Resveratrol Enhances Exercise-Induced Cellular and Functional Adaptations of Skeletal Muscle in Older Men and Women

Abstract

Older men (n = 12) and women (n = 18) 65–80 years of age completed 12 weeks of exercise and took either a placebo or resveratrol (RSV) (500 mg/d) to test the hypothesis that RSV treatment combined with exercise would increase mitochondrial density, muscle fatigue resistance, and cardiovascular function more than exercise alone. Contrary to our hypothesis, aerobic and resistance exercise coupled with RSV treatment did not reduce cardiovascular risk further than exercise alone. However, exercise added to RSV treatment improved the indices of mitochondrial density, and muscle fatigue resistance more than placebo and exercise treatments. In addition, subjects that were treated with RSV had an increase in knee extensor muscle peak torque (8%), average peak torque (14%), and power (14%) after training, whereas exercise did not increase these parameters in the placebo-treated older subjects. Furthermore, exercise combined with RSV significantly improved mean fiber area and total myonuclei by 45.3% and 20%, respectively, in muscle fibers from the vastus lateralis of older subjects. Together, these data indicate a novel anabolic role of RSV in exercise-induced adaptations of older persons and this suggests that RSV combined with exercise might provide a better approach for reversing sarcopenia than exercise alone.

Sarcopenia is predictive of falls and is associated with periods of hospital-associated immobilization (1). This is noteworthy, as aging muscles display an impaired or failed recovery following immobilization-induced muscle atrophy (2,3). Several potential countermeasures to unloading-induced atrophy have been tested, but most have only been partially successful. For example, taurine was shown to prevent the slow to fast myosin shift with unloading (4), but it did not prevent the loss of muscle mass. Insulin growth factor-I overexpression during hindlimb unloading in rodents induced some potential molecular signaling improvements, especially the atrogen Muscle RING-finger protein-1 (MuRF1), but there was no overall change in muscle protein loss or fiber type transition with this treatment (5). In addition, β-hydroxy-β-methylbutyrate, a leucine metabolite, provided modest protection against skeletal loss of muscle mass during disuse in sarcopenic rats (6). We have found that green tea extract increased satellite cell proliferation and differentiation in rat plantaris and soleus muscles during recovery from hindlimb suspension as compared to vehicle-treated muscles and decreased oxidative stress and the abundance of the Bcl-2–associated X protein (Bax), a proapoptotic protein, yet this did not further improve muscle recovery in reloaded muscles over placebo (PL) treatments (7).

In vitro studies have shown that resveratrol (3,5,4′-trihydroxystilbene; RSV) increases protein synthesis (8), inhibits protein degradation, and attenuates atrophy of skeletal muscle fibers (9–12). A high dose of RSV in vivo (400 mg/kg/d) was reported to attenuate muscle fiber atrophy following hindlimb suspension (13) in rodents. We have found that a low dose (12.5 mg/kg/d) of RSV (14) had a trend (p = .06) to blunt the loss of muscle mass during hindlimb suspension in old rats but it did not improve recovery or increase the number of satellite cells or muscle mass after disuse in old animals (15). Nevertheless, other data failed to identify RSV-mediated improvements in metabolic function in rodents (16). Furthermore, while exercise adaptations in the elderly are typically attenuated as compared to responses in young adults (17), it is not known if RSV coupled to exercise provides a stronger stimulus for reducing or reversing sarcopenia in humans (18) than exercise alone.

Thus, the purpose of this study was to determine if 12 weeks of exercise would improve muscle fatigue resistance more than exercise alone in older men and women. We tested the hypothesis that RSV treatment would increase mitochondrial density, which would decrease muscle fatigue resistance, and improve indices for cardiovascular risk as compared to exercise alone in older men and women.

Methods

Detailed methods are provided in Supplementary Methods.

Subjects

Healthy older men (n = 12) and women (n = 18) completed this study. The subjects were randomly assigned in a balanced design to include the same number of males (n = 6) and females (n = 9) in each experimental group. A resting 12-lead ECG and a symptom-limited maximal graded exercise test (GXT) were conducted to rule out heart disease and to establish proper cardiovascular function. The study was approved by the West Virginia University Institutional Review Board for Human Research, and all participants signed informed consent prior to participating in the study.

Body Mass Index and Body Fat

Trained personnel measured height, weight, and waist and hip circumference. Body composition was estimated via air displacement plethysmography (BOD POD, COSMED, Inc.). The body mass index was calculated as weight (kg) divided by height (m) squared.

Blood Indices

Venous blood from the antecubital vein was obtained after a 12-hour overnight fast. Fasted post-training samples were collected a minimum of 48 hours after the last exercise session. Plasma obtained from blood sampling was analyzed at West Virginia University Hospital’s analytical laboratory.

Assessment of Maximal Oxygen Uptake

Oxygen consumed (VO₂), carbon dioxide produced (VCO₂), respiratory exchange ratio = VCO₂/VO₂, ventilation-expired (VE), ventilation-inhaled (VI), ventilatory threshold (VT), ventilatory equivalent for CO₂ (VE/VCO₂), and O₂ (VE/VO₂) were measured using a validated (19), twenty-second averaging breath-by-breath analysis system (TrueOne 2400; ParvoMedics, Sandy, UT).

Muscle Function and Fatigue

Maximal torque

Measures of muscle strength and fatigue were obtained using a velocity-dependent Biodex System 3 dynamometer (Biodex, Inc., Shirley, NY) because function and fatigue are compromised in aging (20). Maximal knee extension torque was evaluated at 60, and 240°/s. Data were calculated for peak torque (Newton-meters), total work in one repetition (Joules), total work over seven repetitions (Joules), work fatigue (%), average power (Watts), and average peak torque over seven repetitions (Newton-meters). Subjects were not provided with encouragement but the investigator counted each repetition loudly during the test.

Isokinetic fatigue

Fatigue data were calculated from the decline in torque over 32 isokinetic contractions at 240°/s on the Biodex unit. This represents a moderate load for older subjects (21). Muscular fatigue was expressed as a fatigue index (fatigue index = [1 − (initial torque − final torque)/initial torque]). These are similar to methods used previously in our labs (22,23).

Muscle Samples

Needle biopsies of the vastus lateralis muscle were collected under local anesthesia with 2% lidocaine using a Bergstrom needle as described previously (17,24,25). The subjects refrained from strenuous activity for 36 hours prior to the biopsy and for 48 hours after the biopsy. Visible blood and connective tissue were removed from the biopsy sample and the tissue was frozen in isopentane cooled to the temperature of liquid nitrogen, and stored at −80°C until analysis.

Muscle Myosin Composition and Morphology

Frozen tissue sections were cut at −20°C and placed on glass slides, air-dried, and incubated overnight at 4°C with antibodies from the Hybridoma Bank against: myosin heavy chain (MHC) IIA (SC-71), MHCI (BAF8S), and MHC IIX (6H1). Fiber area and the minimum Feret diameter was determined from the digital images and quantified by Image J software (NIH) from a minimum of 500 fibers. All analyses were performed with the investigator blinded to treatment (RSV or PL) using established methods (26,27).

Muscle Ultrastructure

Fifty fibers were randomly selected from five specimen blocks for ultrastructural analyses. The electron micrographs were analyzed by a stereological 168-point short-line test grid (28). Morphometric measures were made of the proportion of fibers occupied by the mitochondria. The stereological measures were made without the investigator’s knowledge of the group that was analyzed.

Mitochondria Biogenesis

The relative number of copies of human mitochondrial DNA (mtDNA) was quantified by real-time PCR using a Human mtDNA Monitoring Primer Set (Takara/Clontech). Nuclear DNA (nDNA) content was used as a standard. Two mtDNA targets (ND1, ND5) and two nDNA targets (SLCO2B1, SERPINA1) were used for the detection assay. The manufacturer has designed the primers to avoid amplification of pseudogenes. DNA was purified from skeletal muscle biopsies (Qiagen) and amplified with the primers according to the manufacture’s recommendation. The CT value for each mitochondria and genomic DNA primer was determined and 2^(−ΔΔCT) for the primer pairs and the mtDNA copy number was determined.

PCR Arrays for Mitochondrial Metabolism

RNA was purified from muscle biopsies and mRNA for mitochondrial genes was assessed using the Human Mitochondria RT² Profiler PCR Array (Qiagen). ΔCT was calculated for each gene and the housekeeping genes and the fold change between pre-exercised and post-exercised groups for each gene was calculated using 2^(−ΔΔCT). The fold differences of 2^(−ΔΔCT) between exercise and PL groups were reported as log base 2 fold differences between genes in pre- to post-training conditions, or between pre-exercise and between post-exercise conditions for PL and RSV treatments.

Satellite Cell and Myonuclei

Antibodies against paired box transcription factor 7 (Pax7; Developmental Studies Hybridoma Bank, Iowa city, IA) was used to identify satellite cells and laminin (Developmental Studies Hybridoma Bank) was used to identify the boundaries of muscle fibers. The sections were counter stained with 4′,6-diamidino-2-phenylindole (DAPI) to visualize all of the nuclei in each tissue section (Vector Laboratories, Burlingame, CA). The number of nuclei that were positive for Pax7 and the total number of DAPI-positive nuclei were counted from digital confocal images (Zeiss) by an investigator who was blinded to the experimental groups.

DNA Oxidative Damage

Fixed tissue sections were incubated with 8-hydroxyguanosine (8-OHdG; Santa Cruz) an indication of oxidative damage to skeletal muscle nuclear DNA. To confirm that the nuclei were inside the boundary of the muscle fibers, the tissue sections were also incubated with antibodies against dystrophin and laminin (Developmental Studies Hybridoma Bank). The sections were counter stained with DAPI and then examined by an investigator who was blinded to the experimental groups.

Exercise Training and RSV Treatment

As exercise provides health benefits for older subjects, all subjects were exercised for 12 weeks in this study. Resistance training was given to provide an anabolic effect for muscle growth. Aerobic training was added to improve mitochondrial biogenesis and function, as mitochondria are central to aging-associated signaling for cell death including apoptosis (7,29,30) and autophagy (31,32).

The subjects were randomly assigned in a double-blind manner into either a 12-week exercise group that took 500 mg RSV/d in capsule form (RSV; 80% female) or a 12-week exercise control group that took 500 mg/d of a corn starch PL in capsule form (PL; 80% female). The code identifying the PL and RSV groups was not broken until the data analysis was completed, and then the subjects were placed in the appropriate groups for data analyses.

Statistical Analysis

Pre-exercise variables were compared between groups with a one-way analysis of variance (ANOVA) and post-training data were compared with a two-way repeated measures ANOVA as group (PL vs RSV) and condition (pre- vs post-exercise). A Tukey’s post hoc test was conducted if a main effect or an interaction effect was identified by the ANOVA. A Condition (ie, training) × Group interaction indicates that the response to training differed between the PL and RSV groups. The data are reported as means ± SEM. P≤.05 was defined as statistically significant.

Results

Effects of Exercise Training on Body Composition and Metabolic Biomarkers

One subject in the PL group and four subjects in the RSV group reported a lack of time to commit to training and withdrew from the study and additional subjects were recruited to the subject groups so that 15 subjects completed the study in both PL and RSV groups.

The characteristics of the subjects that completed the study are shown by treatment group in Table 1. Data are separated by sex for each group and are given in Supplementary Table 1. Both PL and RSV groups had similar characteristics at the beginning of the study. Three-way ANOVA analysis showed that there were subject group differences based on the sex of the subjects with men having greater body weight (p < .001), body fat (p = .026), and lean body mass (p < .001) than women, but there were no differences in these variables between women in the two groups or men in the PL or RSV groups (Supplementary Table 1); however, there were no differences between groups when males and females were averaged within each group.

Table 1.

Effects of Exercise Training and Resveratrol on Body Composition, and Metabolic Biomarkers in Older Subjects

	Grouped (male + female)				p Values
	Pre-PL	Post-PL	Pre-RSV	Post-RSV	Condition (pre, post)	Group (PL, RSV)	Condition × Group Interaction
Age (y)	67.9 ± 1.1	67.9 ± 1.1	68.1 ± 1.1	68.1 ± 1.1	1	.513	1
Height (cm)	66.9 ± 1.0	66.9 ± 1.0	65.9 ± 1.0	65.9 ± 1.1	.645	<.001	<.001
Weight (kg)	74.1 ± 3.7	73.2 ± 4.3	73.0 ± 3.8	73.6 ± 4.1	.935	.860	.454
Lean mass (kg)	50.1 ± 2.4	49.5 ± 4.3	48.5 ± 3.1	46.5 ± 3.2	.565	.464	.218
Body fat (%)	32.6 ± 2.1	32.1 ± 2.1	35.2 ± 2.0	33.4 ± 2.0	.572	.117	.111
BMI (kg/m2)	26.2 ± 1.3	25.8 ± 1.4	25.3 ± 1.3	26.4 ± 1.2	.835	.802	.397
Hemoglobin A1C (% of Hb)	6.0 ± 0.3	5.3 ± 0.2	5.4 ± 0.3	5.3 ± 0.2	.182	.124	.18
Triglycerides (mg/dL)	132.8 ± 10.7	124.7 ± 4.5*	117.1 ± 23.7	109.7 ± 11.9*	.002	.112	.942
Total cholesterol (mg/dL)	228.9 ± 10.3	219.6 ± 10.1	214.9 ± 9.2	207.8 ± 9.3	.401	.196	.907
HDL (mg/dL)	44.1 ± 3.7	53.8 ± 4.7	51.1 ± 3.3	53.4 ± 3.4	.094	.363	.281
LDL (mg/dL)	147.1 ± 7.8	122.5 ± 7.1*	131.6 ± 7.2	114.4 ± 7.9*	.007	.120	.624
Insulin (µU/mL)	10.4 ± 1.1	7.9 ± 0.9*	10.8 ± 1.2	8.0 ± 0.8*	.002	.827	.795

Note: BMI = body mass index; HDL = high-density lipoprotein; LDL = low-density lipoprotein; PL = placebo; RSV = resveratrol. Older male and female subjects were given a PL or 500 mg of RSV and exercised for 12 wk. The data are reported for blood triglycerides, total cholesterol, HDL, LDL, as well as body weight, body fat, lean body mass, and oxygen uptake (VO₂ max) (n = 15 PL; n = 15 RSV; 60% of each group was comprised of women: 9 women and 6 men were randomly assigned to each group in a balanced design). *p < .05, PL vs RSV.

Exercise resulted in significant reductions in total triglycerides, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, and serum insulin, but RSV did not change these metabolic variables further (Table 1). This exercise effect and lack of RSV effect was similar in both men and women (Supplementary Table 1). Neither the exercise nor the RSV treatment significantly altered bodyweight, body fat, body mass index, or lean body mass in older subjects (Table 1).

Effects of RSV on Improvements in Oxygen Uptake to 12 Weeks of Exercise

Cardiovascular adaptations to exercise training are best reflected by changes in absolute maximal oxygen uptake as relative measures of oxygen uptake can be affected by changes in bodyweight. Before exercise, absolute VO₂ max was similar in PL- and RSV-treated groups. The ANOVA analysis revealed a significant main effect (p < .05) in absolute VO₂ max on condition (pre- to post-exercise). While the post hoc analyses to identify the source of the main effect showed that the absolute VO₂ max was significantly improved by exercise in only the RSV-treated subjects (p = .028), there appeared to be no central improvement in absolute VO₂ max in the exercise-trained PL group (p = .286). There was no statistically significant interaction (p = .09) between Group (PL vs RSV) × Condition (pre- vs post-exercise). The absolute VO₂ max data are shown in Figure 1A.

Figure 1.

(A) Maximal oxygen uptake (VO₂ max) was determined before (Pre) and after (Post) 12 wk of exercise in subjects who consumed a placebo (PL) or resveratrol (RSV) throughout the training period. The data are expressed as absolute oxygen consumed. Maximal oxygen consumption increased in RSV-treated subjects but not in PL-treated subjects after 12 wk of exercise. (B) Representative electron micrographs of vastus lateral muscle sections prepared for electron microscopy before training (i: scale bar = 500 µm) and after 12 wk of exercise training in PL-treated (ii: scale bar = 500 µm) and RSV-treated (iii: scale bar = 500 µm) samples. A lower magnification of post-exercise RSV-treated muscle (iv: scale bar = 800 µm). Examples of apparently morphologically normal mitochondria are shown with the red arrows, and apparently, abnormal mitochondria are shown with blue arrows. (C) mRNA was extracted from biopsies of the vastus lateralis muscles before and after exercise. Subjects took either a PL or RSV for 12 wk and exercised throughout the study. Gene expression was analyzed by PCR array and the global data set were evaluated from 84 genes. Quantitative analysis of genes that were differentially increased in muscles of RSV- vs PL-treated older subjects. The Heat Map provides a visualization of the fold changes in mitochondrial log2 fold gene expression between the selected groups for every gene with the layout for the Heat Map as follows:

Layout	01	02	03	04	05	06	07	08	09	10	11	12
A	ATP5A	ATP5B	ATP51	ATP5F	ATP51	ATP5	ATP53	ATP5	ATP5I	ATP5	ATP5J	ATP5
B	ATP5O	COX4I1	COX5A	COX5B	COX6A1	COX6A2	COX6B1	COX6C	COX7A2	COX7A2	COX7B	COX8A
C	CYC1	NDUFA1	NDUFA1	NDUFA11	NDUFA3	NDUFA3	NDUFA4	NDUFA5	NDUFA6	NDUFA8	NDUFAB1	NDUFB1
D	NDUFB2	NDUFB3	NDUFB4	NDUFB5	NDUFB6	NDUFB7	NDUFB8	NDUFB9	NDUFC1	NDUFC2	NDUFS1	NDUFS2
E	NDUFS3	NDUFS4	NDUFS5	NDUFS6	NDUFS7	NDUFS8	NDUFV1	NDUFV2	NDUFV3	PPA1	SDHA	SDHB
F	SDHC	SDHD	UQCR11	UQCRC1	UQCRC2	UQCRFS1	UQCRH	UQCRQ	ARRDC3	ASB1	CYB561D	DNAJB1
G	EDN1	GADD45B	HSPA1A	HSPA1B	LRP5L	MitoH1	MitoH2_	MitoH2_	MitoH2_4	MitoH2_	RNU11S	SLC25A25

The legend for the color intensity is shown below each heat map.

(D) Isokinetic measures of knee extensor muscle torque (Newton-meters), average peak torque, and average power (watts) were obtained on a Biodex at 60°/s. The data represent mean data from the subjects in each group. *p < .05, pre-training vs post-training. **p < .05, RSV vs PL treatment. Pre-PL, before training in the placebo group; Post-PL, after 12 wk of training in the placebo group. Pre-RSV, before training in the resveratrol group; Post-RSV, resveratrol-treated muscles after 12 wk of exercise. *p < .05, pre-training vs post-training.

Similarly, the ANOVA analysis showed that there was a significant main effect for condition (pre- vs post- exercise differences; p < .05) in the relative maximal oxygen uptake when the subjects were considered as a single group. There was not a statistically significant interaction between Group (PL vs RSV) × Condition (pre- vs post-exercise) (p = .71). Post hoc analyses identified a significant improvement in relative VO₂ max in RSV-treated subjects (24.0 ± 1.3 mL/kg/min vs 28.7 ± 1.7 mL/kg/min, p < .05) but relative VO₂ max was unchanged in PL-treated subjects who exercised (23.7 ± 1.3 mL/kg/min vs 26.1 ± 1.4 mL/kg/min, p = .13). This improvement in relative maximal oxygen uptake likely reflects a central adaptation in RSV-treated subjects because bodyweight did not differ before or after exercise in RSV-treated subjects, and bodyweight in PL and RSV groups were not changed by treatment or exercise (Table 1).

Mitochondrial Density

There was a significant main effect of condition (pre- vs post-exercise) for mitochondrial volume density (p < .05). Without considering the apparent health of the mitochondria, quantification of the volume density of mitochondria from longitudinal sections was 9.3% ± 1.3% before exercise, and 13.5% ± 2.3% post-training in muscle samples from the PL group but post hoc analyses following the identification of the significant main effect of condition showed that this did not represent a statistically significant improvement in mitochondrial volume density (p = .093). On the other hand, post hoc analysis of the significant main effect of condition revealed that mitochondria volume density in the vastus lateralis significantly increased from 9.8% ± 1.7% pre-training to 16.8% ± 2.5% (p < .05) in the RSV group post-training. The fiber type of the samples could not be determined from the electron micrographs. Of interest, apparently morphologically normal (red arrows) and abnormal mitochondria (blue arrows) could be identified before and after exercise in PL- and RSV-treated samples (Figure 1B), although these were not quantified. Nevertheless, it was apparent that mitochondria with normal morphological characteristics (eg, normal cristae and mitochondria boundaries etc.) were abundant in the RSV-treated longitudinal muscle samples, but abnormal mitochondria (broken cristae etc.) were readily obvious in the post-trained PL-treated samples. However, the two-way ANOVA did not identify a significant effect for group (PL vs RSV) and no significant interaction effect between group and condition were found.

Signaling for Mitochondrial Biogenesis and Mitochondrial Function

Total RNA was isolated from muscle biopsies before and after exercise after PL and RSV treatments. The RNA was used for mitochondrial microarray analyses. The fold difference values of altered mRNAs are presented as heat maps of log2 transformed fold change from pre- to post-exercise (Figure 1C). Scatter plots showing normalized log 10-transformed signal intensities for each comparison show global gene fold differences between PL or RSV groups before and after exercise training (Supplementary Figures 1–4).

RSV promoted proapoptosis signaling

The expression of apoptotic genes BCL2 antagonist/killer 1 (BAK1), BH3-interacting domain death agonist (BID), cyclin-dependent kinase inhibitor 2A (CDKN2A), phorbol-12-myristate-13-acetate-induced protein 1 (PMAIP1), strafin (SFN), dynamin 1-like protein (DNM1L), and prefoldin-like chaperone (UXT) had a 3- to 11-fold greater pre- to post-exercise change in RSV as compared to PL muscle samples (Table 2; Supplementary Results).

Table 2.

Normalized Mitochondrial mRNA Log2 Fold Differences With 2 Fold or Greater Increases in Resveratrol vs Placebo Treatment Pre–Post Exercise

	Gene Symbol	Pre-PL vs Pre-RSV	Post-PL vs Post-RSV	Pre- vs Post-Exercise PL	Pre- vs Post-Exercise RSV
Apoptosis
	BAK1	3.14122	3.92495	5.70290	12.38249
	BCL2	3.29122	−0.81512	3.69288	5.78463
	BCL2L1 (Bcl-xl)	0.74122	1.06497	6.82299	8.24248
	BID	2.59122	1.06497	9.78299	13.05249
	CDKN2A	3.91122	2.08488	−1.01711	4.59464
	PMAIP1	2.07122	2.82488	4.51289	9.02448
	SFN	2.61122	3.09494	7.76291	13.08249
	DNM1L	6.75122	0.71488	1.61289	8.69461
	UXT	11.55122	−0.18512	0.21289	11.19463
Metabolism
	AIP	4.12122	5.84495	2.16289	11.74249
	COX18	3.47122	3.81488	−4.13711	2.76463
	CPT1B	3.93122	1.06497	5.24299	9.85249
	UCP2	5.25122	1.06497	6.70299	12.63249
	UCP3	4.46122	1.06497	8.67299	13.81249
	IMMP2L	1.21122	3.73494	6.88285	11.44249
	LRPPRC	0.53122	7.08495	3.52289	10.75249
	HPRT1	7.21122	8.89495	2.94289	18.66249
Morphology
	MFN1	6.70122	0.59488	0.62289	7.53461
	MFN2	1.80122	1.06497	6.09299	8.57249
	MSTO1	2.54122	1.06497	9.10298	12.32249
Carrier/import
	SLC25A10	4.18122	1.47485	7.32289	12.59430
	SLC25A22	5.75121	1.06497	4.48300	10.91249
	SLC25A27	5.91122	3.06488	0.99289	9.58459
	SLC25A31	3.33122	1.21488	5.03291	9.19458
	SLC25A37	9.65122	−0.90512	−1.67711	6.68463
	SLC25A5	4.92122	1.06497	10.73299	16.33249
	STARD3	1.43122	2.92487	5.41288	9.38450
	TIMM10	2.23122	4.58495	6.05287	12.48249
	TIMM10B	5.02122	6.80495	1.23289	12.67249
	TIMM44	5.09122	0.64488	4.82291	10.17451
	TIMM50	1.59122	1.06497	9.03299	11.30249
	TIMM8A	4.77122	8.02495	2.83289	15.24249
	TOMM34	2.05122	3.44488	3.57289	8.68457
	TOMM40L	5.08122	1.06497	9.45298	15.21249

Note: PL = placebo; RSV = resveratrol.

RSV promoted antiapoptotic changes

Two antiapoptotic proteins, B-cell lymphoma 2 (BCL2) and BCL2-like 1 (Bcl-xL), had approximately twofold greater gene expression differences pre- to post-exercise in RSV as compared to PL treatments.

RSV regulation of mitochondrial metabolic genes

Several mitochondrial metabolic genes were expressed to higher levels from pre- to post-exercise in RSV versus PL subjects. This included carnitine/choline acetyltransferase (CPT1B, 4-fold), COX18 (3-fold), uncoupling protein 2 (UCP2, 6-fold), uncoupling protein 3 (UCP3, 5-fold), and hypoxanthine phosphoribosyltransferase 1 (HPRT1, 16-fold) (Table 2).

RSV regulation of mitochondrial morphology

Mitofusin 1 (MFN1), mitofusin 2 (MFN2), and morphology regulator (MSTO1) had approximately 7-, 2.5-, and 3-fold greater gene level pre- to post-exercise in muscles from RSV versus PL subjects, respectively (Table 2; Supplementary Results).

Carrier/shuttle/import proteins

Compared to PL, RSV had two- to fourfold greater pre- to post-exercise increases in gene expression for solute carrier family 25 (SLC25) members, including SLC25A10, SLC25A22, SLC25A27, SLC25A31, and SLC25A37, and StAR-related lipid transfer domain containing 3 (STARD3) (Table 2). Similarly, genes for members of the translocase of inner mitochondrial membrane (TIMM) family transport protein, including TIMM10, TIMM10B, TIMM44, TIMM50, and TIMM8A, and translocase of outer mitochondrial membrane (TOMM) family members, TOMM34 and TOMM40L, inner mitochondrial membrane peptidase subunit 2 (IMMP2L), and leucine-rich pentatricopeptide repeat containing (LRPPRC) protein had 3- to 13-fold greater gene differences in muscles of RSV-treated subjects as compared to PL-treated subjects (Table 2; Supplementary Results).

Genes That Had a Lower Increase With RSV versus PL Treatments

The genes in which there was a twofold or more lower in RSV-treated versus PL-treated genes are shown in Table 3 and Figure 1C.

Table 3.

Normalized Mitochondrial mRNA Log2 Fold Differences That Were 2 Fold or More Lower in Resveratrol vs Placebo Treatment Pre–Post Exercise

	Gene Symbol	Pre-PL vs Pre-RSV	Post-PL vs Post-RSV	Pre- vs Post-Exercise PL	Pre- vs Post-Exercise RSV
Apoptosis/autophagy	SH3GLB1	1.66122	−5.75510	8.21299	3.73460
	RHOT1	3.78122	−6.12510	7.06298	4.33465
	RHOT2	−1.98879	−4.79518	8.84298	1.67450
Metabolism	COX10	−1.65878	−1.94511	8.17299	4.18469
	CPT2	1.64122	−3.62510	7.15299	4.78477
	GRPEL1	3.05122	−5.59510	10.51299	7.58466
	MIPEP	2.07122	−4.56512	2.62289	−0.25537
Carrier/import	SLC25A13	−4.45878	1.19494	9.86300	6.21252
	SLC25A14	−4.88878	−0.43512	8.10289	2.39473
	SLC25A15	−1.26878	−2.49510	9.75299	5.60457
	SLC25A19	−2.61878	0.62488	5.39290	3.01469
	SLC25A20	−7.17857	1.06497	8.15298	−0.34365
	SLC25A21	−3.56879	−1.07510	7.55298	2.52434
	SLC25A30	−4.12879	1.06497	9.83299	6.38248
	SLC25A4	−0.52878	−0.25512	5.05289	3.88463
	MTX2	−8.16857	1.06497	9.14299	−0.34365
	TIMM23	2.86122	−5.54518	8.00298	4.93468

Note: PL = placebo; RSV = resveratrol.

Apoptosis/autophagy

SH3 domain containing GRB2-like endophilin B1 (SH3GLB1) and Ras homolog family (RHOT) members T1 (RHOT1) and T2 (RHOT2) had a approximately 5-, 2.5-, and 7-fold lower gene differences in RSV versus PL pre- to post-exercise, respectively (Table 3).

Carrier/shuttle/import proteins

Gene expression differences from pre- to post-exercise in SLC25A4, SLC25A13, SLC25A14, SLC25A15, SLC25A19, SLC25A20, SLC25A21, and SLC25A30, mitochondrial import proteins, metaxin 2 (MTX2), and translocase of inner mitochondrial membrane 23 (TIMM23) were two- to eightfold lower in RSV- versus PL-treated muscles (Table 3; Supplementary Results).

Decreased RSV regulation of mitochondrial metabolic genes

The gene expressions levels of COX10, and carnitine palmitoyltransferase 2 (CPT2), GrpE-like 1, and mitochondrial (GRPEL1) and mitochondrial intermediate peptidase (MIPEP) were two- to fourfold lower in RSV-treated muscles as compared to PL-treated muscles.

Muscle Strength and Fatigue

Subjects who exercised and consumed RSV had an 8.5% (p < .05) increase in knee extensor peak torque that was measured at 60°/s as compared to the pre-training data, whereas there was no change in maximal torque production in the subjects who consumed the PL and exercised for 12 weeks. In a similar fashion, average knee extension torque that was measured at 60°/s, increased by 14.1% (p < .05) in subjects that had taken RSV from pre- to post-training, whereas, there was no significant change in the average peak torque at 60°/s in subjects who consumed the PL and trained for 12 weeks.

Muscle power was unchanged in PL-fed subjects from before to after 12 weeks of exercise training. However, RSV-treated subjects had a 14.0% (p < .05) increase in knee extensor muscle power after training as compared to before training (Figure 1E). The interaction between Session (pre- or post-training) × Group (PL vs RSV) was significant for average power in knee extension (Watts) at 60°/s (p = .004) and average peak torque over seven repetitions in extension (Newton-meters) at 60°/s (p = .038) (Figure 1D). Other indices of muscle function were unchanged by RSV treatment as compared to PL treatment (Supplementary Tables 2–7).

Muscle Fiber Characteristics

Fiber type

An example of tissue cross-sections that were used for fiber type and fiber size distribution is shown in Figure 2. Before training, type I fiber distribution was not statistically different in the vastus lateralis of PL-treated (68.3% ± 6.3%) and RSV-treated (51.9% ± 5.5%) groups. Type I fiber composition was not significantly altered after 12 weeks of exercise training in PL-treated (50.4% ± 6.9%) or RSV-treated (43.0% ± 5.5%) subjects. In the PL group, the distribution of type IIA fibers was unchanged before (31.2% ± 6.0%) as compared to after training (48.3% ± 5.9%). Similarly, in the RSV-treated group, the distribution of type IIA fibers was unchanged from pre-training (47.8% ± 5.5%) to post-training (55.6% ± 5.1%). A small number of fibers co-expressed type I and type IIA myosin heavy chain. There was no difference before exercise in the percent of type IIA/I co-expressing fibers between the PL group (0.45% ± 0.32%) and the RSV group (0.33% ± 0.35%). There was no statistically significant change in the percent distribution of type IIA/I co-expressing fibers after exercise in either the PL-treated (1.37% ± 0.96%) or RSV-treated (1.39% ± 0.56%) subjects. Type IIX fibers were rare and made up less than 0.1% of the total population of fibers that were examined in muscle biopsies of the subjects in this study. Consequently, few fibers co-expressed IIA/IIX in this sample. The IIA/IIX fibers made up 0.014% ± 0.01% and 0.05% ± 0.02% of the fiber population in the PL group before and after exercise, respectively. Similarly, the percentage of type IIA/IIX fibers was and 0.017% ± 0.02% and 0.09% ± 0.03% in the RSV group.

Figure 2.

Fibers from placebo (PL)-treated and resveratrol (RSV)-treated muscles. (A) Mean type I fiber area. *p < .05, pre-training vs post-training. **p < .05, RSV vs PL treatment. (B) Mean Feret diameter from type I fibers. *p < .05, pre-training vs post-training. **p < .05, RSV vs PL treatment. (C) Fiber area-fiber frequency histograms of type I fibers from PL-treated muscles. (D) Fiber area-fiber frequency histograms of type I fibers from RSV-treated muscles. (E) Cumulative frequency area histograms of type I fibers from PL treated and RSV treated. (F) Representative fiber sections stained with DAPI to show muscle nuclei type IIA MHC (green), type IIX MHC (red), and type I MHC (purple) fibers. The overlay shows co-expression of type IA and IIX MHC (yellow/green).

Type I fiber size

To test if fiber size were improved with RSV treatment, we obtained fiber cross-sectional area and Feret diameters from tissue sections that had been incubated in antibodies against type I, type IIA, and type IIX myosin heavy chain (Figure 2A). The data suggest that fiber area adaptations to exercise and treatment were fiber type specific. Twelve weeks of exercise significantly improved mean fiber type I area (Figure 2B) and Feret diameter (

There was a shift to the right in the fiber area-frequency histogram of PL-treated subjects especially in the 5,000–6,000 µm² range (Figure 2D). RSV-treated fibers also had a shift to the right of the fiber area-frequency distribution and a significantly greater percentage of fibers larger than 7,000 µm² after training as compared to before training (Figure 2E).

These differences can be seen readily in the cumulative frequency distribution for type I fibers, where there is a small shift towards larger fibers in the PL trained group but a clearer shift to larger fibers was found in the RSV-treated group (Figure 2F; Supplementary Figure 5).

Type IIA fiber size

Unlike the adaptations in type I fibers, 12 weeks of exercise failed to improve mean type IIA fiber area or Feret diameter in the PL group (Figure 3A). In contrast to the PL group, in the subjects treated with RSV, type IIA fiber area and Feret diameter were 32.2% and 11.9% greater, respectively, after 12 weeks of training as compared to pre-training (Figure 3B). Exercise training in the PL group had a rightward shift in the type IIA fiber area-frequency distribution with a greater frequency of fibers with sizes exceeding 6,000 µm² (Figure 3C). Type II fibers from the RSV-treated group had a shift to the right of the fiber area-frequency distribution and a significantly greater percentage of fibers larger than 5,000 µm² after training as compared to before training (Figure 3D). The similarity in the pre- and post-training type IIA cumulative frequency distribution can been readily seen for the PL trained group (Figure 3C), whereas a shift to larger fibers can be seen in the trained RSV treated type IIA fibers as compared to before training (Figure 3E). Although the cumulative frequency curves for the type IIA fibers shifted to the right after exercise as compared to before exercise, the shift to larger fibers (shift to the right) was more pronounced in type IIA fibers of the RSV group (Figure 3D–F; Supplementary Figure 6).

Figure 3.

Mean fiber area (A) and mean Feret diameter (B) in type IIA fibers from placebo (PL)-treated and resveratrol (RSV)-treated muscles fibers. *p < .05, pre-training vs post-training. **p < .05, RSV vs PL treatment. Fiber area-fiber frequency histograms of type IIA fibers from PL-treated (C) and RSV-treated (D) muscles fibers. Cumulative frequency area histograms of type IIA fibers from PL-treated and RSV-treated muscles fibers (E).

Type IIA/type IIX co-expressing fiber size

There did not appear to be any fibers that were solely expressing type IIX myosin heavy chain in the biopsy samples. There were very few fibers that co-expressed type IIA and type IIX myosin heavy chain before exercise in most subjects. Although there was a tendency for more fibers to co-express both type IIA and type IIX myosin heavy chain after exercise, there was considerable variability among the subjects in both groups. Type IIA/IIX co-expressing fibers were relatively small and averaged 1677.8 ± 246 µm² after exercise in the PL group and 1834.4 ± 537.1 µm² after exercise in the RSV-treated group but the low number of fibers and the large variation in size prevented any statistically important observations in this fiber type.

Nuclei and Satellite Cell Content

Although there was no statistical change in the number of nuclei/fiber with the PL treatment after training as compared to the pre-training data, the total number of nuclei (satellite cell + myonuclei) was increased by 20% (p < .05) in the RSV-treated tissue samples of RSV-treated muscles as compared to the pre-training samples. The number of Pax7 positive nuclei (satellite cells) was significantly increased from 0.45 to 1.1 Pax7 satellite cells/fiber (p < .05) in the post-trained sample of the RSV-treated subjects, but total satellite cell content/fiber did not change in the PL-treated samples (Figure 4A–D).

Figure 4.

An example of tissue sections cut from human biopsies. (A) DAPI (blue)-stained nuclei. (B) Pax7-positive satellite cell nuclei are green. (C) The basal lamina of each fiber was identified by a laminin antibody (fuchsia). (D) The overlay of panels A–C. White arrows point to Pax7 nuclei that are located inside the basal lamina. The yellow arrow shows a Pax7 nucleus that is outside of the basal lamina. (E) Total nuclei and satellite cell (Pax7⁺) content were quantified and expressed per fiber. Nuclei that were outside of the basal lamina were assumed to be of non-myogenic origins. Although Pax7 nuclei that were outside of the basal lamina would be expected to be capable of myogenic contributions, they were not counted as it was not possible to tell if they were migrating from another muscle fiber, or if they would contribute to new de novo muscle fibers. *p < .05, pre-training vs post-training. **p < .05, RSV vs PL treatment. (F) 8-Hydroxyguanosine (8-OHdG)-positive nuclei were identified and quantified if they resided inside the basal lamina as an indication of oxidative damage to skeletal muscle nuclear DNA. *p < .05, pre-training vs post-training. (G) A representative cross-section from a 78-year-old female after 12 wk of exercise training using a PL. (i) DAPI-stained nuclei (blue). (ii) 8-OHdG-positive nuclei are green. (iii) The muscle sarcolemma (red) was indicated by an anti-dystrophin antibody. (iv) The basal lamina was identified by an antibody to laminin and this is shown in white. (v) The overlay of the preceding images is shown. White arrows indicate the same 8-OHdG-positive nuclei in each section. Nuclei that were located inside the basal lamina and were 8-OHdG positive were quantified. Most nuclei that sat between the basal lamina and sarcolemma would be expected to be satellite cells, but these cells were only rarely 8-OHdG positive, whereas more frequently, non-muscle nuclei that resided outside of the basal lamina and sarcolemma were identified as 8-OHdG positive (yellow arrow) but these nuclei were not quantified.

Oxidative Damage

8-OHdG-positive nuclei were used as an indication of oxidative DNA damage in nuclei. No nuclei that resided between the basal lamina and sarcolemma were 8-OHdG positive. However, nuclei that were either inside the basal lamina or outside the basal lamina were identified to have oxidative damage. 8-OHdG-positive nuclei that were located outside of the basal lamina were not quantified. Oxidative damage of DNA as measured by 8-OHdG-positive nuclei inside the basal lamina was 20.9% and 18.8% in muscles from the pre-training biopsies of PL and RSV treatment groups. Exercise decreased the number of nuclei that were located inside the basal lamina with oxidative damage in both PL (5.6% of the total nuclei) and RSV (3.7% of the total nuclei) groups, but there was some variability between subjects in each group and there were no statistically significant differences between RSV and PL groups (Figure 4E–G). These data suggest that there is less oxidative damage after 12 weeks of exercise training but that RSV treatment did not reduce the percent of total muscle nuclei (inside the basal lamina) with DNA damage further.

Discussion

VO₂ max and fatigue resistance have been reported to improve when RSV was combined with endurance exercise training (33,34). However, other preclinical long-term studies have also reported that RSV provided modest or no increases in reducing aging-associated decrements in muscle mass or oxidative stress (16,35). In addition, several clinical RSV studies in humans have shown no, or only minor improvements in function (36–38). In contrast, our data suggests that RSV acts synergistically with exercise-induced signaling pathways to enhance cellular and molecular adaptations that lead to improved mitochondria density, oxygen uptake, fiber area, and muscle function in aged persons.

Mitochondria and Oxygen Consumption

Consistent with our hypothesis, we found that RSV improved maximal oxygen consumption and mitochondrial volume density more so than exercise alone (Figure 1). We noticed that in the electron micrographs, some of the mitochondria appeared to be morphologically appropriate, while others appeared to have increased spacing of the cristae and less dense packing of the mitochondrial membrane. The RSV-treated samples appeared to have fewer mitochondria with abnormal morphology as compared to PL-treated muscles. The implications are of course that more mitochondria that are not “healthy” would not be of any advantage to the muscle; whereas if the mitochondria are morphologically sound and presumably fully functional then we might assume that, the increase in mitochondrial density could be of benefit for improved energy production in the muscle cells. Based on the subjective evaluation of the micrographs, we speculate that the better mitochondrial morphology along with greater mitochondrial volume density in RSV-treated muscles after exercise training as compared to PL conditions may have occurred in part to removal or repair of the mitochondria with abnormal morphology. Nevertheless, we have no way of knowing if the appearance of abnormal mitochondria had any negative effect on the mitochondria’s ability to generate ATP or to oxidize substrates.

Consistent with the increase in mitochondria volume density, there was some enhancement of gene expression with RSV treatment in genes associated mitochondrial morphology, and mitochondrial import (Table 2). It was also particularly interesting to note that RSV also increased genes associated with mitochondrial-regulated apoptotic signaling. However, apoptotic signaling might be an important mechanism to remove dysfunctional organelles and this could be part of a process to “clean” up and remove dysfunctional proteins in aging. However, this is not a simple relationship because several apoptotic genes were also decreased as were genes associated with autophagy in RSV- versus PL-treated muscle cells (Table 3). Furthermore, a number of import genes were downregulated in RSV-treated samples pre- to post-exercise as compared to PL-treated samples (Table 3), perhaps as a result of a greater magnitude of proapoptotic signaling and removal of dysfunctional proteins (Table 3). This varied gene regulation of mitochondrial-associated apoptotic and import signaling warrants further investigation, to determine if these changes provide an improved muscle environment for adaptation to exercise with RSV in elderly subjects.

Our results in older subjects in the current study differ somewhat with data from Scribbans and colleagues (39) who reported no additive effects of RSV supplementation on aerobic exercise performance. Gliemann and colleagues (40) have reported that RSV might blunt the effects of aerobic exercise training on low-density lipoprotein, total cholesterol/high-density lipoprotein ratio, and triglycerides. However, we did not observe any blunting of these measures in our study, but rather we found a small synergistic effect of combining RSV with exercise training (aerobic + resistance exercise) on maximal oxygen uptake but no synergistic improvement in the lipid profiles in older subjects (Table 1).

It is unlikely that RSV either blunted or prevented some of the improvements in cardiovascular indices or mitochondrial function that might have been expected as part of exercise-induced adaptations. Otherwise, we would have expected to have poorer functional outcomes (VO₂ max, fatigue resistance, mitochondrial volume etc.), in RSV-treated muscles, but this was not the case.

Our data do not support a RSV-mediated improvement in lipid profiles, or a reduction in oxidative damage that is additive to exercise in older subjects. However, it is possible that the dose that we gave the subjects (500 mg/d) was not sufficient to result in some of the metabolic and/or mitochondrial enhancements that have been reported in preclinical animal models. Metabolic rates differ in humans and rodents, and after a first pass through the liver and perhaps the active RSV per se or active metabolites of RSV may have differed or been proportionally lower in humans who took RSV and exercised.

Muscle Function

Exercise improvements were observed in knee extension strength (peak torque, average peak torque, and average power) in RSV-treated subjects but not in PL-treated subjects who exercised for 12 weeks. As the randomly assigned RSV group tended to have a lower starting torque as compared to the PL group, we examined if the greater difference in the strength of the RSV group post-training could be due at least in part to the lower starting point in this group as compared to the PL group, and the RSV group simply caught up to the PL group by the post-training measurement. However, we observed only small differences from comparing the group means from the pre- and post-training data for each subject in each group (see Supplementary Results). Thus, the data strongly suggest that RSV provided a greater benefit for developing maximal torque as compared to the PL treatment in response to exercise in elderly subjects.

However, the longer time to peak torque along with no differences in power or average absolute torque suggests that while RSV increased torque production, it may have slowed torque development. This conclusion is consistent with the shift towards slower myosin heavy chain characteristics of type I fibers (41,42) and the approximately 30% greater type I fiber area in vastus lateralis muscles of RSV-treated as compared to PL-treated subjects in the current study. Thus, the greater relative area that is comprised of type I fibers may account for the lower rate of torque production after exercise in the muscles of RSV-treated subjects.

Fiber Area and Satellite Cells

The improvement in muscle function (Figure 1E) suggests a new role for RSV-mediated contractile protein accumulation (Figures 2 and 3), to regulate at least in part, muscle force and power. Furthermore, RSV induced an improvement in fiber cross-sectional area of exercised muscles (Figures 2 and 3). Additional evidence for this adaption is observed by a shift to the right in fiber area-fiber frequency histograms (Figures 2 and 3). Of particular interest, type IIA fiber area did not improve with exercise in the PL group, but it increased in the subjects who consumed RSV and exercised. This contrasts to data that showed a RSV-associated blunting of both hypertrophy and satellite cell proliferation in muscles of old mice (43). Nevertheless, our current results are consistent with data from reloading after atrophy in rodents, where improvement in especially type II fiber size was reported in animals given RSV (15).

Along with the larger fibers (Figure 4), RSV induced an increase in the exercise-associated nuclear and satellite cell numbers (Figure 4). While it is clear that muscle satellite cells are critical for successful muscle regeneration/repair (44), the idea that satellite cells are essential for increasing fiber size during muscle hypertrophy (45–49) has been challenged from preclinical rodent models (50). However, it is not necessary for RSV to enhance satellite cell proliferation, as RSV could potentially contribute to a greater total nuclear pool by reducing apoptotic signaling such as proapoptotic proteins Bax, cleaved caspase 3, and cleaved caspase 9 in exercised muscles (15), and therefore reducing the potential for elimination of muscle nuclei. As we did not conduct a time course evaluation, we do not know if the total satellite cell and nuclear abundance was increased early as a result of hypertrophic adaptations to exercise training, or if it were accomplished more slowly and in part by eliminating fewer nuclei through apoptotic pathways. The association of RSV-mediated increase in fiber size and total myonuclei (satellite cells and muscle nuclei) is consistent with the nuclear domain hypothesis, but additional time course studies should be conducted to provide a better documentation of satellite cell and myonuclei accretion with the combined effects of exercise and RSV treatments.

Study Limitations

A potential limitation of this study is that we used a somewhat moderate training protocol, where only small or in some cases, no changes were noted in the variables that where measured. Thus, we cannot rule out the possibility that if we had used a training protocol that had been more intensive and had resulted in greater exercise-induced improvements, that the effects of RSV might have been different than what we report in this study. However, the intensity of the exercise protocol that we used in this study is in our opinion, more “translationally relevant” than if we had used a more aggressive training protocol similar to that used in other studies. This is because our findings will be relevant for average older persons who are interested in increasing their physical activity but who are not interested in engaging in training at high intensity. However, it is exciting to note that even with this modest training approach, we saw some very important structural adaptations in muscle to exercise with RSV treatment.

Conclusion

We propose that exercise-initiated an overall improvement in mitochondrial function that was linked to improved anabolic signaling pathway leading to hypertrophy of type I and IIA fiber sizes, and an increase in total muscle nuclei (satellite cells and myonuclei) that was enhanced in muscles from RSV-exercised versus PL-exercised subjects. The greater fiber sizes account for the RSV-induced elevation in muscle torque and power. These data support a role for RSV treatment combined with exercise as a potential means for reducing or reversing sarcopenia in elderly persons.

Supplementary Material

Supplementary data is available at The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences online.

Funding

This project was supported by the National Institute of General Medical Sciences (Grant/Award Number: 5P20RR016440, P20RR016477, P30RR032138/GM103488, and U54 GM104942).

Conflict of Interest

The authors declare no conflicts of interest.