative fiber type muscle, or in Gas, a mixed fiber-type muscle (Fig. 1). Thus decreases in IκBα levels induced by DIO in skeletal muscle are fiber-type dependent.
Acute HL reproduces the effects of DIO on IκBα levels in skeletal muscle.
Dysregulated lipid metabolism and HL are commonly associated with obesity and have been implicated in the pathogenesis of insulin resistance (2, 19, 25, 36). Furthermore, lipids activate IKK/IκB/NF-κB in macrophages (23), endothelial cells (8), and L6 rat myotubes (32). Thus HL is one possible mechanism mediating decreased skeletal muscle IκBα levels in DIO. To address this hypothesis in the absence of other confounding metabolic abnormalities of obesity, acute HL was induced in Lean animals by an infusion of liposyn II for 6 h. As previously reported (37), this intervention increased plasma FFA levels by ∼20-fold (Table 1) and induced insulin resistance as determined by the glucose clamp method (glucose infusion rate of 31.9 ± 1.1 and 38.5 ± 0.8 mg·kg−1·min−1 in lipid-infused compared with saline-infused animals, respectively, in the final 60 min of a hyperinsulinemic clamp, P = 0.001 lipid vs. saline, n = 6 in each group). Similar to DIO, IκBα levels were decreased in Vas (26 ± 5%, P = 0.006) in lipid-infused compared with saline-infused animals, but were unaltered in Sol (Fig. 2). However, unlike DIO, IκBα levels were also decreased in Gas (15 ± 4%, P = 0.01) compared with controls.
Skeletal muscle triglyceride levels do not correlate with the fiber-type-dependent effects of DIO and acute HL on IkBα levels.
Accumulation of tissue lipid (lipotoxicity) has been implicated in the pathogenesis of insulin resistance and activation of inflammatory pathways. Thus one possible mechanism to explain the fiber-type differences in activation of the NF-κB pathway is fiber-type differences in accumulation of lipid in response to DIO and acute HL. Basal (Lean or saline-infused groups) triglyceride levels differed by fiber type (Sol > Vas, P < 0.05, Fig. 3) and by animal weight (heavier > lighter, P < 0.05, Table 1 and Fig. 3). DIO increased Sol triglycerides compared with Lean controls by 53.7 ± 8.9% (Fig. 3A). Vas triglycerides were increased in DIO to a similar relative extent (44.2 ± 12.4%, Fig. 3A) compared with Lean controls, but the absolute increase (50.6 ± 14.9 μg/mg protein in Sol vs. 22.8 ± 4.9 μg/mg protein in Vas) was less in Vas compared with Sol (Fig. 3A, P = 0.005). In response to acute HL, there was a paradoxical decrease in Sol triglycerides, confirming previous reports (37), and no change in Vas triglycerides. Thus the fiber-type-dependent effects of DIO and acute HL on IκBα levels are not correlated with differential accumulation of triglycerides.
Skeletal muscle JNK activity and p38 MAPK phosphorylation are not changed by DIO.
Similar to IKK/IκB/NF-κB, altered JNK, and p38 MAPK activity has been implicated in mediating alterations in insulin action (6, 14, 15, 22). Thus we next determined the effects of DIO and acute HL on activity of these two pathways in skeletal muscle. DIO did not alter the activity of JNK or p38 MAPK in any of the skeletal muscles examined (Figs. 4 and 5). Acute HL decreased p38 MAPK phosphorylation in Vas, but not in Sol or Gas, and had no effect on JNK activity. JNK activity was not measured in Sol because of lack of tissue.
The primary goal of the present study was to determine the effects of DIO in rats on the activity of inflammatory/stress pathways in skeletal muscle and to begin to address possible mechanisms responsible for activation of these pathways. The rationale for this study was recent observations, implicating altered activity of proinflammatory/stress pathways, specifically IKK/IκB/NF-κB, JNK, and p38 MAPK in altering insulin action. A number of novel observations arise from our studies. We demonstrate that 1) DIO is associated with decreased levels of the IκBα in skeletal muscle; 2) the effects of DIO are fiber-type dependent, as IκBα levels were decreased in fast, glycolytic fibers, but not in slow, oxidative fibers; 3) acute HL reduces skeletal muscle IκBα levels, suggesting one mechanism for the effects of DIO; 4) differential accumulation of skeletal muscle triglycerides does not correlate with the fiber-type-dependent effects of DIO and acute HL on IkBα levels; and 5) DIO has no effect on p38 MAPK or JNK activity in rat skeletal muscle.
The present study demonstrates that both DIO and acute HL result in decreased levels of skeletal muscle IκBα, a situation that is normally associated with increased NF-κB activity. These data are in good agreement with a previous observation in humans of decreased skeletal muscle IκBα levels in response to an acute lipid infusion (18). Recent studies have demonstrated that inhibition of the IKK/IκB/NF-κB pathway by salicylates, an IKK inhibitor, decreases skeletal muscle insulin resistance associated with a high-fat diet (38), acute HL (20), and type 2 diabetes (17). Direct evidence for a role for increased IKK/IκB/NF-κB activity in the pathogenesis of skeletal muscle insulin resistance is more varied. One group reports in mice that a deficiency of IKK activity, and hence NF-κB activity, protects against the development of insulin resistance on a high-fat diet (38) or in response to a lipid infusion (20), but a more recent study failed to reproduce the effects of a high-fat diet (31). The reasons for these discrepant findings are unclear. Constitutively active skeletal muscle IKK activity in mouse induces severe muscle wastage, but insulin sensitivity is unaltered (4). However, it is unclear how insulin sensitivity was determined in this study or whether the potentially confounding variable of muscle wastage influenced insulin action in these animals. In the context of the present study, we can state that IKK/IκB/NF-κB activation is associated with the development of skeletal muscle insulin resistance in two independent rat models (present study and Ref. 20). Furthermore, previous studies in L6 rat myotubes demonstrate that activation of IKK/IκB/NF-κB by fatty acids induces insulin resistance and that blocking of NF-κB activation is sufficient to prevent lipid-induced insulin resistance (32). However, given the studies discussed above, it cannot be stated at this time that NF-κB activation in skeletal muscle is required for the development of insulin resistance in vivo.
An important conclusion arising from our study is that the effects of DIO and acute HL on the levels of IκBα in skeletal muscle are fiber-type dependent. Skeletal muscle fiber types can be characterized based on differences in their oxidative capacity and insulin sensitivity. Thus slow-twitch fibers have a high oxidative capacity and are very insulin sensitive. Conversely, fast-twitch fibers have a lower oxidative capacity and insulin sensitivity. IκBα levels were decreased in DIO in superficial Vas, but not in Gas or Sol muscle. Relatively similar results were obtained with acute HL, although, unlike DIO, IκBα levels were also decreased in Gas. In humans, acute HL activates IKK/IκB/NF-κB in a mixed fiber-type muscle (Vas), confirming that our observations are not restricted to rodents. Perhaps the most unexpected observation was that IKK/IκB/NF-κB was not activated in Sol muscle, the most insulin-sensitive muscle examined, by DIO or acute HL. Previous studies (3, 37) have demonstrated that insulin resistance is induced in Sol muscle by DIO and acute HL. One possible interpretation of our data is that IKK/IκB/NF-κB activity is not required for obesity/lipid-induced insulin resistance in Type I fibers. Alternatively, activation of IKK/IκB/NF-κB may have occurred before the time points examined in this study. In this regard, the IκBα gene is a target of NF-κB activity, resulting in the replenishment of IκBα and dampening of inflammatory signals that activate NF-κB. This negative feedback is commonly considered a critical mechanism for regulating the length and magnitude of inflammatory responses. Lastly, it has been demonstrated that IKK serine phosphorylates insulin receptor substrate-1, in HepG2 cells (11), which would be expected to decrease insulin-stimulated phosphatidylinositol 3-kinase activity, independent of decreases in IκBα.
Although the mechanism(s) responsible for the decrease in IκBα levels in skeletal muscle in DIO are unclear, our demonstration that acute HL decreases IκBα in skeletal muscle in Lean animals is suggestive. In support of a role for dyslipidemia in activation of inflammatory pathways are the observations of dysregulated lipid metabolism in obesity, resulting in substantially elevated plasma levels of FFAs, triglycerides, and lipoproteins, and elevated tissue levels of triglycerides and other lipid metabolites (2, 19, 36). Furthermore, lipid-induced activation of IKK/IκB/NF-κB has been demonstrated in L6 myotubes (32), macrophages (23), and endothelial cells (8). Finally, there is a substantial body of work supporting a role for lipids in the pathogenesis of insulin resistance (2, 19, 36). However, it should be noted that a number of proinflammatory factors that are elevated in obesity are also capable of activating IKK/IκB/NF-κB, suggesting that activation of this pathway may be mediated by a number of mechanisms in obesity.
It has been proposed that accumulation of lipids within tissues may be a mechanism of activation of inflammatory pathways. In the present study, triglyceride levels were increased in both Sol and superficial Vas muscles in DIO compared with Lean controls, whereas IκBα levels were only reduced in the superficial Vas. The effects of acute HL were somewhat more complex. Thus, as previously reported by Yu et al. (37), Sol muscle triglycerides decreased in response to acute HL. Superficial Vas triglycerides were unchanged. What one can conclude from these data is that the fiber-type-dependent effects of DIO and HL on IkBα levels are not explained by differential accumulation of triglycerides in the different fiber types. Furthermore, in a separate study, we measured the accumulation of ceramide and diacylglycerol (two other lipid metabolites implicated in the activation of the NF-κB pathway) in response to acute HL and found accumulations of both metabolites but no differences in the extent of accumulation between superficial Vas and Sol muscles (Dube JJ, Bhatt BA, Dedousis N, Bonen A, and O’Doherty RM, unpublished data).
A somewhat surprising observation in the present study is that the activity of JNK and p38 MAPK was, for the most part, unaltered by DIO or acute HL in the three muscles examined. A previous study (14) in mice demonstrated increased JNK activity in skeletal muscle of DIO mice compared with Lean mice, whereas increased basal activity of p38 MAPK has been reported in skeletal muscle in type 2 diabetes (6, 22). The reason for the differences between our present observations and those from other studies are unclear. In the case of JNK, one can propose species difference (rats vs. mice), but, given the close phylogenetic relationship of the two species, this is a somewhat unsatisfactory explanation. In the case of p38 MAPK, it is more likely that the species differences do play a role, in addition to the fact that Type 2 diabetes is a potentially confounding variable.
In conclusion, the present study demonstrates that both DIO and acute HL decrease skeletal muscle IκBα levels in a fiber-type-dependent manner. Because the effects of DIO and acute HL are similar, we speculate that dyslipidemia may be a mechanism for activation of inflammatory/stress pathways in obesity and type 2 diabetes.
This work was supported by an American Diabetes Association Career Development Award and National Institute of Diabetes and Digestive and Kidney Diseases Grant RO1-DK-58855–01 (both to R. M. O’Doherty). B. A. Bhatt is supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant T32-DK-07052 (Research Training in Diabetes and Endocrinology).
We sincerely thank Dr. Hotamisligil and Umut Ozcan for valued help with the JNK activity assay.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 the American Physiological Society