The effects of resveratrol feeding and exercise training on the skeletal muscle function and transcriptome of aged rats

Ethics statement

All animal care and experimental procedures were approved by the Animal Care and Use Committee of Chongqing Medical University (ACUC/CMU/036/2016) and followed the standards for environmental and housing facilities for laboratory animals in China (Gb/T14925-2001). This study uses the method of Zhou et al. (2018), and the method description partly reproduces their wording.

Animals and treatments

Male SD rats (25 months old) were used as experimental animals. Thirty rats were randomly divided into three groups (n = 10 for each group): one group was set as the control (Old), one group was treated with daily exercise training (Trained), and one group was treated with resveratrol feeding (Resveratrol). Animal rearing was processed as previously described in Zhou et al. (2018). Specifically, the rats were maintained in a pathogen-free room with a 12 h/12 h light/dark cycle. The air temperature was controlled at 20 ± 2 °C, and the relative humidity was controlled at 55 ± 5%. All rats were fed ad libitum with standard rat chow, and diets were refreshed once every day. The Old group remained sedentary in their cages for 6 weeks. The Trained group was trained with exercise on a motorized treadmill once at 9:00 daily. The exercise training program was adapted at a speed of 15 cm s⁻¹ for 15 min during the first week, enhanced to a speed of 20 cm s⁻¹ for 20 min during the second week and then enhanced to 30 cm s⁻¹ for 30 min during the next four weeks. During the 6 weeks, the Resveratrol group was fed with oral resveratrol at a dose of 150 mg kg⁻¹d⁻¹ by mixing the resveratrol with the diet. The rats were weighed once weekly, and the ration levels were regulated.

Muscle function measurement

Muscle function was determined as previously described in Zhou et al. (2018). Specifically, at the end of the six-week period, the forelimb grip strength was determined by an electronic grip tester (Cat.47200, Ugo Basil). The strength measurement of each individual was repeated three times, and the maximal value was used as the absolute grip strength (g). The relative grip strength was calculated by dividing the absolute value (g) by the body mass (g). The rats were subsequently transferred back to their cages for three more days of treatment. Then, the rats were euthanized and weighed at 24 h after the last exercise bout/dose of resveratrol. The gastrocnemius muscles of the hindlimbs were quickly sampled and weighed to 0.01 g. The gastrocnemius muscle index (10⁻³) was calculated by dividing the muscle mass (g) by the body mass (g). Then, the muscle samples from the right hindlimb were frozen in liquid nitrogen and stored in a freezer at −80 °C for transcriptome sequencing.

RNA-Seq library preparation and sequencing analysis

The RNA-Seq library and sequencing analysis were processed as previously described in Zhou et al. (2018). Specifically, muscle samples of three rats from each group were randomly selected for transcriptome sequencing analysis. Total RNA from the muscle tissues was isolated using TRIzol (Invitrogen) and an RNeasy RNA purification kit with DNase treatment (Qiagen). A total amount of 3 g of RNA per sample was used as an input material for the RNA sample preparations. mRNA was purified from the total RNA using poly-T oligo-attached magnetic beads and reverse transcribed into double strand cDNA fragments. Sequencing libraries were generated using the NEBNext^® Ultra^TM RNA Library Prep Kit from Illumina^® (NEB, Ipswich, MA, USA) following the manufacturer’s recommendations, and index codes were added to attribute sequences to each sample. The library quality was assessed on the Agilent Bioanalyzer 2100 system. The clustering of the index-coded samples was performed on a cBot Cluster Generation System using the TruSeq PE Cluster Kit v3-cBot-HS (Illumina). The library preparations were sequenced on an Illumina HiSeq platform, and 125-bp/150-bp paired-end reads were generated.

Raw reads in the fastq format were first processed through in-house Perl scripts by removing reads containing adapters, ploy-Ns and low-quality reads from the raw data. Then, Q20, Q30, and GC content were calculated. The reference genome and gene model annotation files were downloaded from the genome website. The reads that passed the quality control were mapped to the Rattus norvegicus genome using Tophat (v2.0.12). HTSeq (v0.6.1) was used to count the read numbers mapped to each gene. Then, the FPKM of each gene was calculated based on the length of the gene, and the read count was mapped to this gene (Trapnell et al., 2010).

Differentially expressed genes and enrichment analysis

The differences in gene expression among groups were analyzed as previously described in Zhou et al. (2018). Specifically, differential expression analysis was performed using the DESeq R package (1.18.0). DESeq provided statistical routines for determining the difference in the digital gene expression data using a model based on the negative binomial distribution. The resulting P-values were adjusted using Benjamini and Hochberg’s approach for controlling the false discovery rate. Gene Ontology (GO) enrichment analysis of differentially expressed genes was implemented by the GOseq R package, in which the gene length bias was corrected. GO terms with a corrected P-value less than 0.05 were considered to be significantly enriched by differentially expressed genes. KOBAS software was used to test the statistical enrichment of the differentially expressed genes in the KEGG pathways.

Quantitative real-time PCR

Quantitative real-time PCR was performed as previously described in Zhou et al. (2018). Specifically, 1.5 g of extracted total RNA was added to a final volume of 20 l to synthesize cDNA using a PrimeScript^TM RT Reagent Kit with gDNA Eraser (Takara Biomedical Technology, Beijing, China). Eight muscular function-related genes (Prkaa-2, FOXO3, IGF-1, CASK, TRIM67, IL-6, MGMT, SCN5a) were selected, and the coding mRNA quantity was analyzed by the ABI VIIA^TM7 Real-Time PCR system (Applied Biosystems, Foster City, CA) with a quantitative real-time PCR kit (GoTaq^® qPCR Master Mix, Promega Biotech, Beijing, China). Amplification was performed for 1 cycle at 95 °C for 10 min, followed by 40 cycles at 95 °C for 15 s, and 60 °C for 30 s with 1 pM primers (Table 1). qRT-PCR for all genes was performed in triplicate, with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) serving as the reference gene. Relative expression levels were calculated with the 2^−ΔΔCT method and were presented as fold changes compared to the Old group.

Statistical analysis

Parameters of the body and muscle were calculated using Microsoft Excel 2003 (Microsoft Corporation, Redmond, WA, USA) and analyzed using SPSS 11.5 (SPSS Inc., Chicago, IL, USA). Initial body mass, body mass gain, and specific growth rate among the groups were compared using one-way ANOVA followed by Duncan’s test. Final body masses and absolute grip strengths among the groups were also compared with initial body mass as a covariate. A paired-samples t-test was used to compare the initial mass with the final body of each group. Differences were considered significant at the level of P < 0.05. Data are presented as the mean ± s.e.m.

Body mass and muscle function

The final body masses of the Trained group and Resveratrol group were significantly lower than those of the Old group with initial body mass as a covariate (P < 0.001) (Table 2). The final body masses of the Old group were 1.94% larger but not significantly different from their initial body masses (t = 0.921, P = 0.388). Both the Trained group and the Resveratrol group showed decreases in body mass. The final body masses of the Trained group were 12.2% lower than their initial body masses (t =  − 8.763, P < 0.001), while the final body masses of the Resveratrol group were 5.31% lower than their initial body masses (t =  − 2.543, P = 0.039). Body mass gain and specific growth rate were significantly larger in the Old group compared to both the Trained group and Resveratrol group. The gastrocnemius muscle indexes for the left limb and the right limb were 4.81 × 10⁻³ and 4.88 × 10⁻³ for the Old group, 5.10 × 10⁻³ and 5.38 × 10⁻³ for the Trained group, and 4.90 × 10⁻³ and 4.97 × 10⁻³ for the Resveratrol group (Fig. 1). The gastrocnemius muscle index was not significantly different among the three groups (left limb: F = 0.399, P = 0.676; right limb: F = 2.984, P = 0.072). The absolute grip strengths were 1,206, 1,329, and 1,297 g for the Old, Trained, and Resveratrol rats, respectively, and were not significantly different among the three groups (F = 1.203, P = 0.320) (Fig. 2A). However, the analysis of covariates showed significant differences among the three groups, with final body mass as a covariate (F = 6.488, P = 0.008). Similarly, the relative grip strengths were significantly higher in both the Trained group and Resveratrol group than in the Old group (F = 5.019, P = 0.017) and were 2.04, 2.62, and 2.33 g g⁻¹ for the Old, Trained, and Resveratrol rats, respectively (Fig. 2B).

Transcriptome sequencing and assembly

Three cDNA libraries were built from the skeletal muscle of each rat treatment. The parameters of the reads are shown in Table S1, and the reads have been uploaded to the NCBI Sequence Read Archive (accession number: SRA605175). The expression level correlated strongly among treatments (Fig. 3).

Trained vs. Old: Only 21 differentially expressed genes were identified between the Trained group and the Old group (Table 3, Fig. S1), within which 14 genes (e.g., ribosomal RNA processing seven homolog A (Rrp7a), RT1 class I, locus CE9, pseudogene 1 (RT1-CE9-ps1), the rRNA promoter binding protein (LOC257642), protein YIF1B (LOC103689986), and potassium sodium-activated channel subfamily T member 1 (Kcnt1)) were downregulated, while seven genes (e.g., slingshot protein phosphatase 2 (SSH2), polypeptide N-acetylgalactosaminyltransferase 5 (GALNT5), RGD1562339 (RGD1562339), EFR3 homolog B (EFR3b), zinc finger protein 474 (ZFP474), mitochondrial carrier triple repeat 1 (MCART1), and cytidine deaminase-like (LOC100909857)) were upregulated in the Trained group compared to those in the Old group. Seventy GO enrichment terms were enriched in the Trained group vs. the Old group (Table S2). The twenty most enriched GO terms were primarily associated with RNA metabolic processes and transmembrane transporters (Table 4). KEGG analysis identified several significantly upregulated pathways in the Trained group vs. Old group, including mucin-type O-Glycan biosynthesis, drug metabolism, and pyrimidine metabolism, but with only one differentially expressed gene (Table S4). The eight selected genes (Prkaa-2, FOXO3, IGF-1, TRIM67, IL-6, MGMT, and SCN5a) were verified by quantitative real-time PCR, within which only two genes (CASK and TRIM67) showed significant differences in the Trained group vs. the Old group (Fig. 4).

Resveratrol feeding vs. Old: Only 12 differentially expressed genes were identified between the Resveratrol group and the Old group (Table 3, Fig. S2), within which two genes were downregulated and 10 genes (e.g., aldehyde dehydrogenase 1 family, member A1 (ALDH1a1), tripartite motif-containing 67 (TRIM67), synaptotagmin 1 (SYT1), syntaxin-7-like (LOC100910446), synaptosomal-associated protein 25 (SNAP25), transthyretin (TTR), solute carrier family 17 member 7 (SLC17a7), neural EGFL-like 2 (NELL2), apolipoprotein C-I-like (LOC100910181) were upregulated in the Resveratrol group compared to those in the Old group. We found 219 GO terms that were enriched in the Resveratrol group vs. the Old group (Table S3). The most enriched GO terms were mainly associated with neurotransmitter transport and synaptic vesicle (Table 4). KEGG analysis identified several significantly upregulated pathways in the Resveratrol group vs. the Old group, including synaptic vesicle cycle, nicotine addiction, retinol metabolism, insulin secretion, retrograde endocannabinoid signaling, and glutamatergic synapse, but with only one differentially expressed gene (Table S4). Only the synaptic vesicle cycle pathway was significantly enriched with more than one differentially expressed gene (Table 5). Only CASK and TRIM67 showed significant differences between the Resveratrol group and the Old group by quantitative real-time PCR verification (Fig. 4).

Effects of exercise

Even though exercise was suggested as a potential strategy in addressing age-related sarcopenia, previous studies obtained conflicting results on the effect of exercise on muscle mass and function. Many studies have found that exercise can ameliorate the aging-related loss of muscle mass and function and change muscle function-related gene expression (Sakamoto et al., 2005; Thomson & Gordon, 2005; Rolland et al., 2008; Shefer et al., 2010; Cisterna et al., 2016). However, the present study showed no positive effects of exercise on muscle mass and absolute strength (Figs. 1 and 2), which may result from heterogeneity in subject characteristics and the exercise type and intensity (Pattanakuhar et al., 2017). A moderate intensity and duration was used in the exercise training program in our study. More studies are needed to test whether exercise works better with stronger intensity or longer duration. In addition, our results showed that the relative grip strength of the Trained group increased compared to the Old group, which could be primarily due to the remarkable effect of exercise training on weight loss.

One explanation for the loss of muscle mass and function in aging rats is the reduced activities of AMPK, a fundamental regulator of energy metabolism (Salminen, Hyttinen & Kaarniranta, 2011; Hardie & Ashford, 2014; Hardman et al., 2014). It has been reported that exercise may lead to AMPK activation (Sakamoto et al., 2005). However, the exercise-induced phosphorylation status of AMPK was observed only in the soleus muscle but not in the plantaris muscle of aged rats (Thomson & Gordon, 2005). Some studies found no significant exercise-induced upregulation of AMPK expression in the muscle of aged rats (Reznick et al., 2007). The present study showed no positive effects of exercise on the expression of muscle function-related genes, e.g., AMPK, FOX3, and IGF-1 (Figs. 1, 2 and 4; Table 3). In addition, the expression levels of CASK, sodium and calcium channel regulation-related genes (Eichel et al., 2016; Nafzger & Rougier, 2017) were even reduced in the aged rats after exercise training.

Recently, sixteen genes have been identified as grip strength determining genes in humans (Willems et al., 2017); however, none of these genes was regulated in aged rats by exercise training in the present study. In fact, only a limited number of genes were upregulated in the Trained group compared to the Old group, and several transport- or transcript-related genes were even downregulated in the Trained group, e.g., RRP7a, RT1CE9ps1, and KCNT1. Our results suggest that exercise training has limited effects on the expression of genes in the skeletal muscle of aged rats. However, our results presented only data on the transcriptomic response. More studies on other levels, such as protein expression, metabolites and hormone dynamics, are necessary to test the effects of exercise on aging-related sarcopenia.

Effect of resveratrol

Similar to the effects of exercise training, dietary resveratrol has remarkable effects on weight loss in aged rats. Consistently, it has been reported that resveratrol (300 and 750 mg/kg/d) reduced the body weight of rats aged 6 weeks during the 13-week feeding period (Williams et al., 2009). In addition, resveratrol (400 mg/kg/d) reduced the body weight of mice aged 8 weeks during the 9-week feeding period, which was attributed to the increase in energy expenditure (Lagouge et al., 2006). However, resveratrol at a dose of 400 mg/kg/d did not significantly affect the body weight of rats during the 2-week feeding period (Momken et al., 2011), and resveratrol at a dose of 60 mg/kg/d did not affect body weight in mice aged 28 months during the 10-month feeding period (Jackson, Ryan & Alway, 2011). That the body weight was reduced by a relatively higher dose and longer period (Lagouge et al., 2006 and the present results) but not by a lower dose or shorter period (Jackson, Ryan & Alway, 2011; Momken et al., 2011) suggests that the effects of dietary resveratrol on body mass depend on both its dose and duration of administration. Consistently, it has been suggested that the effects of resveratrol on in vitro muscle cell plasticity are dose-dependent: low resveratrol doses promoted in vitro muscle regeneration and attenuated the impact of ROS, while high doses blocked the regenerative process (Bosutti & Degens, 2015).

Resveratrol has been shown to alter fatty acid and glucose metabolism, inhibit protein degradation, and protect against cellular oxidative stress (Naylor, 2009; Kaminski et al., 2012; Gordon, Diaz & Kostek, 2013). In the present study, dietary resveratrol did not change the muscle mass or absolute strength of the aged rats, and the apparent increase in the relative strength should be attributed to the weight loss effects of resveratrol in the aged rats. The effects of dietary resveratrol on skeletal muscle mass and function remain diverse in previous studies. Resveratrol reduced the aging-related declines in muscle mass and function of mice at 18 weeks of age (Murase et al., 2009; Dolinsky et al., 2012) but not of mice at 28 months of age (Jackson, Ryan & Alway, 2011). Additionally, dietary resveratrol attenuated the decrease in muscle mass and strength caused by mechanical unloading in young rats (∼250 g) (Momken et al., 2011) but was unable to attenuate the reductions in muscle mass in aged rats (32 months) (Bennett, Mohamed & Alway, 2013). This finding implies that resveratrol may affect the skeletal muscle differently in the old individuals compared to the young ones and may have limited effects in attenuating sarcopenia.

Resveratrol activates a number of signaling pathways, including the expression of IGF-1, AKT, AMPK, and SIRT1, in myoblasts (Barger et al., 2008; Murase et al., 2009; Naylor, 2009; Park et al., 2012; Montesano et al., 2013; Ligt, Timmers & Schrauwen, 2015). However, our study showed no effects of dietary resveratrol on the expression of muscular function-related genes. Similarly, no effects of resveratrol on gene expression profiles were observed in normal mice, and it was explained that collecting tissue samples from the objectives fasted overnight may mask the effects of resveratrol on its target gene expression (Yoshino et al., 2012), which could be a reason for the unchanged expression of those genes in rats by resveratrol in our study, since the rats were fasted overnight before tissue sampling.

Interestingly, however, resveratrol induced the expression of several genes, including ALDH1a1, TRIM67, SYT1, SNAP25, TTR, SLC17a7, NELL2, LOC100910446, and LOC100910181 (Table 3). Among these genes, most of them (e.g., SYT1, LOC100910446, SNAP25, SLC17a7) are related to neuron synaptosome function (Tucker & Chapman, 2002; Yoshihara & Montana, 2004; Santos, Li & Voglmaier, 2009; Antonucci et al., 2016). NELL2, a gene involved in neural cell growth and differentiation, was also upregulated in aged mice fed with resveratrol. The enhanced expression of these genes suggests an upregulation of synaptic vesicle cycle signaling in aged mice fed with resveratrol. SLC17a7 encodes vesicular glutamate transporter 1 (VGLUT1), which is associated with the functions in transporting the excitatory neurotransmitter glutamate into synaptic vesicles and has important roles in the neuromuscular synapse in nervous systems and neuromuscular junctions of muscles (Varoqui et al., 2002; Kraus, Neuhuber & Raab, 2004). Even though the functional significance of VGLUT1 has not been determined, VGLUT1 expression has been found to be altered by the aging process and several diseases (Jung et al., 2018). A recent study found that retinal VGLUT1 protein and gene expression were decreased in diabetic mice, indicating changes in the signal transmission from photoreceptors to bipolar cells and among postreceptoral neurons (Ly et al., 2014). This study also found that the decreased expression of VGLUT1 could not be normalized by metformin treatment (Ly et al., 2014). The positive effects of resveratrol on SLC17a7 expression imply a potential application of diabetic retinopathy intervention.

An expression of ed inat rosci creased expression of 1. Another finding of the present study was that a tripartite motif family gene (TRIM67) was upregulated in aged mice fed resveratrol. The TRIM protein family has been implicated in many biological processes, including cell differentiation, apoptosis, transcriptional regulation and signaling pathways, and plays an important role in the broader immune response (McNab et al., 2011), especially in the restriction of infection by lentiviruses (Ozato et al., 2008). Overexpression of TRIM proteins can increase the membrane repair capacity of muscular dystrophy and restore muscle function and morphology (Alloush & Weisleder, 2013; Dahl-Halvarsson et al., 2018). TRIM67 is highly expressed in the brain and regulates neuritogenesis, brain development, and behavior (Yaguchi et al., 2012; Boyer et al., 2018). Our results suggest that resveratrol has positive effects on the expression of TRIM67 in the muscle of aged rats, implying that the underlying mechanisms warrant further study.