Dissociation of intramyocellular lipid storage and insulin resistance in trained athletes and type 2 diabetes patients; involvement of perilipin 5?

KEY POINTS: Intramyocellular lipid storage is negatively associated with insulin sensitivity. However, endurance trained athletes and type 2 diabetes mellitus (T2DM) patients store similar amounts of lipids in their muscle; the so-called athlete's paradox. Compared to T2DM, trained athletes possess higher levels of perilipin 5 (PLIN5), a lipid droplet (LD) coating protein. We examined whether coating LD with PLIN5 affects the pattern of muscle lipid (LD size and number) in relation to the athlete's paradox. Despite differences in PLIN5 protein content, we observed that coating the LD with PLIN5 could not explain the observed differences in LD size and number between athletes and T2DM. PLIN5-coated LDs were positively associated with oxidative capacity but not with insulin sensitivity. We conclude that coating of LDs with PLIN5 cannot causally explain the athlete's paradox. ABSTRACT: Intramyocellular lipid (IMCL) hampers insulin sensitivity, albeit not in endurance-trained athletes (Trained). Compared to type 2 diabetes mellitus (T2DM) patients, Trained subjects have high levels of perilipin 5 (PLIN5). In the present study, we tested whether the fraction of PLIN5-coated lipid droplets (LDs) is a determinant of skeletal muscle insulin sensitivity and contributes to the athlete's paradox. Muscle biopsies were taken from eight Trained, Lean sedentary, Obese and T2DM subjects. Trained, Obese and T2DM subjects were matched for total IMCL content. Confocal images were analysed for lipid area fraction, LD size and number and PLIN5+ and PLIN5- LDs were measured. A stepwise linear regression was performed to identify factors explaining observed variance in glucose infusion rate (GIR). Trained and T2DM subjects stored IMCL differently; Trained subjects had a higher number of LDs compared to T2DM subjects (0.037 ± 0.004 μm-2 vs. 0.023 ± 0.003 μm-2 , P = 0.024) that were non-significantly smaller (0.27 ± 0.01 μm2 vs. 0.32 ± 0.02 μm2 , P = 0.197, Trained vs. T2DM). Even though total PLIN5 protein content was almost double in Trained vs. T2DM subjects (1.65 ± 0.21 AU vs. 0.89 ± 0.09 AU, P = 0.004), PLIN5 coating did not affect LD number or size significantly. Of the observed variance in GIR, the largest fraction by far (70.2%) was explained by maximal oxygen uptake. Adding PLIN5 protein content or PLIN5+ LDs increased the explained variance in GIR (74.7% and 80.7% for PLIN5 protein content and PLIN5+ LDs, respectively). Thus, the putative relationship between PLIN5 and insulin sensitivity is at best indirect and is apparent only in conjunction with maximal oxygen uptake. Hence, PLIN5 abundance cannot be causally linked to the athlete's paradox.