Endothelial MRs Mediate Western Diet–Induced Lipid Disorders and Skeletal Muscle Insulin Resistance in Females

Abstract

Consumption of a Western diet (WD) consisting of excess fat and carbohydrates activates the renin–angiotensin–aldosterone system, which has emerged as an important risk factor for systemic and tissue insulin resistance. We recently discovered that activated mineralocorticoid receptors (MRs) in diet-induced obesity induce CD36 expression, increase ectopic lipid accumulation, and result in systemic and tissue insulin resistance. Here, we have further investigated whether endothelial cell (EC)–specific MR (ECMR) activation participates in WD-induced ectopic skeletal muscle lipid accumulation, insulin resistance, and dysfunction. Six-week-old female ECMR knockout (ECMR^−/−) and wild-type (ECMR^+/+) mice were fed either a WD or a chow diet for 16 weeks. ECMR^−/− mice were found to have decreased WD-induced in vivo glucose intolerance and insulin resistance at 16 weeks. Improved insulin sensitivity was accompanied by increased glucose transporter type 4 expression in conjunction with improved soleus insulin metabolic signaling in phosphoinositide 3-kinases/protein kinase B and endothelial nitric oxide synthase activation. Additionally, ECMR^−/− also blunted WD-induced increases in CD36 expression and associated elevations in soleus free fatty acid, total intramyocellular lipid content, oxidative stress, and soleus fibrosis. Moreover, in vitro and in vivo activation of ECMR increased EC-derived exosomal CD36 that was further taken up by skeletal muscle cells, leading to increased skeletal muscle CD36 levels. These findings indicate that in the context of an obesogenic WD, enhanced ECMR signaling increases EC-derived exosomal CD36 resulting in increased uptake and elevated concentrations of CD36 in skeletal muscle cells, contributing to increased lipid metabolic disorders and soleus insulin resistance.

Mineralocorticoid receptors (MRs), the primary receptors for the hormone aldosterone, play a key role in plasma volume, electrolyte homeostasis, and blood pressure (1). Recent data indicate that MRs also exist in tissues outside the kidney, including pancreatic islets, adipose, skeletal muscle, liver, and cardiovascular tissue (1). Inappropriate tissue-specific MR activation engages in the pathophysiology of insulin resistance (1, 2). Our recent studies have shown that consumption of a diet high in fat and refined sugars, a Western diet (WD), increases plasma aldosterone levels (3) and activates MRs (2, 4) to directly impair insulin metabolic signaling in protein kinase B (Akt) phosphorylation/activation. Conversely, inhibition of MRs with spironolactone prevents WD-induced systemic, liver, and skeletal muscle insulin resistance (2, 4).

Our recent data indicates CD36 mediates the activated MR-induced excessive free fatty acid (FFA) uptake and ectopic lipid accumulation associated with systemic and tissue-specific insulin resistance (2, 4). Related to this, CD36 is a fatty acid translocase that enhances cellular FFA uptake and corresponding development and progression of lipid disorders, atherosclerosis, and metabolic syndrome (5, 6). Recent studies have indicated that CD36 in endothelial cells (ECs) is very important due to EC CD36 acting as a gatekeeper for parenchymal cell (skeletal muscle cell) FFA uptake and thus directly affects downstream actions on glucose utilization and insulin action in skeletal muscle (7, 8). Importantly, CD36 also is present in exosomes, which shuttle CD36 between neighboring cells, and have been implicated as mediators in diet-induced lipid metabolic disorders and metabolic syndrome (9, 10). However, the impact of endothelial-specific MR (ECMR) signaling and EC exosomes on skeletal muscle CD36, lipid metabolic disorders, insulin resistance, and underlying molecular mechanisms remains unexplored. Because soleus muscle is composed of 75% slow-twitch muscle fibers, which have a higher content of mitochondria, FFA uptake, and mitochondrial FFA oxidation (11), we investigated whether suppression of ECMR (ECMR^−/−) would prevent overnutrition-induced increases in EC-derived exosomal CD36, skeletal muscle CD36, and associated soleus lipid metabolic disorders and insulin resistance.

Materials and Methods

Experimental Animals

A genomic region encompassing exons 5 and 6 of the MR gene, encoding the hinge region and the N-terminal part of the ligand binding domain, was flanked by loxP sites via homologous recombination in embryonic stem cells, and floxed MR mice (MR^f/f) were generated as previously described (12). ECMR^−/− mice were generated by crossing MR^f/f mice with VE-Cad-Cre⁺ mice as previously described (13). MR^f/f Cad-Cre⁻ littermates (ECMR^+/+) were used as controls. When mice were 6 weeks of age, ECMR^−/− and ECMR^+/+ female mice were fed a WD consisting of a high-fat (46%) and a high-carbohydrate component comprising sucrose (17.5%) and high-fructose corn syrup (17.5%) for 16 weeks. Parallel groups of age-matched females were fed regular mouse chow diet (CD) for the same period. All procedures (#9945) were approved in advance by the Institutional Animal Care and Use Committee of the University of Missouri and mice were cared for according to NIH guidelines.

Structural and Biochemical Profiles and Glucose Tolerance Testing

Mice were interrogated via a noninvasive procedure with a nuclear magnetic resonance imaging whole-body composition analyzer (EchoMRI 4in1/1100; Echo Medical Systems, Houston, TX) at the end of 16 weeks of feeding. The percent body fat, including whole body fat mass, lean mass, and total body water, were investigated as previously described (2). At euthanasia, following a 5-hour fast, fasting blood samples were collected, and plasma chemistry tests, including cholesterol, triglyceride, glucose, and insulin, were determined in the University of Missouri Small Animal Veterinary Clinic. FFA levels were determined by an assay kit (#MAK044) from Sigma-Aldrich (St. Louis, MO). The enzyme-linked immunosorbent assay kits were utilized to determine leptin C (RRID:AB_2922952) (14) and adiponectin (RRID:AB_2922953) (15) levels (#KMC2281, #KMP0041, Thermo Fisher Scientific) according to the manufacturer's instructions. Intraperitoneal glucose tolerance tests (IPGTTs) were performed as previously described. Briefly, dextrose (1.5 g/kg) was injected intraperitoneally following a 5-hour fast and the glucose excursion was evaluated over time and compared between treatment groups. Blood glucose levels were examined by glucose meter (AlphaTRACK, Abbott, IL,) measurements at time 0 and 15, 30, 45, 60, and 120 minutes following dextrose injection. Homeostatic Model for Insulin Resistance (HOMA-IR) = (fasting plasma glucose × fasting insulin levels)/22.5 were also determined (16). The glucose areas under the curve (AUC) were calculated as the glycemic index as previously described (2).

In Vitro Cell Coculture, Exosome Isolation, and Exosome Uptake Assay

Skeletal muscle microvascular ECs isolated from C57BL/6 mice (#C57-6220) and C2C12 (#91031101) were purchased from Cell Biologics (Chicago, IL) and Sigma (St Louis, MO), respectively. In vitro ECs were cultured with exosome-free media and treated with 100 nmol aldosterone with or without 10 µM spironolactone for 24 hours. Plasma EC-derived exosomes were initially isolated with a total plasma exsome isolation kit (#4484450, Thermo Fisher Scientific) and then further isolated by using Dynabeads CD31 (#11155D, Invitrogen), which is a specific marker for ECs (13, 17). In vitro exosomes from ECs were isolated by using a total exosome isolation reagent (#4478359, Invitrogen, Waltham, MA). In vivo and in vitro cellular uptake of exosomes were determined by a PKH67 green fluorescent cell linker kit (MIDI67-1KT, Sigma Aldrich, St Louis, MO) as previously described (18, 19). Briefly, plasma EC exosomes were collected and labeled with PKH67 green fluorescent dye, which were further cocultured in C2C12 cells or were injected into mice via the tail vein for 24 hours. Plasma EC exosome uptake in C2C12 and ex vivo soleus was observed under fluorescence microscopy (18, 19). For exchange assay of EC-derived exosomal CD36 exchange between ECs and C2C12, in vitro ECs were treated with 100 nmol aldosterone with or without 20 µM GW4869 for 24 hours, which is a neutral sphingomyelinase inhibitor and is widely used for blocking exosome generation and release (20). EC exosomes were then isolated and subsequently cocultured in C2C12 for 24 hours (21). CD36 levels in recipient C2C12 were determined by Western blot.

Western Blot

Soleus, EC-derived exosomes, in vitro C2C12 were lysed in lysis buffer containing protease inhibitors, and lysates measured with bicinchoninic acid protein assay kit (Thermo Scientific, Wilmington, DE) as previously described (22, 23). Proteins (10 µg) were loaded and separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes (22, 23). Blots were incubated overnight at 4 °C with primary antibodies against glucose transporter type 1 (Glut1) (RRID:AB_2687899) (24), Glut4 (RRID:AB_823508) (25), phospho (p)-p85-phosphoinositide 3-kinase (PI3K) (RRID:AB_659940) (26)/p85-PI3K (RRID:AB_659889) (27), p-S⁴⁷³ Akt (RRID:AB_329825) (28)/Akt (RRID:AB_329827) (29), CD36 (RRID:AB_2798458) (30), and glyceraldehyde 3-phosphate dehydrogenase (RRID:AB_10622025) (31) (1:1000 dilution, catalog numbers: 12939, 2213, 4228, 4257, 9271, 9272, 14347, 5174s, Cell Signaling Technology, Danvers, MA), Glut3 (RRID:AB_10979336) (32) (1:1000 dilution, #OSG00012W, Invitrogen, Waltham, MA), p-S¹¹⁷⁷endothelial nitric oxide synthase (p-S¹¹⁷⁷eNOS) (RRID:AB_399751) (33)/eNOS (RRID:AB_397691) (34) (1:1000 dilution, #BDB610297, BDB612393, BD Biosciences, San Jose, CA), and CD63 (RRID:AB_627877) (35) (1:1000 dilution, SC-5275, Santa Cruz, Dallas, TX). After rinsing, blots were incubated with secondary antibodies (1:5000 dilution of each antibody) for 1 hour at room temperature. Bands were visualized by chemiluminescence, and images were recorded using a Bio-Rad ChemiDoc XRS image analysis system.

Immunohistochemistry and Transmission Electron Microscopy

Soleus samples were fixed in 3% paraformaldehyde, dehydrated in an ethanol series, paraffin embedded, and transversely sectioned in 5-μm slices (2, 4). Four sections were examined from 5 mice in each group. Sections were dewaxed, rehydrated, and placed in 95 °C citrate buffer for 25 minutes for antigen retrieval. Nonspecific binding cites were blocked with 5% bovine serum albumin and 5% donkey serum. Next, sections were incubated with antibodies to Glut4 (RRID:AB_823508) (25) (1:100 dilution, #2213, Cell Signaling Technology, Danvers, MA) and 3 nitrotyrosine (RRID:AB_310089) (36) (1:200 dilution, #06-284, Millipore, Billerica, MA) (2, 4). Oil Red O staining was performed on frozen soleus sections to detect the presence of lipid accumulation as previously described (2, 4). To evaluate fibrosis sections were stained with picrosirius red for the determination of collagen accumulation. The location and intensity of red color on the images was interpreted as being indicative of collagen deposition (2, 4). The location and intensity for Glut4, 3 nitrotyrosine, lipid accumulation, and fibrosis were assessed under a biphoton confocal microscope (Leica) and quantified as “gray scale intensities” by using MetaVue as described before (2, 4). In brief, 5 randomly selected areas in each slide were captured with a Nikon50i microscope and each image was auto-leveled with Photoshop. Signal intensities were normalized to area and quantified using MetaVue software (Molecular Devices, Sunnyvale, CA). Average values for each group were based on measurements made on 5 randomly selected areas. Data are presented as average gray scale intensities (2, 4).

For electron microscopy imaging, in vitro ECs, isolated exosomes, and soleus samples were fixed in 2% paraformaldehyde and 2% glutaraldehyde in 100 mM sodium cacodylate buffer as previously described (2, 22, 23). These samples were poststained using Sato's triple lead solution stain and 5% aqueous uranyl acetate. A JOEL 1400-EX transmission electron microscope (Joel Ltd. Tokyo, Japan) was utilized to capture images and analyze.

Voluntary Wheel Running

Mice were individually housed in wheel running cages with a wheel diameter of 24.2 cm (Tecniplast, West Chester, PA) as previously described (2). Running wheels were connected to a Sigma BC509 cycling computer (B003BC9PJ6; Sigma Sport) to determine running distance over 5 consecutive days. Data reported were only that which was recorded over the last 24 hours in order to avoid the variability that occurs during the first 4 days due to mice learning and becoming familiar with use of the wheel.

Statistical Analysis

Results are reported as means ± SEM. Differences in measured parameters were determined using 2-way analysis of variance multiple comparison analysis and Gabriel Students–Newman–Keuls post-test and were considered statistically significant when P < .05. All statistical analyses were performed using Sigma Plot (version 12) software (Systat Software).

Results

Baseline Metabolic and Biochemical Parameters

Sixteen weeks of WD induced increases in body weight (39% increase), fat mass (117% increase), and plasma cholesterol (58% increase). These parameters were unaltered between ECMR^+/+ and ECMR^−/− mice fed the WD (Table 1). Meanwhile, WD feeding also induced an increase of plasma leptin (153% increase), an effect that was attenuated in WD ECMR^−/− mice. In relation to this, leptin is primarily produced in white adipose tissue (37) and elevated plasma leptin levels are typically observed in obese individuals. Leptin can increase aldosterone production and MR activation, which further induces systemic and tissue insulin resistance (37). Moreover, there were no significant differences in lean mass, soleus weight, plasma triglycerides, and adiponectin levels between any of the groups (Table 1).

Table 1.

Effects of ECMR on characteristics of mice fed a Western diet

Measures	CD ECMR+/+	CD ECMR−/−	WD ECMR+/+	WD ECMR−/−
Body weight (g)	21.81 ± 0.28	22.18 ± 0.34	30.33 ± 1.23a	29.95 ± 1.85a
Fat mass (g)	6.09 ± 0.44	6.20 ± 0.49	13.22 ± 1.17a	12.83 ± 1.13a
Lean mass (g)	15.46 ± 0.42	16.42 ± 0.40	16.96 ± 0.37	16.85 ± 0.55
Soleus (mg)	10.00 ± 0.47	9.78 ± 0.68	10.56 ± 0.53	10.11 ± 0.39
Cholesterol (mg/dL)	62.13 ± 2.40	65.13 ± 2.24	98.50 ± 4.86a	93.88 ± 3.35a
Triglyceride (mg/dL)	61.75 ± 2.83	65.38 ± 2.70	67.63 ± 4.77	63.88 ± 1.77
Leptin (ng/mL)	6.26 ± 0.15	6.18 ± 0.23	15.84 ± 0.72a	9.90 ± 0.35b
Adiponectin (µg/mL)	7.85 ± 0.26	8.35 ± 0.21	7.78 ± 0.29	7.71 ± 0.33

Values are mean ± SEM. n = 8.

^a P < .01 compared with CD ECMR^+/+.

^b P < .01 compared with WD ECMR^+/+.

ECMR Mediates WD-Induced Glucose Intolerance and Insulin Resistance

As shown in Fig. 1A and 1B, CD-fed groups of ECMR^+/+ and ECMR^−/− mice presented similar fasting plasma, fasting glucose, and insulin levels. However, 16 weeks of WD induced increases in plasma insulin levels, as well as insulin resistance and glucose intolerance determined by HOMA-IR and IPGTT, respectively (Fig. 1B-1E). ECMR^−/− attenuated WD-induced insulin resistance and glucose intolerance (Fig. 1C-1E). To this point, there was not a statistically significant interaction between WD and ECMR^−/− in fasting plasma insulin (F = 2.236, P = .146) (Fig. 1B), HOMA-IR (F = 1.918, P = .77) (Fig. 1C), blood glucose levels at 15 minutes (F = 4.118, P = .052) and 30 minutes (F = 3.408, P = .075) (Fig. 1B) of IPGTT (Fig. 1D), and both WD and ECMR^−/− were the main effects. However, there was a statistically significant interaction between WD and ECMR^−/− at 45 and 60 minutes of IPGTT (Fig. 1D), as well as in the glucose AUC (Fig. 1E).

Figure 1.

ECMR^−/− attenuates WD-induced glucose intolerance and insulin resistance. Fasting plasma glucose (A) and insulin (B) levels were measured as mg/dL and ng/mL. (C) HOMA-IR was calculated based on fasting plasma glucose and insulin values. In vivo glucose tolerance was determined by IPGTT (D) and the glucose areas under the curve (AUC) were calculated as the glycemic index (E). n = 8. †P < .05 vs CD ECMR^+/+. ‡P < .05 vs WD ECMR^+/+ groups.

ECMR Signaling and WD Regulate Glut4 Expression

Expression of Glut1 and Glut3 was similar in ECMR^+/+ and ECMR^−/− mice fed a CD or WD (Fig. 2A and 2B). Meanwhile, no significant differences in Glut4 expression were detected between ECMR^+/+ and ECMR^−/− mice fed a CD (Fig. 2A-2C). However, 16 weeks of WD induced increases in Glut4 expression, which was further elevated in WD ECMR^−/− mice (Fig. 2A-2D). In this regard, there was not a statistically significant interaction between WD and ECMR^−/− in Glut4 expression. While WD was the main effect in Western blot Glut4 expression (F = 1.238, P = .279) (Fig. 2A and 2B), ECMR^−/− was the main effect in Glut4 expression examined by immunostaining (F = 0.542, P = .474) (Fig. 2C and 2D). These data indicate that both ECMR and WD regulate Glut4 expression.

Figure 2.

Expression of Gluts was determined by Western blot (A) with corresponding quantitative analysis (B) (n = 6). (C) Representative images of Glut4 expression in soleus with corresponding quantitative analysis (D). Scales bars = 50 µm (n = 4 or 5). †P < .05 vs CD ECMR^+/+. ‡P < .05 vs WD ECMR^+/+ groups.

ECMR Mediates WD-induced Impairments in Soleus Insulin Metabolic Signaling

There was a statistically significant interaction between WD and ECMR^−/− in PI3K/Akt signaling (Fig. 3). Conversely, there was not a statistically significant interaction between WD and ECMR^−/− in eNOS activity (F = 0.794, P = .383) and both WD and ECMR^−/− were the main effects (Fig. 3). Moreover, 16 weeks of WD induced inactivation of insulin metabolic PI3K/Akt signaling, which correlated with suppression of eNOS activity in soleus. However, these abnormalities in soleus insulin resistance were attenuated in WD ECMR^−/− mice (Fig. 3).

Figure 3.

ECMR^−/− attenuates WD-induced impairment of insulin metabolic signaling in the soleus. ECMR^−/− mice showed attenuation of WD-induced impairments in soleus insulin metabolic signaling in PI3K/Akt signaling and related eNOS inactivation (n = 6). †P < .05 vs CD ECMR^+/+. ‡P < .05 vs WD ECMR^+/+ groups.

ECMR Mediates WD-Induced Increases in Soleus CD36 Expression and Related Lipid Metabolic Disorders

There was not a statistically significant interaction between WD and ECMR^−/− in plasma FFA (F = 0.608, P = .442) and WD was only the main effect (Fig. 4A). However, there was a statistically significant interaction between WD and ECMR^−/− in soleus FFA (Fig. 4B), intramyocellular (IMC) lipid content (Fig. 4C), and CD36 expression (Fig. 4E). Our data further indicated that 16 weeks of WD consumption increased both plasma (Fig. 4A) and soleus FFA levels (Fig. 4B), as well as soleus IMC lipid content (Fig. 4C and 4D), and these increases were associated with increased expression of soleus CD36 protein (Fig. 4E). However, WD-induced increases in soleus FFA (Fig. 4B), total IMC lipid accumulation (Fig. 4C and 4D), and CD36 expression (Fig. 4E) were attenuated in WD ECMR^−/− mice, suggesting that enhanced ECMR signaling participates in increased soleus CD36 and related lipid metabolic disorders in diet-induced obesity.

Figure 4.

ECMR^−/− attenuates WD-induced increases in soleus CD36 expression and lipid metabolic disorders. (A, B) ECMR^−/− prevented WD-induced increases in soleus FFA concentrations without altering plasma FFA levels (n = 8). ECMR^−/− prevented WD-induced increases in soleus lipid accumulation determined by oil red O staining (C) and TEM (D), respectively. Red arrow denotes lipid droplets. Scales bars = 50 µm (n = 4 or 5). (E) 16 weeks of WD increased CD36 expression, but this effect was blunted in WD ECMR^−/− mice (n = 6). †P < .05 vs CD ECMR^+/+. ‡ P < .05 vs WD ECMR^+/+ groups.

Enhanced ECMR Signaling Increases EC-Derived Exosomal CD36 that Contributes to Increase Skeletal Muscle CD36

Exosomes and their cargoes mediate intercellular communication and regulate skeletal muscle physiology and pathophysiology (9, 10). The current study further demonstrated that 16 weeks of WD increased plasma EC-derived exosomal CD36, and these increases in EC exosomal CD36 were inhibited in ECMR^−/− WD mice (Fig. 5A-5C). Of note, CD63 is an exosome marker (9, 10) and was used an endogenous control for exosomal CD36 expression (Fig. 5C). To test our hypothesis that EC-derived exosomal CD36, regulated by ECMR activation, directly contributes to increased skeletal muscle CD36, plasma EC exosomes enriched with CD36 were collected and labeled with PKH67 green fluorescent dye and were further cocultured in C2C12 cells or were injected into mice via the tail vein for 24 hours. As shown in Fig. 5D and 5E, a PKH67 fluorescent (green) exosomal signal was clearly present within in vitro C2C12 cells and ex vivo soleus. To further observe the role of in vitro ECMR activation on EC exosomal CD36 and skeletal muscle CD36, aldosterone and spironolactone were used to activate or inactivate MR in cultured ECs, respectively. As shown in Fig. 5F and 5G, exosomes in cultured ECs and EC culture media were clearly observed by transmission electron microscopy (TEM). Meanwhile, in vitro enhanced ECMR signaling with aldosterone increased EC-derived exosomal CD36 expression, which was inhibited by additional MR antagonist (spironolactone) (Fig. 5H). It was noted that there was a statistically significant interaction between WD and ECMR^−/− in ex vivo plasma (Fig. 5C) and in vitro (Fig. 5H) EC-derived exosomal CD36. Importantly, exchange of EC-derived exosomal CD36 between ECs and C2C12 resulted in an increase in CD36 in recipient C2C12 cells which was inhibited by GW4869 treatment (Fig. 5I). Related to this, there was not a statistically significant interaction between WD and ECMR^−/− in C2C12 CD36 levels (F = 3.667, P = .08) and both WD and ECMR^−/− were main effects. These data indicate that enhanced ECMR signaling increases EC-derived exosomal CD36 which contributes to increased skeletal muscle CD36 levels.

Figure 5.

Enhanced ECMR activation increases EC-derived exosomal CD36 which contributes to increased skeletal muscle CD36. Representative TEM images of plasma EC exosomes (A) and their size distribution (B). (C) ECMR^−/− attenuated WD-induced increases in plasma EC-derived exosomal CD36. n = 5. Representative images of EC-derived exosomes uptake in C2C12 cells (D) and soleus (E). Red arrow denotes uptake of EC-derived exosomes (green) observed in cultured C2C12 cells (D) and soleus (E). Blue color is cell nuclei stained with 4′,6-diamidino-2-phenylindole (DAPI). Red color in E is the sarcolemma of soleus stained with wheat germ agglutinin. Representative TEM images of exosomes in cultured ECs (F) and the EC culture media (G). (H) In vitro inhibition of MR with spironolactone (Sp) attenuated aldosterone-induced increases of exosomal CD36 in cultured ECs. (I) EC-derived exosomes enriched CD36 contributes to an increase of CD36 in C2C12, which was reversed by additional GW4869 treatment. n = 4. †P < .05 vs control. ‡P < .05 vs aldosterone groups.

ECMR Mediates WD-Induced Oxidative Stress and Fibrosis in Soleus

There was a statistically significant interaction between WD and ECMR^−/− in ex vivo soleus oxidative stress and fibrosis (Fig. 6A-6D). Meanwhile, we observed that 16 weeks of WD feeding promoted soleus IMC lipid accumulation and insulin resistance, which were associated with soleus oxidative stress and fibrosis (Fig. 6A-6D). These abnormalities were blunted in WD ECMR^−/− mice (Fig. 6). To determine whether increased ECMR activation, excessive oxidative stress, and tissue fibrosis impacted physical activity levels, multiple measurements of physical activity in voluntary wheel running were recorded. As shown in Fig. 6E, neither WD nor ECMR signaling affected mouse running activity over a 24-hour period.

Figure 6.

ECMR^−/− attenuates WD-induced oxidative stress and fibrosis in soleus. Representative images of immunostaining for soleus 3 nitrotyrosine (brownish color) (A, B) and soleus fibrosis (red color) (C, D) with corresponding quantitative analysis (n = 4 or 5). Scales bars = 50 µm. (E) Distance covered during 24 hours by running in the wheel (n = 6). †P < .05 vs CD ECMR^+/+. ‡P < .05 vs WD ECMR^+/+ groups.

Discussion

Our previous study demonstrated that 16 weeks of WD consumption increased plasma aldosterone levels (3) and MR activation (1, 2, 4), which was associated with systemic and skeletal muscle insulin resistance. The current study further defines the potential role and mechanisms of endothelial specific ECMR in the development of lipid metabolic disorders, insulin resistance, and skeletal muscle dysfunction in diet-induced obesity. First, ECMR^−/− inhibited 16 weeks of WD-induced in vivo glucose intolerance and insulin resistance. Improved insulin sensitivity in ECMR^−/− mice was associated with increased Glut4 expression, and improved insulin metabolic signaling in PI3K/Akt and eNOS activation. Moreover, ECMR^−/− attenuated 16 weeks of WD-induced increases in soleus FFA, IMC lipid content, CD36 expression, oxidative stress, and soleus fibrosis. Furthermore, in vivo and in vitro enhanced ECMR signaling increased EC-derived exosomal CD36 which was further taken up by skeletal muscle cells, resulting in increased skeletal muscle CD36 levels. However, both WD and ECMR signaling did not impair mouse running activity. Collectively, these data suggest that endothelial specific ECMR activation increased EC-derived exosomal CD36, elevated skeletal muscle CD36 levels, promoted soleus lipid metabolic disorders, and increased skeletal muscle insulin resistance in diet-induced obesity.

It is well accepted that overnutrition or obesity is associated with inappropriate activation of the renin–angiotensin–aldosterone system, insulin resistance, and metabolic syndrome (1, 2, 4). Human studies have demonstrated that plasma aldosterone levels are positively correlated with body mass index, dyslipidemia, and indexes of insulin resistance (38). The Framingham Offspring Study also indicated that higher circulating aldosterone levels are associated with development of the metabolic syndrome and with longitudinal change of its components including plasma glucose, triglycerides, and cholesterol (39). Although aldosterone is typically synthesized in the zona glomerulosa of the adrenal cortex, aldosterone is also produced from adipose tissue in obese individuals (1). Indeed, our recent data also demonstrated that 16 weeks of WD feeding significantly increases plasma aldosterone levels to 3166 pmol/L (3) which is much higher than the physiological level (412.5 pmol/L or 15 ng/dL) (40). Aldosterone is typically synthesized in the zona glomerulosa of the adrenal cortex in response to renin–angiotensin system activation or high dietary potassium (1). Aldosterone is also produced by adipose tissue in obesity or under conditions of overnutrition (1). Elevated aldosterone further activates MRs in vascular (41), liver (4), and skeletal muscle tissue (2) with resultant increases in systemic and tissue specific insulin resistance. In this regard, vascular tissues, including both ECs and vascular smooth muscle cells, express the enzyme 11-beta hydroxysteroid dehydrogenase 2, which allows aldosterone to selectively activate MR by inactivating cortisol (42). This contrasts with skeletal muscle and adipose tissue where the main ligand for MRs is cortisol (42). Meanwhile, aldosterone and MRs can induce adipose tissue differentiation. Our recent data found that inhibition of MRs with spironolactone promotes white adipose tissue browning (4). Therefore, there is an interplay between aldosterone, MRs, and adipose tissue in diet-induced obesity. To our knowledge, this is the first report to provide evidence of endothelial specific ECMR activation in diet-induced systemic and soleus insulin resistance.

Skeletal muscle contains several glucose transporters, including Glut1, Glut3, and Glut4 (43). Among those 3 isoforms, Glut4 is the most abundant glucose transporter in skeletal muscle where its mobilization plays an integral role in the uptake of glucose (43). While Glut4 increases to promote muscle glucose uptake as a compensatory mechanism in diet-induced obesity, Glut1 and Glut3 were not involved in WD-induced soleus glucose disposal in this study. Moreover, gene deletion of ECMR further increased Glut4 expression, which results in increased glucose uptake, improved glucose tolerance, as well as increased insulin sensitivity, suggesting that both WD and ECMR signaling impact Glut4 expression and mobilization. WD and MRs regulate Glut4 expression and activation through insulin metabolic signaling (1, 2, 4). Indeed, both excessive nutrient intake and MR activation increase serine phosphorylation of the critical insulin-signaling/docking molecule insulin receptor substrate 1, which impairs insulin metabolic signaling in PI3K/Akt pathways, as well as downstream Glut4 mobilization and expression (1, 2, 4). As indicated by our findings, endothelial-specific ECMR signaling also mediated WD-induced reductions in insulin metabolic signaling in PI3K/Akt pathways. Moreover, impaired insulin metabolic signaling further suppresses eNOS activation and nitric oxide production, leading to reduced insulin-mediated capillary recruitment, glucose delivery, and subsequent muscle glucose-uptake.

Overnutrition or obesity is the primary mechanism leading to the total IMC lipid content and ectopic fat storage in skeletal muscle, playing a key role in the development of insulin resistance, type 2 diabetes, lipid metabolic disorders, and the metabolic syndrome (44, 45). Our recent data further demonstrated that enhanced MR signaling is associated with excessive FFA uptake, IMC lipid accumulation, and hepatic steatosis by increased CD36 (2, 4). Of note, CD36 is a scavenger receptor that functions in high-affinity tissue uptake of FFA and contributes to lipid accumulation and metabolic dysfunction (5, 6). Studies have demonstrated that CD36 is increased in obese rodents and patients (2, 46, 47), and gene deletion of CD36 decreases FFA uptake, IMC lipid accumulation, and insulin resistance (7, 47, 48). Moreover, abnormal microRNA-99a participates in activated MRs induced increases in CD36 expression and skeletal muscle lipid metabolic disorder (2). Thus, MRs may indirectly regulate CD36 expression through miR-99a in a nongenomic manner. Interestingly, EC-specific ECMR activation also mediated WD-induced increases in soleus CD36, FFA uptake, IMC lipid accumulation, and lipid metabolic disorders.

There are about 5% to 15% of EC-derived microparticles in circulating microparticles, and these exosomal cargoes serve as mediators of cell–cell and tissue–tissue communications under normal and pathological conditions (20). For instance, while EC-derived exosomal kinases are delivered to muscle cells (cardiomyocytes) that promote diabetic cardiomyopathy and systemic insulin resistance (49), skeletal muscle–derived exosomes regulate EC functions through nuclear factor κB signaling (50). Moreover, palmitic acid increases CD36 sorting into exosomes which are then endocytosed by hepatocytes, leading to hepatic lipid accumulation and inflammation (9). Elevated functional CD36 in circulating exosomes also promotes FFA uptake in cardiomyocytes in the postprandial state (10). Our data also demonstrated that CD36 enriched plasma EC exosomes can be taken up by in vitro and in vivo skeletal muscle and is an important risk factor in developing skeletal muscle lipid disorders and insulin resistance. In addition to the characterization of in vivo and in vitro activated ECMR increases in EC-derived exosomal CD36, we developed a simple coculture assay to assess the role of ECMR activation/increased EC exosomal CD36 on in vitro C2C12 CD36 levels. While EC exosomal CD36 could be taken up in cultured C2C12 cells, inhibition of exosome (exosomal CD36) formation and release in ECs with GW4869 prevented aldosterone/activated ECMR-induced increases in C2C12 CD36. Therefore, EC exosomes mediate EC–skeletal muscle communication, and increased EC-derived exosomal CD36 contributes to activated ECMR-induced skeletal muscle lipid metabolic disorders and insulin resistance.

One of the noteworthy findings is that cell-specific ECMR activation also mediates WD-induced soleus oxidative stress and fibrosis. Indeed, our previous studies have demonstrated that enhanced MR signaling mediates 16 weeks of WD-induced skeletal muscle metabolic lipid disorders which further promotes mitochondrial dysfunction characterized by mitochondrial disorganization with loss of electron density and cristae (2). These metabolic abnormalities are associated with tissue oxidative stress (2, 4). In addition to increasing tissue mitochondrial dysfunction and oxidative stress, consumption of a WD also increases release of inflammatory cytokines, such as interleukin-6, tumor necrosis factor-α, CD68, and CD86 (4). The resultant pathological responses, induced by WD and lipid metabolic disorders, are well known to play key roles in the development of insulin resistance, type 2 diabetes, and metabolic syndrome (51, 52). It was noted that neither WD nor ECMR signaling affect mouse voluntary running activity. It was likely that 16 weeks of WD feeding did not induce a significant phenotypic difference in skeletal muscle function. There are several limitations to note with this investigation. For instance, only female mice were chosen in this study because our previous studies have shown that premenopausal females with obesity and diabetes lose hormonal metabolic protection and have greater propensity to develop insulin resistance, type 2 diabetes, and metabolic syndrome (53, 54). Meanwhile, we do not know exactly where these EC exosomal CD36 originate from and how these EC exosomes impact insulin sensitivity of other tissues. It was noted that we only evaluated the role of in vitro ECMR signaling on EC exosomal CD36 and C2C12 CD36 expression. An EC specific exosomal CD36 knockout genetic model and a concentration dependence of in vitro aldosterone treatment may further disclose the role of ECMR in systemic and skeletal muscle insulin resistance. Moreover, we cannot exclude other cell MRs and CD36's role in the pathogenesis of diet-induced metabolic disorders. Additionally, sample sizes and data variation may affect our results. Thus, the mechanisms of obesity and ECMR activation in insulin resistance are complex. Further studies are required to directly verify the roles of ECMR and EC exosomal CD36 in diet-induced systemic and skeletal muscle insulin resistance.

Collectively, results of this investigation suggest a pivotal role of ECMR activation in development of skeletal muscle lipid metabolic disorders and insulin resistance. These pathophysiological changes include associated increases in CD36 expression, FFA uptake, excessive IMC lipid accumulation, and soleus oxidative stress and fibrosis. Furthermore, EC-derived exosomal CD36 is associated with activated ECMR-induced increases in skeletal muscle CD36 expression, lipid metabolic disorders, and insulin resistance. Further studies are required to better delineate interactions between ECMR signaling and EC-derived exosomal CD36 in regulation of skeletal muscle metabolic disorders.

Abbreviations

Akt

protein kinase B

AUC

areas under the curve

CD

chow diet

EC

endothelial cell

ECMR

endothelial cell–specific mineralocorticoid receptor

eNOS

endothelial nitric oxide synthase

FFA

free fatty acid

Glut

glucose transporter

HOMA-IR

Homeostatic Model for Insulin Resistance

IMC

intramyocellular

IPGTT

intraperitoneal glucose tolerance test

MR

mineralocorticoid receptor

PI3K

phosphoinositide 3-kinases

WD

Western diet

Funding

This research was supported by the National Institute of Diabetes and Digestive and Kidney Diseases (R01 DK124329) and an American Diabetes Association Innovative Basic Science Award (1-17-IBS-201) to G. Jia. Dr. Sowers received funding from NIH (R01 HL73101-01A and R01 HL107910-01). Drs. Whaley-Connell and Lastra received funding from Veterans Affairs Merit System Grants BX003391 and 5101BX001981, respectively.

Author Contributions

G.J. and J.R.S. conceived and designed research; J.H., C.H., H.N., D.C., and G.J. performed experiments; J.H. and G.J. analyzed data; J.H. and G.J. interpreted results of experiments; G.J., A.W.-C., G.L., and J.R.S. drafted manuscript, edited, and revised manuscript. All authors reviewed, edited, and approved the manuscript. G.J. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Disclosures

No potential conflicts of interest relevant to this article were reported. The authors have nothing to disclose.

Data Availability

The datasets underlying this article will be shared on reasonable request to the corresponding author.