The FASEB Journal article fj.14-266668. Published online March 25, 2015.

The FASEB Journal • Research Communication

Extracellular matrix remodeling and its contribution to protective adaptation following lengthening contractions in human muscle Robert D. Hyldahl,*,1 Brad Nelson,† Ling Xin,‡ Tyson Welling,* Logan Groscost,* Monica J. Hubal,§ Stuart Chipkin,‡ Priscilla M. Clarkson,‡ and Allen C. Parcell* *Department of Exercise Sciences, Brigham Young University, Provo, Utah, USA; †Department of Natural Sciences, Ohio Dominican University, Columbus, Ohio, USA; ‡Department of Kinesiology, University of Massachusetts Amherst, Massachusetts, USA; and §Department of Integrative Systems Biology, George Washington University, Washington, DC, USA This study determined the contribution of extracellular matrix (ECM) remodeling to the protective adaptation of human skeletal muscle known as the repeated-bout effect (RBE). Muscle biopsies were obtained 3 hours, 2 days, and 27 days following an initial bout (B1) of lengthening contractions (LCs) and 2 days following a repeated bout (B2) in 2 separate studies. Biopsies from the nonexercised legs served as controls. In the first study, global transcriptomic analysis indicated widespread changes in ECM structural, deadhesive, and signaling transcripts, 3 hours following LC. To determine if ECM remodeling is involved in the RBE, we conducted a second study by use of a repeated-bout paradigm. TNC immunoreactivity increased 10.8-fold following B1, was attenuated following B2, and positively correlated with LC-induced strength loss (r2 = 0.45; P = 0.009). Expression of collagen I, III, and IV (COL1A1, COL3A1, COL4A1) transcripts was unchanged early but increased 5.7 6 2.5-, 3.2 6 0.9-, and 2.1 6 0.4-fold (P < 0.05), respectively, 27 days post-B1 and were unaffected by B2. Likewise, TGF-b signaling demonstrated a delayed response following LC. Satellite cell content increased 80% (P < 0.05) 2 days post-B1 (P < 0.05), remained elevated 27 days post-B1, and was unaffected by B2. Collectively, the data suggest sequential ECM remodeling characterized by early deadhesion and delayed reconstructive activity that appear to contribute to the RBE.—Hyldahl, R. D., Nelson, B., Xin, L., Welling, T., Groscost, L., Hubal, M. J., Chipkin, S., Clarkson, P. M., Parcell, A. C. Extracellular matrix remodeling and its contribution to protective adaptation following lengthening contractions in human muscle. FASEB J. 29, 000–000 (2015). www.fasebj.org ABSTRACT

Key Words: eccentric exercise • satellite cell • repeated-bout effect muscle damage

•

Abbreviations: B1/2, bout 1/2; B2M, b-2-microglobulin; CK, creatine kinase; COL, collagen types; Ct, cycle threshold; CTGF, connective tissue growth factor; ECM, extracellular matrix; ITGA, integrin-a7; LAMC2, laminin-g2; LC, lengthening contraction; MVIC, maximal voluntary isometric force; NCBI, National Center for Biotechnology Information; RBE, repeated-bout effect; Smad, mothers against decapentaplegic; TGF-bRII, TGF-b receptor type 2; TNC, tenascin C

0892-6638/15/0029-0001 © FASEB

RECOVERY AND ADAPTATION OF HUMAN skeletal muscle following damaging exercise are poorly understood. Investigations aimed at understanding the dynamic nature of muscle recovery and adaptation in humans have induced muscle damage predominantly via standardized exercise protocols that emphasize high-force (eccentric) lengthening contractions (LCs). LC paradigms generally result in functional (i.e., force loss and soreness) and histologic (i.e., sarcomere disruption and necrotic myofibers) evidence of muscle damage, followed by the initiation of remodeling and regenerative events characterized, in part, by infiltration of inflammatory cells and an increase in activated progenitor cells (satellite cells) (1–5). Healthy recovery of skeletal muscle following LC results in an adaptation of the muscle so that it is able to resist future damage following LC, a phenomenon known as the repeated-bout effect (RBE) (6, 7). Although extensive work has described the conditions under which the RBE is present, there is little consensus regarding the underlying mechanisms driving the protective adaptation. An appreciation of the mechanisms of the RBE has implications, not only for the prevention of muscular injury but also in understanding the basis of skeletal-muscle adaptation in the context of muscle disease or wasting. For the most part, studies aimed at elucidating the mechanisms of exercise-induced muscle damage, recovery, and adaptation in humans have focused on the contractile element of skeletal muscle, with some attention given to the potential role for the surrounding extracellular matrix (ECM) (7–9). Skeletal-muscle ECM is a complex matrix composed of proteoglycans, glycoproteins, and collagens, of which the type I and III isoforms predominate. The ECM interfaces with the myofiber sarcolemma at a specialized basement membrane composed primarily of type IV collagen and laminin. The basement membrane, in turn, associates with a costamere structure and a network of membrane-spanning integrins that facilitate the transmission of mechanical force from within the myofiber to the ECM and then onward to the tendon. In this manner, force 1

Correspondence: Department of Exercise Sciences, Brigham Young University, 106 SFH, Provo, UT 84602, USA. E-mail: [email protected] doi: 10.1096/fj.14-266668

1

transfer can occur laterally over the entire perimeter of each myofiber [reviewed in Grounds et al. (10)]. In addition to facilitating the lateral transmission of forces, the ECM provides structural integrity, relays mechanical signals to the myofibers (11, 12), and regulates the behavior of ECMresiding cells, including muscle progenitor cells (13, 14). Mechanical loading influences remodeling and strengthening of skeletal-muscle ECM. Both high-force LC and one-legged kicking exercises (primarily concentric) increase collagen synthesis in human muscle (15, 16). Likewise, matricellular deadhesion proteins [i.e., tenascin C (TNC)] and growth factors [i.e., connective tissue growth factor (CTGF)] are exceptionally sensitive to loading, and their up-regulation is necessary for complete regeneration of damaged muscle (17, 18). In rodents, resistance to contraction-induced injury is associated with greater muscle collagen content (19). Loading-induced ECM remodeling may be dependent on the autocrine/paracrine action of local cytokines and growth factors, most notably, members of the TGF-b family and CTGF. Both animal and human studies have demonstrated loading- and damageinduced changes in the expression of TGF-b1 and -b2 and CTGF, with corresponding changes in expression of skeletal-muscle ECM collagens (9, 20, 21). In vitro studies have provided additional evidence of the regulatory relationship between these autocrine/paracrine factors and stimulation of fibrillar collagen expression (22, 23). Given its clearly defined role in the transmission of forces and its tendency to adapt in accordance to mechanical load, we hypothesized that disruption and subsequent restructuring of the ECM play a role in the functional manifestations (force loss) and the protective adaptations (RBE) that are induced by damaging LC. To address this primary hypothesis, we performed 2 separate studies. In the first, we used a global transcriptomics screen to define early (3 hours post-LC) changes in the expression of ECM-related genes following a single bout of LC. This approach was undertaken to derive testable hypotheses for follow-up analyses. We hypothesized that the results of the screen would be suggestive of an ECM remodeling response, characterized by increased expression of ECM structural, matricellular, and signaling transcripts. For the second study, we used a repeated-bout design to determine how ECM remodeling was associated with functional changes and the protective adaptation following LC. In general, we expected that those transcripts, shown to increase following a single bout of LC, would be attenuated at the protein and transcript level following a second bout of LC, indicating a role for the ECM in muscle adaptation to exercise stress. In this study, we also assessed how repeated bouts of LC influence the proliferation of ECM-dwelling muscle satellite cells.

MATERIALS AND METHODS Study 1 Subjects Subjects for Study 1 have been described previously (24). Thirtyfive healthy men (20.9 6 0.5 yr, 178.9 6 1.2 cm, and 80.6 6 2.9 kg) volunteered for this study. Subjects were informed of all procedures and potential risks and signed a written, informed consent document, approved by the University of Massachusetts Amherst

2

Vol. 29

July 2015

Institutional Review Board. Subjects had no prior history of musculoskeletal injury of the lower extremity and were willing to refrain from participating in new physical activities or taking oral or topical analgesics for the duration of the study. Potential subjects were excluded if they had been involved in a strength-training program in the past 6 months or routinely lifted heavy objects or engaged in activity with a heavy eccentric component. Study design The study design for Study 1 has been described in detail in a previously published report (24). In brief, subjects completed 7 laboratory visits over the course of 1 week. On their first visit, 1 leg was randomly assigned the "exercised leg," and the subjects completed baseline strength measurements of this leg on an isokinetic dynamometer (Biodex Medical Systems, Shirley, NY, USA). The following day, subjects reported to the lab in the fasted state and were provided a standardized breakfast consisting of 400 kcal (;55% carbohydrate, 30% fat, and 15% protein). On this day, subjects completed another strength assessment on the exercise leg that was followed by the LC protocol. Subjects completed 10 sets of 10 repetitions with 10-second rests between repetitions and 1-minute rests between sets. Subjects were instructed to resist the dynamometer lever as it moved from 35° to the subject’s maximal flexion angle at an angular velocity of 30°/s. After completing the LC protocol, subjects rested in the laboratory for 3 hours, after which, a muscle biopsy was taken from the vastus lateralis of both legs. The biopsy from the nonexercised leg was designated as the control. Microarray Details for the transcriptomic experiment and analyses for Study 1 can be found in a previously published paper (24). In brief, total RNA was isolated from control and exercised biopsy samples by use of TRIzol (Life Technologies, Carlsbad, CA, USA). Doublestranded cDNA was created by use of the Ovation Pico WTA system (NuGEN, San Carlos, CA. USA), and Cy3 labeling and hybridization onto an Agilent Whole Genome Microarray (Agilent, Santa Clara, CA, USA) were done at Gene Logic (Gaithersburg, MD, USA), according to the manufacturer’s instructions. Microarray data were log2 transformed and analyzed by use of an analysis of covariance (age and body mass index covariates) with Partek Genomics Suite software. The data outputs were stringently filtered on the basis of P value (P , 0.007). Study 2 Subjects Fourteen healthy adults volunteered to participate in this study, 7 men (23.3 6 2.1 yr, 178.5 6 8.3 cm, and 84.8 6 23.4 kg) and 7 women (25.6 6 2.5 yr, 166.0 6 8.3 cm, and 76.6 6 32.7 kg). Subjects were informed of all procedures and potential risks and signed a written, informed consent document approved by the Brigham Young University Institutional Review Board. The subject’s habitual activity ranged from sedentary to mildly active, but none was actively training in any particular sport nor had they participated in a lower or upper body strength-training program for at least 6 months before participating in the study. All subjects agreed to refrain from participating in new physical activity or taking oral or topical analgesics for the duration of the study. Study design A schematic of the Study 2 design is found in Fig. 1. Subjects performed 2 bouts of LC of the knee extensors with the same randomized leg, separated by 28 days. Muscle biopsies were taken

The FASEB Journal x www.fasebj.org

HYLDAHL ET AL.

Figure 1. Illustration of the study design, depicting time-points for collection of biopsy and blood samples, assessments of strength, and performance of the LCs. 1 day before B1 of exercise (designated Time 0 and used as baseline control) from the nonexercised leg, and at 2 and 27 days postexercise from the exercised leg. On the 28th day post-B1, subjects completed B2 of LC on the same leg, and a muscle biopsy was taken 2 days post-B2. As our study design required that 3 postexercise biopsies be taken from the exercised leg, the preexercise biopsy was taken from the nonexercised leg to reduce the total number of biopsies on the same leg. Isometric strength of the exercised leg was assessed on the day before exercise and at 2, 3, 4, and 5 days following both B1 and B2. Each exercise session consisted of 30 sets of 10 maximal LCs performed on a dynamometer (Biodex Medical Systems). There was a 1 minute rest between sets. Subjects were instructed to resist the dynamometer lever as it moved from 30° (0°, full knee extension) of knee flexion to the subject’s maximal knee flexion angle (;110°) at an angular velocity of 120°/s. During the exercise session, subjects were verbally encouraged to produce maximal force with each contraction. While subjects were enrolled in the study, they were asked to maintain their habitual diet and levels of physical activity. Subjects self-reported to have not participated in any new physical activity between B1 and B2 of LC. Subjects also refrained from use of caffeine, alcohol, or nonsteroidal anti-inflammatory drugs during each week of testing. However, 1 subject reported use of ice, 2 days post-B1 to reduce muscle soreness.

Strength assessments Maximal voluntary force-generating capacity was measured on an isokinetic dynamometer. Before the strength trials, subjects were positioned so that their knee joint aligned with the axis of the dynamometer. The lever arm was then attached to their ankle, 3 cm above the lateral malleolus. The positioning for each subject remained identical throughout the study. Isometric torque was measured at a knee joint angle of 70° (0°, full extension). Subjects completed 3 maximal isometric contractions of 4 seconds each, with a 1-minute rest between each contraction. For each contraction, subjects were instructed to act as forcefully as possible. Consistent verbal encouragement was given for each repetition. Peak isokinetic torque values were defined as the highest attainable value from the 3 trials with an intraindividual coefficient of variation ,5%.

Muscle biopsy procedure Percutaneous needle biopsies were taken from the vastus lateralis muscle. The first biopsy was taken from the nonexercised leg before the exercise interventions. The subsequent 3 biopsies were taken 2 and 27 days post-B1 and 2 days post-B2 (Fig. 1) from the exercised leg. To account for diurnal variation in the expression of genes, each muscle biopsy procedure was carried out at

the same time in the morning for every subject. All biopsies were done between 6:00 AM and 10:00 AM, local standard time. Under local anesthesia (2% lidocaine), a small incision was made into the skin and fascia, and the biopsy needle was inserted into the muscle. With the use of manual suction, 75–150 mg tissue was withdrawn. Muscle samples were then separated from any fatty tissue and divided into 25–50 mg portions. Portions of tissue designated for protein and mRNA homogenates were frozen immediately in liquid nitrogen. Tissue portions designated for sectioning and microscopic analysis were mounted on a cork with tragacanth gum and frozen in isopentane cooled in liquid nitrogen. Biopsy insertions that were performed on the same leg were placed ;5 cm proximal to the last insertion to minimize potential confounding effects from previous biopsies. However, as samples were only taken from the exercised leg postexercise, we cannot rule out that the observed changes may have been affected by repeated biopsies. Immunohistochemistry Cross-sections of muscle biopsy tissue samples (8 mm) were cut by use of a cryostat at 225°C. Samples were mounted to SuperFrost slides and air dried for 30 minutes. Sections were fixed in 2% paraformaldehyde (Sigma-Aldrich, St. Louis, MO, USA) for 4–8 minutes. Following fixation, sections were permeabilized in 0.2% Triton X-100 for 10 minutes and then blocked in a 2% BSA, 5% FBS solution for 60 minutes at room temperature. Sections were incubated in the first primary antibody (Pax7 or TNC) in a humidified chamber overnight at 4°C. Following several washes, sections were then incubated in the appropriate secondary antibody for 30 minutes at 37°C. For double-stained preparations, sections were then fixed and blocked again, as described above, and incubated with the second primary antibody (dystrophin) for 1 hour at room temperature. Following multiple washes, slides were incubated in DAPI and the appropriate secondary antibodies for 30 minutes at 37°C. Stained slides were washed in PBS with Tween 20, dried, and then mounted by use of Fluoroshield histology mounting medium (Sigma-Aldrich). Sections were imaged on an Olympus IX73 fluorescence-capable inverted microscope. All antibody staining was verified by use of appropriate negative controls. The following primary antibodies were used: Pax7, IgG1 (cell supernatant 1:100; Developmental Studies Hybridoma Bank, University of Iowa, Iowa City, IA, USA), TNC, rabbit polyclonal (1:100; EMD Millipore, Billerica, MA, USA), dystrophin, rabbit polyclonal (1:100; ab15277; Abcam, Cambridge, United Kingdom). Secondary antibodies used were: DyLight 488 IgG1 specific (1:100; Jackson ImmunoResearch Laboratories, West Grove, PA, USA), Alexa Fluor 488 goat antirabbit (1:500; Life Technologies), and Cy3 goat anti-rabbit (1:100; Jackson ImmunoResearch Laboratories).

Quantification of immunofluorescent images All quantification of immunofluorescent images was carried out by an investigator that was blind to both condition and time-point. For Pax7 analyses, enumeration was made by use of 4–10 randomly acquired fields from all subjects at all time-points by use of a 320 objective. Fields were acquired randomly for Pax7 analysis by first scanning the image by use of the Cy3 filter (dystrophin). A representative field under this filter would then be chosen and an image taken. The filter was then changed to reveal the Pax7 stain (Alexa Fluor 488). This method was repeated until at least 10 fields were imaged (large sections) or until the entire section had been imaged (smaller sections). An average of 425 6 168 muscle fibers was analyzed/subject/time-point. Figure 2 depicts a muscle cross-section from a subject 27 days post-B1 stained for Pax7 and dystrophin. Enumeration of myonuclei and quantification of

3

Figure 2. Immunohistochemical stain used to determine satellite cell quantity, myonuclei, and cross-sectional area. This cross-section was from a subject 27 days following B1 of LCs. Myofiber boundaries were determined by use of an antibody against dystrophin (red). Satellite cells were identified by use of an antibody against Pax7 (green). Nuclei were stained with DAPI (blue). Arrowheads in merged image denote Pax7+/DAPI+ satellite cells. Arrows denote myonuclei. Original scale bar, 50 mm.

to the manufacturer’s recommendations. RNA concentrations were measured by use of spectroscopy at 260 nm. Absorbance was also measured at 230 and 280 nm to ensure RNA purity via 260/ 280 and 260/230 nm ratios. Total RNA (500 ng) was reverse transcribed by use of a High Capacity RNA-to-cDNA kit (Life Technologies), according to the manufacturer’s instructions. SYBR Green Master Mix (Life Technologies) was used for all PCR protocols. Quantitative RT-PCR reactions were performed in 96well plates with all cDNA samples run in triplicate for each gene of interest along with no template controls. The average cycle threshold (Ct) value for triplicate samples was used for data analysis. Samples were run for 40 cycles of amplification on a MyiQ single-color real-time PCR detection system (Bio-Rad, Hercules, CA, USA). Transcript Ct values ranged from 17 to 35. At the end of each reaction, a melting curve analysis was run to ensure target specificity. Differences in gene expression were determined by the DDCt relative quantification method. Values were normalized to the levels of b-2-microglobulin (B2M), an internal control, previously validated for eccentric exercise studies (25). Forward and reverse primers (IDT, Coralville, IA, USA) for all genes of interest (see Table 1 for sequences) were designed by use of National Center for Biotechnology Information (NCBI) gene sequences with the Primer Express program, v2.0 (Applied Biosystems, Life Technologies). All primers were tested for efficiency via a standard curve and demonstrated efficiencies between 95 and 105% and thus, were considered adequate for analysis via the DDCt relative quantification method. Primer sequences can be found in Table 1. TGF-b magnetic bead multiplex

cross-sectional area were carried out on the Pax7/dystrophinstained images (Fig. 2). For myonuclear enumeration, myonuclei were counted manually on each image. DAPI+ nuclei with a mass center inside of the dystrophin ring were considered myonuclei, and nuclei with their mass center outside of the dystrophin ring were considered nonmyonuclei. Myonuclei are expressed relative to the total number of myofibers. Cross-sectional area measurements were made with the same Pax7/dystrophin-stained sections. Myofibers were traced manually around the dystrophin-stained boundary and quantified by use of Olympus cellSens software. Quantification of TNC was carried out by calculating the total immunoreactive area of the entire section (2–3 310 images that were taken at the same exposure time). For quantification, the total TNC immunoreactive area was expressed relative to the total area of the imaged section. All analyses were done by use of Olympus cellSens software.

Frozen tissue was first homogenized by use of a Total Protein Extraction Kit (EMD Millipore) with added protease and phosphatase inhibitor cocktails (Thermo Scientific, Rockford, IL, USA). Total protein concentrations were determined by use of a Direct Detect Spectrometer (EMD Millipore). Multianalyte profiling of biopsy sample homogenates was carried out on a Luminex MAGPIX multiplexing platform by use of a TGF-b signaling pathway magnetic bead 6-plex, according to the manufacturer’s specifications (EMD Millipore). In brief, antibodyconjugated magnetic beads were incubated with 25 mg tissue homogenate overnight at 4°C. Bead complexes were then washed and incubated in biotinylated detection antibody for 30 minutes on a plate shaker at room temperature. This was followed by incubation in streptavidin-phycoerythrin for 15 minutes on a plate shaker at room temperature. Bead complexes were then read on a MAGPIX multiplex platform (Luminex, Austin, TX, USA). Median fluorescent values, recorded from a minimum of 80 beads, were used for data analysis.

Total RNA isolation and real-time PCR

Statistics

Total RNA was isolated from 25–50 mg tissue from each subject at each time-point by use of TRIzol (Life Technologies), according

A 1-way repeated-measures ANOVA was used for all analyses. Where a significant main effect was found, Dunnett’s multiple

TABLE 1. PCR primer sequences Gene

COL1A1 COL3A1 COL4A1 LAMC2 ITGA7 B2M

Forward primer

Reverse primer

59-TTCTGTACGCAGGTGATTGG-39 59-CCACTTTCTCCTCTGTCACC-39 59-TGAGTCAGGCTTCATTATGTTCT-39 59-CTTAACATTCCTGCCTCAGACC-39 59-GAGCCATGCAGTCTGAGC-39 59-ACCTCCATGATGCTGCTTAC-39

59-GACATGTTCAGCTTTGTGGAC-39 59-TGGATCTCCTGGTGGCAA-39 59-AGAGAGGAGCGAGATGTTCA-39 59-CCTTGTCAGTTGCTCCATGT-39 59-CATGATGTTGAGGAAGGCAGA-39 59-GGACTGGTCTTTCTATCTCTTGT-39

LAMC2, laminin-g2; ITGA, integrin-a7.

4

Vol. 29

July 2015

The FASEB Journal x www.fasebj.org

HYLDAHL ET AL.

comparisons test was used to identify statistical differences between means. Creatine kinase (CK) data were log transformed and analyzed by use of a 1-way repeated-measures ANOVA with SPSS, v22.0 software (IBM, Armonk, NY, USA). All other analyses were carried out by use of Prism 6 (GraphPad Software, San Diego, CA, USA). Gene-expression data are expressed relative to the pre-B1 time-point. Linear regression analysis was used to determine the relationship between force loss and TNC immunoreactivity. Analysis of the microarray data for Study 1 is described in detail in a previously published work (24). Unless otherwise noted, data are presented as means 6 SE. Significance was set a priori at P , 0.05.

RESULTS Study 1 Global gene-expression analysis To assess transcriptional changes in response to LC and to probe for changes in ECM-related transcripts, we performed a global transcriptome screen on muscle samples obtained from the exercised and nonexercised leg of 35 subjects, 3 hours post-LC. Overall, the analysis yielded 463 modified transcripts by use of a cut-off P value of 0.007. The full dataset can be accessed via the NCBI Gene Expression Omnibus database (Accession No. GSE23697). To investigate early changes in ECM-related transcripts, we systematically categorized differentially expressed genes into “physiological networks” that have been defined for skeletal muscle in 3 previous studies (26–28). Figure 3 depicts the network related to ECM structure and function. Overall, there were up- and down-regulated ECM-related transcripts. Genes encoding proteins for basal lamina collagens (COL4A3, COL4A4, COL4A5, COL15A1, and COL15A2) were down-regulated. Genes encoding for the sarcolemmaspanning integrins were generally up-regulated, along with their basal laminar-binding partners LAMC1 and LAMC2. Notably, the highest responding transcripts in the network were those that encoded for proteins with known functions of ECM deadhesion (TNC; 11.6-fold increase) and ECM transcriptional regulation and signaling (CTGF, 7.6-fold increase;

TGFB2, 7.8-fold increase). Overall, with the exception of COL1A1 (1.7-fold increase) and COL10A1 (3-fold decrease), there were few changes in the transcriptional activity of the fibrillar collagens, 3 hours post-LC. Study 2 Exercise performance and muscle function Subjects performed 2 bouts of LC separated by 28 days. There were no significant differences in the amount of total workcompletedbetweenB1(43.269kJ)andB2(45.3617kJ; P = 0.42). The effect of repeated bouts of LC on muscle function was assessed by measuring maximal voluntary isometric force (MVIC)-producing capacity. Following B1, MVIC was significantly (P , 0.05) reduced relative to baseline (0 day) at 2 (24%), 3 (33%), and 4 (27%) days and recovered back to baseline levels by 27 days post-B1. Following B2, MVIC was reduced significantly relative to baseline (0 day) at 2 (19%) and 3 (19%; Fig. 4A) days. Relative to the 27 day time-point (new baseline for B2), MVIC was reduced significantly at 2 and 3 days post-B2. Multiple comparison tests also revealed that subjects lost significantly more force at 3 and 4 days post-B1 compared with those same time-points following B2, indicating a smaller loss in force-producing capacity following B2 (Fig. 4A). As a further indirect marker of muscle damage, we assessed CK activity before and at 2 and 5 days following B1 and B2. There were significant (P , 0.05) increases in circulating CK activity, 2 and 5 days post-B1 relative to the baseline measurement. CK activity returned to baseline by 27 days, and no further increases were found at 2 and 5 days post-B2, relative to any other time-point (Fig. 4B). TNC TNC immunoreactivity was scarcely detectable in the baseline muscle biopsy samples, with only small, punctate portions of the cross-section showing a positive stain (Fig. 5A; 0 day). Two days following B1, TNC immunoreactivity increased markedly and exhibited a ubiquitous

Figure 3. Illustration of fold changes and cellular localization of ECM-related genes that were altered significantly, 3 hours after the performance of LCs. Arrows indicate a known interaction between the gene’s protein product and elements of each level of the ECM. Loxl2, Lysyl oxidase homolog 2; MMP19, matrix metallopeptidase 19; ITGB1/3, integrin-b1/3; ITGB1BP3, integrin-b1 binding protein 3.

5

ECM-related gene expression ECM structural components were measured at the mRNA transcript level to quantify efforts of skeletal muscle to strengthen the ECM following damaging LC (Fig. 6). We measured transcripts that corresponded to each level of the ECM structure (basement membrane, fibrillar ECM and membrane-spanning proteins) or were highly represented in our microarray analysis. The expression of COL1A1, COL3A1, and COL4A1 mRNA was not changed 2 days following B1. Expression of all of the collagen transcripts increased 5.7 6 2.5-, 3.2 6 0.9-, and 2.1 6 0.4fold, respectively, 27 days post-B1 (Fig. 6; P , 0.05). In contrast, B2 of LC had no effect on expression of any of the collagen transcripts. ITGA7, the gene encoding for the a7 subunit of the a7b1 integrin, showed a significant 4.6 6 1.5-fold increase (P , 0.05), 2 days post-B1, and then returned to baseline by 27 days post-B1 with no change following B2 of LC. The laminin-encoding transcript LAMC2 was unchanged in response to B1 and B2. TGF-b signaling

Figure 4. Indirect measures of muscle damage following 2 bouts of LCs. A) Changes in maximal voluntary kneeextension torque before B1 of LC for 4 days post-B1 and 27 days post-B1 and for 4 days following B2 of LC. *P , 0.01, significant difference relative to the pre-B1 (0 day) measure; #P , 0.05, significant difference relative to the 27 day timepoint; †P , 0.05, significant difference between values at the same time-point post-B1 and -B2 (i.e., mean strength loss was significantly greater at 3 days post-B1 compared with 3 days post-B2). Data are means 6 SE. B) Circulating CK activity before (0 days) and 2, 5, and 27 days post-B1 and 2 and 5 days post-B2. *P , 0.05; **P , 0.01). Data were log transformed and are presented on a logarithmic scale. Vertical broken lines represent time of the second bout of LC.

stain localized to the ECM surrounding the majority of muscle fibers in most subjects (Fig. 5A; B1 + 2 days). By 27 days post-B1, TNC immunoreactivity had returned to baseline levels. Following B2, TNC displayed a much more variable response, with 4/14 subjects demonstrating a pervasive stain surrounding many muscle fibers, but the majority of subjects showed little or no evidence of increased TNC immunoreactivity (Fig. 5A, B). Notably, those subjects who demonstrated a high level of TNC immunoreactivity following B2 were those who also showed a high response following B1 (r = 0.54; P = 0.04). Repeated-measures ANOVA revealed a significant difference in the percentage of TNC immunoreactive area between the baseline measure (0.44 6 0.52%) and the 2-day post-B1 time-point (5.2 6 4.3%; Fig. 5B). A significant difference in TNC immunoreactivity was also found between the 2 day post-B1 (5.2 6 4.3%) and 27-day post-B1 measures (0.61 6 0.55%; Fig. 5B). The decrease in strength following B1 LC was associated with the increase in TNC immunoreactivity (r2 = 0.45; P = 0.009; Fig. 5C). 6

Vol. 29

July 2015

TGF-b signaling was assessed at the protein level by use of a multiplexed magnetic bead assay (Fig. 7). The most marked change in TGF-b signaling was noted in expression of TGF-bRII. Expression of the receptor was not changed 2 days post-B1 but was increased significantly by 60%, 27 days post-B1, and 111%, 2 days following B2 (P , 0.05), relative to the baseline measurement. Post hoc testing also revealed a significant difference between the 2 day post-B1 and 2 day post-B2 time-points. Total Smad4 protein content was also elevated significantly (P , 0.05) by 35%, 2 days post-B2, relative to the baseline measurement. Phosphorylated Smad2 protein was unchanged following B1 but then increased by 27 days and remained elevated following B2. There was no change in the quantity of phosphorylated Smad3 following B1 or B2. Satellite cells, cross-sectional area, and myonuclei Quantification of satellite cell content, muscle crosssectional area, and myonuclei at each sampling timepoint is found in Table 2. Pax7+ satellite cells increased significantly, 2 days post-B1, and remained elevated significantly relative to baseline at the 27 and 2 days post-B2 time-points. Interestingly, B2 of LC did not promote further expansion of the satellite cell pool relative to the initial increase following B1. As expected, given that only 2 bouts of exercise were completed, we found no changes in myofiber cross-sectional area or the number of myonuclei per myofiber at any time-point. DISCUSSION This is the first study to systematically assess acute and prolonged changes to ECM components following multiple bouts of maximal voluntary LC in human muscle. Our findings are generally consistent with the ECM-related changes reported by Mackey et al. (9)

The FASEB Journal x www.fasebj.org

HYLDAHL ET AL.

Figure 5. TNC immunoreactivity following 2 bouts of LCs. A) Muscle cross-section (8 mm) depicting TNC immunoreactivity at each time-point. B) Quantification of TNC immunoreactive area at each time-point. Data are presented with a scatter plot and mean bars. **P , 0.001, significant differences. C) Positive relationship between the percentage of strength loss and TNC immunoreactivity, 2 days after B1 of LCs.

following repeated bouts of electrical stimulation. The data generally support the concept that acute and prolonged changes to ECM components are associated with the functional outcomes of repeated bouts of voluntary LC in human subjects. Force loss following B1 of LC was associated with a general deadhesion response in the muscle, as indicated by a robust TNC response. Several weeks following B1 of LC, we observed elevated TGF-bR

content to support the increased expression of collagen mRNAs. This apparent strengthening of the ECM was coupled with an attenuated damage response following B2 of LC, indicated by a faster recovery of maximal force, reduced CK activity, and a reduced TNC response. Collectively, the findings of the present study suggest a role for the ECM in muscle remodeling and the RBE.

Figure 6. mRNA expression of COL1A1, COL3A1, COL4A1, LAMC2, and ITGA7, measured at each time-point. Data are displayed as fold change relative to the pre-exercise sample (0 day) and are means 6 SE. **P , 0.01; *P , 0.05.

7

Figure 7. TGF-b signaling in skeletal muscle following 2 bouts of LCs, assessed by a multiplexed magnetic bead assay. Shown are the median fluorescent values for total TGF-b receptor type 2 (TGF-bRII) and mothers against decapentaplegic homolog 4 (Smad4) protein and phosphorylated Smad2 and Smad3 protein. Data are means 6 SE. **P , 0.01; *P , 0.05.

Our global transcriptome screen indicated widespread changes to the ECM in the early hours following a bout of muscle-damaging LC. The results of the screen drove our hypotheses that alterations to the ECM following LC were 1) related to the magnitude of functional deficits (i.e., force loss) that accompany muscle damage and 2) a potential source of adaptation responsible for the RBE. Whereas we were unable to follow up on all of the altered genes identified in Study 1 as a result of tissue-sample constraints, we chose a subset of those genes to follow up on in Study 2 based on their magnitude of change in Study 1, their known relationship to muscle adaptation, or their relative abundance in skeletal-muscle ECM. Although our analysis did not include any direct histologic measures of muscle damage (i.e., z-line disruption, inflammatory cell infiltration), our strength-loss data indicate a moderate level of muscle damage after the initial bout of LC. Reductions (20–50%) in peak force-generating capacity are consistently associated with direct markers of damage, including inflammatory cell infiltration and altered immunohistochemical staining of structural proteins (i.e., desmin and dystrophin) (1, 3, 4, 8). More importantly, our repeatedbout design successfully induced a protective adaptation,

evidenced by attenuated circulating CK and a faster recovery of force-generating capacity following B2 of LC. This allowed us to evaluate how changes in the ECM related to muscle function and adaptation following damaging exercise. Early ECM deadhesion In agreement with previous studies, we found that a bout of LC markedly increased the expression of TNC at the mRNA (3 hours postexercise) and protein level (2 days post) in untrained individuals (29, 30). TNC functions as a facilitator of cell deadhesion, a process characterized by a reduction in ECM/integrin affinity and a restructuring of the focal adhesion complexes (31). Deadhesion is thought to be an important intermediate state that facilitates cytokinesis during tissue regeneration and remodeling (18, 31). Mice deficient in TNC display a defective muscleregeneration program following injurious mechanical loading (18). With the exception of 4 high-responding subjects, the increase in TNC immunoreactivity following the damaging B1 of LC was attenuated after B2. Our TNC observation paralleled 3 other variables: strength loss, CK

TABLE 2. Assessment of satellite cell number, myonuclei, and cross-sectional area at each time-point Measures

0 day (pre-exercise)

Satellite cell/myofiber Myonuclei/myofiber Cross-sectional area (mm)2 Data are means 6

8

Vol. 29

July 2015

SD.

a

0.05 6 0.02 2.40 6 0.54 4237 6 1262

2 days post-B1

27 days post-B1

2 days post-B2

0.09 6 0.04a 0.08 6 0.02a 0.07 6 0.02a 2.35 6 0.44 2.67 6 0.78 2.46 6 0.66 4410 6 1367 4723 6 1469 4368 6 1337

P , 0.05 relative to the pre-exercise time-point.

The FASEB Journal x www.fasebj.org

HYLDAHL ET AL.

activity, and satellite cell proliferation. Indeed, strength loss, CK activity, and satellite cell proliferation showed significant changes following B1 of LC and showed blunted responses after B2 of LC. Collectively, these findings lend support to the hypothesis that cellular deadhesion is an important early event and possible trigger in skeletalmuscle remodeling. Interestingly, 3 out of the 4 subjects who demonstrated greater TNC immunoreactivity following B2 of LC also demonstrated an increase in strength loss, serum CK activity, and satellite cell-proliferative response following B2 relative to B1. This observation suggests that these subjects may not have been damaged sufficiently following B1 of LC to induce a RBE and lends further support to the assertion that cell deadhesion may be an important mediating factor for muscle remodeling. Given the similar time course of TNC expression and muscle-functional changes and in light of the important role of the ECM in the transmission of force, we also asked whether there was a relationship between cellular deadhesion and force loss following LC. Regression analysis revealed a significant relationship between TNC immunoreactivity and relative force loss following B1 of LC. Thus, it is possible that deadhesion of the ECM may contribute to the loss of force following a novel bout of damaging exercise. Nevertheless, this relationship was lost following B2. Why the relationship between strength loss and TNC expression was evident following B1 but not B2 of LC is somewhat unclear. As the majority of subjects displayed very little strength loss and TNC immunoreactivity following B2, it is possible that the decreased range of both variables influenced our ability to detect a relationship. Furthermore, there were 4 subjects that demonstrated an abnormally high TNC response (Fig. 5B) yet showed very little strength loss following B2. When we removed the TNC “high-responding” subjects from the analysis of the B2 data, the significant relationship was restored. Thus, it is possible that deadhesion may contribute to the loss of force following a single bout of damaging exercise, but its contribution is unclear beyond the initial, unaccustomed bout. Nevertheless, our data, taken with previous work, support the hypothesis that increases in TNC are an early responding event, corresponding with strength loss, as well as satellite cell activation. Given its previously described role as a deadhesion protein, TNC appears to create a favorable environment for subsequent ECM remodeling by disassembling adhesion complexes that may promote satellite cell survival (9). ECM remodeling Our transcriptomic analysis indicated widespread changes in the expression of genes encoding for ECM-related structural components, proteases, and cell-signaling molecules. Whereas we cannot rule out the possible contribution of ECM-related changes that occurred as a result of vascular damage and/or remodeling by use of our current approach, a volumetric scaling of absolute amounts would bias toward ECM changes associated with the myofiber. Regarding the structural components, the analysis showed that collagen expression, primarily represented by the basal laminar collagen IVs, was generally suppressed or not changed in the 3 hours post-LC. Thus, we supposed that

the synthesis of collagens and subsequent strengthening of the ECM structural matrix would occur at some later timepoint. In support of this assertion, collagen synthesis and ECM remodeling were shown to be delayed following muscle damage induced by electrical stimulation (9). Consistent with this idea, we demonstrated in Study 2 that expression of ECM collagens was unchanged, 2 days following voluntary LC, but was eventually up-regulated by 27 days postexercise. In contrast, collagen mRNA expression was not altered in the short term (2 days) following B2. This observation was consistent across fibrillar (types I and III) and basal lamina (IV) collagen types. Our analysis presents 2 time-points, 2 and 27 days, and indicates that ECM structural reinforcement following damage is delayed at least 2 days and that ECM adaptation appears to be a prolonged process that clearly manifests by Day 27 post-LC. The absence of a collagen response following B2 of LC may indicate that further adaptation (i.e., additional collagen support) is not needed or that B2 of LC did not produce a significant stimulus for adaptation. However, as we only assessed collagen expression out to 2 days following B2 of LC, it may be that the second collagen response could have occurred later. A study designed to assess collagen expression during a more prolonged time course following B2 would help resolve this question. Expression of ECM collagens is regulated by the actions of several signal transduction pathways, the most prominent and well-studied of these being the TGF-b/ Smad pathway. TGF-b is a cytokine that binds to types I and II transmembrane receptors, which, in turn, phosphorylate intracellular Smad signaling proteins, leading to transcriptional activation of target genes. TGF-b/ Smad signaling is known to target several collagenspecific genes, such as COL1A1, COL3A1, COL5A2, and COL6A1 (32). Previous research has established that acute and chronic exercise up-regulate the expression of TGF-b1 mRNA in skeletal muscle (17, 21, 33, 34). Given that TGF-b is an upstream signal for mechanical loadinginduced collagen synthesis, we probed the TGF-b signaling pathway via a multiplexed bead assay that assessed changes in phosphorylated and total proteins involved in the TGF-b signaling pathway. We found no changes in total or phosphorylated protein levels of any component of the TGF-b signaling pathway, 2 days following B1. However, the majority of the TGF-b signaling pathway elements (phosphorylated Smad2, Smad4, TGF-bRII) increased relative to baseline by 27 days post-B1, with an even further augmented response in TGF-bRII protein levels, 2 days following B2 of LC. In agreement with the collagen-expression data, these data indicate that TGF-b signaling is likewise delayed after an initial bout of LC. Collectively, the TGF-b signaling data support the notion of a delayed collagen synthesis response in the ECM from a damaging insult. In contrast, expression of integrins was increased early following LC. Genes encoding integrin subunits were highly represented in the microarray dataset, and in Study 2, integrin expression was increased significantly, 2 days following B1 of LC but not at later time-points. Boppart et al. (35) showed that a single bout of damage-inducing exercise results in marked increases in ITGA mRNA in rodent skeletal muscle. Furthermore, they showed that muscle from a7 null mice is equally damaged following an 9

initial and repeated bout of exercise, suggesting a role for the a7 in the RBE (35). However, this is the first study to demonstrate rapid integrin expression following damaging LC in humans. The ubiquitous up-regulation of integrin transcripts following B1 but not B2 of exercise in human subjects from both studies supports the hypothesis that enhanced integrin support following damage acts to protect the muscle from damage after a second exposure. Satellite cell content Satellite cells reside in and move about within the skeletalmuscle ECM. They constitute an important component of the ECM niche, and cross-talk between satellite cells and matrix constituents have been proposed to regulate satellite cell activity following damage (36). We observed that satellite cell content increased 2 days following a bout of LC, remained elevated at 27 days, and was unaffected by B2 of LC. Time-course studies in human muscle have consistently demonstrated expansion of the satellite cell pool from 1 to 8 days post-LC (37, 38). Previously, long-term (28 days) persistence of the expanded satellite cell pool was noted following damaging electrical stimulation in human muscle (9). Our present study confirms those previous findings and provides novel data on the extended time course of their proliferation following LC. Additionally, we report for the first time that B2 of LC does not induce an additional augmentation of the satellite cell pool. This finding is interesting and may be explained by several possible mechanisms, one of which might be the reduced deadhesive environment (i.e., attenuated TNC expression), noted following B2. Indeed, a strong adhesive state is associated with cell quiescence (39), whereas a deadhesive environment is more conducive to cell motility and proliferation and is associated with tissue remodeling (31, 40). It must be noted that our satellite cell histologic analysis did not include a basement membrane maker (i.e., laminin), which disallowed the discrimination of satellite cells that were residing in their basal lamina niche versus those that may have been migrating between fibers (41). Lastly, in our analysis of satellite cells, we were unable to detect a change in the number of myonuclei, despite a significant increase in satellite cell proliferation. Whereas our data do not address this directly, 2 possibilities exist. First, fusion may have occurred, but we were unable to detect the increased myonuclei, possibly because there were too few in number to detect. Alternatively, despite satellite cell activation, the LC did not result in fusion of satellite cells to muscle fibers over the course of the study. Therefore, if fusion indeed does not occur, an interesting question that arises from this observation is the physiologic significance of satellite cell proliferation following a bout of LC. It has been proposed that some satellite cells may become activated by lower threshold stimuli (exercise), not to repair damage as they would in instances of severe injury (i.e., muscle strain or tear) but rather, to replenish the satellite cell pool as a preemptive measure, enhancing the regenerative and growth capacity of the muscle (42, 43). CONCLUSIONS This is the first study to provide a potential link between ECM alterations and the protective adaptation of the RBE 10

Vol. 29 July 2015

following voluntary LC. We have described an ordered response of ECM remodeling following voluntary LC, characterized by early cell deadhesion and increased integrin expression, followed by expression of structural components and activity of ECM-related signaling pathways. Relative to the 27-day sampling time-point, none of the measured variables was altered significantly in the short term (2 days) following B2 of LC. The lack of response following B2 hints at a role for ECM remodeling in the RBE. However, a longer time course of changes is needed to determine if the response to B2 of LC may simply be delayed. The association of TNC expression with the loss of maximalforce production indicates that cell deadhesion may contribute to functional deficits following LC. Finally, our quantification of satellite cell content provides novel information on the extended time course of satellite cell pool enhancement following LC. Moreover, we demonstrate for the first time that in contrast to the response following an unaccustomed bout of LC, a repeated bout of LC does not induce additional satellite cell proliferation.

REFERENCES 1. Hubal, M. J., Chen, T. C., Thompson, P. D., and Clarkson, P. M. (2008) Inflammatory gene changes associated with the repeated-bout effect. Am. J. Physiol. Regul. Integr. Comp. Physiol. 294, R1628–R1637 2. Hyldahl, R. D., and Hubal, M. J. (2014) Lengthening our perspective: morphological, cellular, and molecular responses to eccentric exercise. Muscle Nerve 49, 155–170 3. Paulsen, G., Crameri, R., Benestad, H. B., Fjeld, J. G., Mørkrid, L., Hall´en, J., and Raastad, T. (2010) Time course of leukocyte accumulation in human muscle after eccentric exercise. Med. Sci. Sports Exerc. 42, 75–85 4. Beaton, L. J., Tarnopolsky, M. A., and Phillips, S. M. (2002) Contraction-induced muscle damage in humans following calcium channel blocker administration. J. Physiol. 544, 849–859 5. Hyldahl, R. D., Olson, T., Welling, T., Groscost, L., and Parcell, A. C. (2014) Satellite cell activity is differentially affected by contraction mode in human muscle following a work-matched bout of exercise. Front. Physiol. 5, 485 6. Nosaka, K., and Clarkson, P. M. (1995) Muscle damage following repeated bouts of high force eccentric exercise. Med. Sci. Sports Exerc. 27, 1263–1269 7. McHugh, M. P. (2003) Recent advances in the understanding of the repeated bout effect: the protective effect against muscle damage from a single bout of eccentric exercise. Scand. J. Med. Sci. Sports 13, 88–97 8. Stupka, N., Tarnopolsky, M. A., Yardley, N. J., and Phillips, S. M. (2001) Cellular adaptation to repeated eccentric exercise-induced muscle damage. J. Appl. Physiol. (1985) 91, 1669–1678 9. Mackey, A. L., Brandstetter, S., Schjerling, P., Bojsen-Moller, J., Qvortrup, K., Pedersen, M. M., Doessing, S., Kjaer, M., Magnusson, S. P., and Langberg, H. (2011) Sequenced response of extracellular matrix deadhesion and fibrotic regulators after muscle damage is involved in protection against future injury in human skeletal muscle. FASEB J. 25, 1943–1959 10. Grounds, M. D., Sorokin, L., and White, J. (2005) Strength at the extracellular matrix-muscle interface. Scand. J. Med. Sci. Sports 15, 381–391 11. Boppart, M. D., Burkin, D. J., and Kaufman, S. J. (2006) Alpha7beta1integrin regulates mechanotransduction and prevents skeletal muscle injury. Am. J. Physiol. Cell Physiol. 290, C1660–C1665 12. Kjaer, M. (2004) Role of extracellular matrix in adaptation of tendon and skeletal muscle to mechanical loading. Physiol. Rev. 84, 649–698 13. Thomas, K., Engler, A. J., and Meyer, G. A. (2015) Extracellular matrix regulation in the muscle satellite cell niche. Connect. Tissue Res. 56, 1–8 14. Urciuolo, A., Quarta, M., Morbidoni, V., Gattazzo, F., Molon, S., Grumati, P., Montemurro, F., Tedesco, F. S., Blaauw, B., Cossu, G.,

The FASEB Journal x www.fasebj.org

HYLDAHL ET AL.

15.

16. 17.

18. 19.

20.

21.

22.

23. 24.

25.

26. 27.

28. 29.

Vozzi, G., Rando, T. A., and Bonaldo, P. (2013) Collagen VI regulates satellite cell self-renewal and muscle regeneration. Nat. Commun. 4, 1964 Miller, B. F., Olesen, J. L., Hansen, M., Døssing, S., Crameri, R. M., Welling, R. J., Langberg, H., Flyvbjerg, A., Kjaer, M., Babraj, J. A., Smith, K., and Rennie, M. J. (2005) Coordinated collagen and muscle protein synthesis in human patella tendon and quadriceps muscle after exercise. J. Physiol. 567, 1021–1033 Mackey, A. L., Donnelly, A. E., Turpeenniemi-Hujanen, T., and Roper, H. P. (2004) Skeletal muscle collagen content in humans after high-force eccentric contractions. J. Appl. Physiol. (1985) 97, 197–203 Heinemeier, K. M., Olesen, J. L., Haddad, F., Langberg, H., Kjaer, M., Baldwin, K. M., and Schjerling, P. (2007) Expression of collagen and related growth factors in rat tendon and skeletal muscle in response to specific contraction types. J. Physiol. 582, 1303–1316 Fl¨uck, M., Mund, S. I., Schittny, J. C., Klossner, S., Durieux, A. C., and Giraud, M. N. (2008) Mechano-regulated tenascin-C orchestrates muscle repair. Proc. Natl. Acad. Sci. USA 105, 13662–13667 Lapier, T. K., Burton, H. W., Almon, R., and Cerny, F. (1995) Alterations in intramuscular connective tissue after limb casting affect contraction-induced muscle injury. J. Appl. Physiol. (1985) 78, 1065–1069 Heinemeier, K. M., Olesen, J. L., Haddad, F., Schjerling, P., Baldwin, K. M., and Kjaer, M. (2009) Effect of unloading followed by reloading on expression of collagen and related growth factors in rat tendon and muscle. J. Appl. Physiol. (1985) 106, 178–186 Heinemeier, K. M., Bjerrum, S. S., Schjerling, P., and Kjaer, M. (2013) Expression of extracellular matrix components and related growth factors in human tendon and muscle after acute exercise. Scand. J. Med. Sci. Sports 23, e150–e161 Lindahl, G. E., Chambers, R. C., Papakrivopoulou, J., Dawson, S. J., Jacobsen, M. C., Bishop, J. E., and Laurent, G. J. (2002) Activation of fibroblast procollagen alpha 1(I) transcription by mechanical strain is transforming growth factor-beta-dependent and involves increased binding of CCAAT-binding factor (CBF/NF-Y) at the proximal promoter. J. Biol. Chem. 277, 6153–6161 Pan, X., Chen, Z., Huang, R., Yao, Y., and Ma, G. (2013) Transforming growth factor b1 induces the expression of collagen type I by DNA methylation in cardiac fibroblasts. PLoS ONE 8, e60335 Hyldahl, R. D., Xin, L., Hubal, M. J., Moeckel-Cole, S., Chipkin, S., and Clarkson, P. M. (2011) Activation of nuclear factor-kappaB following muscle eccentric contractions in humans is localized primarily to skeletal muscle-residing pericytes. FASEB J. 25, 2956–2966 Mahoney, D. J., Carey, K., Fu, M. H., Snow, R., Cameron-Smith, D., Parise, G., and Tarnopolsky, M. A. (2004) Real-time RT-PCR analysis of housekeeping genes in human skeletal muscle following acute exercise. Physiol. Genomics 18, 226–231 Wang, Y., Winters, J., and Subramaniam, S. (2012) Functional classification of skeletal muscle networks. I. Normal physiology. J. Appl. Physiol. (1985) 113, 1884–1901 Mukund, K., Mathewson, M., Minamoto, V., Ward, S. R., Subramaniam, S., and Lieber, R. L. (2014) Systems analysis of transcriptional data provides insights into muscle’s biological response to botulinum toxin. Muscle Nerve 50, 744–758 Smith, L. R., Meyer, G., and Lieber, R. L. (2013) Systems analysis of biological networks in skeletal muscle function. Wiley Interdiscip. Rev. Syst. Biol. Med. 5, 55–71 Crameri, R. M., Langberg, H., Teisner, B., Magnusson, P., Schroder, H. D., Olesen, J. L., Jensen, C. H., Koskinen, S., Suetta, C., and Kjaer,

30.

31. 32.

33. 34.

35.

36.

37. 38.

39.

40.

41.

42.

43.

M. (2004) Enhanced procollagen processing in skeletal muscle after a single bout of eccentric loading in humans. Matrix Biol. 23, 259–264 Raastad, T., Owe, S. G., Paulsen, G., Enns, D., Overgaard, K., Crameri, R., Kiil, S., Belcastro, A., Bergersen, L., and Hall´en, J. (2010) Changes in calpain activity, muscle structure, and function after eccentric exercise. Med. Sci. Sports Exerc. 42, 86–95 Murphy-Ullrich, J. E. (2001) The de-adhesive activity of matricellular proteins: is intermediate cell adhesion an adaptive state? J. Clin. Invest. 107, 785–790 Verrecchia, F., Chu, M.-L., and Mauviel, A. (2001) Identification of novel TGF-beta/Smad gene targets in dermal fibroblasts using a combined cDNA microarray/promoter transactivation approach. J. Biol. Chem. 276, 17058–17062 Gavin, T. P., and Wagner, P. D. (2001) Effect of short-term exercise training on angiogenic growth factor gene responses in rats. J. Appl. Physiol. (1985) 90, 1219–1226 Breen, E. C., Johnson, E. C., Wagner, H., Tseng, H. M., Sung, L. A., and Wagner, P. D. (1996) Angiogenic growth factor mRNA responses in muscle to a single bout of exercise. J. Appl. Physiol. (1985) 81, 355–361 Boppart, M. D., Volker, S. E., Alexander, N., Burkin, D. J., and Kaufman, S. J. (2008) Exercise promotes alpha7 integrin gene transcription and protection of skeletal muscle. Am. J. Physiol. Regul. Integr. Comp. Physiol. 295, R1623–R1630 Murphy, M. M., Lawson, J. A., Mathew, S. J., Hutcheson, D. A., and Kardon, G. (2011) Satellite cells, connective tissue fibroblasts and their interactions are crucial for muscle regeneration. Development 138, 3625–3637 Macaluso, F., Brooks, N. E., Niesler, C. U., and Myburgh, K. H. (2013) Satellite cell pool expansion is affected by skeletal muscle characteristics. Muscle Nerve 48, 109–116 McKay, B. R., Toth, K. G., Tarnopolsky, M. A., and Parise, G. (2010) Satellite cell number and cell cycle kinetics in response to acute myotrauma in humans: immunohistochemistry versus flow cytometry. J. Physiol. 588, 3307–3320 Cao, Y., Zhao, Z., Gruszczynska-Biegala, J., and Zolkiewska, A. (2003) Role of metalloprotease disintegrin ADAM12 in determination of quiescent reserve cells during myogenic differentiation in vitro. Mol. Cell. Biol. 23, 6725–6738 Palecek, S. P., Loftus, J. C., Ginsberg, M. H., Lauffenburger, D. A., and Horwitz, A. F. (1997) Integrin-ligand binding properties govern cell migration speed through cell-substratum adhesiveness. Nature 385, 537–540 Watt, D. J., Morgan, J. E., Clifford, M. A., and Partridge, T. A. (1987) The movement of muscle precursor cells between adjacent regenerating muscles in the mouse. Anat. Embryol. (Berl.) 175, 527–536 Sambasivan, R., Yao, R., Kissenpfennig, A., Van Wittenberghe, L., Paldi, A., Gayraud-Morel, B., Guenou, H., Malissen, B., Tajbakhsh, S., and Galy, A. (2011) Pax7-expressing satellite cells are indispensable for adult skeletal muscle regeneration. Development 138, 3647–3656 Paulsen, G., Mikkelsen, U. R., Raastad, T., and Peake, J. M. (2012) Leucocytes, cytokines and satellite cells: what role do they play in muscle damage and regeneration following eccentric exercise? Exerc. Immunol. Rev. 18, 42–97 Received for publication November 11, 2014. Accepted for publication March 9, 2015.

11