Ceramide analog C2-cer induces a loss in insulin sensitivity in muscle cells through the salvage/recycling pathway

Abstract

Ceramides have been shown to play a major role in the onset of skeletal muscle insulin resistance and therefore in the prevalence of type 2 diabetes. However, many of the studies involved in the discovery of deleterious ceramide actions used a nonphysiological, cell-permeable, short-chain ceramide analog, the C2-ceramide (C2-cer). In the present study, we determined how C2-cer promotes insulin resistance in muscle cells. We demonstrate that C2-cer enters the salvage/recycling pathway and becomes deacylated, yielding sphingosine, re-acylation of which depends on the availability of long chain fatty acids provided by the lipogenesis pathway in muscle cells. Importantly, we show these salvaged ceramides are actually responsible for the inhibition of insulin signaling induced by C2-cer. Interestingly, we also show that the exogenous and endogenous monounsaturated fatty acid oleate prevents C2-cer to be recycled into endogenous ceramide species in a diacylglycerol O-acyltransferase 1-dependent mechanism, which forces free fatty acid metabolism towards triacylglyceride production. Altogether, the study highlights for the first time that C2-cer induces a loss in insulin sensitivity through the salvage/recycling pathway in muscle cells. This study also validates C2-cer as a convenient tool to decipher mechanisms by which long-chain ceramides mediate insulin resistance in muscle cells and suggests that in addition to the de novo ceramide synthesis, recycling of ceramide could contribute to muscle insulin resistance observed in obesity and type 2 diabetes.

Keywords: Akt PKB; cell signaling; diacylglycerol; lipid signaling; lipogenesis; lipotoxicity; metabolism; oleate; signal transduction.

Figure 1

Effect of C2-cer on endogenous ceramide species synthesis in C2C12 myotubes.A, C2C12 myotubes were incubated for 16 h with 0.75 mmol/l palmitate in the presence or not in the presence of 10 μmol/l myriocin followed by 100 nmol/l insulin 10 min before cell lysis. B, C2C12 myotubes were incubated with 100 μmol/l C2-cer for 2 h followed by 100 nmol/l insulin 10 min before cell lysis. Cell lysates were immunoblotted with the indicated antibodies. C, C2C12 myotubes were incubated with 50 μmol/l C2-cer from 30 min to 2 h and stained for ceramide using an anti-ceramide antibody (green). In parallel, nuclei were stained with DAPI (blue). Scale bar represents 50 μm. D, C2C12 myotubes were incubated with 100 μmol/l C2-cer for 2 h. Following these incubations, muscle cells were harvested to assess ceramide content as described in the Experimental procedures section. Results are mean ± SEM (n = 5–6). ∗ represents significant change p ≤ 0.05 relative to the untreated control myotubes. E, C2C12 myotubes were incubated with 100 μmol/l C2-cer for 2 h or with 0.75 mmol/l palmitate for 16 h. Then, DAG content was assessed. Results are mean ± SEM (n = 3). ∗ represents significant change p ≤ 0.05 relative to the untreated control. C2-cer, C2-ceramide; DAG, diacylglycerol.

Figure 2

Effect of ceranib-2 on endogenous ceramide synthesis and insulin response.A, C2C12 myotubes were incubated with 50 μmol/l ceranib-2 for 2 h, and ceramide species content was assessed. Results are mean ± SEM (n = 4). ∗ represents significant change p ≤ 0.05 relative to the untreated control. B, C2C12 myotubes were incubated with 50 μmol/l ceranib-2 for 2 h and with or without 100 nmol/l insulin for 10 min before being lysed. Cell lysates were immunoblotted with the indicated antibodies. Scanning densitometry was performed to quantify changes in ⁴⁷³Ser Akt and ³⁰⁸Thr Akt abundance in cell lysates. Bars represent mean ± SEM (n = 3–4). ∗ represents significant change p ≤ 0.05 relative to the untreated control myotubes.

Figure 3

Effect of ceranib-2 on C2-cer–induced endogenous ceramide synthesis and insulin response.A, C2C12 myotubes were incubated with 100 μmol/l C2-cer for 2 h in the presence or not in the presence of 50 μmol/l ceranib-2 (added 30 min before). Sphingosine content was analyzed. Results are mean ± SEM (n = 4). ∗ represents significant change p ≤ 0.05 relative to the untreated control. ^a represents significant change p≤ 0.05 relative to the C2-cer–treated cells. B, C2C12 myotubes were incubated with 100 μmol/l C2-cer for 2 h. Total ceramide content was assessed. Results are mean ± SEM (n = 4). ∗ represents significant change p ≤ 0.05 relative to the untreated control. C, C2C12 myotubes were incubated with 100 μmol/l C2-cer for 2 h in the presence of 50 μmol/l ceranib-2, and total ceramide levels were assessed. Results are mean ± SEM (n = 4). ∗ represents significant change p ≤ 0.05 relative to the untreated control. D, C2C12 myotubes were incubated with 100 μmol/l C2-cer for 2 h in the presence of 50 μmol/l ceranib-2, followed by 100 nmol/l insulin 10 min before cell lysis. Cell lysates were immunoblotted with the indicated antibodies. Scanning densitometry was performed to quantify changes in ⁴⁷³Ser Akt and ³⁰⁸Thr Akt abundance in cell lysates. Bars represent mean ± SEM (n = 3). ∗ represents significant change p ≤ 0.05 relative to the untreated control myotubes. C2-cer, C2-ceramide.

Figure 4

Effect of C2-cer and FB1 on the insulin signaling pathway in C2C12 myotubes. C2C12 myotubes were incubated with 100 μmol/l C2-cer for 2 h in the presence or not in the presence of 1 μmol/l FB1 (A) or of 10 μmol/l myriocin (B) (both FB1 or myriocin added 30 min before), and total ceramide content was analyzed. Results are mean ± SEM (n = 3). ∗ represents significant change p ≤ 0.05 relative to the untreated control myotubes. C, C2C12 myotubes were incubated with 100 μmol/l C2-cer for 2 h in the presence or not in the presence of 1 μmol/l FB1, followed by 100 nmol/l insulin for 10 min before cell lysis. Scanning densitometry was performed to quantify changes in ⁴⁷³Ser Akt and ³⁰⁸Thr Akt abundance in cell lysates. Results are mean ± SEM (n = 3). ∗ represents significant change p ≤ 0.05 relative to the untreated control myotubes. D, C2C12 myotubes were incubated with 100 μmol/l C2-cer for 2 h in the presence or not in the presence of 1 μmol/l FB1, followed by 100 nmol/l insulin for 20 min before cell harvesting and subcellular fractionation as described in the Experimental procedures section. Isolated plasma membranes were immunoblotted with the indicated antibodies. Scanning densitometry was performed to quantify changes in GLUT4 abundance in plasma membranes. Results are mean ± SEM (n = 3). ∗ represents significant change p ≤ 0.05 relative to the untreated control myotubes. E, C2C12 myotubes were incubated with 100 μmol/l C2-cer for 2 h in the presence or not in the presence of 1 μmol/l FB1, followed by 100 nmol/l insulin for 20 min. At the end of the incubation period, 2-DG uptake was assessed as described in the Experimental procedures section. Results are mean ± SEM (n = 4). ∗ represents significant change p ≤ 0.05 relative to the untreated control myotubes. C2-cer, C2-ceramide; FB1, fumonisin B1.

Figure 5

Inhibition of lipogenesis prevents the deleterious action of C2-cer on insulin signaling in C2C12 myotubes.A, C2C12 were cultured in either 5 mmol/l or 25 mmol/l glucose-containing medium for 16 h before addition of 100 μmol/l C2-cer for 2 h and 100 nmol/l insulin for the last 10 min. Then, cells were harvested and lysates were immunoblotted with the indicated antibodies. Scanning densitometry was performed to quantify changes in ⁴⁷³Ser Akt and ³⁰⁸Thr Akt abundance in cell lysates. Bars represent mean ± SEM (n = 3). ∗ represents significant change p ≤ 0.05 relative to the untreated control myotubes. C2C12 myotubes were incubated with 100 μmol/l C2-cer for 2 h with or without 4 μg/ml TOFA (B) or 25 μmol/l C-75 (C) (both added 30 min before) and with 100 nmol/l insulin for the last 10 min. Then, cells were harvested and lysates were immunoblotted with the indicated antibodies. Scanning densitometry was performed to quantify changes in ⁴⁷³Ser Akt and ³⁰⁸Thr Akt abundance in cell lysates. Bars represent mean ± SEM (n = 3). ∗ represents significant change p ≤ 0.05 relative to the untreated control myotubes. D, C2C12 myotubes were incubated with 100 μmol/l C2-cer for 2 h with or without 4 μg/ml TOFA, and total ceramide content was analyzed. Results are mean ± SEM (n = 3). ∗ represents significant change p ≤ 0.05 relative to the untreated control myotubes. ^a represents significant change p ≤ 0.05 relative to the C2-cer–treated myotubes in the absence of TOFA. E, C2C12 myotubes were treated with 100 μmol/l C2-cer for 2 h in the presence or not in the presence of 4 μg/ml TOFA, followed by 100 nmol/l insulin for 20 min. At the end of the incubation period, 2-DG uptake was assessed as described in the Experimental procedures section. Results are mean ± SEM (n = 6). ∗ represents significant change p ≤ 0.05 relative to the untreated control myotubes. C2-cer, C2-ceramide; TOFA, 5-tetradecyloxy-2-furoic acid.

Figure 6

Oleate prevents the deleterious action of C2-cer on insulin signaling in C2C12 myotubes.A, C2C12 myotubes were incubated with 100 μmol/l C2-cer for 2 h with or without 0.3 mM oleate (added 30 min before) and with 100 nmol/l insulin the last 10 min. Then, cells were harvested and lysates were immunoblotted with the indicated antibodies. Scanning densitometry was performed to quantify changes in ⁴⁷³Ser Akt, ³⁰⁸Thr Akt, and ²⁰²Thr/²⁰⁴Tyr ERK abundance in cell lysates. Bars represent mean ± SEM (n = 3). ∗ represents significant change p ≤ 0.05 relative to the untreated control myotubes. ^a represents significant change p ≤ 0.05 relative to insulin-treated myotubes. B, C2C12 myotubes were incubated with or without 10 μmol/l C2-cer, 0.3 mmol/l oleate, or 0.3 mmol/l stearate for 2 h, or with or without 2 μmol/l A939572 for 16 h, before to be harvested. Fatty acids were assessed as described in the Experimental procedures section. Bars represent mean ± SEM (n = 6). ∗ represents significant change p ≤ 0.05 relative to the untreated control myotubes. C, C2C12 myotubes were incubated with 2 μmol/l A939572 for 16 h before to add 1 to 100 μmol/l C2-cer for 2 h. After 10 min insulin treatment (100 nmol/l), cells were lysed and lysates were immunoblotted with the indicated antibodies. Scanning densitometry was performed to quantify changes in ⁴⁷³Ser Akt, ³⁰⁸Thr Akt, and ²⁰²Thr/²⁰⁴Thr ERK abundance in cell lysates (n = 3). ∗ represents significant change p ≤ 0.05 relative to A939572-untreated myotubes. C2-cer, C2-ceramide.

Figure 7

Oleate channels fatty acids towards storage and prevents C2-cer to act on insulin signaling in C2C12 myotubes.A, C2C12 myotubes were incubated with 0.3 mmol/l oleate, in the presence or not in the presence of 10 μmol/l A922500 for 16 h before to fix the cells and to label TG (red) and nuclei (blue). Scale bar represents 50 μm. B, C2C12 myotubes were incubated with 10 μmol/l A922500 15 min before to add 0.3 mmol/l oleate. Thirty minutes later, cells were incubated for 2 h with 100 μmol/l C2-cer. After 10 min insulin treatment (100 nmol/l), muscle cells were harvested and lysates were immunoblotted with the indicated antibodies. Scanning densitometry was performed to quantify changes in ⁴⁷³Ser Akt, ³⁰⁸Thr Akt, and ²⁰²Thr/²⁰⁴Thr ERK abundance in cell lysates. Bars represent mean ± SEM (n = 3). ∗ represents significant change p ≤ 0.05 relative to the untreated control myotubes. ^a represents significant change p ≤ 0.05 relative to oleate/C2-cer/insulin–treated myotubes. C, C2C12 myotubes were incubated with 0.3 mmol/l oleate. Thirty minutes later, cells were incubated for 2 h with 100 μmol/l C2-cer. Ceramide content was assessed as described in the Experimental procedures section. Results are mean ± SEM (n = 3–4). ∗ represents significant change p ≤ 0.05 relative to untreated control myotubes. ^a represents significant change p ≤ 0.05 relative to C2-cer–treated myotubes. D, C2C12 myotubes were incubated with 10 μmol/l A922500 15 min before to add 0.3 mmol/l oleate. Thirty minutes later, cells were incubated for 2 h with 100 μmol/l C2-cer. Ceramide content was assessed as described in the Experimental procedures section. Results are mean ± SEM (n = 3). ∗ represents significant change p ≤ 0.05 relative to C2-cer/oleate–treated myotubes. E, C2C12 myotubes were incubated or not incubated with 10 μmol/l A922500 15 min before to add 0.3 mmol/l oleate for 2 h. DAG content was assessed as described in the Experimental procedures section. Results are mean ± SEM (n = 3). ∗ represents significant change p ≤ 0.05 relative to untreated control myotubes. ^a represents significant change p≤ 0.05 relative to oleate-treated myotubes. C2-cer, C2-ceramide; DAG, diacylglycerol.