Gymnastik- och idrottshögskolan, GIH

Altered ceramide accumulation contributes to skeletal muscle insulin resistance, but mechanisms underlying fibre-type-specific susceptibility remain unclear. We hypothesized that fibre-type-specific ceramide metabolism governs vulnerability to lipid-induced insulin resistance. Lipidomics and quantification of ceramide-pathway enzymes were performed in mouse skeletal muscles with distinct fibre-type composition (oxidative, mixed and glycolytic) from control-diet (n = 12) and high-fat-diet (HFD; n = 12) mice. In humans, lipidomics and enzyme profiling were done in vastus lateralis biopsies from 36 adults stratified into oxidative or glycolytic phenotypes; insulin sensitivity was determined by glucose tolerance testing. siRNA-mediated silencing of SGMS1 and SGMS2 followed by lipidomics probed sphingomyelin-ceramide cycling in human myoblasts. In mouse muscle, ceramide composition rather than total content, differed by fibre type: oxidative muscle was enriched in very-long-chain ceramides, whereas glycolytic and mixed muscles contained higher C18-ceramides, paralleled by fibre-type-specific expression of enzymes involved in de novo synthesis and sphingomyelin-ceramide cycling. HFD induced ceramide remodelling, with C18-ceramides accumulating in oxidative and mixed muscles and very-long-chain species decreasing in glycolytic muscle; among all assessed enzymes, only SGMS2 was significantly downregulated in oxidative muscle. In humans, an oxidative phenotype associated with higher very-long-chain ceramides and insulin sensitivity, whereas a glycolytic phenotype displayed higher C16-18 ceramides, higher SGMS1 and SMPD2 expression, and lower insulin sensitivity. Elastic net regression identified C16-18 ceramides and galactosylceramides as negative predictors of insulin sensitivity. SGMS2 silencing caused broader ceramide accumulation than SGMS1 silencing, supporting a central role for SGMS2-mediated sphingomyelin-ceramide cycling in limiting ceramide burden.

Exercise training (ET) has demonstrated beneficial effects in autoimmune and neurological disorders, including multiple sclerosis and its animal model, experimental autoimmune encephalomyelitis (EAE). ET modulates the gut microbiota, which influences neuroimmune interactions via the microbiota-gut-brain and microbiota-gut-immune system axes. However, the role of gut microbiota in mediating ET's protective effects in autoimmune neuroinflammation remains unclear. We investigated whether gut microbiota mediates the beneficial effects of ET on EAE development. Healthy mice underwent high-intensity continuous training (HICT). Fecal microbiota from HICT and sedentary mice were transplanted into naïve recipients, followed by proteolipid protein (PLP) immunization to induce EAE. Disease severity, gut microbial composition (16S rDNA sequencing), short-chain fatty acid (SCFA) levels (LC-MS), and autoreactive T-cell proliferation (flow cytometry) were assessed. Faecal microbiota transplantation (FMT) from HICT donors significantly reduced EAE severity, delaying onset and decreasing CNS inflammation, demyelination, and axonal damage. These effects correlated with distinct microbial signatures, including increased Faecalimonas and Escherichia genera, and decreased Mucispirillum genus. HICT-FMT mice exhibited higher Faecalimonas abundance and reduced serum SCFA levels. PLP-reactive T-cell proliferation was suppressed in HICT-FMT recipients. Gut microbiota from HICT mice confers protection against EAE development, associated with microbial-metabolic shifts and modulation of autoreactive T-cell responses.

Brain-derived neurotrophic factor (BDNF) is essential for neuroplasticity. Exercise caninduce increases in forearm venous plasma and serum BDNF, often assumed to be indicativeof release from the brain. We investigated the effects of exercise on circulating levels of matureBDNF (mBDNF) and its precursor proBDNF. Sixteen healthy, physically fit adults (20–40 years old)cycled for 20 min at 40, 60 and 80% of V˙O2 max, separated by 30 min of rest. BDNF was analysed in blood samples from the brachial artery, internal jugular vein, femoral vein and antecubital vein. Brain/skeletal muscle exchange of BDNF, calculated as arterial-venous differences in BDNF multiplied by blood flow in the middle cerebral artery/common femoral artery, was measured simultaneously with blood sampling. Exercise intensity-dependent increases were observed in blood platelet count, forearm venous serum mBDNF and plasma proBDNF, but not in forearm venous plasma mBDNF. Brain release (or uptake) was not detected for either plasma mBDNF, serum mBDNF or plasma proBDNF. However, muscle uptake of plasma mBDNF and release of plasma proBDNF were observed after high-intensity exercise. Our findings demonstrate that exercise-dependent increases in serum mBDNF are not derived from the brain or the exercised skeletal muscle. Rather, the source of the increase appears to be the increase in platelets that are enriched with mBDNF. Furthermore, in physically fit adults, BDNF is not released from the brain into the bloodstream, after exercise, regardless of exercise intensity. Finally, changes in plasma proBDNF after exercise appear to be dependent on exercised skeletal muscle rather than brain release. KEY POINTS: Previously shown exercise-induced increases in forearm venous brain-derived neurotrophic factor (BDNF) are often assumed to be indicative of release from the brain. We investigated whether exercise-induced changes in forearm venous mature BDNF (mBDNF) and precursor proBDNF are paralleled by concomitant changes in BDNF exchange over the brain and skeletal muscle. We observed exercise intensity-dependent increases in platelet count, forearm venous serum mBDNF and plasma proBDNF, but not in forearm venous plasma mBDNF. We found muscle uptake of plasma mBDNF and release of plasma proBDNF after high-intensity exercise but no exercise intensity-dependent brain exchange of either plasma mBDNF, serum mBDNF or plasma proBDNF. Our findings suggest that acute exercise-induced increases in circulating serum mBDNF may be solely a result of increased platelet count, probably due to splenic platelet release; and that exercised skeletal muscle, and not the brain, responds to high-intensity exercise by releasing plasma proBDNF.

CONTEXT: Insulin resistance (IR) is a major risk factor for the development of several diseases that have reached epidemic proportions worldwide, including hypertension, obesity and type 2 diabetes. In many diseased states, IR is associated with fasting hyperinsulinemia/excessive glucose-stimulated insulin secretion. However, it is not known whether hyperinsulinemia precedes/leads to the natural development of IR or vice versa.

OBJECTIVE: Here, we assess the relationship between hyperinsulinemia and insulin sensitivity in a cohort of healthy young lean men and women, where IR is observed in those who exhibit a low expression of type I skeletal muscle fibers and a high resting heart rate.

METHODS: Biopsies were obtained from the vastus lateralis muscle, followed by an intravenous glucose tolerance test. Insulin secretion and whole-body insulin sensitivity were calculated.

RESULTS: In this young population of normoglycemic, glucose-tolerant individuals, insulin sensitivity was significantly and negatively associated with fasting levels of plasma insulin, as well as insulin secretion in response to glucose infusion. Surprisingly, however, all the correlations became stronger when calculated in women, but became insignificant when calculated in men. In contrast, insulin sensitivity was significantly correlated with expression of type I skeletal muscle fibers and resting heart rate to similar extents in both sexes.

CONCLUSIONS: In the natural development of IR in men, it appears that hyperinsulinemia is a compensatory adaptation to peripheral IR rather than its cause.

OBJECTIVE: Here we use skeletal muscle fiber composition to investigate whether defects in amino acid metabolism are involved in the early development of IR in healthy young individuals before onset of clinical manifestations.

DESIGN: Two groups consisting of healthy young men and women, insulin-sensitive and insulin resistant, were studied using a cross-sectional design.

METHODS: Biopsies were obtained from the vastus lateralis muscle and an intravenous glucose tolerance test was performed. Plasma and muscle tissue were analyzed by metabolomics.

RESULTS: Subjects in group 1 (n=20; age 28±5 yrs; body mass index 22.3±2.7 kg/m2) had an expression of type I muscle fibers and whole-body insulin sensitivity, respectively, of 58.8±5.7% and 1.8±0.7 units. Subjects in group 2 (n=16; age 25±6 yrs; body mass index 22.6±3.0 kg/m2) had an expression of type I muscle fibers and whole-body insulin sensitivity, respectively, of 29.8±6.6% and 0.8±0.3 units (P<0.001 vs. group 1 for both). Anserine and β-alanine contents in muscle were significantly higher and taurine lower in group 2 vs. 1, consistent with the differences in muscle fiber composition between groups. Taurine correlated well with insulin sensitivity and expression of type I muscle fibers (r=0.63; P<0.001 for both). In contrast, there were no significant differences in plasma or tissue contents of glutamine, arginine, or branch-chain amino acids between groups.

CONCLUSIONS: These data demonstrate that the early development of IR is not a consequence of defects in amino acid metabolism. Rather, defects in amino acid metabolism in diseased states are more likely a consequence of IR.

Insulin resistance (IR) is a risk factor for the development of several major metabolic diseases. Muscle fiber composition is established early in life and is associated with insulin sensitivity. Hence, muscle fiber composition was used to identify early defects in the development of IR in healthy young individuals in the absence of clinical manifestations. Biopsies were obtained from the thigh muscle, followed by an intravenous glucose tolerance test. Indices of insulin action were calculated and cardiovascular measurements, analyses of blood and muscle were performed. Whole-body insulin sensitivity (SI_galvin) was positively related to expression of type I muscle fibers (r=0.49; P<0.001) and negatively related to resting heart rate (HR, r=-0.39; P<0.001), which was also negatively related to expression of type I muscle fibers (r=-0.41; P<0.001). Muscle protein expression of endothelial nitric oxide synthase (eNOS), whose activation results in vasodilation, was measured in two subsets of subjects expressing a high percentage of type I fibers (59±6%; HR = 57±9 beats/min; SI_galvin = 1.8±0.7 units) or low percentage of type I fibers (30±6%; HR = 71±11; SI_galvin = 0.8±0.3 units; P<0.001 for all variables vs. first group). eNOS expression was: 1. higher in subjects with high type I expression; 2. almost two-fold higher in pools of type I vs. II fibers; 3. only detected in capillaries surrounding muscle fibers; and 4. linearly associated with SI_galvin. These data demonstrate that an altered function of the autonomic nervous system and a compromised capacity for vasodilation in the microvasculature occur early in the development of IR.

Exercise promotes brain plasticity partly by stimulating increases in mature brain-derived neurotrophic factor (mBDNF), but the role of the pro-BDNF isoform in the regulation of BDNF metabolism in humans is unknown. We quantified the expression of pro-BDNF and mBDNF in human skeletal muscle and plasma at rest, after acute exercise (+/- lactate infusion), and after fasting. Pro-BDNF and mBDNF were analyzed with immunoblotting, enzyme-linked immunosorbent assay, immunohistochemistry, and quantitative polymerase chain reaction. Pro-BDNF was consistently and clearly detected in skeletal muscle (40-250 pg mg^-1 dry muscle), whereas mBDNF was not. All methods showed a 4-fold greater pro-BDNF expression in type I muscle fibers compared to type II fibers. Exercise resulted in elevated plasma levels of mBDNF (55%) and pro-BDNF (20%), as well as muscle levels of pro-BDNF (∼10%, all P < 0.05). Lactate infusion during exercise induced a significantly greater increase in plasma mBDNF (115%, P < 0.05) compared to control (saline infusion), with no effect on pro-BDNF levels in plasma or muscle. A 3-day fast resulted in a small increase in plasma pro-BDNF (∼10%, P < 0.05), with no effect on mBDNF. Pro-BDNF is highly expressed in human skeletal muscle, particularly in type I fibers, and is increased after exercise. While exercising with higher lactate augmented levels of plasma mBDNF, exercise-mediated increases in circulating mBDNF likely derive partly from release and cleavage of pro-BDNF from skeletal muscle, and partly from neural and other tissues. These findings have implications for preclinical and clinical work related to a wide range of neurological disorders such as Alzheimer's, clinical depression, and amyotrophic lateral sclerosis.

The impact of high-intensity interval training (HIIT) on the central nervous system (CNS) in autoimmune neuroinflammation is not known. The aim of this study was to determine the direct effects of HIIT on the CNS and development of experimental autoimmune encephalomyelitis (EAE). Healthy mice were subjected to HIIT by treadmill running and the proteolipid protein (PLP) transfer EAE model was utilized. To examine neuroprotection, PLP-reactive lymph-node cells (LNCs) were transferred to HIIT and sedentary (SED) mice. To examine immunomodulation, PLP-reactive LNCs from HIIT and SED donor mice were transferred to naïve recipients and analyzed in vitro. HIIT in recipient mice did not affect the development of EAE following exposure to PLP-reactive LNCs. HIIT mice exhibited enhanced migration of systemic autoimmune cells into the CNS and increased demyelination. In contrast, EAE severity in recipient mice injected with PLP-reactive LNCs from HIIT donor mice was significantly diminished. The latter positive effect was associated with decreased migration of autoimmune cells into the CNS and inhibition of very late antigen (VLA)-4 expression in LNCs. Thus, the beneficial effect of HIIT on EAE development is attributed solely to systemic immunomodulatory effects, likely because of systemic inhibition of autoreactive cell migration and reduced VLA-4 integrin expression.

There is a debate on whether lipid-mediated insulin resistance derives from an increased or decreased capacity of muscle to oxidize fats. Here we examine the involvement of muscle fiber composition in the metabolic responses to a 3-day fast (starvation, which results in increases in plasma lipids and insulin resistance) in two groups of healthy young subjects: 1, area occupied by type I fibers = 61.0 ± 11.8%; 2, type I area = 36.0 ± 4.9% (P<0.001). Muscle biopsies and intravenous glucose tolerance tests were performed after an overnight fast and after starvation. Biopsies were analyzed for muscle fiber composition and mitochondrial respiration. Indices of glucose tolerance and insulin sensitivity were determined. Glucose tolerance was similar in both groups after an overnight fast and deteriorated to a similar degree in both groups after starvation. In contrast, whole-body insulin sensitivity decreased markedly after starvation in group 1 (P<0.01), whereas the decrease in group 2 was substantially smaller (P=0.06). Non-esterified fatty acids and β-hydroxybutyrate levels in plasma after an overnight fast were similar between groups and increased markedly and comparably in both groups after starvation, demonstrating similar degrees of lipid load. The capacity of permeabilized muscle fibers to oxidize lipids was significantly higher in group 1 vs. 2, whereas there was no significant difference in pyruvate oxidation between groups. The data demonstrate that loss of whole-body insulin sensitivity after short-term starvation is a function of muscle fiber composition and is associated with an elevated rather than a diminished capacity of muscle to oxidize lipids.