A Single Bout of Ultra-Endurance Exercise Reveals Early Signs of Muscle Aging in Master Athletes

nd master endurance athletes exhibit similar physical performance and long-term muscle adaptation to aerobic exercise. Nevertheless, we hypothesized that the short-term plasticity of the skeletal muscle might be distinctly altered for master athletes when they are chal- lenged by a single bout of prolonged moderate-intensity exercise. Six middle-aged (37Y) and ve older (50Y) master highly-trained athletes performed a 24-h treadmill run (24TR).Vastus lateralismus- cle biopsies were collected before and after the run and assessed for proteomics, ber morphometry, intramyocellular lipid droplets (LD), mitochondrial oxidative activity, extracellular matrix (ECM), and micro-vascularisation. Before 24TR, muscle ber type morphometry, intramyocellular LD, oxidative activity, ECM and micro-vascularisation were similar between master and middle-aged runners. For 37Y runners, 24TR was associated with ECM thickening, increased capillary-to- ber interface, and an 89% depletion of LD

after the run and assessed for proteomics, ber morphometry, intramyocellular lipid droplets (LD), mitochondrial oxidative activity, extracellular matrix (ECM), and micro-vascularisation. Before 24TR, muscle ber type morphometry, intramyocellular LD, oxidative activity, ECM and micro-vascularisation were similar between master and middle-aged runners. For 37Y runners, 24TR was associated with ECM thickening, increased capillary-to- ber interface, and an 89% depletion of LD in type-I bers. In contrast, for 50Y runners, 24TR did not alter ECM and capillarization and poorly depleted LDs. Moreover, an impaired succinate dehydrogenase activity and functional class scoring of proteomes suggested reduced oxidative phosphorylation post-24TR exclusively in 50Y muscle. Collectively, our data support that middle-aged and master endurance athletes exhibit distinct transient plasticity in response to a single bout of ultra-endurance exercise, which may constitute early signs of muscle aging for master athletes. Keywords:aging; exercise; skeletal muscle; capillaries; lipid droplets; extracellular matrix 1. Introduction Skeletal muscle is central for not only coordinated movements and postural control but also for energy metabolism [1] and myokines secretion [2]. Skeletal muscle is the most abundant tissue in the adult body and a major storage site for amino acids (in the form of myo brillar proteins) and glucose (in the form of glycogen). Skeletal muscle thereby plays an important role in many physiological processes, including carbohydrate metabolism, fatty acid oxidation and thermogenesis [3]. A striking physiological characteristic of the skeletal muscle is also its capacity to progressively modulate local blood ow, substrate utilization, energy production and con- tractile proteins in response to long-term endurance training [4]. Such adaptative changes are important to blunt the homeostatic threats generated by exercise challenges and to promote optimal performance. The predominant fuels used during endurance exercise are fats and carbohydrates. Carbohydrates are stored as muscle and liver glycogen, and fats Int. J. Mol. Sci.2022,23, 3713.

Int. J. Mol. Sci.2022,23, 3713 2 of 16 are stored as reserves in subcutaneous and visceral adipose tissue. Smaller quantities of fats are present in circulating lipoprotein particles and in lipid droplets inside muscle bers. For endurance athletes, long-term adaptations of the skeletal muscle to aerobic training are re ected by gradual enhancements in capillary density and transport capacity of fatty acids and glucose, increases in metabolic enzyme activities and mitochondrial density, an accumulation of intramyocellular lipid droplets (LD), and adaptations of contractile pro- teins to slow type isoforms [5]. Changes in functional and morphological characteristics of the skeletal muscle also involve a remodeling of the extracellular matrix (ECM) embedding muscle bers [6]. Such adaptive changes of the skeletal muscle to long-term endurance training nonethe- less enable short-term plasticity. Even for highly-trained endurance athletes, a single bout of exercise still elicits transient changes in myocellular processes, re ecting metabolic and functional adaptation capacities. Accordingly, current evidence demonstrates short-term malleability of intramyocellular lipids [7], and remodeling of muscle mitochondrial and proteolytic pathways [8] in response to a single bout of aerobic exercise. Whether such short-term plasticity also occurs for the ECM and capillarization remains to be established. The loss of skeletal mass and function during the aging process (sarcopenia) is one of the most dramatic changes affecting the human body. Most previous studies investi- gating human sarcopenia relied on the comparison between young (e.g., 20–30 years) and old subjects (65–75 years). They provided major information about changes in ber mor- phology, oxidative metabolism, lipid droplets [9], capillarization, ECM brosis [10], ions and oxylipins homeostasis [11,12], and modulations of the muscle proteome [13,14] and transcriptome [15]. However, muscle mass varies over a lifetime; it reaches maximal levels in middle-aged adults (up to 40 years of age) and progressively declines then after [16]. Regular exercise is a primary preventive approach against age-related muscle wasting [4]. There is an increased participation of master athletes (i.e., > 40 years old) in endurance and ultra-endurance (lasting more than 6 h) events [17]. Given their level of performance, master athletes represent a model of successful aging. Aging,

40 years of age) and progressively declines then after [16]. Regular exercise is a primary preventive approach against age-related muscle wasting [4]. There is an increased participation of master athletes (i.e., > 40 years old) in endurance and ultra-endurance (lasting more than 6 h) events [17]. Given their level of performance, master athletes represent a model of successful aging. Aging, nonetheless, results in a decrease in endurance performance. Indeed, for both non-elite and elite endurance athletes, peak endurance performance is maintained in middle-aged adults until 35 years of age but is followed by a modest decrease for master runners until 50–60 years of age, with progressively steeper declines thereafter [18]. Despite no evidence for major differences in performance and in long-term muscle remodeling between middle-aged and master endurance athletes, we hypothesized that the short-term plasticity of the skeletal muscle might be altered for master athletes when they are challenged by a single bout of prolonged aerobic exercise. The present study focuses on 24-h treadmill ultra-endurance running (24TR), and we assessed if muscle proteome, ber morphometry, intramyocellular LD, oxidative activity, ECM and micro-vascularisation differ between master and middle-aged runners. 2. Results 2.1. Subject Clinical Characteristics Table involved in the present study. Body weight, body mass index, ultra-endurance experience, training volume and 24TR performance were similar between middle-aged (37Y) and master (50Y) athletes. The physiological variables, VO2max, velocity at VO2max, average speed sustained over the 24TR, velocity associated with the lactate in ection point (V 4mmol), and running economy (RE) at 8 km/h also did not differ between 37Y and 50Y runners.

Int. J. Mol. Sci.2022,23, 3713 3 of 16 Table 1.Subjects characteristics. 37Y (n= 6) 50Y ( n= 5) Age (yr) 37.0 0.4 50.3 2.3 * Body weight (kg) 72.3 2.4 76.3 5.2 BMI (kg/m 2 ) 1 22.9 0.6 24.6 1.4 Ultra-endurance experience (yr) 7.5 2.0 6.5 1.4 Training volume (km/week) 87.5 6.7 73.6 3.8 24TR Performance (km) 150.5 5.1 139.3 8.6 Effective running time (h) 20.5 0.6 21.4 0.6 VO 2max(mL/min/kg) 53.4 2.5 49.3 3.5 V VO2max(km/h) 18.8 0.6 17.1 0.7 Velocity during running (%V VO2max) 39.5 2.2 38.0 0.4 V 4mmol(%V VO2max) 88.0 1.1 88.0 2.5 RE (mL/min/kg) 28.5 0.9 28.6 0.9 1 BMI, body mass index; VVO2max, velocity associated with VO2max; V 4mmol, the velocity at blood lactate 4 mM; RE, running economy, i.e., VO2at 8 km/h. Data are presented as means SEM. *p< 0.05 vs. 37Y. 2.2. No Age-Related Effect of 24TR on Fiber Type Distribution The human skeletal muscles are of mixed ber-type composition, as they comprise slow-oxidative (type-I), fast-oxidative-glycolytic (type-IIA), and fast-glycolytic (type-IIX), together with hybrid bers. Myosin heavy chain speci c antibodies were used to assess contractile types in the human biopsies (FigureA), and on average, 232 (114–397) bers per individual pre-24TR (PRE) and post-24TR (POST) were analyzed for contractile type, cross-sectional area (CSA), perimeter and shape. As shown in FigureB, 37Y or 50Y runners exhibited similar PRE ber type distribution. No difference in ber type distribution was observed between PRE or POST muscles for either 37Y or 50Y runners. For all runners, type-I and then type-IIA bers were the most abundant, while type-IIX and hybrid bers (type I-IIA and type IIA-IIX) were scarce and not further investigated. 2.3. Age-Related Effect of 24TR on Fiber Type Morphometry While the proportion of the different fiber types remained constant, the mean fiber cross-sectional area (CSA) changed in response to 24TR, though only for 50Y subjects. As shown in Figure between 37Y and 50Y runners, and they remained similar POST for 37Y subjects. However, for 50Y subjects, type-I CSA tended to, and type-IIA CSA did decrease in response to 24TR. A frequency histogram of type-I and

fiber types remained constant, the mean fiber cross-sectional area (CSA) changed in response to 24TR, though only for 50Y subjects. As shown in Figure between 37Y and 50Y runners, and they remained similar POST for 37Y subjects. However, for 50Y subjects, type-I CSA tended to, and type-IIA CSA did decrease in response to 24TR. A frequency histogram of type-I and type-IIA fiber CSA confirmed that for 50Y athletes, 24TR was associated with a shift to fibers with a smaller CSA (Figure CSA could be associated with an altered shape, such as the flattening of fibers. To assess this point, the perimeter of each fiber was measured to calculate a shape factor. As shown in the supplementary figures, no change in type-I or type-IIA fiber perimeter (Figure S1A) and shape (Figure S1B) could be detected in response to 24TR for either 37Y or 50Y runners. 2.4. Age-Related Effect of 24TR on Mitochondrial Enzymatic Activities In addition to morphological and contractile properties of muscle bers, 24TR may affect intramyocellular organelles, such as mitochondria and lipid droplets. Succinate dehydrogenase (SDH) and cytochrome c oxidase (COX) are classically used to assess mitochondrial citric cycle and oxidative activities, respectively. For each runner PRE and POST, an average of 220 (125–308) bers were analyzed for SDH and COX activity. As shown in Figure S1C, COX activity remained unaffected by 24TR in types I and IIA bers in both 37Y and 50Y runners. However, SDH activity decreased after 24TR in both types I and IIA bers in 50Y athletes (Figure S1D). 2.5. Age-Related Effect of 24TR on Muscle Lipid Droplets Oil red O staining of neutral lipids was used to assess intramyocellular lipid droplets (LDs) and to calculate a ber type-speci c lipid content index (LI) (FigureA). PRE-24TR,

Int. J. Mol. Sci.2022,23, 3713 4 of 16 intramyocellular lipids accumulated more in slow-oxidative bers, as LI was 3–4 times higher in type-I than in type-IIA bers. There was no group difference before 24TR, as ber type-I and type-IIA speci c LIs were similar between 37Y and 50Y runners (FigureB). However, 24TR sharply affected intramyocellular lipids and this in an age-related fashion. For 37Y runners, 24TR strongly decreased type-I ber LI, but not type-IIA ber LI. In contrast, for 50Y runners, 24TR poorly altered muscle lipid droplets, as there was only a trend for 24TR to decrease type-I ber LI (p= 0.097). Figure 1. Fiber type-speci c distribution and morphometry.Vastus lateralisbiopsies were from 37Y and 50Y runners before (PRE) and after (POST) 24TR. (A) Representative images of cross-sections labeled for myosin heavy chain I (green) and counter-stained for laminin- 1 (red) to outline the bers; the scale bar represents 50 m. (B) The ber type proportion was measured PRE (black bars) and POST (gray bars) 24TR for middle-aged (37Y) and master (50Y) runners. Different letters indicate a signi cant difference (p< 0.05) between ber types. (C) Mean cross-section area (CSA) of type-I (left) and type-IIA (right) bers for 37Y and 50Y runners PRE- and POST-24TR; statistical interaction was p= 0.049 between time and group for type-IIA CSA. (D) Frequency histogram of CSA for type-I (left) and type-IIA (right) bers PRE- and POST-24TR for 50Y runners. *p< 0.05 between PRE and POST.

Int. J. Mol. Sci.2022,23, 3713 5 of 16 Figure 2. Intramyocellular lipid content. (A) Representative images of cross-sections stained with Oil red O for PRE and POST muscles of 37Y and 50Y runners; the scale bar represents 20 m. (B) Lipid content index (LI) expressed as a percentage of ber area occupied by lipid droplets (LD); statistical interaction wasp= 0.037 between time and group for type-I ber. (C) Mean number of LD per ber in type-I (left) and type-IIA ber (right) for PRE (black bars) and POST (gray bars) muscles of middle-aged (37Y) and master (50Y) runners; statistical interaction wasp= 0.045 between time and group for type-I ber. (D) Frequency histogram of droplet area for type-I bers in PRE (black bars) and POST (gray bars) muscles of 37Y (left) and 50Y (right) runners. *p< 0.05 between PRE and POST; #p< 0.05 between 37Y and 50Y runners; §p< 0.05 between type-I and type-IIA bers. Modi cations in ber-speci c LI could be due to changes in LD number and/or LD area. Image analysis was then used to assess 20,000–60,000 LDs per subject. These data indicated that 24TR decreased the LD number in type-I bers for both 37Y and 50Y runners (FigureC left). However, this decrease in LD number was twice as important for 37Y than for 50Y runners (89% vs. 37%, respectively). No 24TR-dependent change in the LD

Int. J. Mol. Sci.2022,23, 3713 6 of 16 number (FigureC right) or mean-area (not shown) was observed for type-IIA bers. The frequency histogram of the area of all individual LDs in type-I bers indicated that 24TR was associated with a shift to smaller droplets for 37Y, but not for 50Y runners (FigureD). 2.6. Age-Related Effect of 24TR on Muscle Extracellular Matrix Besides muscle bers, the functional assembly of the skeletal muscle is governed by the extracellular matrix (ECM) [19]. ECM was then investigated by using Sirius red that labels major ECM constituents (collagens I and III) [20] and using image analysis to distinguish the ECM endomysium from the perimysium (FigureA). The ECM is critical to maintaining structures and to transfer forces, and interestingly, the regression analysis that included all PRE-24TR subjects indicated that the endomysium area was negatively correlated with running economy (RE) (r= 0.87,p= 0.004). As shown in FigureB, 24TR increased the endomysium's thickness for 37Y but not for 50Y runners. Figure 3. Muscle extracellular matrix (ECM) endomysium. (A) Representative images of cross- sections stained with Sirius red indicating endomysium (arrow) and perimysium (arrowhead) for middle-aged (37Y) and master (50Y) runners PRE- and POST-24TR; the scale bar represents 50 m. (B) ECM endomysium mean thickness for 37Y and 50Y runners PRE (black bars) and POST (gray bars) 24TR.). Statistical interaction wasp= 0.031 between time and group. §p< 0.05 between type-I and type-IIA bers; *p< 0.05 between PRE and POST muscles. 2.7. Age-Related Effect of 24TR on Muscle Capillarization The ECM contains various stromal cells, including stem cells, immune cells, adipocytes and capillaries. The remodeling of the ECM that occurs with 24TR for 37Y runners might be associated with alterations in capillarization. To investigate this point, we further assessed blood capillaries using an anti-CD31 that recognizes a trans-membranous glycoprotein (PECAM-1) speci cally expressed by vascular endothelial cells [21]. On average, 232 (153–307) capillaries per participant were analyzed PRE- and POST-24TR (FigureA). As shown in Table S1, 37Y and 50Y runners exhibited similar capillarization indexes PRE- 24TR. Moreover, 24TR did not modify the capillary density (CD) and ber type-speci c indices,

capillaries using an anti-CD31 that recognizes a trans-membranous glycoprotein (PECAM-1) speci cally expressed by vascular endothelial cells [21]. On average, 232 (153–307) capillaries per participant were analyzed PRE- and POST-24TR (FigureA). As shown in Table S1, 37Y and 50Y runners exhibited similar capillarization indexes PRE- 24TR. Moreover, 24TR did not modify the capillary density (CD) and ber type-speci c indices, such as capillary-to- ber ratio (CAF), individual capillary-to- ber ratio (C/Fi), and capillary-to- ber perimeter (CFPE) in 37Y and 50Y muscles.

Int. J. Mol. Sci.2022,23, 3713 7 of 16 Figure 4. Muscle capillarization. (A) Representative images of cross-sections labeled with anti-CD31 antibody for 37Y and 50Y runners PRE- and POST-24TR; the scale bar represents 20 m. (B) The functional surface of exchange between capillaries and muscle bers (LC/PF) is expressed as a percentage of ber perimeter in contact with capillaries for middle-aged (37Y) and master (50Y) runners PRE (black bars) and POST (gray bars) 24TR. Statistical interaction between time and group wasp= 0.001 for type-I andp= 0.003 for type IIA bers. (C) Frequency histogram of the length of contact of each capillary (LCi) for type-I (left) and type-IIA bers (right) in PRE (black) and POST (gray) muscles of 37Y runners. *p< 0.05 between PRE and POST. However, 24TR did change the functional surface of exchange between capillaries and muscle bers (LC/PF), and this is an age-related fashion. Speci cally, for 37Y runners, LC/PF signi cantly increased POST compared to PRE, while no signi cant change occurred for 50Y runners. Moreover, increases in LC/PF of 37Y runners were not dependent on the type of muscle ber around which capillaries were observed, as it occurred for both type I and IIA bers (FigureB). The increase in LC/PF in the POST group of 37Y runners is mostly due to an increase in the total length of contact (LC) since PF was unchanged (Figure S1A). The frequency histogram of the length of contact for each capillary (LCi) con rmed that for 37Y runners, 24TR was associated with a shift to capillaries with longer capillary-to- ber contacts for both type-I and type-IIA bers (FigureC).

Int. J. Mol. Sci.2022,23, 3713 8 of 16 2.8. Global Indexes of Oxidative Metabolism, Lipid Droplets and Capillarization Further studies were then performed to assess coordinated regulations between capil- larization, intramyocellular lipids, and mitochondrial energy metabolisms at a ber type level. As our studies were performed with serial cross-sections, many bers could be matched between analyses, and for each ber, we computed global indices of capillariza- tion (LC, the total length of contact with capillaries), intramyocellular lipids (åLD, the sum of LD area) and mitochondrial oxidative capacity (Int-SDH, the spatially integrated SDH activity), and assessed ber-to- ber inter-correlations. As shown in FigureA,C, for both PRE-24TR age groups, all three indices were inter- related for type-I bers, indicating substantial coordination between vascular, lipid, and mitochondrial processes. However, correlations with the lipid index were less preserved in type-IIA bers that rely less on lipid metabolism. Figure 5. Coordination between microvascular, lipid, and mitochondrial processes. Linear Pearson's correlations between global indices of capillarization (LC, the total length of contact with capillaries), intramyocellular lipids (åLD area, the sum of LD area), and mitochondrial oxidative capacity (Int- SDH) for type I (upper left) and type IIA (lower right) bers of (A) PRE and (B) POST vastus lateralis of 37Y (top) and 50Y (bottom) runners. Correlations coef cients (r) are indicated when p< 0.05. (C) Examples of linear regressions between Int-SDH,åLD area, and LC for 37Y (top) and 50Y (bottom) runners. Most correlations vanished POST-24TR, suggesting a loss in coordination between vascu- lar, lipid, and mitochondrial processes (Figure 50Y muscles, as they maintained correlations between vascular and mitochondrial indexes. 2.9. Age-Related Effect of 24TR on the Muscle Proteome Label-free quantitative protein pro ling was nally used to compare the muscle proteomes of 37Y and 50Y runners PRE- and POST-24TR. A total number of 626 proteins were identi ed and quantitation analyses were performed for proteins detected in all samples. PRE-24TR, 419 proteins were compared between 37Y and 50Y athletes, and only 4 proteins were differentially expressed and downregulated in 50Y muscles (Table S2): cytoplasmic isoform 3 of malate dehydrogenase (MDH1) is important for the NADH/NAD+ shuttle

This study investigates muscle aging signs in master athletes after ultra-endurance exercise.