Human skeletal muscle mitochondrial dynamics in relation to oxidative capacity and insulin sensitivity

Abstract

Aims/hypothesis: Mitochondria operate in networks, adapting to external stresses and changes in cellular metabolic demand and are subject to various quality control mechanisms. On the basis of these traits, we here hypothesise that the regulation of mitochondrial networks in skeletal muscle is hampered in humans with compromised oxidative capacity and insulin sensitivity.

Methods: In a cross-sectional design, we compared four groups of participants (selected from previous studies) ranging in aerobic capacity and insulin sensitivity, i.e. participants with type 2 diabetes (n = 11), obese participants without diabetes (n = 12), lean individuals (n = 10) and endurance-trained athletes (n = 12); basal, overnight fasted muscle biopsies were newly analysed for the current study and we compared the levels of essential mitochondrial dynamics and quality control regulatory proteins in skeletal muscle tissue.

Results: Type 2 diabetes patients and obese participants were older than lean participants and athletes (58.6 ± 4.0 and 56.7 ± 7.2 vs 21.8 ± 2.5 and 25.1 ± 4.3 years, p < 0.001, respectively) and displayed a higher BMI (32.4 ± 3.7 and 31.0 ± 3.7 vs 22.1 ± 1.8 and 21.0 ± 1.5 kg/m2, p < 0.001, respectively) than lean individuals and endurance-trained athletes. Fission protein 1 (FIS1) and optic atrophy protein 1 (OPA1) protein content was highest in muscle from athletes and lowest in participants with type 2 diabetes and obesity, respectively (FIS1: 1.86 ± 0.79 vs 0.79 ± 0.51 AU, p = 0.002; and OPA1: 1.55 ± 0.64 vs 0.76 ± 0.52 AU, p = 0.014), which coincided with mitochondrial network fragmentation in individuals with type 2 diabetes, as assessed by confocal microscopy in a subset of type 2 diabetes patients vs endurance-trained athletes (n = 6). Furthermore, lean individuals and athletes displayed a mitonuclear protein balance that was different from obese participants and those with type 2 diabetes. Mitonuclear protein balance also associated with heat shock protein 60 (HSP60) protein levels, which were higher in athletes when compared with participants with obesity (p = 0.048) and type 2 diabetes (p = 0.002), indicative for activation of the mitochondrial unfolded protein response. Finally, OPA1, FIS1 and HSP60 correlated positively with aerobic capacity (r = 0.48, p = 0.0001; r = 0.55, p < 0.001 and r = 0.61, p < 0.0001, respectively) and insulin sensitivity (r = 0.40, p = 0.008; r = 0.44, p = 0.003 and r = 0.48, p = 0.001, respectively).

Conclusions/interpretation: Collectively, our data suggest that mitochondrial dynamics and quality control in skeletal muscle are linked to oxidative capacity in humans, which may play a role in the maintenance of muscle insulin sensitivity. CLINICAL TRIAL REGISTRY: numbers NCT00943059, NCT01298375 and NL1888 Graphical abstract.

Keywords: Fission, FIS1; Fusion; HSP60; Insulin sensitivity; Mitochondria; OPA1; Oxidative phosphorylation; Skeletal muscle.

Figures

Graphical abstract

Fig. 1

Type 2 diabetic individuals show lower expression of proteins involved in mitochondrial dynamics. Representative western blots (a) and associated quantifications (b–f) of mitochondrial dynamics regulatory proteins (MFN1, MFN2, OPA1, DRP1 and FIS1) in human skeletal muscle biopsies. T2DM, participants with type 2 diabetes (n = 11); O, obese individuals without type 2 diabetes (n=12); L, lean individuals (n=10); A, endurance-trained athletes (n=12). Values are normalised to Revert staining and expressed as mean ± SD. *p<0.05, **p<0.01

Fig. 2

Type 2 diabetic individuals display a more fragmented skeletal muscle mitochondrial network. Representative confocal images of the mitochondrial network in type I muscle fibres. The cellular membrane was stained in green using laminin as a marker, in red the mitochondrial network is visualised using TOMM20 as a marker. In the overview images (a, b and e, f) scale bar, 20 μm; in the zoomed images (c, d and g, h) scale bar, 5 μm. (a–d) Representative overview and zoomed images of two separate individuals with type 2 diabetes. (e–h) Representative overview and zoomed images of two individual trained athletes. Note the longitudinal pattern of interconnected mitochondria alongside the myofibrils (yellow arrows) and the cross-striated pattern reflecting mitochondria near the Z-line (blue arrows) in the athletes whereas this pattern is virtually absent in the type 2 diabetic participants and the networks present a more punctate dot-like pattern, reflecting disconnected mitochondria

Fig. 3

Type 2 diabetic individuals show lower mitochondrial content. Representative blots (a) and associated quantifications (b–g) of mitochondrial density markers VDAC1 and OXPHOS complex I (CI-NDUFB8), complex II (CII-SDHB), complex III (CIII-UQCRC2), complex IV (CIV-COX II) and complex V (CV-ATP5A) in human skeletal muscle. T2DM, participants with type 2 diabetes (n=11); O, obese individuals without type 2 diabetes (n=12); L, lean individuals (n=10); A, endurance-trained athletes (n=12). Values are normalised to Revert staining and expressed as mean ± SD. OXPHOS complexes are expressed as fold change, with T2DM set to 1. *p<0.05, **p<0.01, ***p<0.001

Fig. 4

Mitonuclear protein balance differs in lean individuals and athletes vs obese and type 2 diabetic participants. Representative blots (a) and associated quantifications of (b) the nuclear-encoded complex IV (CIV) subunit COX4I1, (c) the mtDNA-encoded complex IV subunit COX II and (d) mitonuclear protein balance (COX4I1/COX II) in human skeletal muscle. Both COX4I1 and COX II were detected on the same blots. T2DM, participants with type 2 diabetes (n=11); O, obese individuals without type 2 diabetes (n=12); L, lean individuals (n=10); A, endurance-trained athletes (n=12). Values are expressed as mean fold change ± SD, with T2DM set to 1. *p<0.05, **p<0.01, ***p<0.001, †p=0.06

Fig. 5

Endurance-trained athletes show highest protein levels of the mitochondrial chaperone HSP60. Representative western blot analysis (a) and quantification of (b) mitochondrial chaperone HSP60 and (d, e) mito-/autophagy-related proteins in human skeletal muscle biopsies. (c) Mitonuclear protein balance (COX4I1/COX II) is associated with HSP60 protein levels. T2DM, participants with type 2 diabetes (n=11); O, obese individuals without type 2 diabetes (n=12); L, lean individuals (n=10); A, endurance-trained athletes (n=12). Values are normalised to Revert staining and expressed as mean ± SD. *p<0.05, **p<0.01. Correlations are computed with Pearson’s r

Fig. 6

Proteins involved in mitochondrial dynamics and quality control positively correlate with metabolic variables in human skeletal muscle. Association between protein expression levels of OPA1, FIS1 and HSP60, normalised to Revert staining, with V˙O2max (a–c) and insulin sensitivity (d–f). Correlations are computed with Pearson’s r