Myopathy with trabecular muscle fibers

Neuromuscular Disorders 9 (1999) 208–214

Myopathy with trabecular muscle fibers Boaz Weller a, Stirling Carpenter b, Hanns Lochmu¨ller c, George Karpati d ,* a

The Technion, Haifa, Israel b Porto, Portugal c Genzentrum, Munich, Germany d Neuromuscular Research Group, Montreal Neurological Institute, 3801 University Street, Rm 633, McGill University, Montreal, Quebec H3A 2B4, Canada Received 19 October 1998; received in revised form 1 December 1998; accepted 4 December 1998

Abstract A systematic review of muscle biopsies over a 15 year period in a large neurological hospital revealed 21 cases (7% of the total of noninflammatory myopathies) with a distinctive pattern of myopathology and a limb-girdle clinical phenotype. The muscle pathology was dominated by a large prevalence (20–90%) of trabecular or lobulated fibers in which maldistribution of intermyofibrillar mitochondria produced a lobulated pattern of oxidative enzyme activity on transverse sections. The clinical picture was characterized by adult onset, slowly progressive muscle weakness affecting mainly proximal limb musculature, although mild distal weakness was also present in 60% of the cases. The trabecular pattern of oxidative enzyme reaction reflects maldistribution of the intermyofibrillar mitochondria; this may be caused by malfunction of a putative anchoring mechanism. While trabecular fibers can occur as a nonspecific alteration of muscle fibers in many diverse myopathies, the high prevalence of trabecular fibers as the dominant pathology in trabecular fiber myopathy makes it a distinctive (though not necessarily etiologically homogeneous) clinico-pathological entity.  1999 Elsevier Science B.V. All rights reserved. Keywords: Trabecular fibers; Lobulated fibers; Lacey fibers; Mitochondria; Maldistribution of mitochondria; Cytoskeleton; Trabecular fiber myopathy; Limb-girdle syndrome; Limb-girdle dystrophy

which has been usually grouped in the general category of ‘atypical limb-girdle dystrophy’.

1. Introduction Trabecular, (or lobulated or lacey) muscle fibers have been described in various neuromuscular diseases [1–9]. Trabecular fibers (TF) are characterized by a peculiar pattern of oxidative enzyme reaction on histochemical preparations of muscle biopsies [10] reflecting an abnormal spatial distribution of the intermyofibrillar mitochondria. As a result of careful analysis of an extensive muscle biopsy material of 2199 cases, here we describe a syndrome in 21 patients characterized by slowly, progressive, mainly proximal muscle weakness and a high prevalence of TF as the predominant pathological change in the muscle biopsy. This syndrome, which we call trabecular fiber myopathy (TFM), is a relatively common, adult-onset muscle disease,

* Corresponding author.

0960-8966/99/$ - see front matter PII: S09 60-8966(98)001 30-8

2. Material and methods 2.1. Patients Initially, all muscle biopsy reports of children and adults issued at the Montreal Neurological Hospital between l981 and l996 were carefully reviewed. For this period, detailed clinical and muscle biopsy documentation of all cases was available, as well as frozen muscle biopsy material preserved at −70°C. In this review, all cases that were clinically diagnosed as limb-girdle dystrophy and those found to have trabecular fibers in their muscle biopsy were identified. Their case histories and biopsy material were subsequently analyzed by two of the investigators (B.W. and G.K).

 1999 Elsevier Science B.V. All rights reserved.

B. Weller et al. / Neuromuscular Disorders 9 (1999) 208–214

2.2. Muscle biopsies 2.2.1. Histochemistry and immunocytochemistry: The original biopsy material was used to prepare a panel of histological and histochemical preparations that included transverse and longitudinal sections stained with hematoxylin and eosin, modified Gomori trichrome, and a battery of histochemical techniques including NADH tetrazolium reductase, myosin ATPase at pH 10.2, 4.5 and 3.9 preincubation, cytochrome C oxidase (COX) and succinic dehydrogenase (SDH). Subsequently, immunocytochemical preparations were produced using specific antibodies for three domains of dystrophin (NovoCastra, Newcastle, UK), N-terminus of utrophin (NovoCastra, Newcastle, UK), a-sarcoglycan (a generous gift from Dr. K. Campbell, University of Iowa), a2-laminin (Chemicon International, California, CA), laminin b2 (Sigma) and N-CAM (Beckton Dickinson). 2.2.2. Morphometry of muscle biopsies On a 15 inch monitor, color camera-generated images of transverse muscle sections stained for COX, a2-laminin (for capillary staining) and myosin ATPase with pH 4.6 preincubation were analyzed with the NIH Image Program version 1.59. The following parameters were measured (1) the percentage of trabecular fibers in the total fiber number which was between 200 and1000 (2) the mean cross-sectional area of trabecular and non-trabecular fibers (3) the total area occupied by mitochondria (COX stain) on the cross-section as a percentage of the total transverse fiber area in trabecular and non-trabecular fibers (4) the fiber type representation among trabecular and non-trabecular fibers (5) the ratio of the mean number of capillaries per muscle fiber or per mm2 of transverse cross-sectional area of muscle fibers, calculated from 200 to 1000 fibers per biopsy 2.2.3. Electron microscopy Glutaraldehyde-fixed upon embedded semithin and ultrathin sections were examined by phase and electron microscopy, respectively. 2.2.4. Immunoblot analysis Fifty consecutive 10-mm thick cryostat sections from all TFM patients were suspended in a 100 ml sample buffer. SDS-PAGE was prepared in the usual manner [11]. In addition to the TFM group, muscles from five histologically normal muscle biopsies and three muscle biopsies of occulopharngeal dystrophy patients were used. The immunoblots were reacted with the following monoclonal antibodies: vimentin 53–57 kD (Boehringer–Mannheim) at 1:500 dilution, a and b tubulin both 55 kD (Sigma) at 1:1000 dilution, a and b actin 42 kD (Sigma) at 1:1500 and 1:1000 dilution and desmin 50–55 kd (NovoCastra Laboratories) at 1:1000 dilution. The reaction product was visualized by a peroxidase conjugated antimouse

209

IgG (Dako) and subsequently visualized by an chemiluminescent detection kit (ECL Amersham).

3. Results For the analysis of the results, TF were defined as those showing an irregular pattern of oxidative enzyme reaction on cross-section, in which the normal lattice pattern of the reaction was replaced by an irregular trabecular and coarse granular appearance. As a result, large areas of muscle fibers are devoid of mitochondrial reaction product, and there is often a conspicuous ring-like excess of the reaction product on the fiber periphery. These changes are best visualized by SDH, COX or NADH-tetrazolium reductase reactions. Trabecular fiber myopathy (TFM) was defined as a syndrome in which more than 20% TF in the muscle biopsy constituted the major pathological change, while no other pathological alterations were present in the biopsy, except for smallness of TF. The 20% ‘cut-off’ was used, since in patients with 20% or more TF had no other significant myopathology and showed a non-specific limb-girdle clinical phenotype. The prevalence of TF in biopsies of patients with a known muscle disease was maximum 13%. Of the 2199 total biopsy reports evaluated, 301 were identified as non-inflammatory myopathy. In this group, 21 patients (7%) were found to have TFM according the above criteria. Seventeen other patients were found to have 3–13% TF along with other distinctive myopathological changes or other evidence of a specific muscle disease. This group included a non-specific necrotizing myopathy, a vacuolar myopathy, adult-onset nemaline myopathy and a myotonic myopathy. Of the 21 patients who satisfied the myopathological criteria of TFM, 11 patients (47%) were clinically diagnosed as limb-girdle dystrophy or atypical limb-girdle dystrophy prior to the biopsy. Thirty-one of the 301 patients had the clinical diagnosis of limb-girdle dystrophy prior to biopsy, and 11 (30%) of these turned out to have TFM by the muscle biopsy criteria as defined above. Clinical features of the TMF cohort are given in Table 1. Of the 21 patients, 48% were females and 52% males. The average age at the time of the muscle biopsy was 47.8 ± 16.1 years (range 23 to 79). The onset of symptoms averaged 38.1 years. Only 14% of the patients had a suggestively positive family history. Consanguineous marriage was not present in any of the families. Cramps were reported in four patients, while six patients had signs or symptoms of fatigability. All patients, had limb muscle weakness predominantly in the proximal distribution. Three cases had only upper limb weakness, whereas one patient only had lower limb weakness. One patient showed a mild bulbar involvement. No ptosis or extraocular muscle weakness was noted. Mild respiratory and trunk muscle weakness was present in two cases. Mild neck muscle flexor weakness was present in five cases. Laboratory data is shown in Table 2.

210

B. Weller et al. / Neuromuscular Disorders 9 (1999) 208–214

Table 1 Clinical features of 21 patients with trabecular fiber myopathy Gender Age of diagnosis (years) Age of onset (years) Family history Cramps Fatigability of muscles as a symptom and/or sign

10 (48%) Males, 11 (52%) females Range 23 to 79 (mean 47.8 ± 16.1) Range 5 to 77 (mean 38.1 ± 22.3) Three patients (14%) Four patients (19%) Six patients (29%)

Weakness Proximal limbs Distal limbs Neck flexors Trunk Respiratory muscle weakness Bulbar weakness

20 Patients (95%) 13 Patients (62%) Five patients (24%) Two patients (10%) Two patients (10%) One patient (5%)

Severity of weaknessa

18 Patients (86%); 4/5; three patients (14%): 2–3/5

a

Average score of major muscle groups on MRC scale.

Serum CK level was high in 13/21 patients averaging 1004 IU/l. All patients were euthyroid. 16/21 patients had EMG studies; 12 patients showed small amplitude and polyphasic motor unit potentials with hyperrecruitment. Spontaneous activity was rare. Of the four cases who had 31P invivo MRS, three showed minor delay of exercise-induced phosphocreatine regeneration while one was normal. 3.1. The following is a representative history of a patient with trabecular myopathy A 56-year-old male first noted mild generalized weakness at age 47. Over the subsequent years, the weakness gradually worsened in the proximal muscles of the upper and lower extremities. Examination showed normal extraocular movements without ptosis. The strength of the facial, neck flexor and extensor and respiratory muscles was normal. There was a relatively mild muscle weakness (Grade 4 Medical Research Council (MRC) scale) in the proximal and distal upper and the proximal lower extremity distributions. Stretch reflexes were slightly diminished. Serum creatine kinase activity was 1374-IU/l. EMG showed myopathic motor unit potentials and 31P in-vivo magnetic resonance spectroscopy at rest and on exercise was normal.

3.2. Muscle biopsy findings 3.2.1. Histological and histochemical data None of the patients had abnormal immunostaining with dystrophin, utrophin, a-sarcoglycan, desmin, and a2 laminin. According to the prevalence of the trabecular fibers, the biopsies could be divided into three groups: very high (60% or higher, ten patients), high (40–60%, four patients) and medium (20–40%, seven patients). The TF showed the typical characteristics by histochemistry (Fig. 1) and electron microscopy (Fig. 2) as defined earlier. There was no bias of the percentage of TF in the biopsied muscles (biceps, quadriceps, deltoids). The muscle morphometry results are shown in Table 3. Seventeen patients had biopsies suitable for morphometry. The mean area of trabecular fibers was significantly smaller (69%) than the non-trabecular ones. The mean aggregate mitochondrial volume (deduced from the mitochondrial area) was significantly increased in TF versus non-TF. The overwhelming majority of the TF were of histochemical type 1, while most non-TF were type 2. Necrotic or regenerating fibers or excessive connective tissue were not seen. The capillarity per individual muscle fiber was 22% higher in the trabecular group than in normal controls,

Table 2 Laboratory data Serum creatine kinase activity Normal Abnormal

Six patients (0–150 IU/l) 13 Patients (range 247 to 4009, mean: 1004 IU/l)

EMG Myopathic Non-conclusive

12 Patients Four patients

31

P magnetic resonance spectroscopy (four cases) Normal Non-specific abnormalities

One patient Three patients

B. Weller et al. / Neuromuscular Disorders 9 (1999) 208–214

211

Fig. 1. Transverse cryostat sections reacted for succinic dehydrogenase of a representative case of TFM in which the TF constitute about 65% of the total number of fibers. (a) Typical trabecular fibers are substantially smaller in transverse area than non-TF ( × 63). (b) Typical trabecular fibers show the conspicuous maldistribution of mitochondria (×350). (c) Trabecular fiber pattern is shown on longitudinal view (×425). (d) In many fibers, the intermyofibrillar mitochondria appear displaced to the subsarcolemmal position, creating conspicuous peripheral rings of mitochondria (×100).

which is probably explained by the smaller size of TF than normal ones. On the other hand, the capillarity per mm2 muscle fiber area was 69% higher in the trabecular group versus the normal control group (N = 3) and the oculopharyngeal dystrophy group (Table 4), which probably reflected the type 1 fiber predominance in TFM. 3.2.2. Resin histology and electron microscopy On longitudinal sections, myofibrils showed no Z disc streaming, although there was a tendency for braiding of bundles of myofibrils separated by bands of mitochondria. Mitochondria tended to be elongated, and many were not in

the normal para-Z-disc position. On transverse sections the surface outlines of many fibers were irregular. Clusters of mitochondria were seen in the edges of fibers, often forming triangular masses pointing towards the center of the fiber. On the other hand, fairly extensive myofibrillar areas did not show any mitochondrial profiles. The usual size difference between subsarcolemmal and intermyofibrillar mitochondria was not seen. The mitochondria had normal ultrastructure. 3.2.3. Immunoblot analysis Western blots of vimentin, a-tubulin, b-tubulin, a-actin,

Fig. 2. These electron micrographs show the maldistribution of otherwise normal appearing mitochondria. (a) ×11 000 (b) ×75 600.

212

B. Weller et al. / Neuromuscular Disorders 9 (1999) 208–214

Table 3 Morphometric analysis of muscle biopsies. Trabecular fibers as percentage of total fiber, 20–90 (median, 47). Statistics are by Student’s paired t-test. Trabecular fibers Mean cross sectional area (m2) 2237 ± 1249* Mean mitochondrial area on 98 ± 12*** transverse section (arbitrary units) Fiber type distribution of trabecular fibers (mean + SD) Type I (% of total) 72 ± 16 Type II (% of total) 28 ± 16

Non-trabecular fibers 3246 ± 1099** 89 ± 12****

*Versus**, P , 0.001; *** versus****, P , 0.001.

b-actin and desmin were done on five TFM muscles, four normal controls and four mitochondrial myopathy cases. Tubulin, actin, desmin and porin were normal in the TFM group. Fig. 3 shows the Western blot of vimentin and a tubulin, where vimentin in four of the five TFM patients is in excess, in comparison with the normal controls and the disease controls. A lesser degree of overexpression of a-tubulin was also present in three TMF as compared with five normal controls but not compared with the disease controls.

4. Discussion This paper touches upon four general issues. (i) The nature and the cause of the cellular/molecular defect in TF. (ii) The probable deleterious effects of mitochondrial maldistribution on muscle fiber function. (iii) The issue of specificity of TF. (iv) The question of clinical distinctiveness of TFM. The histochemical and fine structural features of TF indicate that the trabecular (lobulated) pattern is caused by maldistribution of mitochondria. The shape and fine structure of mitochondria by EM showed no significant abnormality, and there was no indirect evidence that the mitochondrial oxidative phosphorylation, as such, was compromised. The higher total aggregate mitochondrial area (measured in arbitrary units) of TF in non-trabecular ones can be explained by the smaller size of TF fibers. The significantly higher mitochondrial area per mm2 muscle fiber area in trabecular myoTable 4 Capillarity of muscle. Statistics are by Student’s paired t-test.

Number of patients Ratio of capillaries per fiber Ratio of capillaries per mm2

Normal control

Trabecular myopathy

Disease control

3 1.61 ± 0.18*

5 1.96 ± 0.36**

4 1.36 ± 0.14***

333 ± 35 +

562 ± 169 + +

271 ± 98 + + +

* Versus **, NS; ** versus ***; P , 0.01; + versus ++, P , 0.05; ++ versus +++ , P , 0.01.

Fig. 3. Immunoblot of muscles from normal controls (N = 5), trabecular myopathy patients (N = 5) and oculopharyngeal dystrophy (N = 4). The vimentin shows an increase in four cases of TFM and a-tubulin also appears higher in both the TFM and disease-control muscles.

pathy muscles versus controls probably reflects type 1 predominance in TFM. Mitochondria in muscles fibers are normally found in four major locations: subsarcolemmal, intermyofibrillar at the para-Z-disc position, perinuclear and at the endplate [12]. In TF, it appears that the intermyofibrillar mitochondria are only situated in irregularly distributed large clusters or bands, creating conspicuous mitochondria-free and mitochondria-rich areas in the fibers. In addition, some of the intermyofibrillar mitochondrial population seems to be displaced to the subsarcolemmal area creating a focal overabundance that appears as conspicuous mitochondrial rings. Alternatively, this may have been due to an absolute increase in organelle number. There is no clear evidence indicating that the perinuclear and the endplate mitochondrial populations also suffer a distributional abnormality. The mechanism by which the intermyofibrillar mitochondria are normally kept in a precise position at the 2 sides of the Z disc is unclear. It is unlikely that this positioning is a result of a passive process. We envisage a cytoskeletal anchoring system that is responsible for the normal spatial mitochondrial distribution. However, microscopic evidence for putative anchoring structures has not been produced. Maldistribution of mitochondria could conceivably occur as a result, either of an absence or abnormality of such a putative mitochondrial anchoring system or of an abnormality of the outer mitochondrial surface to which potential anchoring filaments might be attached [13]. We have investigated candidate molecules that might be present in the putative mitochondrial anchoring system, including desmin, vimentin, b actin and a and b tubulin. None of these molecules showed any major or consistent deficiency or an abnormal mobility on Western blot. There was a significant excess of vimentin which we attributed to a secondary hypervascularization of the muscle but other possible causes could not be ruled out. Vita et al. [5] also studied cytoskeletal molecules in TF, including talin and vinculin and found no abnormalities by microscopic immunocytochemistry.

B. Weller et al. / Neuromuscular Disorders 9 (1999) 208–214

The maldistribution of mitochondria that occurs in TF can be differentiated from and compared to other situations in which the mitochondrial staining pattern shows conspicuous irregularities. These include targets, central cores and multicores, ‘motheaten’ areas, oligomitochondrial fibers and ragged red (blue) fibers. These changes are best visualized by the histochemical reactions for NADH-tetrazolium reductase, succinic dehydrogenase, or cytochrome oxidase. In fibers with targets or cores or multicores, there are circumscribed areas, usually in the central portion, in which mitochondria are scarce or absent [14] and Z disc streaming occurs. In so-called ‘motheaten’ fibers, oxidative enzyme reactions show irregular multiple areas with scarcity or lack of mitochondria but, unlike TF, there are no areas where there is an excess of mitochondria [15]. This pattern is, thus, not necessarily a result of mitochondrial maldistribution. The most extensive ‘motheaten’ changes occur in the reinnervated type 1 fibers that occur in type groups. In oligomitochondrial myopathy, there is an absence or decrease of mitochondria over large areas of many muscle fibers without an increase of mitochondrial staining [16] in any part of the cross sectional area. In ragged red fibers, an excessive number of mitochondria creates a pattern that might be confused with maldistribution. However, in these fibers, conspicuous mito-chondriafree areas are not present, and many of the mitochondria have an abnormal ultrastructure. Lipid globules are generally increased. In many cases the abnormality is not detectable with NADH-tetrazolium reductase since this part of the electron transport chain is defective in the supernumerary mitochondria. The maldistribution of mitochondria in TF seems to have a deleterious effect on muscle fibers and presumably it is the cause of the clinical phenotype. TF are significantly smaller than non-trabecular ones. While individual mitochondria are assumed to show adequate oxidative phosphorylation function, the maldistribution of the mitochondria could create an unevenness of the availability of chemical energy throughout the myofibrillar space. This uneven availability of ATP is presumed to compromise the maintenance of an optimal level of creatine phosphate, most of which is probably concentrated at the M line [17]. This, in turn, could produce deleterious consequences for muscle structure and function. It is also possible that some other aspect of mitochondrial function is involved. The mitochondrial maldistribution is the most conspicuous in type 1 fiber and the TFM muscle shows a significant numerical preponderance of the type 1 fibers, which would enhance the negative effect of mitochondrial maldistribution. The increased capillarity of the muscle actually reflected the numerical type 1 fiber preponderance. Comprehensive muscle or cardiovascular exercise tests of TFM patients could further reveal the impact of mitochondrial maldistribution in type 1 fibers on aerobic mus-

213

cle function [18] and could also determine if graded aerobic training could improve muscle function in this disease [19]. Since trabecular fibers may occur in a wide variety of myopathies, it is clearly not a strict disease-specific alteration of muscle fibers. In our material, TF occurred with a 3– 13% frequency in cases of dermatomyositis, necrotic myopathies, myotonic myopathy and adult onset acid maltase deficiency. Others described TF in facioscapulohumoral dystrophy, distal myopathy, spinal muscular atrophy, late onset acid maltase deficiency and other myopathies [6–8]. It does not appear that any common pathogenic mechanism, such as regeneration or atrophy, predisposes to TF formation. 4.1. Is TFM a distinctive clinico-pathological entity? In the TFM group of patients, the median percentage of TF was 47, and in these patients TF was the only pathological change apart from the smallness of TF. These patients were considered to have TFM. By contrast, the other group of patients whose biopsies contained TF had varied other myopathological features and the maximum percentage of TF was only 13%. In these cases the TF are considered secondary. This explains the 20% dividing value between the two groups: non-specific TF (below 20%) and distinctive TFM (above 20%). In fact, only one TFM patient had 20% TF, most of the values were 40–90%. The association of the distinctive TFM pathological picture with a benign limb-girdle syndrome was first observed by Bethlem et al. [1]. In our material, of the 21 patients with biopsy evidence of TFM, only half were clinically designated as limb-girdle dystrophy before biopsy. Of the 31 cases that carried the clinical diagnosis of limb-girdle dystrophy, 11 (30%) proved to have TFM. These patients showed no evidence of other causes of limb-girdle dystrophy (i.e. sarcoglycanopathy [20,21] or other specific pathological features (dermatomyositis, glycogen storage disease, FSH, etc)). While calpain deficiency [22] was not tested, it is a very rare cause of limb-girdle syndrome. Therefore, we suggest that this cohort of patients represented a distinct entity that may be called TFM. The clinical phenotype is fairly stereotyped. The typical pattern consists of an adult male or female with an insidious onset and slowly progressive course of, mainly, proximal limb muscle weakness causing relatively little functional deficit even many years after the onset. The presence of mild distal limb muscle weakness in about 60% of the cases and a lack of calf hypertrophy, as well as the relatively frequent fatigability of muscle tend to distinguish these cases from the main bulk of limb-girdle dystrophy patients. There is no cardiomyopathy or involvement of extraocular muscles. Serum CK levels may be normal or moderately elevated. Despite the apparently distinctive clinico-pathological phenotype, there still may be an etiologic heterogeneity in TFM. The etiology is not likely to be a directly genetic one,

214

B. Weller et al. / Neuromuscular Disorders 9 (1999) 208–214

since a possible familial occurrence was only present in just over 10% of the cases. Acknowledgements Dr. Douglas Arnold, MD, interpreted the in-vivo MRS results; Dr. Eric Shoubridge PhD, provided helpful advice; Carol Allen provided excellent technical assistance. References [1] Bethlem J, Van Wijngaarden GK, De Jong J. The incidence of lobulated fibers in the facioscapulohumeral type of muscular dystrophy and the limb girdle syndrome. J Neurol Sci 1973;18:351–358. [2] Guerard MJ, Sewry CA, Dubowitz V. Lobulated fibers in neuromuscular diseases. Neurol Sci 1985;69:345–356. [3] Durcan L, Sripathi N, Carpenter S, Karpati G. Myopathy with trabeculated fibers: a distinctive subgroup of limb-girdle dystrophy. Neurology 1994;44 (Suppl. 2):A230–A231. [4] Yamanouchi Y, Arikawa E, Arahata K, Ozawa E, Nonaka I. Limbgirdle muscular dystrophy: clinical and pathologic reevaluation. J Neurol Sci 1995;129:15–20. [5] Vita G, Monici M, Toscano A, Messina C. Lobulated muscle fibers have normal cytoskeletal network. Muscle Nerve 1996;19:931. [6] Fukunaga H, Imaguma M, Taniguchi H, Nakashima H, Osame M, Igata A. A case of nemaline myopathy with core/targoid fibers and lobulated fibers (in Japanese). No To Shinkei 1980;32:965–970. [7] Watanabe S, Kumamoto T, Teramoto H, Okamoto H, Araki S. An adult with type III glycogenosis with lobulated fibers (in Japanese). Rinsho Shinkeigaku 1986;26:801–808. [8] Matsumoto W, Umehara F, Fukunaga H, Okadome T, Osame M. A case of scapuloperoneal atrophy with rigid spine has lobulated fibers in muscle biopsy (in Japanese). Rinsho Shinkeigaku 1991;31: 417– 421. [9] Kumamato T, Ueyama H, Fujimoto S, Nagao S, Tsuda T. Clinicopathologic characteristics of polymyositis patients with numerous tissue eosinophils. Acta Neurol Scand 1996;94:110–114.

[10] Banker BQ, Engel AG. Basic reactions of muscles. In: Engel AG, Franzini-Armstrong C, editors. Myology. New York: McGraw Hill, 1994:832–888. [11] Acsadi G, Lochmu¨ller H, Jani A, et al. Dystrophin expression in muscles of mdx mice after adenovirus-mediated in vivo gene transfer. Hum Gene Ther 1996;7:129–140. [12] Carpenter S, Karpati G. Pathology of skeletal muscle. New York: Churchill Livingston, 1984:176–193. [13] Leterrier JF, Rusakov DA, Nelson BD, Linden M. Interactions between brain mitochondria and cytoskeleton: evidence for specialized outer membrane domains involved in the association of cytoskeleton-associated proteins to mitochondria in situ and in vitro. Microsc Res Tech 1994;27:233–261. [14] Carpenter S, Karpati G. Pathology of skeletal muscle. New York: Churchill Livingston, 1984:434–442. [15] Carpenter S, Karpati G: Pathology of skeletal muscle. New York: Churchill Livingston, 1984:105. [16] Genge A, Karpati G, Shoubridge EA, Carpenter S. Familial myopathy with conspicuous depletion of mitochondria in muscle fibers; a morphologically distinct disease. Neuromusc Disord 1995;5:139– 144. [17] Shoubridge EA, Radda GK. A gated 31P NMR study of tetanic contraction in rat muscle depleted of phosphocreatine. Am J Physiol 1987;252:C532–C542. [18] Haller RG, Lewis SF, Estabrook RW, DiMauro S, Servidei S, Foster DW. Exercise intolerance, lactic acidosis, and abnormal cardiopulmonary regulation in exercise associated with adult skeletal muscle cytochrome C oxidase deficiency. J Clin Invest 1989;84:155–161. [19] Taivassalo T, Matthews PM, De Stefano N, et al. Combined aerobic training and dichloroacetate improve exercise capacity and indices of aerobic metabolism in muscle cytochrome oxidase deficiency. Neurology 1998;50:1055–1060. [20] Hoffman EP. Clinical and histopathological features of abnormalities of the dystrophin-based membrane cytoskeleton. Brain Pathol 1996;6:49–61. [21] Ozawa E, Noguchi S, Mizuno Y, Hagiwara Y, Hoshida M. From dystrophinopathy to sarcoglycanopathy: evaluation of a concept of muscular dystrophy. Muscle Nerve 1998;21:421–438. [22] Beckmann JS, Bushby KM. Advances in the molecular genetics of the limb-girdle type of autosomal recessive progressive muscular dystrophy. Curr Opin Neurol 1996;9:389–393.