Effect of eccentric vs concentric exercise training on mitochondrial function. Marie-Eve Isner-Horobeti1,2 MD-PhD, Laurence Rasseneur 2 , Evelyne Lonsdorfer-Wolf2 MD-PhD, Stéphane Pascal Dufour2 PhD, Stéphane Doutreleau2 MD-PhD, Jamal Bouitbir2 PhD, Joffrey Zoll2 PhD, Sophia Kapchinsky3, Bernard Geny2 MD-PhD, Frédéric Nicolas Daussin5PhD, Yan Burelle4 PhD, Ruddy Richard6 MD-PhD. 1. Strasbourg University, Physical and Rehabilitation Medicine Department, Strasbourg University Rehabilitation Institute, France. 2. Strasbourg University, Fédération de Médecine Translationnelle de Strasbourg (FMTS), EA 3072 "Mitochondrie, stress oxydant et protection musculaire", France. 3. Mc Gill University Health Centre Department of Kinesiology, RECRU, Montreal Chest Institute, Montreal, Quebec, Canada. 4. University of Montreal, Department of Kinesiology, Montreal, Québec, Canada. 5. University of Lille North of France and UDSL, EA 4488: Physical Activity -MuscleHealth, Lille-France 6. Department of Sport Medicine and Functional Explorations, CHU Clermont-Ferrand, France.
Acknowledgements The authors would like to thank Dr Tanja Taivassalo PhD, Respiratory and Epidemiology Clinical Research Unit, Montreal Chest Institute - McGill University Health Center, Quebec, Canada, for her helpful suggestions regarding the content of the manuscript. We also thank the whole laboratory staff from Physiology and Functional Explorations Department and the Equipe d’Accueil 3072 for their daily technical support. The assistance of Dr Eric Sauleau has been very much appreciated in the statistical analyses of the data.
This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process which may lead to differences between this version and the Version of Record. Please cite this article as an ‘Accepted Article’, doi: 10.1002/mus.24215
Muscle & Nerve
Name and postal email addresses for the corresponding author: Marie-Eve Isner-Horobeti Université de Strasbourg, Institut Universitaire de Réadaptation Clémenceau-Strasbourg France F-67000 Strasbourg France. Tel: +33 3 88 21 16 39 Fax: +33 3 88 21 16 34 Email: [email protected]
Running title: Eccentric exercise and mitochondria
John Wiley 2& Sons, Inc.
Page 2 of 32
Page 3 of 32
Muscle & Nerve
Effect of eccentric vs concentric exercise training on mitochondrial function. Abstract
Introduction: The effect of eccentric (ECC) vs concentric (CON) training on metabolic
properties in skeletal muscle is understood poorly. We determined the responses in oxidative
capacity and mitochondrial H2O2 production after eccentric (ECC) vs concentric (CON) training performed at similar mechanical power.
Methods: 48 rats performed 5- or 20-day eccentric (ECC) or concentric (CON) training programs. Mitochondrial respiration, H2O2 production, citrate synthase activity (CS), and skeletal muscle damage were assessed in gastrocnemius (GAS), soleus (SOL) and vastus intermedius (VI) muscles. Results: Maximal mitochondrial respiration improved only after 20 days of concentric (CON) training in GAS and SOL. H2O2 production increased specifically after 20 days of eccentric ECC training in VI. Skeletal muscle damage occurred transiently in VI after 5 days of ECC training. Conclusions: Twenty days of ECC vs CON training performed at similar mechanical power output do not increase skeletal muscle oxidative capacities, but it elevates mitochondrial H2O2 production in VI, presumably linked to transient muscle damage. Keywords: eccentric exercise, skeletal muscle, mitochondria, reactive oxygen species, oxidative capacities
John Wiley 3& Sons, Inc.
Muscle & Nerve
Page 4 of 32
Introduction Since the pioneering publication by Abott and Bigland 1, eccentric (ECC) exercises have been investigated widely, mainly in the fields of muscle damage/regeneration and resistance exercise training for muscle strengthening. In addition to high levels of muscle force, eccentric (ECC) muscle work is also characterized by its very low metabolic cost and cardiorespiratory stress 2-4. Most of studies have focused on skeletal muscle force and structure, but the literature is scarce regarding the metabolic adaptations that occur in skeletal muscle after ECC exercise5-9. Conversely, the beneficial effect of regular concentric (CON) exercise training on skeletal muscle mitochondrial function is well documented in animals
. In view
of these results, it is of interest to examine further the physiological processes involved in skeletal muscle metabolic adaptations, i.e. mitochondrial function, after CON and ECC exercise training programs performed at similar mechanical power. Among the potential signals involved in the CON training-induced mitochondrial adaptations, cellular reactive oxygen species (ROS) might play a key role elevations, as evaluated by mitochondrial H2O2 production
, possibly via moderate
. Moreover, it has been clearly
shown that ECC exercise, performed at high intensity, can also induce greater skeletal muscle oxidative stress and damage than CON exercise 17. However, mitochondrial H2O2 production with low-load, endurance-like ECC exercise training and its role in the training-induced responses in skeletal muscle oxidative capacity is unclear. Our study was designed to compare CON vs ECC training programs performed at similar mechanical power (endurance-like training programs) to explore the interplay between training-induced adaptations in skeletal muscle oxidative capacities and mitochondrial H2O2 production. We hypothesized that CON vs ECC training programs promote different responses in skeletal muscle oxidative capacity associated with different levels of
John Wiley 4& Sons, Inc.
Page 5 of 32
Muscle & Nerve
mitochondrial H2O2 production. The main objectives of this work were: 1) to determine whether low-load, endurance-like ECC vs CON training improves skeletal muscle oxidative capacity, and 2) to establish the effects of ECC vs CON training programs on mitochondrial
Material and Methods Ethical approval Experiments were performed in accordance with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH publication number 85-23, revised 1996). The protocol was approved by the local animal care and use committee (agreement number AL/01/15/08/09). Experimental Design Forty eight male Wistar rats weighing 250-300 g (aged 6-8 weeks) were housed at 23°C in a light-controlled room and were given free access to food (commercial rat chow) and water ad libitum. All animals were assigned randomly to 5 experimental groups: a control (CTRL) group (n=8), 2 CON training groups (n=20), and 2 ECC training groups (n=20). The training rats were further assigned to either a 5-day (5D) or 20-day (20D) training protocol (n=10 per group in CON and ECC), using 5 training sessions per week. Rats were included progressively in the study in order to keep uniform age and body weight at time of death, even for control rats. The study was completed within 4 weeks, and the order of training was randomized over the entire intervention period. Experimental procedures Training Before inclusion in the study, all rats were familiarized initially with treadmill running for 2 consecutive days (2 min at 10m/min followed by 4 min at 25m/min with 0° slope) in a 5-lane treadmill apparatus (Panlab®, Barcelona, Spain).
John Wiley 5& Sons, Inc.
Muscle & Nerve
After the habituation period, animals were assigned randomly to 1 of the 5 experimental groups. All training sessions were performed with a slope of +15°(CON) or -15°(ECC), which was kept constant throughout the training period. Each training session started with a 2 min warm-up period at 10 m/min, after which the speed was increased to 20 m/min. For the 5D groups, rats exercised for 20 min during each session, whereas for the 20D groups, the exercise duration was increased progressively from 20 min (first week) to 40 min (last week). Vertical mechanical work and power of the training session were calculated as follows: Vertical mechanical work = v × (t ÷ 60) × sin α × m × g Vertical mechanical power = vertical mechanical work ÷ t Where: v = treadmill speed in m/s; t = training duration in s; α = treadmill slope in °; m = animal body weight in kg, and g = gravitational constant in m/s². Total training duration, running distance, total vertical distance, mechanical work, and power are listed for each experimental group in Table 1. Tissue collection Trained animals were anesthetized 12h following their last training session. Animals were anesthetized using 3% isoflurane (Aerrane®, Baxter, Maurepas, France ) and 1 L·min-1O2 in an induction box. Rats were removed from the box, placed on a heated pad and maintained under anesthesia using a face mask (1.5% isoflurane, 1 L·min-1 O2). The superficial white portion of the medial gastrocnemius (GAS) (glycolytic), soleus (SOL), and vastus intermedius (VI) (oxidative) muscles of the right leg were removed and used immediately to prepare skinned fibers to measure mitochondrial respiration and H2O2 release 18
. In accordance with previous studies19
, the 3 muscles were chosen because of their
differential recruitment in CON vs. ECC treadmill running; GAS and SOL muscles are
recruited preferentially during CON running, while VI is recruited specifically during
downhill running 21,22.
John Wiley 6& Sons, Inc.
Page 6 of 32
Page 7 of 32
Muscle & Nerve
The GAS, SOL, and VI from the left leg were frozen in liquid nitrogen and stored at -80°C for subsequent biochemical assays. After a blood sample was obtained (3 ml), animals were euthanized immediately with intracardiac injection of pentobarbital (50mg·kg-1). Metabolic measurements Blood lactate concentration Blood lactate concentration was determined using a lactate pro-LT device (Lactate Pro LT1710, Arkay®, Paris, France) in a sample (5µl) collected from the tip of the tail before the first training session and immediately (within 15s) after the first, fifth, and twentieth training sessions. Mitochondrial respiration in situ Mitochondrial respiratory function was assessed in a 3ml water-jacketed oximeter (Strathkelvin Instruments®, Glasgow, Scotland) equipped with a Clark-type electrode in a room air saturated oxygen concentration buffer. After recording baseline oxygen content in the chamber, a bundle of 1-2mg wet weight of permeabilized myofibers was placed into the chamber, which was sealed shut. Mitochondrial respiration in state 2 was assessed in the presence of the complex I substrate glutamate-malate (10:5mM, VGM). Then, mitochondrial respiration rates in state 3 (VGM-ADP and VADP-Succ) were assessed using the following sequential additions: ADP (2mM, VGM-ADP), the complex I blocker amytal (2mM), and the complex II substrate succinate (25mM, VADP-Succ). At the end of each assay, muscle fibers were harvested and dried for 15 min at 150°C. Rates of O2 consumption were expressed in µmoles O2.min-1.g-1 dry weight. All respiratory experiments were performed in duplicate for each animal. Citrate Synthase (CS) Activity Pieces of frozen muscle (5-10 mg wet weight) were homogenized with a vibrating microbead homogenizer (Mikro-dismembrators, Sartorius®, Palaiseau, France) in a ratio (W/V) of 1/30
John Wiley 7& Sons, Inc.
Muscle & Nerve
with a homogenized buffer (in mM: 250 sucrose, 40 KCl, 2 EGTA and 20 Tris-HCl). The homogenate was then supplemented with 0.1% Triton X-100 and incubated on ice for 1h. After centrifugation for 8 min at 10,000g, CS activity was determined through spectrophotometry (Versamax, Molecular Devices® Sunnyvale, USA) using a 96-well plate. CS activity was determined as described by Srere 23. Values were reported in mU· mg-1 of wet weight. Oxidative stress H2O2 production rates in permeabilized muscle fibers Mitochondrial H2O2 production rates were determined as described previously24 in permeabilized muscle fibers using Amplex Red reagent (Invitrogen). Change in fluorescence per second (∆F) was measured continuously with a spectrofluorometer (Fluoromax 4 Jobin Yvon®, Edison, New Jersey, USA) with magnetic stirring. After baseline, ∆F (reactants only) was established, and the reaction was initiated by addition of a permeabilized bundle of muscle fibers from GAS, SOL, and VI to 600 µl of buffer Z containing (in mM) 110 K-MES, 35 KCl, 1 EGTA, 10 K2HPO4, and 3 MgCl2 with Amplex Red (5µM) and HRP (0.3U/ml). The following additions were then made sequentially: the complex I substrate glutamatemalate (GM) (5:2.5 mM, H2O2
the complex I blocker amytal (2mM), the complex II
substrate succinate (5mM, Succ), ADP (2 mM, H2O2 Succ-ADP). The rates of H2O2 production were calculated after subtracting the background from a standard curve established in the same experimental conditions, except that fibers were absent. At the end of each assay, muscle fibers were harvested and dried for 15 min at 150°C. H2O2 measurements were expressed in minute per mg dry weight (pmoles· min-1· mg-1 dry weight). Mitochondrial free radical leak H2O2 production and O2 consumption were measured under similar conditions using permeabilized fibers from GAS, SOL, and VI. This allowed calculation of the fraction of
John Wiley 8& Sons, Inc.
Page 8 of 32
Page 9 of 32
Muscle & Nerve
electrons which reduced O2 to H2O2 in the respiratory chain (percentage of free radical leak) instead of reaching cytochrome oxidase to reduce O2 to water. Since 2 electrons are needed to reduce 1 mole of O2 to H2O2, whereas 4 electrons are transferred in the reduction of 1 mole of O2 to water, the percentage was calculated as the rate of H2O2 production divided by 2 times the rate of O2 consumption and multiplied by 100
. The free radical leak (FRL) was
calculated in the presence of glutamate+malate, amytal, succinate, and ADP. Muscle damage Plasma creatine kinase (CK) activity Blood samples were taken through cardiac puncture and centrifuged immediately. The catalytic activity of creatine kinase (CK) in plasma was determined at 37 °C according to the IFCC method on an Advia 1650 analyser using CK-NAC reagents (Siemens, France). Values were reported in U·L-1. Hematoxylin and eosin (H&E) staining Muscle damage was determined by the hematoxylin and eosin staining method; nuclei were stained blue, and cytoplasm was stained pink. Staining was randomized on GAS, SOL and VI. Muscles were cryosectioned at 8µm, and sections were air dried for 30 min. Slides were then placed in Harris Modified Hematoxylin solution for 5 min, followed by 3 changes of distilled water and 1 change of tap water. They were then placed in Eosin Y solution for 5 min, followed by 5 changes of 100% anhydrous ethyl alcohol. Slides were finally placed in 2 changes of Xylene solution for 5 min each and then mounted for imaging (The ArcturusXT™ system, Applied Biosystems/ Life Technologies).
Statistics Data were expressed as mean ± standard error of the mean (SEM). For statistical inference, we used Bayesian Gaussian mixed models, systematically including a subject random effect with a zero mean normal prior and a 0.5 precision. For lactate concentrations the model also
John Wiley 9& Sons, Inc.
Muscle & Nerve
included training effect (control, CON or ECC), time effect (rest, 1, 5, or 20 days) and their interactions. We then retrieved the posterior distributions of several differences between predicted means of lactate concentrations; at each time point, comparisons of concentrations between training groups and inside each training group, and comparisons of concentrations between time. For all other variables of interest but lactate concentration, training effect, muscle effect (GAS, SOL, or VI), and their interactions were included in the models. All analyses were performed using OpenBUGS software. We used Markov chains Monte Carlo techniques with a 50,000 iterations burn-in and a 25 thinning along the 50,000 following iterations for retrieving 2,000 values in posterior distributions. We used noninformative priors (zero mean and 0.01 precision for normal prior on effects and flat gamma prior distributions on the main precision of the variables of interest). With the aim of simplifying the reading of results, Bayesian P-values were added by mimicking a frequentist Wald test on posterior mean and deviation of the estimated effects.
Results Body mass and training properties No differences in body mass were observed after 5 days in both training modalities compared to CTRL (table 2; P=0.13 vs ECC and P=0.24 vs CON). In the 20D groups, body mass was 11% and 14% lower in ECC (P