Aberrant REDD1-mTORC1 responses to insulin in skeletal muscle from Type 2 diabetics

Abstract

The objective of this study was to establish whether alterations in the REDD1-mTOR axis underlie skeletal muscle insensitivity to insulin in Type 2 diabetic (T2D), obese individuals. Vastus lateralis muscle biopsies were obtained from lean, control and obese, T2D subjects under basal and after a 2-h hyperinsulinemic (40 mU·m⁻²·min⁻¹)-euglycemic (5 mM) clamp. Muscle lysates were examined for total REDD1, and phosphorylated Akt, S6 kinase 1 (S6K1), 4E-BP1, ERK1/2, and MEK1/2 via Western blot analysis. Under basal conditions [(-) insulin], T2D muscle exhibited higher S6K1 and ERK1/2 and lower 4E-BP1 phosphorylation (P < 0.05), as well as elevations in blood cortisol, glucose, insulin, glycosylated hemoglobin (P < 0.05) vs. lean controls. Following insulin infusion, whole body glucose disposal rates (GDR; mg/kg/min) were lower (P < 0.05) in the T2D vs. the control group. The basal-to-insulin percent change in REDD1 expression was higher (P < 0.05) in muscle from the T2D vs. the control group. Whereas, the basal-to-insulin percent change in muscle Akt, S6K1, ERK1/2, and MEK1/2 phosphorylation was significantly lower (P < 0.05) in the T2D vs. the control group. Findings from this study propose a REDD1-regulated mechanism in T2D skeletal muscle that may contribute to whole body insulin resistance and may be a target to improve insulin action in insulin-resistant individuals.

two-thirds of the united states has a 25% or greater prevalence of obesity (54), which places increased demand and cost on the health care system. Of the numerous comorbidities associated with obesity, the development of insulin resistance and Type 2 diabetes (T2D) are major concern for obese individuals, leading to overall reductions in metabolic flexibility (50). This becomes more of a concern, since the population of obese individuals is expanding in all age groups. Findings from the National Health and Nutrition Examination Survey data (68, 69) show that skeletal muscle mass is fundamental for insulin sensitivity. Skeletal muscle mass is lower in the obese when expressed relative to total body mass vs. a lean individual and has a high pervasiveness of ectopic lipid within skeletal muscle that contributes to reduced insulin action (27, 50). Among other factors, elevated endogenous glucocorticoid concentrations contribute to insulin resistance in the obese and T2D (58, 74).

Insulin activation of the insulin receptor substrate 1 (IRS-1), phosphatidyl-inositol-3 kinase (PI3K), Ak strain transforming (Akt), and the ERK 1/2 pathways are essential for the regulation of cellular glucose uptake, proliferation, and growth. Akt and ERK1/2 activation increases GTPase function of the Ras homolog enriched in brain (Rheb) protein toward the mechanistic target of rapamycin (mTOR; also known as mammalian target of rapamycin) by inactivating the repressive effects of the tuberous sclerosis complex (TSC). The TSC complex remains central to the integration of signaling inputs from Akt, ERK1/2, and AMPK (32, 44, 60) for metabolic homeostasis, cell growth, and adaptation (26, 48). The protein regulated in development and DNA damage responses 1 [REDD1; also known as DNA-damage-inducible transcript 4 (DDIT4)], dexamethasone-induced gene 2 (Dig2) and RTP801 act to repress mTOR signaling by promoting an active TSC1:TSC2 complex formation, promoting an increase in GDP-bound Rheb. REDD1's mechanism of action on TSC2 is by sequestering the modulatory protein, 14-3-3, which alters protein and kinase function through serine/threonine motif phosphorylation-dependent binding, as well as recruiting serine-threonine protein phosphatase 2A (PP2A) to dephosphorylate T308 on Akt (13).

Despite the principal role of REDD1 in glucocorticoid function and the prevalent use of glucocorticoids by insulin-resistant cohorts (e.g., obese, T2D, aged) (56, 57), little has been established about REDD1 in relation to insulin action. Moreover, the same stressors that upregulate REDD1 expression, such as glucocorticoids (75), DNA damage (42), endoplasmic reticulum stress (61, 76), and hypoxia (7, 65), are the same stressors that are observed in obese individuals (1, 16, 47) and inhibit mTOR function (7, 23, 75). Work from our laboratory (20, 79) and others (14, 47) support the emerging role of REDD1 in metabolism and insulin action. Specifically, we report that obesity-associated elevations in REDD1 expression blunts skeletal muscle Akt and mTOR signaling in response to nutrients (79). In this work (79) and other work from our laboratory (17), we also show that the mTOR substrate ribosomal protein S6 kinase 1 (S6K1) is hyperactive under basal conditions in muscle from obese rodents, which may negatively regulate insulin action through the inhibition of IRS-1 (73). Therefore, we sought to determine whether alterations in the REDD1-mTOR axis underlie skeletal muscle insensitivity to insulin in obese, T2D individuals. We hypothesized that aberrant REDD1 expression is associated with repressed insulin activation of mTOR signaling in obese, T2D skeletal muscle compared with lean individuals used as a benchmark for optimal health.

METHODS

Subjects.

Ten young, healthy control subjects and eight older, T2D sedentary volunteers were recruited from the local population to undergo metabolic testing and a hyperinsulinemic-euglycemic clamp that included skeletal muscle biopsies. All volunteers underwent comprehensive medical screening prior to participation. Baseline subject characteristics are presented in Table 1. The study was approved by the Institutional Review Board of the University of Illinois-Chicago, and all subjects provided oral and written informed consent.

Table 1.

Baseline characteristics

	Control	T2DM
Sex (F/M)	6/4	3/5
Age, yr	26.8 ± 1.0	59.6 ± 3.8*
Weight, kg	64.3 ± 3.9	108.4 ± 7.7*
Height, cm	169.6 ± 2.8	176.0 ± 4.1
BSA, m2	1.7 ± 0.1	2.2 ± 0.1*
BMI, kg/m2	22.3 ± 0.8	35.1 ± 2.2*
FM, kg	15.5 ± 1.5	45.4 ± 5.6*
FFM, kg	48.8 ± 3.1	63.0 ± 3.7*
Fasting glucose, mg/dl	81.5 ± 6.1	199.0 ± 37.2*
Fasting insulin, μU/ml	5.3 ± 1.0	8.0 ± 1.0*
HbA1c, %	5.26 ± 0.1	7.4 ± 0.7*

T2DM, Type 2 diabetes mellitus; BSA, bovine serum albumin; BMI, body mass index; FM, fat mass; FFM, fat-free mass.

^*P < 0.05.

Clinical assessments.

An oral glucose tolerance test was used to screen healthy subjects for abnormal glucose tolerance and confirm diagnosis of T2D for each of the respective study groups. Following an overnight fast of ∼10–12 h, a polyethylene catheter was inserted into an antecubital vein for sampling of venous blood. The subject then consumed a solution containing exactly 75 g of anhydrous glucose within 5 min. Blood samples were then collected every 30 min for 120 min for the determination of glucose and insulin. The World Health Organization and the American Diabetes Association criteria were used to determine the presence or absence or abnormal glucose tolerance and Type 2 diabetes (2). Aside from measures of body weight and height to calculate body mass index (BMI) and body surface area, body composition was measured by dual-energy X-ray absorptiometry (model iDXA; Lunar, Madison, WI) and was used to determine whole body fat mass.

Metabolic control period.

For 3 days prior to the metabolic testing day, subjects were counseled to eat a balanced diet that contained at least 200 g of carbohydrate per day to stabilize muscle and liver glycogen stores (67). During these days, subjects were asked to record the time, content, and volume of foods and beverages consumed using daily food logs. Assessments of peripheral insulin sensitivity and skeletal muscle biopsy were performed in the University of Illinois at Chicago Clinical Research Center (CRC) following the 3-day diet stabilization period. The evening prior to the clamp procedure, participants arrived at ∼6 PM to the CRC and were provided a metabolic meal [55% carbohydrate (≥ 250 g), 35% fat, 10% protein] in the metabolic kitchen. After meal consumption, the subjects were fasted overnight for ∼ 10–12 h. In addition, subjects were asked to refrain from physical activity outside of their normal activities of daily living for 24 h prior to metabolic testing. Subjects were asked to document their physical activity during the 3-day metabolic control period using a physical activity log. Subjects were also asked to refrain from consuming foods and beverages that contain caffeine for a period of 12 h prior to metabolic testing and from consuming alcohol 48 h prior to testing. This approach has been successfully used previously in studies of insulin sensitivity and metabolism to control for the influence of diet and physical activity (29, 30, 34).

Hyperinsulinemic-euglycemic clamp procedure.

On the morning of the clamp experiment, subjects arrived at the CRC at ∼5:30 AM and then rested quietly supine in a darkened room for ∼30 min to return the body to a resting state. Following the 10–12-h overnight fast described above, a polyethylene catheter was inserted into an antecubital vein for infusion of insulin and glucose. A second catheter was inserted retrograde into a dorsal hand vein, and the hand was warmed in a heated box for sampling of arterialized venous blood. Once the catheters were placed, baseline blood samples were obtained for determination of glucose (glucose-lactate analyzer; Yellow Springs Instruments, Yellow Springs, OH), insulin (Roche e411 automated platform), percentage of glycosylated hemoglobin; HbA1c (Beckman automated platform), and cortisol (Caymen no. 500360). Once baseline samples were obtained, a primed-continuous infusion (40 mU·m⁻²·min⁻¹) of human insulin (Humulin-R, Eli Lilly, Indianapolis, IN) was initiated and maintained for a period of 120 min. Glucose levels were clamped at 5.0 mM (90 mg/dl) by use of a variable glucose infusion (20% dextrose). All infusions were administered via Harvard PhD 2000 precision infusion pumps (Harvard Apparatus, Holliston, MA). Blood glucose was measured every 5 min via YSI glucose-lactate analyzer, and blood samples were obtained at 120 min for determination of the analytes described above. Glucose infusion rates were calculated according to the method of DeFronzo et al. (12), and glucose disposal rates (mg/kg/min) were subsequently calculated after space correction (67).

Skeletal muscle biopsy and processing.

Two skeletal muscle punch biopsies were obtained from each subject. Tissue was obtained from the vastus lateralis muscle following local anesthetic (lidocaine HCl 1%) using a 5-mm Bergstrom needle (6) with suction during the baseline and insulin-stimulated period (120 min) of the hyperinsulinemic-euglycemic clamp procedure described above. Biopsies were performed on contralateral legs. Muscle tissue was quickly trimmed of excess connective tissue and fat, blotted with gauze to remove blood, immediately flash frozen in liquid nitrogen, and subsequently stored at −80°C until future analysis. Approximately 10 mg of skeletal muscle tissue was weighed at −20°C and then homogenized by ceramic beads (FastPrep Lysing Matrix D; MP Bio, Cleveland, OH) in 20 volumes of homogenization buffer (no. 9803; Cell Signaling, Danvers, MA) supplemented with protease/phosphatase inhibitor cocktail (MSSAFE; Sigma, St. Louis, MO). Total protein concentration was determined by BCA assay (Pierce, Rockford, IL). Equal volumes of 2× sodium dodecyl sulfate loading buffer was added to the samples and boiled for 5 min; then Western blot analysis was performed.

Western blot analysis.

Per our previously published methods (17, 19, 79), whole muscle protein (35 μg) was resolved using SDS-PAGE, and then transferred onto PVDF membrane (Bio-Rad Protean). After blocking in 5% milk in TBS plus 0.1% Tween-20 (TBS-T) for 1 h at room temperature, membranes were incubated with primary antibody [REDD1 (no. 10638-1) from Proteintech, phospho-Akt S473 (no. 4060), S6K1 T389 (no. 9205), 4E-BP1 T37/46 (no. 9459), MEK1/2 S217/221 (no. 9154), ERK1/2 T202/Y204 (no. 9101), and GAPDH (no. 2118)] from Cell Signaling Technology for 16 h at 4°C in TBS-T. Membranes were washed and incubated with a mouse or rabbit horseradish peroxidase (HRP)-containing secondary antibody (Cell Signaling Technology) for 1 h in a 5% milk/TBS-T solution at room temperature. Then membranes were washed in TBS-T and prepared for imaging. Protein immunoblot images were visualized following the addition of ECL reagent and captured (Bio-Rad ChemiDoc MP Imager). Density measurements for the images were quantified (Bio-Rad Image Lab) and were normalized to GAPDH. Each sample was then normalized to the control group, for the respective blot, and expressed as a mean percentage of the control group between blots.

Statistical analysis.

Statistics were performed using IBM SPSS version 22.0 software. A one-way ANOVA with a least significant difference post hoc test (Control and T2D groups under basal and insulin conditions), two-tailed t-test (control vs. T2D basal-to-insulin percent delta), or a Pearson correlation comparing all groups and conditions, (−) insulin only (both Control and T2D groups), or (+) insulin only (both Control and T2D groups) was used to determine significance between groups. The significance level was set with a priori at P < 0.05. The results are expressed as the means ± SE.

RESULTS

Basal.

Expectedly, the T2D subjects weighed more; had higher BMI and fat mass; and presented with elevated fasted blood glucose, insulin (P = 0.06), and HbA1c compared with the Control subjects (Table 1). Basal activation of the major insulin-sensitive pathway, Akt, was similar (Fig. 1) between groups, whereas ERK1/2 was significantly higher (Fig. 2B; P < 0.05) in skeletal muscle from the T2D−Insulin group vs. the Control−Insulin group. Fasted circulating concentrations of the negative regulator of insulin action, cortisol, were significantly higher (Fig. 3, P < 0.05) in the T2D−Insulin group vs. the Control−Insulin group. Cortisol also positively correlated with body mass index (r = 0.88; P < 0.05) and fat mass (r = 0.84; P < 0.05). Although not significant, REDD1 protein expression was slightly higher in the T2D-Insulin skeletal muscle (Fig. 4). Basal phosphorylation of the mTORC1 substrate, S6K1, was also significantly higher (Fig. 5A; P < 0.05) in T2D−Insulin skeletal muscle compared with the Control−Insulin group. Accordingly, there was a significant relationship (Fig. 6A; P < 0.05) observed between the basal REDD1 expression and S6K1 phosphorylation. Conversely, activation of the eukaryotic initiation factor 4E-binding protein 1 (4E-BP1) was significantly lower (Fig. 5B; P < 0.05 vs. Control−Insulin) in the T2D−Insulin group under basal conditions.

Fig. 1.

Limited insulin activation of Akt in Type 2 diabetes (T2D) human skeletal muscle. A vastus lateralis skeletal muscle biopsy was obtained from lean, control and obese, or T2D subjects under basal and after a 2-h hyperinsulinemic (40 mU·m⁻²·min⁻¹)-euglycemic (5 mM) clamp, then frozen in liquid nitrogen. A total homogenate of this muscle biopsy was analyzed by Western blot analysis for phospho-Akt S473 and normalized to GADPH relative to control and percent change (Δ denotes percent change or delta) of basal-to-insulin phospho-Akt S473 for the respective group. Representative blots are shown. *P < 0.05 vs. Control group.

Fig. 2.

Aberrant MEK-ERK signaling under basal and insulin-stimulated conditions in T2D human skeletal muscle. A vastus lateralis skeletal muscle biopsy was obtained from lean, control and obese, T2D subjects under basal and after a 2-h hyperinsulinemic (40 mU·m⁻²·min⁻¹)-euglycemic (5 mM) clamp, then frozen in liquid nitrogen. A total homogenate of this muscle biopsy was analyzed by Western blot analysis for phospho-MEK1/2 S217/221 (A) and ERK1/2 Y202/204 (B), and was normalized to GADPH relative to Control group and percent change (Δ denotes percent change or delta) of basal-to-insulin phospho-MEK1/2 S217/221 or phospho-ERK1/2 Y202/204 for the respective group. Representative blots are shown. *P < 0.05 vs. Control.

Fig. 3.

Elevated blood cortisol in T2D humans. A blood sample was obtained from lean, control and obese, T2D subjects under basal and after a 2-h hyperinsulinemic (40 mU·m⁻¹·min⁻¹)-euglycemic (5 mM) clamp. The blood was analyzed by ELISA (Caymen cat. no. 500360) for cortisol concentrations. *P < 0.05 vs. Control.

Fig. 4.

Aberrant REDD1 expression in T2D human skeletal muscle. A vastus lateralis skeletal muscle biopsy was obtained from lean, control and obese, T2D subjects under basal and after a 2-h hyperinsulinemic (40 mU·m⁻²·min⁻¹)-euglycemic (5 mM) clamp, then frozen in liquid nitrogen. A total homogenate of this muscle biopsy was analyzed by Western analysis for total REDD1 protein expression and normalized to GADPH relative to Control and percent change (Δ denotes percent change or delta) of basal-to-insulin REDD1 for the respective group. Representative blots are shown. *P < 0.05 vs. Control.

Fig. 5.

Aberrant mTORC1 signaling under basal and insulin-stimulated conditions in T2D human skeletal muscle. A vastus lateralis skeletal muscle biopsy was obtained from lean, control and obese, T2D subjects under basal and after a 2-h hyperinsulinemic (40 mU·m⁻²·min⁻¹)-euglycemic (5 mM) clamp, then frozen in liquid nitrogen. A total homogenate of this muscle biopsy was analyzed by Western blot analysis for phospho-S6K1 T389 (A) and 4E-BP1 T37/46 (B) and normalized to GADPH relative to Control and percent change (Δ denotes percent change or delta) of basal-to-insulin phospho-S6K1 T389 and phospho-4E-BP1 T37/46 for the respective group. Representative blots are shown. *P < 0.05 vs. Control.

Fig. 6.

Relationship of basal skeletal muscle REDD1 expression with S6K1 phosphorylation; r = 0.78; P = 0.002 (A), postclamp circulating cortisol with skeletal muscle REDD1 expression; r = 0.68; P = 0.005 (B), and postclamp circulating cortisol with skeletal muscle S6K1 phosphorylation; r = −0.67; P = 0.0018 (C).

Insulin.

Despite slightly higher (P < 0.05) circulating insulin concentrations during the clamp [87.5 ± 6.5 vs. 68.0 ± 3.7 (μU/ml); T2D vs. Controls, respectively], glucose disposal rates (GDR) were significantly lower (P < 0.05) in the T2D vs. the Control group [4.5 ± 0.6 vs. 9.3 ± 0.6 (mg·kg⁻¹·min⁻¹), respectively]. Insulin infusion did promote a significant increase in Akt (Fig. 1; P < 0.05) in the Control+Insulin vs. the Control−Insulin group, although it remained unchanged in the T2D+Insulin muscle. This resulted in a significant (P < 0.05) basal-to-insulin stimulated percent delta increase in muscle Akt (Fig. 1) phosphorylation for the Control group compared with a significant decrease in the T2D group. Likewise, insulin infusion stimulated a significant increase in MEK1/2 and ERK1/2 phosphorylation (Fig. 2, A and B; P < 0.05) in the Control+Insulin vs. the Control−Insulin group, although it remained unchanged in the T2D+Insulin muscle. Correspondingly, there was a significant (P < 0.05) basal-to-insulin stimulated percent delta increase in skeletal muscle MEK1/2 and ERK1/2 phosphorylation for the Control group vs. no change in the T2D group (Fig. 2, A and B). REDD1 protein expression did not change in the Control+Insulin vs. the Control−Insulin group, but trended higher in the T2D+Insulin group (Fig. 4; P = 0.08 vs. Control). Consistent with the lack of insulin-stimulated mTOR activation in the T2D muscle, a significant increase (P < 0.05) in the basal-to-insulin stimulated percent delta in REDD1 expression was observed in the T2D group compared with the Control group (Fig. 4). Correspondingly, S6K1 phosphorylation (Fig. 5A; P < 0.05) was higher in the Control+Insulin vs. the Control−Insulin group, although it remained unchanged in the T2D+Insulin muscle. This resulted in a significant (P < 0.05) basal-to-insulin-stimulated percent delta increase in muscle S6K1 (Fig. 5A) phosphorylation for the Control group compared with a decrease in the T2D group. 4E-BP1 phosphorylation remained low in the T2D group even after the insulin infusion (Fig. 5B). Again, supportive of the REDD1 data, circulating cortisol concentrations remained elevated through to the end of the insulin clamp in the T2D groups (Fig. 3; P < 0.05) compared with the Control groups. Significant (P < 0.05) correlations were observed between circulating cortisol and REDD1 expression (Fig. 6B; P < 0.05), phosphorylated S6K1 (Fig. 6C; P < 0.05), and GDR (r = −0.48; P < 0.05).

DISCUSSION

The objective of this study was to establish whether alterations in the REDD1-mTOR axis underlie skeletal muscle insensitivity to insulin in Type 2 diabetic, obese individuals. Here, we report that elevations in REDD1 protein expression, coupled with attenuated S6K1 activation, which was associated with high circulating concentrations of the REDD1 agonist cortisol, were observed in insulin-stimulated human skeletal muscle from obese, T2D vs. healthy participants. In addition, significant attenuation of insulin-stimulated Akt and ERK1/2 signaling was observed in the T2D group (see Fig. 7 for pathway). To our knowledge, these are the first human data to show elevated blood cortisol and skeletal muscle REDD1 expression in relation to repressed insulin-stimulated mTOR signaling. We acknowledge that our data were in younger, lean healthy subjects compared with older obese, T2D individuals because we believe this represents the optimal comparison to demonstrate proof-of-principal. Comparison to the lean, healthy group here represents an implied state of good health, and allows for a benchmark of “normal” signaling.

Fig. 7.

Elevated cortisol and REDD1 may contribute to a lack of insulin-mediated activation of Akt, ERK1/2, mTORC1 signaling in obese/insulin resistant (IR), Type 2 diabetic (T2D) skeletal muscle.

Glucose uptake into tissues upon insulin binding to the insulin receptor, GLUT4 translocation to the plasma membrane is promoted through IRS-1-mediated PI3-kinase activation of Akt, and subsequently the Akt substrate of 160 kDa (AS160) and ubiquitin-specific peptide 6 (Trc-2)/budding uninhibited by benzimidazole/cell division cycle (TBC) domain family member 1, TBC1D1. The current findings show that the GDR, during the hyperinsulinemic-euglycemic clamp, was lower in the T2D group vs. the control group. The T2D group had slightly higher circulating insulin concentrations vs. the Control group during the clamp, which would not limit GDR. On the contrary, this further supports the insulin-resistant state of the T2D skeletal muscle.

An explanation for the observed reduction in insulin-mediated skeletal muscle Akt and/or S6K1 phosphorylation in the T2D group may be due to the increased change in REDD1 expression by the end of the clamp. This draws support from a previous report that observed an elevation in the expression of REDD1 from liver tissue of morbidly obese humans (22). Glucocorticoids are elevated in the obese under both fasted and fed conditions (3, 49, 80). Wang et al. (75) were the first to show that treatment of rodents with glucocorticoids (i.e., dexamethasone) promoted an increase in skeletal muscle REDD1 expression and subsequent inhibition of mTOR. Consistent with these data and the increase in skeletal muscle REDD1 expression in the T2D subjects, circulating concentrations of the glucocorticoid cortisol was also significantly elevated under basal and insulin-stimulated conditions. Moreover, elevated cortisol concentrations in the blood can promote REDD1 expression in skeletal muscle, which can be inhibited by the glucocorticoid receptor antagonist, RU486 (38). Endogenous (33, 39, 51) or exogenous (24) approaches to promote REDD1 expression results in skeletal muscle atrophy. REDD1 requires a release of 14-3-3 proteins from TSC2 (7, 15, 66) to repress mTOR, though the exact mechanism of action on mTOR remains unclear. Recent findings (13) suggest that REDD1 promotes PP2A-dependent Akt T308 dephosphorylation and reduced TSC2 phosphorylation, which reduces Rheb GTPase loading and mTORC1 activation.

We observed a positive relationship of REDD1 expression with S6K1 phosphorylation under fasted, basal conditions, and REDD1 expression was further elevated (P = 0.08) by the end of the clamp. Longer treatment or higher concentrations of insulin or insulin-like growth factor 1 promotes REDD1 expression (25, 63) through a PI3K-regulated pathways. The experimental hyperinsulinemia achieved via the clamp procedure has been shown to reasonably reflect postprandial insulin concentrations and reflects the compensatory hyperinsulinemia experienced with insulin resistance and impaired glucose tolerance (12). However, in models in which insulin is low (e.g., Type 1 diabetes or fasting) vs. high (e.g., insulin treatment of Type 1 diabetic or feeding), REDD1 is conversely expressed (47). This coupled with the rapid half-life of REDD1 protein [∼5–10 min; (35)], support the possibility that REDD1 expression could increase over the experimental timeframe. An increased change in REDD1 expression following insulin infusion may serve to either limit an already elevated mTORC1 or act to induce insulin resistance.

Insulin also regulates cellular protein synthesis and adaptation through the rapamycin-sensitive, mTOR complex 1 (mTORC1) containing the regulatory associated protein of mTOR (raptor) by promoting the activation of the Akt and ERK1/2 pathways. Akt and ERK1/2 phosphorylate TSC2 and promote an increase in the GTPase activator protein complex with TSC1:TSC2 and TBC1D7. Consequently, the active form of the TSC1:TSC2 complex is inhibited during treatment with growth-promoting stimuli. Now, in a GTP-bound form, Rheb promotes mTOR activity and subsequent mTOR phosphorylation of the substrates, S6K1 and 4E-BP1, to promote translation initiation. Increased glucose uptake and glycolytic flux can be regulated by hypoxia-inducible factor-1 through an mTORC1 (rapamycin)-sensitive mechanism (4). A handful of studies suggest that much like amino acids that are sensed by the Rag-GTPases to direct mTOR to lysosomes, glucose can also be sensed by Rag-GTPases through lysosomal v-ATPases (21).

Consistent with the current finding that skeletal muscle from obese, T2D humans exhibit hyperactive mTORC1 signaling, our laboratory (18, 19) and others (53, 64, 72, 73) have reported similar findings in skeletal muscle from obese or insulin-resistant mouse models. Similarly, findings have been reported in adipocytes from lean vs. obese individuals (55). However, to our knowledge, there are no other data in obese human skeletal muscle to suggest that mTOR is hyperactive during the basal state. Recent data from Elena Volpi's laboratory (46) reported that reductions in basal protein synthesis are related to basal mTORC1 hyperactivation in nonobese, older individuals. Therefore, the current data may suggest that advancing age drives insulin resistance and aberrant REDD1-mTOR more so rather than obesity/T2D. However, given the strong preclinical support, obesity cannot be discounted as a contributor to the REDD1-mTOR responses and needs to be further examined in human models. Although the contribution of age vs. obesity vs. T2D remains fundamental to understanding mechanisms of skeletal muscle insulin resistance in the current study, to directly address this would require studies on a larger scale that are beyond the scope of the current study.

Obese individuals undergoing a hyperinsulinemic-hyperaminoacidemic-hyperglycemic clamp demonstrated normal protein synthesis and less suppression of breakdown compared with lean subjects, which was associated with lower IRS-1, Akt, and S6K1 phosphorylation (9). It has been proposed that hyperactivation of mTORC1 through S6K1 negatively feeds back to IRS-1, downregulating insulin signaling (73). However, recent findings (31) suggest that S6K1 may not promote serine phosphorylation on IRS-1. Phosphorylation of the mTORC1 substrate and translational repressor, 4E-BP1, were lower in the T2D group independent of treatment, which would suggest that mRNA translation initiation was inhibited through greater binding to the mRNA cap protein, eIF4E (5, 36). Interestingly, hyperactive S6K1 phosphorylation has also been observed in mice that lack 4E-BP1, which also exhibit an increased sensitivity to a high-fat diet (41). Reduced translation initiation, stemming from dephosphorylated 4E-BP1, reduces protein synthesis (36). Findings ranging from reduced to no difference in skeletal muscle protein synthesis have been reported in obese (9, 11, 43) and T2D (59, 71) individuals compared with lean individuals. In most cases, individuals diagnosed with T2D are slightly older, which can exacerbate skeletal muscle protein synthesis (10, 52) and mass (70). The current findings support the differential response of mTORC1 substrates to insulin resistance and obesity and may serve as points of regulation and intervention during insulin resistance.

The consequence of hyperactive ERK1/2 is less apparent in skeletal muscle than hyperactive S6K1. However, previous findings from Williamson et al. (77) showing that skeletal muscle from older men exhibits significantly higher ERK1/2 phosphorylation under basal conditions vs. young men. Tumor necrosis factor-alpha (TNF-α), which is known to be elevated in aged and obese in blood and skeletal muscle (28, 37), may be a means of ERK1/2 dysregulation. Temporal upregulation of ERK1/2 by TNF-α contributes to the inhibition of insulin-stimulated protein synthesis in muscle cultures (78). Consistent with the elevated activation of ERK1/2 and S6K1 in the T2D muscle, ERK1/2 upregulation can repress TSC2 by phosphorylated S644, inactivating the TSC complex, and the autoinhibitory domain (T421/S424) on S6K1, promoting mTORC1 signaling activation (40). Also, despite a similar lack of insulin response for both MEK1/2 and ERK1/2 in the T2D group, we did observe a differentially low and high basal phosphorylation of MEK1/2 and ERK1/2, respectively. MEK1:MEK2 heterodimers are subject to negative feedback by ERK phosphorylation of MEK1 on T292, dephosphorylating S217/S221 on MEK1. Because of a lack of MEK1 or an inability to bind to MEK2, negative feedback is disabled and ERK activation can be prolonged (8). Lastly, constitutively active ERK signaling limits REDD1 degradation (62), which may provide a mechanism by which obese/T2D subjects limit insulin action. Thus, the MEK-ERK pathway may represent an underappreciated means of regulation during insulin resistance and may require additional exploration in the future.

Perspectives and Significance

This work proposes a mechanism in Type 2 diabetic, obese human skeletal muscle that may contribute to insulin resistance. An elevation in REDD1 protein expression in Type 2 diabetic, obese muscle is associated with reduced insulin-mediated glucose disposal and reduced activation of signaling that regulates cellular glucose uptake, lipid metabolism, and growth (see Fig. 7 for pathway). Targeting REDD1 may provide a novel means to regulate glucose homeostasis and improve insulin action in insulin-resistant individuals, as recent findings show that reducing cortisol or glucocorticoid receptor function can improve insulin sensitivity in Type 2 diabetic humans (45).

GRANTS

Support for this study was provided by a CTSA Award ULRR029879 (to J. M. Haus) and an American Diabetes Association Junior Faculty Award (to J. M. Haus).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Author contributions: D.L.W., A.M.M., J.T.M., B.K.B., and J.M.H. conception and design of research; D.L.W., C.M.D., A.M.M., J.T.M., B.K.B., and J.M.H. analyzed data; D.L.W., C.M.D., A.M.M., J.T.M., B.K.B., and J.M.H. interpreted results of experiments; D.L.W., C.M.D., and J.M.H. prepared figures; D.L.W. and J.M.H. drafted manuscript; D.L.W., C.M.D., and J.M.H. edited and revised manuscript; D.L.W., C.M.D., A.M.M., J.T.M., B.K.B., and J.M.H. approved final version of manuscript; A.M.M., J.T.M., B.K.B., and J.M.H. performed experiments.