Strength training increases the size of the satellite cell pool in type I and II fibres of chronically painful trapezius muscle in females

Non-technical summary

Exercise has been shown to greatly reduce pain in female computer users suffering from neck pain (myalgia), but understanding of differences between healthy and myalgic muscle at the cellular level is poor. We report a large increase in muscle stem cell number in types I and II fibres with heavy strength training in this population. These findings contribute to understanding of the potential of exercise to influence the activity of cells found in the painful skeletal muscle of individuals suffering from myalgia.

Abstract

Abstract

While strength training has been shown to be effective in mediating hypertrophy and reducing pain in trapezius myalgia, responses at the cellular level have not previously been studied. This study investigated the potential of strength training targeting the affected muscles (SST, n = 18) and general fitness training (GFT, n = 16) to augment the satellite cell (SC) and macrophage pools in the trapezius muscles of women diagnosed with trapezius myalgia. A group receiving general health information (REF, n = 8) served as a control. Muscle biopsies were collected from the trapezius muscles of the 42 women (age 44 ± 8 years; mean ± SD) before and after the 10 week intervention period and were analysed by immunohistochemistry for SCs, macrophages and myonuclei. The SC content of type I and II fibres was observed to increase significantly from baseline by 65% and 164%, respectively, with SST (P < 0.0001), together with a significant correlation between the baseline number of SCs and the extent of hypertrophy (r = −0.669, P = 0.005). SST also resulted in a 74% enhancement of the trapezius macrophage content (P < 0.01), accompanied by evidence for the presence of an increased number of actively dividing cells (Ki67⁺) post-SST (P < 0.001). GFT resulted in a significant 23% increase in the SC content of type II fibres, when expressed relative to myonuclear number only (P < 0.05). No changes in the number of myonuclei per fibre or myonuclear domain were detected in any group. These findings provide strong support at the cellular level for the potential of SST to induce a strong myogenic response in this population.

Introduction

In the working population the prevalence of neck pain over a 12-month period has been estimated to be in the range of 27–48% (Fejer et al. 2006; Haldeman et al. 2008); the highest values have been reported among females above 45 years engaged in monotonous repetitive work such as computer work (Sjogaard et al. 2006), and the most common clinical diagnosis among these workers is trapezius myalgia – i.e. painful chronic tenderness and tightness of the upper trapezius muscle (Juul-Kristensen et al. 2006). Despite intensive research, understanding of pathomechanisms at a cellular level is poor. Recently, we reported an altered distribution of satellite cells (SCs) in the trapezius muscle of women suffering from chronic neck pain compared with healthy age-matched controls (Mackey et al. 2010a). The higher density of SCs in the type I fibres of myalgic muscle (Mackey et al. 2010a) compliments previous suggestions of a chronic loading of type I fibres with myalgia (Hägg, 1991; Kadi et al. 1998). In contrast, the type II fibres of the myalgic muscle demonstrated a lower SC content when compared with control muscle (Mackey et al. 2010a). This fibre type distribution pattern of SCs has previously only been reported for elderly individuals in the eighth decade of life (Verdijk et al. 2007; Verney et al. 2008). Whether these findings of an accelerated ageing SC distribution phenotype at a relatively young stage are the result of reduced levels of vigorous physical activity or long term exposure to repeated low-intensity work remains unclear. Other studies have shown that strength training restores the imbalance in the fibre type specific content of SCs, such that the number of SCs associated with type II fibres increases to the level detected for type I fibres (Verney et al. 2008; Verdijk et al. 2009), but it is not known if myalgic muscle can respond at a cellular level in a similar way to the healthy elderly individuals of these previous studies.

Strength training specifically targeting the neck/shoulder muscles effectively reduces pain in individuals diagnosed with trapezius myalgia (Andersen et al. 2008a,c, 2010; Blangsted et al. 2008) and has been reported to result in 20% hypertrophy of type II fibres and a tendency for type I fibres to increase in size by 9% (Andersen et al. 2009). Since the first observations of SCs (Bintliff & Walker, 1960; Mauro, 1961), they have been shown to be vital for successful regeneration of postnatal and adult skeletal muscle (Kuang et al. 2006; Lepper et al. 2009). Similarly, it has traditionally been believed that SC proliferation is required to provide new myonuclei for hypertrophying adult muscle. However evidence for increased SC content under conditions where hypertrophy and damage are negligible (Charifi et al. 2003; Mackey et al. 2007b, 2010b; Verney et al. 2008) has led this to be questioned and raises the possibility that expansion of the SC pool represents a basic physiological adaptation to exercise in healthy skeletal muscle. Myalgic muscle is, however, subject to a different local environment to healthy muscle, both at rest and while performing work (Sjogaard et al. 2010). Furthermore elevated levels of the pro-inflammatory cytokine tumour necrosis factor α have been detected in the skeletal muscle of rats performing low force high repetition tasks for 12 weeks (Elliott et al. 2010), suggesting that chronic repetitive office work could initiate an inflammatory response in the working muscles. Despite findings of similar numbers of macrophages in myalgic and healthy muscle (Mackey et al. 2010a), analyses of human myalgic muscle have revealed several abnormalities such as the presence of type I megafibres, poor capillarisation and evidence of disturbed mitochondrial organisation (Henriksson et al. 1982; Kadi et al. 1998; Larsson et al. 2004; Andersen et al. 2008d), reviewed elsewhere (Hägg, 1991; Visser & van Dieen, 2006). Given the influential role of the immediate environment on SC activity (Kuang et al. 2008), together with the disturbed distribution of SCs reported for myalgic muscle (Mackey et al. 2010a), we hypothesised that, following a period of specific strength training, the SC pool of myalgic muscle would not expand to a magnitude similar to previous reports of healthy individuals and would remain unchanged following general fitness training. In particular the type I fibres targeted to a lesser extent than type II fibres by heavy resistance training may be unresponsive. We further hypothesised that a positive correlation would exist between the number of SCs associated with type II fibres at baseline and subsequent hypertrophy of type II fibres. In line with evidence for a strong influence of macrophages on myogenic cell activity (Arnold et al. 2007), we hypothesised that the SC response to heavy resistance training would mirror the macrophage response.

Methods

Ethical approval

All subjects were informed about the purpose and content of the project and gave written informed consent to participate in the study, which conformed to the Declaration of Helsinki, and was approved by the local ethics committee of Copenhagen, Denmark (KF 01-138/04). The study qualified for registration in the International Standard Randomised Controlled Trial Number Register: ISRCTN87055459.

Study design and participants

A randomised controlled trial was performed in Copenhagen, Denmark. In total, 42 women with clinically diagnosed trapezius myalgia participated (44 ± 8 years, 165 ± 6 cm, 72 ± 15 kg; means ± SD). Exclusion criteria were previous trauma, life threatening diseases, whiplash injury, cardiovascular diseases or arthritis in the neck and shoulder. The participants were active in the labour market and recruited from workplaces with monotonous and repetitive work tasks. All participants went through a clinical investigation of the neck and shoulder, performed by trained clinical personnel who worked together as a calibrated team as described (Juul-Kristensen et al. 2006). Briefly, the main criteria for a clinical diagnosis of trapezius myalgia were (1) chronic or frequent pain in the neck area, (2) tightness of the upper trapezius muscle, and (3) palpable tenderness of the upper trapezius muscle.

Interventions

The first group (specific strength training, SST, n = 18) performed high-intensity strength training with five dumbbell exercises specifically for the shoulder and neck muscles (1-arm row, shoulder abduction, shoulder elevation, reverse flyes and upright row) for 20 min three times a week. The specificity and high level of muscle activation of these exercises have been documented previously (Andersen et al. 2008b). During each session three of the five different exercises were performed for three sets of each exercise with relative loadings of 8–12 repetitions maximum in a periodised and progressive manner.

The second group (general fitness training, GFT, n = 16) performed general fitness training on a bicycle ergometer with relative loadings of 50–70% of the maximal oxygen uptake for 20 min three times a week. The loading was estimated based on relative workload = (working heart rate – resting heart rate)/(max heart rate – resting heart rate), where resting heart rate was set to 70 bpm and max heart rate was estimated as 220 – age. The subjects performed leg-bicycling in an upright position with relaxed shoulders.

The third group (REF, n = 8) was a reference group that received information about health-promoting activities for a total of 1 h per week but were not offered any physical training. Unfortunately, the time-wise successive balanced recruitment resulted in a somewhat smaller REF group compared with the two other groups, e.g. due to withdrawal of participants who initially stated they would volunteer for the study.

Muscle biopsies

Muscle biopsies were extracted under local anaesthesia (1% lidocaine) with a Bergstrom biopsy needle from the upper trapezius muscle at the midpoint between the 7th cervical vertebra and the acromion. Prior inspection with ultrasonography was carried out every time to determine the exact biopsy site. On extraction of the specimen, the fibres were aligned, embedded in Tissue-Tek (Sakura Finetek Europe, Zoeterwoude, The Netherlands) and frozen by immersion in isopentane, precooled by liquid nitrogen. Samples were stored at –80°C. All biopsy samples were assigned a unique identification number, thus blinding the investigator to the participant's identity. Serial transverse sections (10 μm) were cut at –24°C using a cryostat and picked up onto SuperFrost Plus glass slides (Menzel-Gläser, Braunschweig, Germany). The same person carried out all subsequent analysis. Biopsies were analysed in batches, arranged by an investigator not involved in the analysis, such that each batch contained biopsies from all three groups and the pre and post samples from the same individuals.

Immunohistochemistry

All immunohistochemical analyses were carried out by microscopic evaluation of cryo-sections stained with the same batches of primary and secondary antibodies as described in detail recently (Mackey et al. 2010a). Alexa Fluor 488 goat anti-rabbit (Molecular Probes cat. no. A11034; Invitrogen A/S, Taastrup, Denmark) and Alexa Fluor 568 goat anti-mouse (Molecular Probes cat. no. A11031; Invitrogen A/S) secondary antibodies combined with 4′,6-diamidino-2-phenylindole (DAPI) in the mounting medium (Molecular Probes ProLong Gold anti-fade reagent, cat. no. P36935; Invitrogen A/S) rendered the two primary antibody targets red or green and the nuclei blue.

Satellite cells and myonuclei

The primary antibodies for Pax7 (cat. no. MO15020; Neuromics, Edina, MN, USA), Type I myosin (cat. no. A4.951; Developmental Studies Hybridoma Bank, Iowa, IA, USA), and laminin (cat. no. Z0097; Dako Norden A/S, Glostrup, Denmark) were applied to the same section. Sites of Pax7 antigenicity were visualised by diaminobenzidine, visible by light microscopy (Fig. 1). The number of Pax7 cells associated with type I (A4.951⁺) or type II (A4.951⁻) fibres was counted separately and expressed relative to the total number of type I or type II fibres included in the assessment. The same sections were used to determine the number of myonuclei associated with type I and type II fibres.

Figure 1. Immunohistochemical detection of Pax7 cells, type I myosin, and laminin on a cross section of trapezius muscle from a patient suffering from myalgia, before (pre) and after (post) a period of specific strength training (SST).

In these image series, Pax7 cells are visible by light microscopy (brown), and fluorescent staining indicates whether they are associated with type I fibres (A4.951⁺; red) or type II fibres (unstained). Laminin staining (green) defines the fibre borders. Scale bars, 50 μm.

To investigate the relationship between the number of SCs and potential for hypertrophy, a correlation was performed on the number of Pax7 cells associated with type II fibres at baseline and the extent of hypertrophy demonstrated with SST. The fibre area data used in this analysis are from an earlier study of the same muscle biopsies (Andersen et al. 2009).

Macrophages

Macrophages were identified using the mouse anti-CD68 (cat. no. M0718; Dako Norden A/S) and laminin (cat. no. Z0098; Dako Norden A/S) antibodies, as described earlier (Mackey et al. 2010a). The relative number of macrophages was quantified by counting the number of CD68⁺ cells on the cross-section and expressing this number relative to the total number of muscle fibres (×100).

Cellular activity

Double staining with antibodies against CD56 (cat. no. 347740; Becton Dickinson, San Jose, CA, USA) and Ki67 (cat. no. CP249; Biocare Medical, Concord, CA, USA) facilitated the enumeration of cells in the active phase of the cell cycle (Ki67⁺, see Fig. 2), as recently described (Mackey et al. 2009). Ki67⁺ cells were identified and whether these cells were CD56 positive (Ki67⁺CD56⁺) or negative (Ki67⁺CD56⁻) was recorded. The values were expressed relative to the number of fibres in the cross-section as cells per 100 fibres. The total number of Ki67⁺ cells per 100 fibres ([(Ki67⁺CD56⁺) + (Ki67⁺CD56⁻)]× 100) was also calculated from these data.

Figure 2. Triple immunohistochemical staining to detect cells in the active phase of the cell cycle (Ki67⁺) on a cross section of trapezius muscle from a patient suffering from myalgia.

In this series of images, 3 satellite cells are visible (CD56⁺; red) containing a nucleus (DAPI, blue), one of which is Ki67⁺. Scale bar, 50 μm.

Fibre remodelling

Myofibre remodelling was assessed by staining three serial sections with antibodies recognising embryonic myosin (F1.652; Developmental Studies Hybridoma Bank), neonatal myosin (NCL-MHCn; Novocastra, Newcastle upon Tyne, UK), or CD56. The number of embryonic myosin, neonatal myosin, and CD56 positive fibres was counted from the respective stainings and expressed as a percentage of the total number of fibres on that section.

Statistics

The investigator performing the biopsy analyses was blinded. Subsequently, the data were unblinded and statistically analysed by an investigator not involved in the muscle biopsy analyses. Differences in the main variables were tested between the three intervention groups REF, GFT and SST using the Mixed procedure in SAS (SAS Institute, Cary, NC, USA). Factors included in the model were group (REF, GFT and SST), time (pre and post), and group by time. Statistical significance was accepted at the 0.05 level. Data are presented as means ± standard error of the mean (SEM) unless stated otherwise. The fibre remodelling variables (CD56, MHCn and F1.652) and the Ki67 data were analysed by Fisher's exact test to test the frequency of biopsies containing either no positive cells, or one or more positive cells.

Results

The number of fibres included in the assessment of Pax7 cells, myonuclei and macrophages is presented in Table 1. The total number of fibres presented for the macrophage analysis also reflects the size of the sections for the Ki67 cell and myofibre remodelling analyses, performed on adjacent sections.

Table 1.

The number of fibres included in biopsy analyses

	Pax7/F		Myonuclei/F		Macrophages
	Type I fibres	Type II fibres	Type I fibres	Type II fibres	Mixed fibres
REF
Pre	404 ± 146	257 ± 192	30 ± 2	28 ± 2	696 ± 395
Post	378 ± 304	159 ± 98	31 ± 2	26 ± 3	609 ± 363
GFT
Pre	429 ± 214	151 ± 61	30 ± 2	28 ± 4	631 ± 311
Post	350 ± 204	200 ± 170	33 ± 5	25 ± 5	529 ± 344
SST
Pre	410 ± 153	222 ± 153	30 ± 2	29 ± 3	660 ± 265
Post	345 ± 175	150 ± 90	32 ± 5	26 ± 3	514 ± 227

Displayed is the number of type I and type II fibres included in the assessment of satellite cells per fibre (Pax7/F) and myonuclei per fibre (Myonuclei/F). Since no distinction between fibre type was made for the macrophage analysis, the number of fibres presented for this variable includes both type I and II fibres. All analyses were performed on cross sections of trapezius muscle biopsies from women with chronic neck pain before (pre) and after (post) a period of general health information (REF), general fitness training (GFT) or specific strength training (SST). Values are means ± SD.

Satellite cells

As displayed in Table 2, there was a significant group by time interaction for the number of Pax7 cells per fibre associated with both type I fibres (P < 0.05) and type II fibres (P < 0.01). Post hoc analyses revealed a 65% increase in the number of Pax7 cells associated with type I fibres (P < 0.0001) and a 164% increase in the number of Pax7 cells associated with type II fibres (P < 0.0001) with SST only. A similar pattern was observed for Pax7 cells expressed as a proportion of myonuclei. No significant changes with time were observed with REF or GFT except for a 23% increase in Pax7 cells per myonuclear number associated with type II fibres in the GFT group (P < 0.05). Data expressed relative to baseline values are presented in Fig. 3. A significant negative relationship was observed in the SST group between the number of SCs associated with type II fibres at baseline and the extent of hypertrophy demonstrated over the period of training (Fig. 4; Pearson's r = –0.669, P = 0.005, n = 16).

Table 2.

The number of satellite cells expressed relative to the number of myonuclei or fibres

	Relative to myonuclear number			Relative to fibre number
	Type I fibres	Type II fibres	Mixed fibres	Type I fibres	Type II fibres	Mixed fibres
REF
Pre	4.18 ± 0.42	2.96 ± 0.47	3.86 ± 0.32	0.092 ± 0.014	0.053 ± 0.010	0.078 ± 0.010
Post	3.41 ± 0.53	1.80 ± 0.63	2.74 ± 0.43	0.085 ± 0.016	0.050 ± 0.012	0.074 ± 0.012
GFT
Pre	4.10 ± 0.39	2.09 ± 0.47	3.70 ± 0.30	0.124 ± 0.012	0.051 ± 0.009	0.103 ± 0.009
Post	3.90 ± 0.53	2.56 ± 0.63b	3.79 ± 0.43	0.131 ± 0.014	0.073 ± 0.011	0.111 ± 0.010
SST
Pre	3.08 ± 0.28	2.17 ± 0.31	2.90 ± 0.21	0.083 ± 0.010	0.043 ± 0.007	0.069 ± 0.007
Post	4.62 ± 0.29a	4.18 ± 0.34a	4.77 ± 0.23a	0.127 ± 0.010a	0.098 ± 0.007a	0.122 ± 0.007a

Satellite cell number was determined separately for type I and II fibres from cross sections of trapezius muscle biopsies from women with chronic neck pain before (pre) and after (post) a period of general health information (REF), general fitness training (GFT) or specific strength training (SST). A combined value is also provided (mixed fibres). Values are estimated means ± SEM. Group × time interaction (P < 0.01) for all variables.

^aPost hoc analysis, P≤ 0.0001 vs. pre.

^bPost hoc analysis, P≤ 0.05 vs. pre.

Figure 3. Illustration of the change in satellite cell number per fibre (Pax7/F) in myalgic trapezius muscle with a period of general health information (REF), general fitness training (GFT), or specific strength training (SST).

Data are expressed as percentage change from baseline. *P≤ 0.0001 vs. pre from statistical analysis on absolute values.

Figure 4. Plot of the number of satellite cells associated with type II fibres at baseline against the extent of hypertrophy (change in fibre area) subsequently demonstrated by the type II fibres of myalgic trapezius muscle over a period of resistance training.

A significant negative correlation was detected. The dashed lines denote the 95% confidence limits.

Myonuclei

No change in the number of myonuclei per fibre or in myonuclear domain size was detected for any fibre type (Table 3). Similarly, the proportion of fibres presenting central nuclei remained constant between groups and biopsy time points (Table 4).

Table 3.

The number of myonuclei per fibre and the area per myonucleus (myonuclear domain)

	Myonuclear domain			Myonuclei per fibre
	Type I fibres	Type II fibres	Mixed fibres	Type I fibres	Type II fibres	Mixed fibres
REF
Pre	2277 ± 147	1804 ± 135	2056 ± 128	2.13 ± 0.22	1.81 ± 0.22	1.97 ± 0.19
Post	2232 ± 179	1659 ± 178	1988 ± 162	2.48 ± 0.27	2.00 ± 0.26	2.25 ± 0.22
GFT
Pre	2008 ± 130	1719 ± 126	1891 ± 113	2.97 ± 0.20	2.21 ± 0.20	2.64 ± 0.17
Post	1984 ± 160	1709 ± 159	1904 ± 145	2.87 ± 0.24	1.98 ± 0.23	2.47 ± 0.20
SST
Pre	1924 ± 94	1591 ± 87	1777 ± 82	2.50 ± 0.14	2.00 ± 0.14	2.25 ± 0.12
Post	1960 ± 101	1685 ± 95	1831 ± 89	2.61 ± 0.15	2.14 ± 0.15	2.40 ± 0.13

Myonuclear number and domain size was determined separately for type I and II fibres from cross sections of trapezius muscle biopsies from women with chronic neck pain before (pre) and after (post) a period of general health information (REF), general fitness training (GFT) or specific strength training (SST). A combined value is also provided (mixed fibres). Values are estimated means ± SEM.

Table 4.

Central nuclei and macrophages content

	Central nuclei (% fibres)	Macrophagesab (number/100 fibres)
REF
Pre	3.3 ± 1.1	24 ± 5
Post	5.2 ± 1.1	27 ± 5
GFT
Pre	4.2 ± 0.9	27 ± 4
Post	4.2 ± 0.9	37 ± 4
SST
Pre	3.3 ± 0.8	27 ± 3
Post	3.6 ± 0.8	47 ± 3c

The percentage of fibres with one of more centrally located nuclei was determined from cross sections of trapezius muscle biopsies from women with chronic neck pain before (pre) and after (post) a period of general health information (REF), general fitness training (GFT) or specific strength training (SST). The number of macrophages was also assessed in these biopsies. Values are estimated means ± SEM.

^aMain effect of time, P≤ 0.01.

^bMain effect of group, P≤ 0.05.

^cPost hoc analysis, P = 0.001 vs. pre.

Macrophages

A main effect of time and group, but no interaction, was observed for the number of macrophages (Table 4). Post hoc analysis uncovered a significant 74% enhancement of the number of macrophages observed in the SST group after the training period when compared with baseline numbers (P = 0.001). No significant change was detected for REF or GFT.

Cellular activity

Table 5 displays the outcome of the frequency analysis performed for the IHC staining of Ki67. Fisher's exact test detected a significantly greater number of biopsies containing one or more Ki67⁺ cells in the SST group post- vs. pre-training (76.5%vs. 16.7%, respectively; (P < 0.001). A similar pattern was observed for Ki67⁺CD56⁺ and Ki67⁺CD56⁻ cells. No differences were seen in the REF or GFT groups.

Table 5.

Active (Ki67⁺) cells

	Ki67+ CD56−		Ki67+ CD56+		Total Ki67+
	N	n (%)	N	n (%)	N	n (%)
REF
Pre	9	3 (33.3)	9	1 (11.1)	9	3 (33.3)
Post	8	5 (62.5)	8	4 (50.0)	8	5 (62.5)
GFT
Pre	15	3 (20.0)	15	1 (6.7)	15	4 (26.7)
Post	14	6 (42.9)	14	1 (7.1)	13	5 (38.5)
SST
Pre	18	2 (11.1)	18	1 (5.6)	18	3 (16.7)
Post	18	11 (61.1)b	18	7 (38.9)a	17	13 (76.5)c

Cells have been characterised into CD56⁺ or CD56⁻, indicating whether the active cell is a satellite cell (Ki67⁺ CD56⁺) or not (Ki67⁺ CD56⁻). Analysis was performed on cross sections of trapezius muscle biopsies from women with chronic neck pain before (pre) and after (post) a period of general health information (REF), general fitness training (GFT) or specific strength training (SST). The table displays the total number of subjects for each time point (N), and the number (n) and percentage (%) of subjects whose biopsies contained one or more active cells.

^aP < 0.05 vs. Pre.

^bP < 0.01 vs. Pre.

^cP < 0.001 vs. Pre.

Fibre remodelling

No differences were observed over time for the number of biopsies containing one or more fibres positive for CD56 or embryonic myosin (F1.652). A significantly greater number of neonatal myosin (MHCn) fibres was detected in the GFT group post- vs. pre-training (P < 0.05), but not for REF or SST. Data are presented in Table 6, where the mean percentage of affected fibres (calculated from only those biopsies containing affected fibres) is also provided.

Table 6.

Remodelling fibres

	CD56+ fibres			MHCn+ fibres			F1.652+ fibres
	N	n (%)	% fibres mean (SD)	N	n (%)	% fibres mean (SD)	N	n (%)	% fibres mean (SD)
REF
Pre	9	8 (88.9)	1.47 (1.85)	9	0 (0)	n/a	9	2 (22.2)	0.17 (0.12)
Post	8	8 (100)	1.21 (0.76)	8	0 (0)	n/a	7	2 (28.6)	0.49 (0.48)
GFT
Pre	15	11 (73.3)	1.06 (0.86)	15	0 (0)	n/a	15	2 (13.3)	0.16 (0.09)
Post	13	9 (69.2)	1.66 (1.78)	12	4 (33.3)a	0.30 (0.26)	13	1 (7.7)	0.62 (n/a)
SST
Pre	18	14 (77.8)	1.22 (1.31)	18	0 (0)	n/a	18	3 (16.7)	0.39 (0.24)
Post	16	13 (81.3)	0.81 (0.70)	17	2 (11.8)	0.12 (0.02)	17	3 (17.7)	0.44 (0.29)

Markers of myofibre remodelling CD56, neonatal myosin (MHCn) and embryonic myosin (F1.652) were investigated on cross sections of trapezius muscle biopsies from women with chronic neck pain before (pre) and after (post) a period of general health information (REF), general fitness training (GFT) or specific strength training (SST). The table displays the total number of subjects for each time point (N), and the number (n) and percentage (%) of those subjects whose biopsies contained one or more affected fibres.

^aP < 0.05 vs. Pre.

The mean percentage (and SD of this percentage) of affected fibres per biopsy, calculated only from those biopsies containing affected fibres (n), is also provided.

Discussion

The main finding of this study is a strong expansion of the SC pool and macrophage content of the SST-trained myalgic trapezius muscle of women with chronic neck pain. In addition, a lesser increase in SC number, when expressed relative to myonuclear number, was observed in the non-exercised trapezius muscle of the group performing leg GFT. These findings indicate that proliferation of SCs in myalgic muscles is well functioning despite an altered distribution of SCs at baseline when compared to healthy individuals (Mackey et al. 2010a).

SST effects on satellite cells in type I fibres

There are numerous reports of increased SC number with heavy resistance training in healthy young and old individuals (Kadi & Thornell, 2000; Kadi et al. 2004; Olsen et al. 2006; Mackey et al. 2007a, 2010b; Petrella et al. 2008; Verney et al. 2008) and in rats (Schiaffino et al. 1972), although only two studies in their enumeration of SCs have made the distinction between type I and II fibres (Verney et al. 2008; Verdijk et al. 2009). These two studies were performed on healthy elderly men in their early 70s and, while both confirmed increases in the number of SCs associated with type II fibres, no significant increase in SC number was observed in type I fibres in either study. The 65% enhancement of SC number detected in type I fibres in the present study is thus the first time that resistance training has been shown to increase SC content specifically in this fibre type. This finding is contrary to our hypothesis. Based on our previous report of elevated numbers of SCs associated with type I fibres of myalgic muscle vs. healthy control muscle (Mackey et al. 2010a), we hypothesised that the chronic myogenic activity in myalgic muscle would make the SCs unresponsive to loading. It is possible that our current findings can be explained by the lower amount of physical activity performed by the myalgic patients before commencement of training, when compared with age-matched healthy controls, such that the training was a stronger stimulus, relatively, for the myalgic individuals. This is further supported by a tendency for a 9% increase in type I fibre area, measured in the same biopsies (Andersen et al. 2009), which underlines the potency of strength training to target type I fibres starting from a low baseline level. In relation to trapezius myalgia in particular, the finding of such a substantial enhancement of the SC pool of type I fibres in this population in response to a physiological stimulus does not support the presence of a deficiency in SC activation in myalgic muscle.

SST effects on satellite cells in type II fibres

With regard to the SC content of type II fibres, increases of 76% (Verdijk et al. 2009) and 73% (Verney et al. 2008) have been reported earlier in healthy old men following a period of resistance training. Thus the mean increase of 164% in the number of SCs per type II fibre is substantially greater than any previous reports of SC pool expansion. It should be noted, however, that on examination of absolute values, in addition to our low baseline values, our SST post-training level of 0.098 SCs per fibre lies between the 0.084 and 0.117 values reported earlier (Verney et al. 2008; Verdijk et al. 2009). This comparison also suggests that the SCs of the myalgic muscle in our study responded in a similar manner to previous reports for SCs in healthy elderly. At this point equivalent data for young healthy individuals is not available for comparison.

The second major finding relating to SCs associated with type II fibres was the negative correlation in the SST group between the number of SCs associated with type II fibres at baseline and the subsequent hypertrophy exhibited by type II fibres by the end of the 10 week programme. This was in contrast to our hypothesis, which was based on a theory put forward in 2008 by Petrella and colleagues (Petrella et al. 2008). In that study, a group of 66 young and old healthy men and women were retrospectively classified into three groups based on the extent of hypertrophy demonstrated over a 16-week period of resistance training. An interesting observation from that analysis was that the muscle in the extreme responders group, demonstrating the greatest hypertrophy, exhibited a higher SC content at baseline than the non- or moderate responders (Petrella et al. 2008). This finding led to the suggestion that some individuals may have a greater inherent myogenic potential due to a higher availability of SCs (Petrella et al. 2008). While this group was predominantly young men (9 out of 17), both age and sex groups were represented. Since this theory has not been tested in other groups, we were keen to see if the type II fibres of myalgic muscle, characterised by a lower SC content than healthy trapezius muscle (Mackey et al. 2010a), would exhibit this same relationship. Our finding of an opposite relationship though indicates that those with the lowest numbers of SCs in their type II fibres demonstrated the greatest extent of hypertrophy. It can be argued that the low number of SCs in the type II fibres was due to the lack of vigorous activity performed by these women and the potentially consequent smaller type II fibre size (Mackey et al. 2010a). Starting from such a low start point, these individuals were thus likely to demonstrate the greatest gains in muscle fibre size. Whether it is the lack of vigorous activity or the myalgia condition that is behind our negative correlation cannot be evaluated in the current study. Taken together, though, the findings of such a great increase in the SC content of type II fibres with SST, together with 20% myofibre hypertrophy (Andersen et al. 2009), supports the existence of a responsive and well functioning pool of myogenic precursor cells in type II fibres of this population.

GFT effects on non-exercised muscles

Unexpectedly, a significant increase in the SC content of type II fibres, when expressed relative to myonuclear number, was also noted in the GFT group in this study, accompanied by the finding of a significantly greater number of biopsies containing MHCn fibres after GFT. GFT has previously been observed, in the same subjects, to result in improved oxygenation locally in the trapezius muscle while performing repetitive work (Sogaard et al. 2011), which supports the potential of systemic changes induced by bicycle training of the lower extremities to exert local effects on the non-exercised trapezius muscles. In further support of this, the regenerative capacity of SCs from old mice has been shown to improve on exposure to the circulation of young mice (Conboy et al. 2005), providing strong evidence for the potential of systemic factors to influence cell activity locally. Exercise in particular is known to alter circulating levels of cytokines and hormones (Suzuki et al. 2000), and, as an example, testosterone alone has been shown to be a potent stimulator of SC proliferation in the vastus lateralis muscle of healthy elderly men (Sinha-Hikim et al. 2006). The presence of MHCn is usually indicative of muscle fibre development, especially in the early stages (Butler-Browne & Whalen, 1984), and has been reported in adult human muscle subjected to many years of heavy training (Kadi et al. 1999). However, our finding with GFT (and not with SST) is difficult to explain and may be a chance finding. Considering that the change in SC number with GFT was only observed when data were expressed relative to myonuclear number and not to fibre number, and that the magnitude of the change in SC content is minor when compared to that observed with SST, further studies are required to confirm whether systemic changes induced by aerobic exercise have the potential to activate SCs in non-working muscles.

Considerations in the assessment of satellite cells

From a methodological perspective, two of the major differences between this and previous studies are the muscle group (deltoid, vastus lateralis, or trapezius) and SC marker (CD56 or Pax7) used. The identification of SCs on the basis of their expression of CD56 (Petrella et al. 2008; Verney et al. 2008; Verdijk et al. 2009) is in contrast to Pax7 employed in the present study. CD56, also known as Leu19 or NCAM, is the marker used in the majority of studies investigating SCs in human skeletal muscle at the light microscopy level since the initial report of the expression of this protein on SCs (Schubert et al. 1989). Pax7 has also been used successfully by many groups (Reimann et al. 2004; Crameri et al. 2007; Verdijk et al. 2007; Carlson et al. 2009; Lindström & Thornell, 2009; Mackey et al. 2009; 2010a; 2011; Mikkelsen et al. 2009; McKay et al. 2010), and it appears that these two markers generally label the same SCs, any differences being attributed to the different staining profiles, recently investigated in detail (Lindström & Thornell, 2009; Mackey et al. 2009). This is the first study to evaluate SC content of trained human muscle on the basis of Pax7 expression according to type I and II fibres, and the findings are in general agreement with previous reports based on CD56 analysis.

Macrophages

A greater proportion of biopsies containing Ki67⁺CD56⁻ cells was also observed with SST, suggesting that proliferation of cells other than SCs was also required at this time point. While the identity of these cells was not investigated it can be speculated that they include endothelial cells, fibroblasts and macrophages (Schiaffino et al. 1972). In support of the latter, significantly elevated numbers of macrophages were detected post-training in the SST group, in agreement with our hypothesis. Although no other data exist on the macrophage response to resistance training in human skeletal muscle, depletion of macrophages in overloaded mouse muscle has been reported to attenuate hypertrophy (DiPasquale et al. 2007), providing convincing evidence for an essential role for macrophages in loading-induced hypertrophy, reviewed recently (Koh & Pizza, 2009). We had previously reported similar numbers of macrophages in rested trapezius muscle of myalgic and healthy controls (Mackey et al. 2010a), indicating that myalgia does not represent an inflammatory condition in the muscle, at least as assessed by CD68 expression. Furthermore, the increase detected in this study with SST is in line with the animal studies and contributes novel information on the macrophage response to longer periods of regular strength training in humans.

The new satellite cells and macrophages

In addition to reduced pain (Andersen et al. 2008a,c, 2010; Blangsted et al. 2008) and significant hypertrophy (Andersen et al. 2009) in the affected muscles, the findings presented here further underline the beneficial effects of this type of training on myalgic muscle. While our SC findings are in line with studies performed in healthy individuals, it is important to highlight that resting and contracting myalgic muscle is subject to a different local environment from healthy muscle (Sjogaard et al. 2010), and we therefore cannot exclude the possibility that other factors are at play. For example, although one of the fates of new SCs is believed to be incorporation into muscle fibres as new myonuclei, we did not detect any changes in the number of myonuclei per fibre with strength training, suggesting that the 10 week training stimulus served mainly to renew and replenish the SC pool in this population. As well as an increase in absolute number, the Ki67 analysis uncovered a significantly greater proportion of biopsies containing actively dividing SCs (Ki67⁺CD56⁻) after the SST intervention. This suggests that the training stimulus at this 10 week time point is still sufficient to maintain cell proliferation. It is not known at this time how long such an enhancement of the SC pool would be retained, but a previous report of a 70% elevation in SC number in the trapezius muscle of elite power lifters compared to healthy control trapezius muscle (Kadi et al. 1999) strongly supports the idea that continued delivery of a heavy loading stimulus, to which the muscle is accustomed, over many years does not result in a return of muscle SC content to pre-training levels.

A recent report of a sustained elevation of SC and macrophage numbers 30 days after human muscle injury (Mackey et al. 2011) could lead to speculation that the elevated numbers of cells observed in the current study at the 10 week time point are due to damage events occurring earlier in the training period. However in contrast to the recent study (Mackey et al. 2011), no signs of damage or regeneration, as assessed by the presence of central nuclei or fibres positive for developmental myosins, could be observed in the biopsies examined in the present study. It can therefore be argued that the increased cellularity is due to the repeated and accumulated stimuli delivered to the muscle with each training session. While the means by which macrophages facilitate muscle hypertrophy have not been clarified, it appears for example that these cells are a major source of insulin-like growth factor-1 (DiPasquale et al. 2007; Lu et al. 2011), suggesting a function as a producer of muscle anabolic agents. Such a role for SCs cannot at this point be ruled out. Furthermore, in vitro evidence points to a role for macrophages in stimulating myogenic cell activity and promoting muscle fibre repair after injury (Arnold et al. 2007), providing strong support for an interactive relationship between macrophages and SCs. It should also be considered that an increased presence of SCs and macrophages has been reported in the muscle of patients suffering from a variety of muscle diseases (Wakayama et al. 1979; Thornell et al. 2009; Dalakas, 2010). Depending on the disease type, concern has been raised that continuous cycles of degeneration and regeneration could eventually exhaust the reservoir of repair cells, leaving the muscle with a poorer regenerative capacity (Decary et al. 2000), or a lower proliferative capacity of SCs (Renault et al. 2000; Thornell et al. 2009). Given our earlier findings of an elevated number of SCs in myalgic vs. healthy trapezius muscle (Mackey et al. 2010a), it is interesting to note that we did not detect a lower proliferative response following SST, although it can be argued that myalgia does not represent a significant pathological state for the muscle and thus poses a substantially lesser regenerative demand on the muscle than a true inflammatory myopathy. While a clear limitation of the present study is the lack of a healthy cohort of women performing the GFT and SST, it is possible to speculate on our findings in relation to reports in the literature on healthy individuals. We suggest that, while the macrophage response to resistance training in healthy individuals remains to be established, the increase in macrophage content of the SST trained myalgic muscle is a necessary physiological response to resistance training. Furthermore, in line with a potential interaction with SCs, the 74% expansion of macrophage cell number in this study follows the pattern of SC pool expansion.

In conclusion, we demonstrate that two cell populations, SCs and macrophages, are well functioning in myalgic trapezius muscle of women with chronic neck pain, as demonstrated by a vigorous expansion of cell numbers in response to only 10 weeks of SST. This is despite an altered basal distribution of SCs and local environment in the myalgic muscle when compared to healthy individuals. Together with evidence for reduced pain and significant hypertrophy of the affected muscles when targeted with SST, the present findings provide support for the potential of SST to initiate muscle cellular responses in this population that are similar to reports in the literature for healthy individuals subjecting their muscle to comparable forms of loading.

Glossary

Abbreviations

GFT

general fitness training

REF

reference group

SC

satellite cell

SST

specific strength training

Author contributions

A.L.M. was responsible for the analysis and interpretation of data, and drafting the manuscript. L.L.A. was responsible for collection, statistical analysis and interpretation of data, and revising the manuscript critically for intellectual content. U.F. was involved in the analysis and interpretation of data and revising the manuscript critically for intellectual content. G.S. was responsible for the conception and design of the experiment, collection, analysis and interpretation of data, and revising the manuscript critically for intellectual content. All authors have read and approved the final version of the manuscript.