Intramyocellular Lipid Droplet Size Rather Than Total Lipid Content is Related to Insulin Sensitivity After 8 Weeks of Overfeeding

Abstract

Objective: Intramyocellular lipid (IMCL) is inversely related to insulin sensitivity in sedentary populations, yet no prospective studies in humans have examined IMCL accumulation with overfeeding.

Methods: Twenty-nine males were overfed a high-fat diet (140% caloric intake, 44% from fat) for 8 weeks. Measures of IMCL, whole-body fat oxidation from a 24-hour metabolic chamber, muscle protein extracts, and muscle ceramide measures were obtained before and after the intervention.

Results: Eight weeks of overfeeding did not increase overall IMCL. The content of smaller lipid droplets peripherally located in the myofiber decreased, while increases in larger droplets correlated inversely with glucose disposal rate. Overfeeding resulted in inhibition of Akt activity, which correlated with the reductions in smaller, peripherally located lipid droplets and drastic increases in ceramide content. Additionally, peripherally located lipid droplets were associated with more efficient lipid oxidation. Finally, participants who maintained a greater number of smaller, peripherally located lipid droplets displayed a better resistance to weight gain with overfeeding.

Conclusions: These results show that lipid droplet size and location rather than mere IMCL content are important to understanding insulin sensitivity.

Trial registration: ClinicalTrials.gov NCT01672632.

Figure 1. Overfeeding associated increases in the content of large myofiber lipid droplets is related to reductions in insulin sensitivity

A) Total intramyocellular lipid (IMCL) content in the soleus and tibialis anterior via magnetic proton resonance spectroscopy and B) in the vastus lateralis using oil red o staining in frozen tissue from biopsy did not increase with overfeeding (N=29 soleus and vastus lateralis, N=28 tibialis anterior). C) Lipid droplet size from the vastus lateralis were determined as centrally vs. peripherally located droplets, and the content of small lipid droplets located peripherally decreased (p=0.005, N=25). D) Representative image of histological cross-sections of the myofibers stained with oil red o from the same participant before and after overfeeding. Red lined borders indicate the demarcation between central and peripherally location and the outer border of the myofiber. Arrows indicate the large lipid droplets that accumulated with overfeeding in this participant. E) Inverse relationship between the percent change in larger lipid droplets throughout the entire fiber and the percent change in glucose disposal rate (GDR) with overfeeding (r=-0.47, p=0.02, N=23). Graphs represents mean±SEM, *p<0.05.

Figure 2. Alterations in lipid droplet size is related to changes in lipid oxidation

A) Baseline content of small, peripherally located lipid droplets are more associated with whole body fat oxidation as evident by inverse relationships with 24 hour respiratory quotient (RQ, r = -0.35, p = 0.09, N = 24) and B) sleep RQ (r = -0.57, p = 0.004, N = 24). C) Changes in smaller lipid droplet content tended to correlate positively with changes in 24 hr. respiratory quotient (RQ) (r = 0.38, p = 0.07, N = 23), and D) correlated inversely to ex vivo total palmitate oxidation (r = -0.58, p = 0.04, N = 12). E) Smaller, peripherally located lipid droplets at baseline are inversely associated with palmitate produced mitochondrial reactive oxygen species (ROS) production (r = -0.72, p = 0.01, N = 11). F) Changes in the content of smaller, peripherally located lipid droplets inversely correlated to changes in palmitate stimulated reactive oxygen species (ROS) production ex vivo from isolated mitochondria (r = -0.65, p = 0.03, N = 11).

Figure 3. Alterations in lipid droplet size is related to changes in perilipin 2 and perilipin 3 protein expression with overfeeding

A) Perilipin 3 (PLIN3) protein content decreased with overfeeding (p=0.04, N=27). B) Baseline human primary myotubes treated with 200μM of palmitate showed that PLIN2 is expressed later over a time course of 2 hours. C) Changes in PLIN2 protein content correlated inversely with glucose disposal rate (GDR, r=-0.43, p=0.03, N=25). Graphs represent mean±S.E.M, *p<0.05.

Figure 4. Decreases Akt activity is associated with increased accumulation of ceramide content

A) Decreases in the activity of protein kinase B (PKB/Akt) before and after overfeeding; p=0.05, N=22). B and C) Muscle total ceramide content and ceramide species increased with overfeeding (p<0.001, all subspecies, N=29). D) The percent change in total ceramide content with overfeeding inversely correlated to the change in Akt activity with overfeeding (r=-0.47, p=0.04, N=22). E) The protein content of mammalian target of rapamycin (mTOR) increased significantly in skeletal muscle with overfeeding (p=0.05, N=27). Graphs represent mean±S.E.M, *p<0.05; ***p<0.001.

Figure 5. Pre-intervention levels of smaller, peripherally located lipid droplets in muscle was associated with less weight gain during overfeeding

A) Participants who had higher pre-intervention levels of smaller, peripherally located lipid droplets gained less weight with overfeeding, indicating that participants who have smaller, peripherally located lipid droplets are resistant to weight gain with overfeeding (r=-0.44, p=0.03, N=25). B) Schematic diagram examining the effects of diet-induced alterations in skeletal muscle lipid droplet morphology.