Journal of Applied Physiology


The primary aim of this investigation was to evaluate the effect of training on the immune activation in skeletal muscle in response to an acute bout of resistance exercise (RE). Seven young healthy men and women underwent a 12-wk supervised progressive unilateral arm RE training program. One week after the last training session, subjects performed an acute bout of bilateral RE in which the trained and the untrained arm exercised at the same relative intensity. Muscle biopsies were obtained 4 h postexercise from the biceps brachii of both arms and assessed for global transcriptom using Affymetrix U133 plus 2.0 microarrays. Significantly regulated biological processes and gene groups were analyzed using a logistic regression-based method following differential (trained vs. untrained) gene expression testing via an intensity-based Bayesian moderated t-test. The results from the present study suggest that training blunts the transcriptional upregulation of immune activation by minimizing expression of genes involved in monocyte recruitment and enhancing gene expression involved in macrophage anti-inflammatory polarization. Additionally, our data suggest that training blunts the transcriptional upregulation of the stress response and the downregulation of glucose metabolism, mitochondrial structure, and oxidative phosphorylation, and it enhances the transcriptional upregulation of the extracellular matrix and cytoskeleton development and organization and the downregulation of gene transcription and muscle contraction. This study provides novel insight into the molecular processes involved in the adaptive response of skeletal muscle following RE training and the cellular and molecular events implicating the protective role of training on muscle stress and damage inflicted by acute mechanical loading.

  • transcription profile
  • repeated bout effect
  • inflammation
  • macrophage

resistance exercise (re) can impart multiple health benefits to individuals, especially those experiencing diminished muscle mass and function. Declines in muscle mass and strength are associated with the aging process and accompany the progression of various chronic diseases such as type II diabetes mellitus, kidney disease, cancer, osteoarthritis, neuromuscular disorders, HIV, and chronic obstructive pulmonary disease (12). RE among these populations has been recommended as a superior modality for increasing muscle volume and strength. Additionally, recent studies (12, 14, 30) provide convincing evidence supporting the benefits of RE for improving metabolic and mental health. Although the cellular and molecular mechanisms underlying the muscle's acute and chronic response to RE and the associated health benefits are not completely understood, shifting inflammatory milieu from a pro- to anti-inflammatory environment is believed to play a critical role. Chronic systemic inflammation is a common characteristic underlying aging and much of the aforementioned chronic diseases. Yet, regular RE may have the potential to be an anti-inflammatory therapy for both healthy individuals and patients with inflammatory diseases (32). For example, Ogawa et al. (29) reported that following a 12-wk low-intensity RE program in elderly women, circulating inflammatory markers were decreased in direct association with increases in muscle thickness. Moreover, in patients with autoimmune inflammatory myopathy, 7 wk of RE increased muscle performance, which was accompanied by an improvement in muscle inflammatory profiles as indicated by a coordinated reduction of proinflammatory and increase in anti-inflammatory gene transcripts (28).

Although it is generally accepted that chronic RE can ameliorate inflammation in chronic inflammatory states, paradoxically, a single bout of RE appears to promote inflammation as indicated in some, but not all, studies (see review by Ploeger et al. 32). Increased mRNA levels of proinflammatory genes such as tumor necrosis factor-α (TNFα), interleukin (IL)-1β, and IL6 (6), as well as the invasion of immune cells (31) have been consistently observed in skeletal muscle following acute RE. Recent studies (18, 40) now indicate that inflammation is an integral process of RE-induced muscle hypertrophy. Suppression of inflammation through anti-inflammatory drug use (18), and experimentally reducing monocyte/macrophage influx into the injured muscle (40), impair muscle repair and regeneration after injury. Neutrophil invasion in the stressed muscle is traditionally believed to be responsible for the “secondary damage” to tissue, yet current evidence suggests that some neutrophil function may be necessary for tissue repair (24). The current opinion is that both the temporal response and communication between neutrophils and macrophages appear to be essential for proper regeneration of injured or stressed muscle tissue (7). However, it remains unknown whether the inflammatory response associated with acute RE, particularly involving eccentric contractions that result in muscle damage, will exacerbate ongoing inflammation in patients with chronic inflammation-associated diseases (32). The current understanding of the inflammatory response of muscle following acute RE has been built on findings from studies using untrained muscle. The training status has the potential to impact the acute response of muscle possibly due to the “repeated bout effect” phenomenon (41). This event occurs when muscle damage from unaccustomed exercise is diminished following a second bout of the same exercise performed 1–6 wk later (41).

Consequently, comparing trained and untrained muscle following an acute bout of RE can provide important information on the inflammatory responses expected after training adaptation has occurred. As such, a primary goal of the present study was to define the molecular mechanisms underlying the potential anti-inflammatory effect of RE training. As suggested by Febbraio (13) and Tidball and Villalta (43), understanding the inflammatory response to exercise can be exceedingly complex. Cytokine interactions work through a series of complicated networks and also involve immune cells playing a role in both destruction and repair processes. Consequently, we used microarray, as a high-throughput gene-scanning tool, to acquire a global view of the primary and secondary molecular and cellular processes associated with the inflammatory response to acute RE in the trained and untrained state. Additionally, the use of whole genome microarrays within this paradigm can also provide a comprehensive understanding of other complex processes involved in muscle adaptation (9).



Seven healthy young men (n = 4) and women (n = 3) were recruited following participation in the FAMuSS study. FAMuSS was a multicenter, National Institutes of Health-funded study designed to identify genetic factors contributing to baseline muscle phenotypes and the response to exercise training in humans. Methodological details about the FAMuSS study have been previously published (15, 42). To be eligible for participation in this study, subjects met the following inclusion criteria: 1) between the ages of 18 and 40 yr, 2) no chronic diseases, 3) no prior resistance training history, and 4) no medications or dietary supplements which may affect musculature. Subject characteristics are presented in Table 1. Each subject signed an informed consent document, and all procedures were approved by an institutional review board for research with human subjects.

View this table:
Table 1.

Characteristics of subjects

Training protocol and acute exercise.

The FAMuSS study utilized a 12-wk (2 training sessions/week) supervised progressive unilateral arm RE training program, which induced significant increases in both muscle mass and muscle strength in the trained arm with minimal or nonsignificant changes in the untrained (control) arm (15). The present study employed an acute RE session 1 wk after completion of the FAMuSS study. The acute RE protocol was identical to the final training session of the FAMuSS program with the exception that both arms (bilateral) were exercised at the same relative intensity. Briefly, during the acute bilateral arm RE session, subjects performed five upper arm exercises including the biceps preacher curl, biceps concentration curl, standing biceps curl, overhead triceps extension, and triceps kickback. Each exercise was performed for three sets of six repetitions using the six repetition maximum (RM) weight. Before the first biceps and the first triceps exercise, subjects completed two warm-up sets and then rested for 3 min. Subjects rested for 2 min between sets. Muscular strength of each arm was assessed using a one-repetition maximum (1RM) at baseline and 48–72 h after the final training session of FAMuSS (3 or 4 days before the acute bilateral exercise session) using a standard preacher curl exercise.

Muscle biopsies.

Four hours after completion of the bilateral arm RE, muscle tissue samples were obtained from the biceps brachii of both arms using a percutaneous needle biopsy technique as previously described (22). Briefly, ∼3–5 cc of lidocaine hydrochloride was used to desensitize the incision area. A ¼-inch diameter University College Hospital (UCH) biopsy needle, with accompanying suction, was used to harvest the tissue. Up to three passes were conducted to obtain a total of 100 mg of muscle tissue. All biopsy samples were immediately weighed and snap frozen in liquid nitrogen-cooled isopentane and stored at −80°C for subsequent analyses.

RNA isolation and cDNA synthesis.

RNA preparation and cDNA synthesis has been previously described (22). Briefly, total RNA was extracted from frozen tissue with a polytron homogenizer (Brinkman, Westbury, NY) and Trizol reagent (Invitrogen, Rockville, MD) and purified with an RNAse kit (Qiagen, Santa Clara, CA). Total RNA was used as a template for double-stranded cDNA synthesis (Superscript Double-Stranded cDNA synthesis kit, Invitrogen). Biotin-labeled cRNA was generated for microarray hybridization (EnzoBioarray high yield RNA transcription labeling kit; Affymetrix, Santa Clara, CA).

Microarray preparation and data processing.

Microarray preparation and data processing were described in detail in a previous publication (22). Briefly, biotin-labeled cRNA was hybridized to Affymetrix Human Genome U133 Plus 2.0 arrays according to manufacturer's instructions. Following hybridization, the probe arrays were washed and stained. The intensity of bound dye was measured with an argon laser confocal scanner (GeneArray Scanner; Agilent), and the stored images were analyzed using the GeneChip software Microarray Analysis Suite (MAS) 5.0 (Affymetrix). Overall 54,675 probe sets representing 20,080 annotated genes were profiled.

Cell intensity files were generated using MAS 5.0. The array quality was confirmed by using the R package affyQCreport to generate quality control reports from the cell intensity files. Raw intensity values of all samples were preprocessed and normalized by RMA using Bioconductor in R. Differentially expressed genes were tested by using an intensity-based Bayesian moderated t-statistic (IBMT; Ref. 36). IBMT has proven advantages in processing microarray data from limited number of samples, owing to improvements in the accuracy and stability of variance estimation (22, 36).

The data discussed in this study have been deposited in the NCBI Gene Expression Omnibus (GEO, website and are accessible through GEO Series accession number GSE28998.

Functional enrichment testing.

Enriched Gene Ontology (GO) terms (3) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways (16) were tested by a logistic regression-based method (LRpath) that allows the use of a paired statistical test and has been shown to perform well for experiments with small sample sizes (22, 35). Enriched GO/KEGG terms were defined as those having a false discovery rate (FDR) <0.01. We used a directional LRpath test to distinguish between up- and downregulated gene groups with −log(P value) calculated if the fold change is up and +log(P value) if the fold change is down.

Directly related GO terms that had considerable overlap in identified genes were considered redundant. The strategy to reduce redundancy is similar to that reported previously (22). Briefly, if only the child and parent term were enriched for a similar group of genes, the child term was retained; if the sibling terms and parent term were all involved, the more generalized parent term was retained.

Quantitative real-time polymerase chain reaction.

Five genes encoding macrophage surface markers were selected for validation via quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR). Muscle biopsy samples from one male and one female subject were not adequate for this experiment, and the analysis was conducted using the remaining three male and two female subjects. The detailed experimental procedure of qRT-PCR was published previously (22). In brief, total RNA was isolated from these samples with TRIzol reagent (Invitrogen). The quality and quantity of the total RNA samples were checked using a spectrophotometer (NanoDrop, ND-1000; Thermo Fisher Scientific, Wilmington, DE) before reverse transcription. Random hexamers were used to generate single-stranded cDNA using Moloney murine leukemia virus reverse transcriptase according to the manufacturer's protocol (Invitrogen). cDNA was cleaned using the QIAquick PCR purification kit (Qiagen). The qRT-PCR was performed using an Applied Biosystems Taqman Gene Expression Assay on the Applied Biosystems 7500 Real-Time PCR System (Applied Biosystems, Life Technologies, Carlsbad, CA). The selected genes and TaqMan assays are displayed in Table 2. The reaction was prepared according to the standard Taqman gene expression assay protocol in a total volume of 20 μl (Applied Biosystems). All samples were analyzed in duplicate. 18s rRNA was included as an internal control to calibrate the expression levels of the target gene. After amplification, all data were normalized to 18s first by calculating ΔCt as Cttarget − Ct18s, and the fold changes in gene expression between trained and control muscle were calculated further using the 2∧ΔΔCt method (23), in which ΔΔCt was calculated as ΔCttrained − ΔCtcontrol.

View this table:
Table 2.

Genes validated via RT-PCR and the corresponding TaqMan assay

Statistical analyses.

The physical characteristics of subjects are reported as median (range). Differences in muscle mass and strength between trained and untrained arms were tested by Wilcoxon rank sum test, and statistical significance was set at P < 0.05. Differentially expressed genes were tested using IBMT, and enrichment analysis for GO terms and KEGG pathways was conducted using LRpath. Multiple testing correction was performed using a FDR approach (5, 39), and significantly enriched concepts (GO or KEGG) were defined as having an FDR <0.01.

To validate the microarray data, a correlation was tested for fold changes measured by the microarray and qRT-PCR for all the selected genes using linear regression analysis after log2 transformation. For the genes represented with multiple probes in the microarray, the probe with the highest average signal intensity (generally associated with lower error) was used. For qRT-PCR, we used the mean fold change, calculated as 2∧ΔΔCt, of the duplicates of each gene. To test for the significance of the differential expression of the five target genes measured by qRT-PCR, a Wilcoxon signed-rank test was used. P < 0.05 was accepted as significant in both tests. IBMT and LRpath were performed using R. All other analyses were conducted using SAS 9.1 (SAS Institute, Cary, NC).


Subjects and intervention.

The subjects' ages ranged from 19 to 35 yr, and their body mass index ranged from 19 to 34.7 kg/m2. The 12-wk unilateral arm RE program increased muscle strength (1RM) [male, 30.9% (6.7 ∼ 52.4%); female, 38.5% (25 ∼ 85.7%)], arm muscle cross sectional area [male, 9.4% (0.5 ∼ 11.4%); female, 14.5% (−1.2 ∼ 22.9%)], and arm muscle volume [male, 9.3% (2.4 ∼ 11.5%); female, 10.7% (−0.5 ∼ 22.3%)] (Table 1).

Minimal differences between males and females.

We first studied mRNA expression changes within subjects (paired analyses) separately for males (n = 4) and females (n = 3). The global transcriptional profiles in trained vs. untrained muscle 4 h after the acute bout of RE showed remarkable similarity between males and females, as demonstrated in Fig. 1. Gene functional analysis was also performed separately for males and females, and the biological themes reflected by significantly enriched GO and KEGG terms were very similar. Additionally, in our previously published investigation (22), we observed that the majority of biological processes transcriptionally regulated by acute RE in males and females at 4 h postexercise was quite common. Consequently, in the final analysis for the present study, we combined males and females.

Fig. 1.

Differential gene expression (false discovery rate <0.05) in trained vs. untrained skeletal muscle following bilateral arm resistance exercise (RE) in healthy young men and women. Heat map included 824 genes with significantly different mRNA levels between trained and untrained arm muscle in either females or males, as tested using intensity-based Bayesian moderated t-statistic. Genes were filtered using separate analysis results for males and females to display their strikingly similar expression responses in an unbiased manner. Log2-fold changes for each individual were clustered using a Bayesian model-based clustering program GIMM (26) to calculate the distance matrix followed by average-linkage hierarchical clustering. Red color indicates higher and green color indicates lower mRNA levels in trained vs. untrained muscle. First 3 columns (F123, F133, and F134) represent female subjects, and the other 4 columns (M1327, M1330, M1335, and M1342) represent male subjects. All genes and their differential expression in trained vs. untrained muscle in females and males are included in Supplemental Table S2.

Training effect on skeletal muscle transcriptional responses to acute RE.

Forty-six GO terms/KEGG pathways were significantly enriched with genes having higher transcript levels in trained compared with untrained muscle following acute RE, and they consistently reflected three primary biological themes including extracellular matrix (ECM) and cytoskeleton development and organization, mitochondrial structure and oxidative phosphorylation, and glucose metabolism (Table 3). Seventy-eight GO terms/KEGG pathways were significantly enriched with genes having lower transcript levels in trained compared with untrained muscle. These pathways consistently reflected seven biological themes including the response to unfolded protein, neurological system process, intra/intercellular signaling; muscle contraction, transcription, immune activation, and apoptosis (Table 4).

View this table:
Table 3.

GO and KEGG significantly enriched with genes having higher mRNA levels in the trained vs. the untrained skeletal muscle following acute bout of resistance exercise

View this table:
Table 4.

GO and KEGG significantly enriched with genes having lower mRNA levels in the trained vs. the untrained skeletal muscle following acute bout of resistance exercise

Real-time qRT-PCR.

Given that our primary aim was to investigate the training influence on immune activation in response to acute RE, we found that the divergent changes in two macrophage subpopulations (M1 and M2) in trained and untrained muscle of particular interest. Consequently, we selected five representative macrophage genes for real-time qRT-PCR validation, including a global macrophage marker, CD68 molecule (CD68); M1 macrophage markers [CD40 molecule (CD40) and Fc fragment of IgG, low affinity IIIb, and receptor (FCGR3B)]; and M2 markers [mannose receptor, C type 1 (MRC1), and CD163 molecule (CD163)] (34). To validate the findings from the microarray study, we calculated fold changes of the selected genes in the trained relative to the untrained muscle for each individual subject obtained from both the microarray and qRT-PCR and tested the correlation between these two sets of data. Our results indicated a robust consistency between microarray and qRT-PCR (R2 = 0.66, P < 0.0001; Fig. 2).

Fig. 2.

Correlation between quantitative (q)RT-PCR and microarray on differential expression of 5 macrophage genes. Each blue diamond represents 1 gene of an individual subject (n = 25).

We also tested the significance of the differential expression of the five target genes based upon qRT-PCR results. The data indicated that following bilateral arm RE, compared with the untrained muscle, the trained muscle had lower FCGR3B (log2fold = −1.34; P = 0.31), higher CD163 (log2fold = 1.17; P < 0.06), and MRC1 (log2fold = 1.49; P = 0.06; Fig. 3). These results were consistent with the microarray, but due to a decrease in sample size (n = 5 in qRT-PCR vs. n = 7 in microarray), differential expression of these genes was marginally significant with the exception of FCGR3B.

Fig. 3.

Differential expression of macrophage genes in trained and untrained muscle following acute RE. Y-axis represents fold changes (log2): positive values indicate upregulation in trained vs. untrained muscle, and negative values indicate downregulation. *P < 0.05, significantly different from the untrained muscle. †P < 0.10, trend towards a differential expression in trained vs. untrained muscle.


In the present study, we used microarrays to profile the global transcriptome of skeletal muscle undergoing an acute RE bout in the trained and the untrained state. We demonstrate that training has a profound influence on the muscle's acute transcriptional responses to RE during the early phase of recovery (4 h postexercise). Compared with untrained muscle, trained muscle had lower transcript levels for genes involved in the stress response, immune activation, and intra- and intercellular signaling as well as gene transcription and muscle contraction. Training also resulted in higher transcript levels for genes involved in the ECM and cytoskeleton development and organization, glucose metabolism, and mitochondrial structure and function following acute RE.

To extend the interpretation of our findings, we compared the results from this study to our recently published microarray data (22), where we defined the transcriptional pathways responsive to acute RE in trained muscle. Both studies recruited subjects after completing the FAMuSS 12-wk unilateral RE program. In these studies, the trained arm was acutely exercised and compared with the contralateral arm that was either acutely exercised (present study) or a nonexercised control (previous study). Furthermore, both studies employed the same approaches for acquiring and processing the samples and conducting the data analysis. Therefore, comparison between the two studies is valid and helpful for gaining a comprehensive understanding of the effect of training on the acute exercise response. Overall, we found that training had differential effects on specific pathways, either blunting the changes induced by an exercise bout or enhancing them (Table 5). We defined our findings as follows: 1) training-enhanced upregulation if the gene functional groups or biological processes (simplified as biological processes in the rest of the study) displayed upregulation in the trained vs. the untrained muscle following both bilateral (present study) and unilateral exercise (previous study); 2) training-blunted upregulation if the biological processes displayed downregulation in the trained muscle vs. the untrained muscle following bilateral exercise but upregulation following unilateral exercise; 3) training-blunted downregulation if the biological processes displayed upregulation in the trained vs. the untrained muscle following bilateral exercise but downregulation following unilateral exercise; and 4) training-enhanced downregulation if the biological processes displayed downregulation in the trained vs. the untrained muscle following both bilateral and unilateral exercise. Consequently, the RE-regulated biological processes significantly influenced by training were grouped into two categories (and 4 subcategories), i.e., training-enhanced responses (including both up- and downregulation) and training-blunted responses (including both up- and downregulation; Table 5).

View this table:
Table 5.

Significant biological processes influenced by training in RE-induced acute transcriptional responses

With the use of these categories, our data suggest that RE training influences acute exercise transcriptional regulation in the following manner: 1) blunts the stress response, immune activation and inflammation, and intra- and intercellular signaling seen in untrained muscle; 2) enhances the upregulation of the ECM and cytoskeleton development and organization and muscle development; 3) blunts the downregulation of glucose metabolism and mitochondrial structure and function; and 4) enhances the downregulation of gene transcription and muscle contraction.

Training blunts RE-induced transcriptional upregulation of the stress response and immune activation.

Exercise is a potent activator of MAPK and NF-κB signaling in skeletal muscle. Both pathways play significant roles in coupling alterations in cellular stress with training adaptations (19). Moreover, training also has been shown to minimize the activation of these pathways (19). However, the majority of these findings were made by investigating the phosphorylation cascade within these signaling pathways. Studies from our laboratory indicate that the corresponding regulation of these pathways at the transcriptional level is also involved in the acute RE response and further influenced by training status. Genes related to the stress response, such as heat shock proteins, exhibited a consistent and robust reduced expression in trained vs. untrained muscle following RE. These genes included heat shock 90, 70. and 27 kDa proteins (Supplemental Table S1; Supplemental Material for this article is available online at the J Appl Physiol website). Thompson et al. (41) previously reported that protein levels of heat shock protein 27 and 70 were both elevated 48 h following sequential bouts of eccentric exercise, suggesting no repeated bout effect (i.e., no difference in the magnitude of increase was detected between the first and the second bout). This discrepancy with our data could be attributed to differences in exercise modality, timing of sampling, and assessment entity (e.g., mRNA vs. protein). It is also conceivable that several weeks of training might further modify the acute response.

Unaccustomed exercise, especially the eccentric contraction-biased exercise, causes muscle damage including sarcolemmal disruption, disorganization of contractile components, and cytoskeletal damage (21). Adaptation occurs following the initial bout of damage, making the muscle cell more resistant to the stress and damage in the subsequent bout of the same or similar exercise. Less severe changes are observed in muscle morphology and biochemical indicators of myofibrillar damage such as muscle stiffness, strength deficit, plasma creatine kinase activity level, and focal myofibrillar disruption following subsequent sessions (21, 25).This phenomenon is referred to as the repeated bout effect (21, 25). Although we did not assess muscle damage following acute bilateral exercise, it is conceivable that less stress and myofibrillar disruption occurred in the trained compared with the untrained muscle. Considering that a single bout of RE can offer protection against muscle damage that may last several weeks to 6 mo (25), it is reasonable to expect that transcriptional alterations associated with the stress response and muscle damage would be blunted in trained muscle. Indeed, in the present study, the transcriptional profile of the trained muscle compared with the untrained demonstrated depressed immune activation and inflammation, stress response (response to unfolded proteins), and MAPK signaling and diminished neurological system processes.

An acute bout of unaccustomed RE can illicit inflammation in skeletal muscle; however, long-term RE training is associated with anti-inflammatory benefits (8). Although the exact mechanism that underlies this seemingly “inverse” adaptation process has yet to be delineated, the activity and function of immune cells, such as neutrophils and macrophages, appear to be involved (7). Accumulating evidence indicates that the invasion of neutrophils and macrophages into skeletal muscle following exercise can confer protection against damage in subsequent exercise bouts (24, 31) and is critical for muscle repair and regeneration (7, 10, 18, 38). Alternatively, RE training may induce adaptations in resident neutrophils and macrophages (7); however, this hypothesis has yet to be explored in skeletal muscle. Interestingly, our finding based on 4 h postexercise transcriptional profiles indicates that although the upregulation of muscle monocyte chemotaxis regulators was suppressed in trained compared with untrained muscle after exercising at the same relative intensity, a persistent decrease in proinflammatory macrophage subtype M1 and an increase in anti-inflammatory macrophage subtype M2 were observed in the trained compared with untrained muscle across two exercise conditions (bilateral and unilateral RE). For example, chemotaxis regulators, chemokine (C-C motif) ligand 2 (CCL2, also known as MCP-1) and plasminogen activator, urokinase (PLAU), were lower in trained vs. untrained exercised muscle (3.26- and 1.79-fold, respectively) but higher in trained than untrained nonexercised muscle (2.46- and 1.31-fold, respectively). The M1 gene FCGR3B (CD16b) decreased expression in the trained muscle compared with either exercised (2.14-fold) or nonexercised (2.97-fold) untrained muscle. On the contrary, the M2 genes CD163 and MRC1 (also known as CD206) increased expression in the trained muscle compared with either exercised (1.61- and 1.67-fold, respectively) or nonexercised (1.22- and 1.20-fold, respectively) untrained muscle. These findings suggest that RE training has the potential to minimize monocyte recruitment, decrease M1, and augment M2 polarization of macrophages. A similar discovery was recently observed in the adipose tissue of high-fat-diet-induced obese mice following exercise training (17). Kawinishi et al. (17) found that endurance exercise training inhibited inflammation both by suppression of macrophage infiltration and acceleration of phenotypic switching from M1 to M2 macrophages. Furthermore, Prokopchuk et al. (33) reported that 6 wk of RE training increased M2 activation related cytokines including IL-4, IL-13, IL-4Ra and IL-13Ra1 in skeletal muscle. Interestingly, we observed a modest decrease in expression of IL23A, an M1 cytokine, and increase in TGFB1 expression, an M2 cytokine, in trained muscle compared with either exercised or nonexercised untrained muscle. To confirm our findings and to clarify the contribution of M1 and M2 macrophages in RE-induced health benefits, characterizing macrophage phenotypes in skeletal muscle following exercise training using histological techniques is warranted in future studies.

Training enhances acute RE-induced upregulation of ECM and cytoskeleton development and organization.

Our postexercise transcriptional data from trained and untrained muscle suggest that training promotes the RE-induced upregulation of a group of biological processes associated with the ECM and cytoskeleton development and organization. This finding is consistent with the results regarding macrophages (discussed above). It has been known that M2 macrophages can facilitate RE-enhanced fibroblast proliferation and differentiation leading to increased production of ECM and cytoskeleton constituents such as collagens (7). Concurrently, we also observed that lysosome constituents were expressed at higher levels in trained muscle following RE. These included cathepsin K, B, and O, which were expressed at higher levels in trained muscle compared with untrained muscle (exercised or nonexercised). This finding might suggest that the increased lysosome activity parallels a concurrent increase in ECM and cytoskeleton proteins for targeted degradation. This process may facilitate ECM and cytoskeleton remodeling and reorganization (4). It has been suggested that an alteration in the mechanical properties of the musculoskeletal system is one mechanism underlying the repeated bout effect (21). Mechanical adaptations in the noncontractile elements of the musculoskeletal system might involve cytoskeletal proteins, responsible for maintaining the alignment and structure of the sarcomere, and intramuscular connective tissue (extracellular matrix). It has also been suggested that remodeling of cytoskeletal proteins, such as desmin and titin, could provide mechanical reenforcement against excessive sarcomere strain and a further increase in connective tissue would result in an improved ability to dissipate myofibrillar stress (25). Induction of ECM and cytoskeleton gene expression and protein production by RE has been documented before (11). However, it is noteworthy that the information provided by our study suggests that training enhances the sensitivity and responsiveness of the muscle ECM and cytoskeleton system to exercise stimuli and that this process might play a fundamental role to muscle adaptation to RE training.

Training blunts acute RE-induced downregulation of glucose metabolism and mitochondrial phosphorylation.

Our data also suggest that with training the transient transcriptional downregulation of mitochondrial structure and oxidative phosphorylation, as well as glucose metabolism, was suppressed in skeletal muscle. This is likely related to changes in muscle fiber type composition. As reported previously (2, 37), RE training can cause transformation of the fast-twitch fiber subtypes from myosin heavy chain (MHC) IIb isoform to more metabolically oxidative fiber MHCIIa. Consistent with these reports, our gene expression data indicated that compared with untrained muscle, in both the nonexercised and exercised state, trained muscle had significantly lower (2 ∼ 5-fold) mRNA levels of MHCIIb encoding gene MYH1 and slightly higher (∼1.1-fold) MYHIIa encoding gene MYH2. Nevertheless, the finding that training ameliorates acute RE-induced downregulation of glucose metabolism and mitochondrial phosphorylation adds to the growing evidence that RE can benefit muscle oxidative capacity and glucose metabolism. In fact, numerous studies now demonstrate that RE can improve muscle metabolic homeostasis through mechanisms distinct from endurance exercise. However, currently our understanding of how this precisely occurs is limited. It has been suggested that there are factors exhibiting dual functionality in the process of myofiber growth and metabolism, such as peroxisome proliferator-activated receptor-γ, coativator 1α (PPARGC1, also known as PGC1α), which is the master regulator of mitochondrial biogenesis and ancillary programs in skeletal muscle (20). Interestingly, we observed significantly higher mRNA levels of PPARGC1 (>2 fold) in trained muscle compared with either exercised or nonexercised untrained muscle.

Training enhances acute RE-induced transcriptional downregulation of gene transcription and muscle contraction.

In the early phase (4 h postRE) of muscle recovery following acute RE, our data suggest that genes involved in gene transcription, such as RNA processing and metabolism, and muscle contractile machinery are downregulated, and with training these processes are further downregulated. Previous reports regarding the influence of RE (acute or chronic) on RNA metabolism and muscle contraction are scarce. This finding may relate to improved energy efficiency following training. During the early phase of recovery from intense RE, the less immediate need for protein production and muscle contraction is deferred in favor of energy conservation to support recovery of homeostasis of the cellular environment. It has been suggested that gene transcription itself is energy demanding and protein production through transcriptional regulation lags behind the elevation in protein production via the mammalian target of rapamycin pathway (1). Similarly, genes involved in maintaining muscle tone, the continuous and passive partial contraction of the muscles, are downregulated following acute RE in the untrained and more so in the trained muscle most likely to conserve muscle energy to promote recovery following exercise.

We recognize that the present study had a limited sample size. However, as discussed in our previous study (22), we utilized a within-subjects design (one arm of an individual subject was trained and the contralateral arm served as control), powerful analytical methods (i.e., LRpath combined with IBMT) that excel in small sample studies, and thus reduced the error variance associated with individual differences and increased the accuracy of the variance estimation, consequently maximizing the detection power. As indicated in the previous study (22), we had ≥91% power to detect a 1.5-fold change at the 0.01 α-level for 90% of the probes with a minimum of three subjects. For the mid to high expressed genes, the power was even greater due to their lower variance.

Also, given the limits with acquiring human muscle samples, we did not have ample tissue to confirm our findings on gene expression with protein measurements. Future studies are needed to confirm our findings by including an assessment of protein levels. Finally, we examined only one postexercise time point, which limits our ability to identify potential differences in the time course of muscle transcriptional regulation that may result from training. A previous study by Yang et al. (44) indicated that gene regulation following exercise generally peaks 4–8 h postexercise. Moreover, in our previous investigation (22), we were able to confirm that the majority of the muscle transcriptional responses to RE, including many novel findings, occurred at 4 h postexercise for both males and females. Thus we are confident that studying the 4 h postexercise time point allowed us to capture maximal transcriptional responses in trained and untrained muscle and to detect any major influence of training status. However, we cannot exclude the possibility that some of the observed differential gene expression may be due, in part, to potential differences in the time course of recovery between trained and untrained muscle. Although difficult with human studies, future investigations using multiple biopsy sampling time points over the course of exercise recovery would provide added clarity. Nevertheless, we systematically assessed the differences in the global gene expression following acute RE in trained and untrained muscle, providing insight into the molecular basis of the adaptive response of skeletal muscle to acute RE. Further, these data comprehensively expand our understanding of the mechanisms behind the protective role of RE training by minimizing muscle stress and damage associated with acute mechanical loading.


Funding for this study was provided by National Institute of Neurological Disorders and Stroke Grant RO1-NS40606-01A1 and the National Center for Medical Rehabilitation Research-Integrated Molecular Core.


No conflicts of interest, financial or otherwise, are declared by the author(s).


Author contributions: P.M.G., G.A.N., and E.P.H. conception and design of research; P.M.G., H.B.I., E.E.P., L.G., and G.A.N. performed experiments; P.M.G., D.L., M.A.S., and G.A.N. analyzed data; P.M.G., D.L., and E.P.H. interpreted results of experiments; P.M.G. and D.L. drafted manuscript; P.M.G., D.L., H.B.I., E.E.P., G.A.N., and E.P.H. edited and revised manuscript; P.M.G., D.L., M.A.S., H.B.I., E.E.P., L.G., G.A.N., and E.P.H. approved final version of manuscript; D.L. and M.A.S. prepared figures.


We appreciate the assistance of Integrative Biostatistics and Informatics Core of the Michigan Metabolomics and Obesity Center.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
View Abstract