Muscle lipogenesis balances insulin sensitivity and strength through calcium signaling

Abstract

Exogenous dietary fat can induce obesity and promote diabetes, but endogenous fat production is not thought to affect skeletal muscle insulin resistance, an antecedent of metabolic disease. Unexpectedly, the lipogenic enzyme fatty acid synthase (FAS) was increased in the skeletal muscle of mice with diet-induced obesity and insulin resistance. Skeletal muscle-specific inactivation of FAS protected mice from insulin resistance without altering adiposity, specific inflammatory mediators of insulin signaling, or skeletal muscle levels of diacylglycerol or ceramide. Increased insulin sensitivity despite high-fat feeding was driven by activation of AMPK without affecting AMP content or the AMP/ATP ratio in resting skeletal muscle. AMPK was induced by elevated cytosolic calcium caused by impaired sarco/endoplasmic reticulum calcium ATPase (SERCA) activity due to altered phospholipid composition of the sarcoplasmic reticulum (SR), but came at the expense of decreased muscle strength. Thus, inhibition of skeletal muscle FAS prevents obesity-associated diabetes in mice, but also causes muscle weakness, which suggests that mammals have retained the capacity for lipogenesis in muscle to preserve physical performance in the setting of disrupted metabolic homeostasis.

Figures

Figure 1. Induction of skeletal muscle FAS by HFD feeding and FASKOS mouse generation.

(A) FAS mRNA, protein, and enzyme activity in soleus muscles from mice fed chow or HFD (n = 7 per group). (B) Western blot analyses of FAS protein expression in various muscles (top) and tissues (bottom). GW, gastrocnemius (white); GR, gastrocnemius (red); TA, tibialis anterior; Epi, epitrochlearis; Vast, vastus lateralis; Diaph, diaphragm; COXIV, mitochondrial cytochrome c oxidase subunit IV. (C) FAS Western blots from muscles (top) and other tissues (bottom) in control and FASKOS mice. (D) FAS mRNA (n = 6 per genotype), protein (n = 7 per genotype), and enzyme activity (n = 10 per genotype) in control and FASKOS mice. Data are mean ± SEM.

Figure 2. Phenotyping of FASKOS mice.

(A) Body weight and composition of control and FASKOS mice (n = 8 per genotype). (B and C) Soleus and EDL weights (B) and liver and adipose tissue weights (C) in HFD-fed control and FASKOS mice (n = 7 per genotype). (D) Serum chemistries. (E) Serum hormones. Data are mean ± SEM.

Figure 3. Skeletal muscle metabolites in HFD-fed control and FASKOS mice.

(A) Triglycerides, (B) free fatty acids, (C) malonyl-CoA, (D) DAG species, and (E) ceramide species (n = 4 per group). Data are mean ± SEM.

Figure 4. FASKOS mice are protected from diet-induced whole-body glucose intolerance and insulin resistance.

(A) VO2 in control and FASKOS mice. (B) RQ. (C–F) Glucose tolerance tests (GTT; C and E) and insulin tolerance tests (ITT; D and F) in mice fed chow (C and D) or HFD (E and F). n = 5 (A and B), 10 (C, D, and F), or 12 (E) per genotype. Data are mean ± SEM.

Figure 5. Hyperinsulinemic euglycemic clamp studies in HFD-fed control and FASKOS mice.

(A) Glucose infusion rate (GIR), (B) glucose disposal rate (GDR; during clamp phase), (C) insulin-stimulated glucose disposal rate (IS-GDR; calculated difference between clamp and basal phases), and (D) suppression of hepatic glucose production (HGP) (n = 10 [control]; 8 [FASKOS]). (E) Circulating free fatty acids during basal and clamp phase (n = 7 per genotype). (F) Western blot analyses of soleus from clamp studies. Data are mean ± SEM.

Figure 6. Isolated soleus muscle studies.

(A) 2-deoxyglucose (2DG) uptake, (B) pAMPKThr172, (C) pACCSer79, (D) pAktThr308, and (E) pAS160Thr642 (n = 6 per group). Data are mean ± SEM. Representative blots from muscles are also shown.

Figure 7. Muscle FAS deletion does not affect PPARα activity, but increases cytosolic calcium.

(A) Western blot analyses of scrambled (SC) or FASKD C2C12 myocytes. (B) PPARα luciferase activity in scrambled or FASKD C2C12 myocytes with or without the PPARα agonist WY14643 (n = 8 replicate measurements per bar). (C) Western blot analyses of scrambled or FASKD C2C12 myocytes with or without WY14643. (D) Cytosolic calcium in scrambled and FASKD C2C12 myocytes measured using fluo-8 AM. Original magnification, ×100. (E) Quantified fluo-8 AM fluorescence in scrambled and FASKD C2C12 myocytes (n = 12 culture wells per condition; horizontal lines denote means). Data are mean ± SEM.

Figure 8. Muscle FAS deletion induces calcium signaling and activates AMPK through CaMKKβ.

(A) Calcium signaling cascade affected by muscle FAS deletion. PLB, phospholamban. (B–D) Western blot analyses for calcium signaling proteins (n = 6 per group). Data are mean ± SEM. Representative blots are also shown. (B) C2C12 myocytes. Phospholamban is phosphorylated by CaMKII on Thr17 and by PKA on Ser16. (C) Muscle from clamp studies of HFD-fed animals. (D) Isolated soleus muscle. (E) Western blot analyses of C2C12 myocytes after FASKD and CaMKKβ knockdown. (F) Incubation of C2C12 myocytes with the CaMKK inhibitor STO-609 (10 μg/ml for 6 hours) eliminated FASKD-induced activation of AMPK signaling.

Figure 9. Muscle FAS deletion decreases SERCA activity.

(A) Differential centrifugation of mouse hindlimb muscles. PM, plasma membrane; GM130, 130-kDa cis-Golgi matrix protein. (B) Soleus co-IP experiments. FAS, but not ACC, coimmunoprecipitated with SERCA1 or SERCA2 antibody; SERCA1 and SERCA2 coimmunoprecipitated with FAS antibody, but not with ACC antibody. (C) Confocal images of C2C12 cells. Original magnification, ×630. Boxed regions are shown enlarged (×4-fold) at far right. (D) Rescue of the FASKD-induced reduction in SERCA enzyme activity by expression of myc-SERCA1 (n = 8 replicates per condition). Moreover, SR SERCA abundance with vector treatment was not altered by FASKD (bottom). (E) Expression of myc-SERCA1 reversed FASKD-induced increase in pAMPK. (F) SERCA activity in soleus from chow- or HFD-fed control and FASKOS mice (n = 6 per group). SR SERCA abundance is also shown. Data are mean ± SEM.

Figure 10. Muscle FAS deletion alters SR phospholipid composition.

Mass spectrometry quantification of (A) relative PE and PC abundance, (B and C) total PE and PC, and (D and E) PE and PC species in isolated SR from hindlimb muscles of chow- or HFD-fed control and FASKOS mice (n = 4–6). Data are mean ± SEM.

Figure 11. FASKOS mice are weak.

(A) Forelimb strength in chow- or HFD-fed control and FASKOS mice (n = 7 per group). (B and C) Total running time and distance of HFD-fed mice with treadmill sprint interval exercise tests (n = 6 per group). Data are mean ± SEM.

Figure 12. FAS-driven modulation of SERCA activity and cytosolic calcium leads to increased insulin sensitivity and muscle weakness.

Increased cytosolic calcium after FAS inactivation likely impairs relaxation of the actin-myosin junction at the sarcomere, promoting muscle weakness. Calcium also likely activates a signaling cascade involving CaMKKβ, AMPK, and AS160 that enhances insulin-stimulated glucose uptake in muscle.