Histochemistry and Cell Biology

, Volume 124, Issue 2, pp 167–175

Skeletal muscle morphology in power-lifters with and without anabolic steroids

Authors

  • Anders Eriksson
    • Department of Integrative Medical Biology, Section for AnatomyUmeå University
    • Department of Health Care, Section for Medical ScienceLuleå University
    • Department of musculoskeletal ResearchGävle University
  • Fawzi Kadi
    • Department of Physical Education and HealthÖrebro University
  • Christer Malm
    • Department of Integrative Medical Biology, Section for AnatomyUmeå University
    • WinternetLuleå university
    • Department of Integrative Medical Biology, Section for AnatomyUmeå University
    • Department of musculoskeletal ResearchGävle University
Original paper

DOI: 10.1007/s00418-005-0029-5

Cite this article as:
Eriksson, A., Kadi, F., Malm, C. et al. Histochem Cell Biol (2005) 124: 167. doi:10.1007/s00418-005-0029-5

Abstract

The morphological appearance of the vastus lateralis (VL) muscle from high-level power-lifters on long-term anabolic steroid supplementation (PAS) and power-lifters never taking anabolic steroids (P) was compared. The effects of long- and short-term supplementation were compared. Enzyme-immunohistochemical investigations were performed to assess muscle fiber type composition, fiber area, number of myonuclei per fiber, internal myonuclei, myonuclear domains and proportion of satellite cells. The PAS group had larger type I, IIA, IIAB and IIC fiber areas (p<0.05). The number of myonuclei/fiber and the proportion of central nuclei were significantly higher in the PAS group (p<0.05). Similar results were seen in the trapezius muscle (T) but additionally, in T the proportion of fibers expressing developmental myosin isoforms was higher in the PAS group compared to the P group. Further, in VL, the PAS group had significantly larger nuclear domains in fibers containing ≥5 myonuclei. The results of AS on VL morphology in this study were similar to previously reported short-term effects of AS on VL. The initial effects from AS appear to be maintained for several years.

Introduction

Investigating the mechanisms of long-term supplementation of AS on human trapezius muscle, Kadi (2000) showed that the enlargement of muscle fibers is accompanied by an increase in the myonuclear number and satellite cell content. In addition the PAS group have a significant higher number of cells expressing fetal myosin heavy chain (MyHC) than power lifters not using anabolic steroids, indicating ongoing new fiber generation.

The alterations in the number of myonuclei and satellite cells in response to AS were later confirmed in a study on the short term effects (20 weeks) of combined treatment with gonadotropin-releasing hormone (GnRH) agonist and testosterone on the vastus lateralis muscle (Sinha-Hikim et al. 2002, 2003). However, the magnitude of muscle fiber hypertrophy was higher in trapezius from athletes with long-term AS supplementation than in muscles from athletes after 20 week of AS supplementation. Another difference was the increased number of centrally located myonuclei in the trapezius muscle (Kadi 2000). These findings might reflect discrepancies in either the response to AS or differences in contraction pattern during exercise between the trapezius and the vastus lateralis muscles.

The vastus lateralis is of importance preferentially in the squat event, when the lifter from an upright position and with the barbell resting across the back of the shoulders, sits or “squats” down by doing a flexion in the knee and hip-joints to a required depth. The lifter then attempts to stand up again, returning to the original position. Such a lift takes approximately 2–5 s and requires a maximal explosive force.

In a dead lift the lifter takes the barbell from the floor to an upright standing position until validated by the referee. In this lift the trapezius muscle performs an isometric contraction for a total time of approximately 8–10 s. These differences in utilization between vastus lateralis and trapezius might be reflected at the muscle fiber level.

The aim of the present investigation was to study the long-term effects of AS on fiber type composition, fiber area, number of myonuclei, internal nuclei, nuclear domains and the number of satellite cells in human vastus lateralis. We also compared the long-term effects of AS on vastus lateralis to (1) the long-term effects of AS on the trapezius muscle, previously reported by Kadi (2000), Kadi et al. (1999b) and (2) to the short-term effects (20 weeks) of AS on the vastus lateralis muscle (Sinha-Hikim et al. 2003).

Material and methods

Subjects

Nineteen power-lifters participated in the present study. Nine of the subjects (31.4±3.3 years) have reported the use of a wide variety of high doses of testosterone and anabolic steroids for a period of 9±3.3 years, whereas ten other power-lifters (27.7±7.5 years), have never used these substances. The drug free group had signed a contract with the local club and Swedish power lifting federation that committed them never to use any drugs. Four of them participated voluntarily as controls in another project aiming to find more effective methods to detect drugs. We were able to perform this study because one of the authors (Anders Eriksson, European and Nordic champion in power lifting 1988) has personal contacts with the Swedish elite power-lifters.

All athletes were highly competitive and participated regularly in Swedish and/or international competitions in power events. They trained regularly four to six times a week, two to three hours per session. The sessions consisted in four to seven sets of exercises and three to twelve repetitions per set.

The nine drug using subjects have been individually interviewed regarding their steroid usage.

Testosterone was used in combination to a variety of anabolic steroids (nandrolone, stanozolol, primobolan, oxymetholone, mastoron, proviron and durobolan). A mean dosage of 938±527 mg testosterone and anabolic steroids were self-distributed each week. In addition, three subjects had used IGF1 (mean dosage 40 mg/day) and one subject growth hormone. Two of them have been caught in regular drug testing. The steroid regimen included both “staking‘’, or simultaneous use of several types at high doses, and “cycling”, a drug-free period followed by times when the doses and the types of drugs taken were increased to a maximum to anticipate peak performance.

All subjects gave their informed consent to participate in the present study. The Ethical Committee of Umeå University approved this work. Written consent in accordance with the policy statement regarding the use of human subjects was obtained from all the subjects.

Muscle samples

Biopsies were taken from the upper ventral part of vastus lateralis muscle using forceps. The samples were mounted in embedding medium (Tissue tek, Miles laboratories, Naperville Ill) and quickly frozen in propane cooled in liquid nitrogen and stored at −80°C until analyzed.

Enzyme histochemistry

Serial 10-μm thick cross-sections were cut at −20°C using a cryostat microtome, mounted on glass cover slips, and air-dried at room temperature. The sections were stained for the demonstration of myofibrillar ATPase (EC 3. 6. 1. 3) after alkaline (pH 10.4 and 9.4) and acid (pH 4.6 and 4.3) preincubations (Dubowitz 1985). Visualization of muscle fiber myonuclei were performed with a Mayer’s hematoxylin (HTX) staining.

Fiber classification

The muscle fibers were classified into types I, IIA, IIAB, IIB and IIC according to their staining intensity for myofibrillar ATPase at alkaline and acid pH. For details see (Kadi 2000).

Immunohistochemistry

Five micrometers thick cross-sections, serial to those used for enzyme histochemistry, were used for immunohistochemical analysis. Myosin heavy chain (MyHC) expression was assessed using well-characterized monoclonal antibodies (mAbs) against human MyHC I (mAb A4.840) and MyHC I& IIA (mAb N2.261) (Hughes et al. 1993). The mAb A 4.840 strongly stained type I fibers, whereas type IIA, IIAB and IIB were unstained. The mAb N2.261 stained strongly type IIA fibers, whereas type I and IIAB were equally weakly stained and type IIB fibers were unstained. Type IIC fibers were strongly stained with mAb N2.261 and moderately stained with mAb A4.840. Type IM fibers exhibited the reversed pattern. For detection of developmental MyHC isoforms sections were stained with mAb against fetal (NCL-MHCn) MyHCs (Ecob-Prince et al. 1989). The antibodies were purchased from Novocastra Laboratories.

Identification of the cell border of the muscle fibers and capillaries was performed with mAb 5H2 against laminin a2 chain (Sewry et al. 1995). Monoclonal Ab 5H2 labels the basement membrane of muscle fibers strongly.

For the identification of myonuclei and satellite cells, sections were double stained with mAb CD 56 against the Leu 19 antigen and Mayer’s hematoxylin. The antibodies were obtained from Becton Dickinson. Leu 19 antigen is a cell–cell recognition molecule expressed during the early stages of fiber formation and in satellite cells. Myonuclei were stained blue whereas a brown rim inside the muscle fiber identified satellite cells (Kadi et al. 1999a). The MHC antibodies were purchased from the American Type Culture Collection.

The proportion of fibers containing fetal MyHCs was calculated as follows: [(number of fibers containing fetal myosin)/(total number of myofibers) × 100].

Immunohistochemical visualization of bound antibody in mAbs against Leu-19, and the MyHCs was performed using indirect peroxidase-antiperoxidase (PAP) staining (Sternberger 1979).

Morphometric analysis

Fiber area, fiber types, myonuclei and satellite cells were analyzed with a light microscope (Zeiss Axiophot, Carl Zeiss, Oberkochen, Germany) connected to an image analysis system (IBAS, Kontron elektronic GMBH, Eching, Germany). Two to four randomly chosen areas from each biopsy were scanned. For analysis of fiber area each biopsy was scanned in sections stained for mAbs against laminin (5H2) and the circumference of each fiber was traced along the periphery of the basement membrane. Counting of number of myonuclei per fiber cross-section and frequency of fibers containing internal myonuclei were performed on sections stained for HTX. Satellite cell frequency was calculated on sections stained for Leu 19 as follows: [(satellite cell number)/(myonuclear number)+satellite cell number) × 100]. This method has previously been used by Kadi et al. (1999a) and Kadi and Thornell (2000). The proportion of developmental MyHCs was measured on whole muscle cross-sections. Classification and measurement of mATPase fiber types and myonuclei were performed with a 20× objective. Counting of fetal MyHC containing fibers was performed with a 10× objective.

Statistical analysis

Data are presented as means and standard deviations. The statistical significance of the differences between the two groups was determined using a t-test for unpaired data. The correlation coefficient (r) was used to determine the degree of relationship between two variables. The statistical significance of correlations between two parameters was determined by using Fichers r to z test.

p-values <0.05 were considered statistically significant.

Results

Fiber types

A mosaic pattern of fiber types in cross-sectioned biopsies was observed in most subjects. The mean values for all fiber types in both groups were very similar but the individual variation was large (Table 1, Fig. 1). Type I and type IIA fibers were the most frequent fiber types in most subjects but their frequency varied from 28 to 62% for type I fibers and from 4 to 69% for type IIA fibers. The proportions of type IIAB and type IIB fibers varied even more. Type IIAB was found in proportions from 0 to 33% and type IIB from 0 to 50%. Interestingly, type IIB fibers occurred preferentially in two subjects in the P group. Fiber type grouping occurred in several of the biopsies from subjects in the PAS group (Fig. 2) but also to some extent in biopsies from subjects in the P group.
Table 1

Mean fiber type and fiber area distribution. No significantly differences were found between the groups

Fibertype

Type I

Type IIA

Type IIAB

Type IIB

Type IIC

Fibers expressing fetal myosin

Proportional fiber type distribution

PAS

40±12

40±16

14±11

0.7±1.6

5.0±2.0

6.7±11

P

46±9

35±20

9±11

7±16

3±3

0.6±0.6

Proportional fiber area distribution

PAS

32±16

45±21

15±13

0.7±1.5

5.1±2.5

 

P

36±9.1

40±21

11±15

9.1±20

2.7±2.5

 
https://static-content.springer.com/image/art%3A10.1007%2Fs00418-005-0029-5/MediaObjects/418_2005_29_Fig1_HTML.jpg
Fig. 1

Immunostained crossections of one PAS subject (A and C) and two P subjects (B, D, E and F). A and B. Most fibers in the PAS (A) and the P (B) subject were of type IIA as evidenced by staining with mAbs N2261 which stains type IIA fibers strongly (IIA) and type I fibers weakly (I). C and D. The outlines of the fibers are visualized with an antibody against laminin. Note the large variation in size of the fibers. Staining within fibers (arrows) indicate some split fibers. E and F. Serial sections stained with mAbs A4840 (E) and N2261 (F) of the subject showing the highest proportion of type IIB fiber. Three fibers are of type I (stained in both E and F (I)) and one fiber is of type II A (strongly stained in F (IIA)) the rest are type IIB and type IIAB fibers (IIAB, IIB). Bar: A, B, E and F 25 μm; C and D 50 μm

https://static-content.springer.com/image/art%3A10.1007%2Fs00418-005-0029-5/MediaObjects/418_2005_29_Fig2_HTML.jpg
Fig. 2

Serial cross sections from PAS subject 1(A–H) and PAS subject 3 (I–P). The sections were stained with mAbs N2.261 (A, E, I and M), A4.840 (B, F, J and N) and NCL-MHC (C, G, K and O) and for NADH dehydrogenase activity (D, H, I and P). Marked areas in (A–D) and (I–L) are shown in higher magnification in (E–H) and (M–P). In both subjects, there was a preponderance of type IIA fibers (strongly stained in A; E; I and M and unstained in B; F; J and N). Note type grouping (*) (group of fibers show the same staining) in I; J and L. In PAS subject 1 a medium sized (enlarged in G) and a number of small fibers (arrows in C) show staining for fetal myosin. In PAS subject 3(K) a high number of small sized fibers are stained (arrows) and sometimes form groups of fibers (area enlarged in O). Sections A–D and I–L: Bar 10 μm; Sections E–H and M–P: Bar 25 μm

The number of fibers expressing fetal MyHC varied considerably between the subjects. In the P group and in five subjects of the PAS group the proportion of fibers expressing fetal myosin was between 0 and 2.0% whereas four subjects in the PAS group had 3.8, 4.1, 12 and 36% of their fibers expressing fetal myosin. The mean values for percent fibers expressing fetal MyHC were 6.7±11% in the PAS group and 0.6±0.6% in the P group (p = 0.12) (Fig. 2, Table 1). No statistically significantly differences were observed in the proportions of fiber types between the PAS and the P groups.

Fiber area

Type IIA and type IIAB fibers had the largest mean fiber area in both groups, but the PAS group had 44% larger type II fibers than the P group (p<0.001) (Table 2). The mean fiber area of both type IIA and IIAB fibers was significantly different between the groups (p<0.05). Similarly, the mean type I fiber area was 61% larger in the PAS group than in the P group (p<0.01) (Table 2).
Table 2

Mean fiber areas for each subject and group values. The PAS group had significantly larger fiber areas for type I, IIA, IIAB and type IM fibers compared to those of the P group (p<0.05)

PAS

Subject

Type I

SD

Type IIA

SD

Type IIAB

SD

Type IIB

SD

Type IIC

SD

Type IM

SD

1

4,453

1,122

14,382

4,334

10,159

2,193

0.0

0.0

0.0

0.0

11,575

8,945

2

11,588

3,308

12,159

3,396

9,300

2,956

0.0

0.0

0.0

0.0

12,412

1,616

3

6,033

1,773

10,062

2,604

0.0

0.0

0.0

0.0

0.0

0.0

13,920

4,402

4

6,401

3,572

10,966

5,460

11,851

4,285

13,714

0.0

0.0

0.0

11,300

4,364

5

6,592

1,479

6,581

1,500

6,382

1,791

5,934

395

1,993

4,616

6,791

0.0

6

7,913

1,777

8,449

1,971

8,689

2,336

0,0.

0.0

12,057

274

6,182

2,465

7

5,088

2,480

10,679

4,756

10,378

2,805

0.0

0.0

10,875

1,511

11,769

6,122

8

11,251

5,276

14,709

3,902

13,044

3,640

0.0

0.0

0.0

0.0

11,212

1,769

9

9,889

3,906

11,231

2,439

10,417

1,394

0.0

0.0

0.0

0.0

8,690

2,413

Mean

7,690*

2,637

11,025*

2,589

10,028*

2,012

9,826

5,498

8,309

5,501

10,428*

2,620

P

1

6,188

1,590

6,109

1,850

9,347

0.0

0.0

0.0

0.0

0.0

0.0

0.0

2

4,394

1,337

8,787

1,995

8,482

1,205

7,877

1,053

0.0

0.0

0.0

0.0

3

4,216

901

7,869

1,245

7,319

2,479

8,134

2,566

0.0

0.0

0.0

0.0

4

3,948

1,124

7,146

2,079

7,954

1,662

7,167

702

0.0

0.0

5,785

0.,0

5

3,030

1,222

6,001

1,236

6,673

1,709

6,852

2.0

3,401

2,347

0.0

0.0

6

5,361

1,259

7,571

2,057

7,829

1,368

0.0

0.0

5,232

0.0

7,934

4,576

7

5,461

1,626

7,428

1,901

0.0

0.0

0.0

0.0

0.0

0.0

6,220

1,738

8

3,506

931

5,739

1,246

5,212

1,236

0.0

0.0

7,063

0.0

5,763

1,610

9

6,544

1,280

9,827

3,699

0.0

0.0

0.0

0.0

0.0

0.0

8,854

3,187

10

5,044

1,541

6,621

1,478

0.0

0.0

0.0

0.0

0.0

0.0

6,910

3,549

Mean

4,770

1,145

7,310

1,293

7,545

1,331

7,508

598

5,232

1,831

6,911

1,256

Myonuclei

The mean number of myonuclei/fiber was 32% higher (p = 0.0001) in the PAS group (6.0±2.2) compared to the P group (4.5±1.6) (Table 3).
Table 3

Mean number of myonuclei per fiber, fibers expressing internal myonuclei and proportion of satellite cells for each subject. The abbreviation (*) means significantly different (p<0.05) between the groups

PAS

Subject

Nuclei/fiber

% fibers with internal nuclei

Satellite cells % of total nuclei

1

7.7

57.7

6.3

2

7.6

19.2

15.5

3

4.8

15.2

11.1

4

5.3

8.3

15.0

5

5.1

6.3

3.8

6

7.0

37.0

9.2

7

5.8

41.5

8.5

8

6.1

42.9

7.7

9

6.8

36.1

6.3

Mean

6.2±1.1*

9.1±5.5*

9.3±4.0

P

1

4.9

19.2

7.3

2

3.7

20.8

8.4

3

4.6

3.8

8.0

4

4.4

2.3

13.8

5

5.2

5.0

13.1

6

6.3

29.5

11.0

7

3.7

0.0

10.9

8

3.8

2.0

6.0

9

6.4

2.6

4.8

10

3.6

8.1

11.1

Mean

4.7±1.0

5.4±5.8

9.4±3.0

Compared to the P group, the PAS group had significantly higher proportion of fibers containing internal myonuclei (29±18 vs 9.3 ±10%) (p = 0.007) (Table 3).

When the number of myonuclei from both the PAS and P groups was plotted against fiber area, a highly significant correlation was found between the fiber area and the number of myonuclei/fiber (r = 0.73, p = 0.0002) (Fig. 3). The mean nuclear domain was 1,531±209 μm2 in the PAS group and 1,348±267 μm2 in the P group. The difference was significant in fibers containing ≥5 myonuclei per fiber (Fig. 4) (p = 0.011). When both groups were combined, a statistically significant correlation was found between the proportion of fibers expressing internal myonuclei and fiber area (r = 0.68, p = 0.0009) (Fig. 5).
https://static-content.springer.com/image/art%3A10.1007%2Fs00418-005-0029-5/MediaObjects/418_2005_29_Fig3_HTML.gif
Fig. 3

Relationship between the mean myonuclei number per fiber and the cross-sectional area in type I and type II fibers from the PAS and the P groups. Correlation coefficient r = 0.73, p<0.01

https://static-content.springer.com/image/art%3A10.1007%2Fs00418-005-0029-5/MediaObjects/418_2005_29_Fig4_HTML.gif
Fig. 4

Muscle fibers separated into two classes on basis of their mean number of myonuclei. The size of the nuclear domains were calculated as a range between mean fiber area and number of nuclei per fiber. One PAS subject (no standard deviation) and seven P subjects formed the class with less than five myonuclei/fiber) whereas eight PAS subjects and three P subjects formed the class with five myonuclei or more. In this latter class, the PAS group had significantly larger myonuclear domains compared to the P group (p = 0.011)

https://static-content.springer.com/image/art%3A10.1007%2Fs00418-005-0029-5/MediaObjects/418_2005_29_Fig5_HTML.gif
Fig. 5

Relationship between the proportion of fibers expressing internal myonuclei and the mean fiber area in the vastus lateralis muscles from the PAS and the P groups. Correlation coefficient r = 0.68, (p<0.01)

Satellite cells

The proportions of satellite cells in the PAS and P groups were practically identical (9.3±4.0 vs 9.4±3.0%, respectively) (Table 3).

Discussion

To our knowledge, this is the first comparative morphological study of the human vastus lateralis muscle in two groups of high-level power-lifter athletes. One group admitted supplementation with testosterone, anabolic steroids and other banned substances for nearly ten years and the other group affirmed never taken banned substances. As we previously have studied the trapezius muscle from the same individuals (Kadi et al. 1999b) it allows us now to compare the AS influence on two muscles used in different ways in power lifting.

The vastus lateralis muscle demonstrated larger muscle fiber areas than normal in both groups (Staron et al. 2000). However the PAS subjects differed from the P subjects in having significantly larger fiber areas, more myonuclei per fiber and more internal myonuclei. No differences were seen with regard to fiber type proportions and frequency of satellite cells. All subjects in the PAS group and seven in the P group had small fibers expressing fetal myosin but there was no significant difference between the groups.

The PAS group had 61% larger type I fiber area and 44% larger type II fiber area than the P group. In the trapezius, the areas were 58 and 33% larger in type I and type II fibers, respectively (Kadi et al. 1999b). In fact, AS supplementation, even without strength training has been reported to induce hypertrophy in human skeletal muscles (Bhasin et al. 1996; Sinha-Hikim et al. 2002). In the study by Sinha-Hikim et al. (2002) the muscle fiber hypertrophy (fiber area) after 20 weeks of AS supplementation (600 mg/week) was 49% in type I fibers and 36% in type II fibers compared to baseline. The dramatic hypertrophic effect on muscle fibers in subjects supplementing with AS is in accordance with the current conception on the effects of testosterone and anabolic steroids, for review see (Herbst and Bhasin 2004). Maximal force of a muscle is related to the muscle fiber area, the total muscle area and the fiber types (Bruce et al. 1997; Bamman et al. 2000).

In the present study we observed that there was a highly significant positive correlation between fiber area and total number of myonuclei in the vastus lateralis muscle (Fig. 3). The PAS subjects, who had the largest fibers, also had the largest numbers of myonuclei, both subsarclemmal and centrally located. The same correlation was observed in our study on the trapezius muscle (Kadi et al. 1999b). In the study from Sinha-Hikim et al. (2002) the number of myonuclei was significantly increased and was correlated to fiber area. Altogether, these results support the idea that the number of myonuclei plays a mechanistic role in muscle fiber hypertrophy (Edgerton and Roy 1991; Allen et al. 1999; Kadi 2000).

It is well known that each nucleus supports a certain volume of the cytoplasm with mRNA for turnover of proteins. This volume is often referred to as a nuclear domain (Cheek 1985). If a myonucleus can expand its nuclear domain by increased synthesis of mRNA or by more efficient transport of mRNA has been discussed (Sinha-Hikim et al. 2003). Recently Kadi et al. (2004) reported that satellite cells are plastic in response to resistance training and that moderate changes in skeletal muscle fiber area can be achieved without addition of new myonuclei. However, our data suggest that addition of myonuclei is a prerequisite for more substantial muscle fibers hypertrophy (Kadi et al. 1999b).

In myopathology it is considered pathological when more than 3% of the fibers contain centrally located nuclei (Greenfield 1957). Centrally located nuclei in muscle fibers of strength-trained subjects might be a phenomenon of adaptation. The centrally located nuclei might be needed to support extremely large fibers, preferentially present in the PAS group. Centrally located nuclei will reduce the diffusion distances from a nucleus to central parts of the myofiber. The observation of large number of fibers with internal nuclei, 29% in the PAS group and 9% in the P group, is of interest in this context. Our previous results on internal nuclei in the trapezius (25% in the PAS group and 5% in the P group (Kadi et al. 1999b)) support this concept. We thus propose that the presence of internal nuclei reflects the limited volume of each nuclear domain, although the mechanisms of internalization of myonuclei remain unknown.

The PAS group had larger nuclear domains in fibers containing more than five myonuclei per fiber compared to the P group (1,531±209 vs 1,348±267 μm2, p<0.01) (Fig. 4). Further, if combining the results from both vastus and trapezius, the difference is 13% (1,656±254 μm2 in the PAS group and 1,463±325 μm2 in the P group) (p<0.05). The suggestion that AS also affects the size of the myonuclear domains is supported by the significantly larger nuclear domains in subjects treated with 300 or 600 mg of testosterone per week compared to controls (Sinha-Hikim et al. 2003). Also, strength training will increase myonuclear domains, at least in older men (Hikida et al. 2000).

Myonuclei in mature muscle fibers are not able to divide, which means that an increase in myonuclei number must come from an external source, reviewed by (Allen 1999). It is generally accepted that these additional nuclei comes from satellite cells (SC) and/or stem cells (for review see Morgan 2003). In both the PAS and P groups we observed a larger proportion of SCs than has been reported for control subjects (for review see Hawke and Garry 2001). However, no significantly differences were observed between the PAS and the P groups, neither in the vastus nor the trapezius muscle. One possible explanation for the higher number of nuclei per fiber in the PAS group can be an increased turnover rate of the satellite cells. A significant increase in satellite cell number has also been observed in young men after supplementation with 300 and 600 mg of testosterone per week for 20 weeks, even without exercise (Sinha-Hikim et al. 2003). Thus, supplementation of testosterone and anabolic steroids as well as high-level resistance training increases the number of satellite cells.

In this study, the fiber typing was based on the myofibrillar ATPase activity after alkaline and acid preincubation (Hughes et al. 1993). The accuracy of the fiber typing was secured on basis of immunohistochemical staining for the different MyHC’s, which is the basis for the myofibrillar ATPase activity (Schiaffino and Reggiani 1994). Human muscle type I fibers contain slow MyHC, the type IIA fibers MyHC IIA and IIB fibers MyHC IIX. The MyHC IIB, which is the fastest MyHC in rodents, is not present in human limb muscles (Smerdu et al. 1994; Ennion et al. 1995). In humans, the order of contraction force and speed of the MyHC’s is I-IIA-IIX where I is the slowest and weakest and IIX is the fastest and strongest (Hilber et al. 1997; Larsson et al. 1997).

The strength-trained athletes, both the PAS and the P subjects, had a high frequency of type II fibers and mainly type IIA fibers. A variation in fiber type ATPase activity was commonly observed, which has been related to the fibers content of different MyHCs (Staron 1997; Pette and Staron 2001). This is typical for highly trained subjects, indicating transformation of fibers to optimize performance (For review see Pette and Staron 2001). In a previous study, we did not observe a significant difference in proportion of fiber type distribution between the PAS and the P groups in the trapezius muscle (Kadi 2000). No change in the relative proportion of fiber types was observed in subjects treated with testosterone for 20 weeks (Sinha-Hikim et al. 2002). Thus, testosterone and anabolic steroids do not seem to affect the relative fiber type distribution in human skeletal muscles. From the view of being successful in sports, examining the individual data for the subjects revealed that each individual athlete actually showed a personal muscle profile. A person with higher proportion of type IIB fibers would, from a theoretical point of view, have a greater chance of becoming a successful power lifter. Interestingly, the subject who had the highest percentage of type IIB fibers (Fig. 1E, F), with approximately half of the muscle fibers being type IIB, had at the time for the biopsy the world record in his weight class in squat and he was not taking AS. Conversely, the subjects with the largest fiber areas were found in the PAS group, which means that these subjects in that respect had an advantage over those in the P group with smaller areas. Our study also shows that several subjects in the PAS group had a large amount of small fibers expressing fetal myosin. Adult muscle fibers do not normally express fetal myosin. The presence of developing myosin isoforms has been interpreted as signs of hyperplasia (McCormick and Schultz 1992; McCormick and Thomas 1992; Antonio and Gonyea 1993). A significant increase in fibers stained for developing isoforms of myosin, has been demonstrated with strength training (Antonio and Gonyea 1993; Kadi and Thornell 1999; McCall et al. 1996; Kadi 2000). These fibers might reflect newly formed fibers or abortively regenerated fibers. In the latter case, failed innervation can cause degeneration and the new fibers would be of no use for the athletes. It can be speculated that the increased number of fibers with fetal myosin is a result from using AS for almost ten years.

In conclusion, these results suggest that the action from AS is similar in the vastus lateralis, both long-term and 20 weeks supplemented, and in the trapezius muscles despite differences in contraction pattern. The results are in agreement according to larger fiber areas, correlation between myonuclei number per fiber and fiber area and to an increased number of myonuclei and satellite cells.

Acknowledgements

We thank Margareta Enerstedt, Mona Lindström and Lena Carlsson for excellent technical assistance. This study was supported by grants from the Swedish National Centre for research in sports (90/98, 79/99), the Swedish Research Council (12X-03934) and the Medical faculty of Umeå University.

Copyright information

© Springer-Verlag 2005