Ultrastructural fiber typing in normal and diseased human muscle

Journal Of the neurological Sciences, 1975, 25 : 99-108

99

~' Elsevier Scientific Publishing Company, A m s t e r d a m - Printed in The Netherlands

Ultrastructural Fiber Typing in Normal and Diseased Human Muscle CLAIRE M. PAYNE, LAWRENCE Z. STERN, RICHARD G. CURLESS AND LINDA K. HANNAPEL

Departments of Pathology (C.M.P.), Neurology (L.Z.S.; L.K.H.) and Pediatrics (R.G.C.), University qf Arizona College of Medicine, Tucson, Ariz. (U.S.A.) (Received 6 November. 1974)

INTRODUCTION

Histochemical analysis of muscle biopsies is well-established as being essential for the definitive evaluation of neuromuscular disorders (W. K. Engel 1962). The usefulness of electron microscopy as both a diagnostic and research tool in the study of diseased skeletal muscle has also been demonstrated (A. G. Engel 1966 ; Stern, Payne and Hannapel 1974). Reliable cross-correlation between histochemical and electronmicroscopic findings would appear to be useful. Based on the variations of the histochemical reaction for myofibrillar ATPase, Brooke and Kaiser (1970) found Types i, IIA and IIB to be the three constant categories in normal human skeletal muscle. Although a greater number of fiber types have been described histochemically (Romanul 1964), the three-fiber classification seems more practical. An attempt has therefore been made to define the ultrastructural correlates of each of these three major fiber types based on the measurements of Z-line and M-line widths. Certain muscles show a predominance of one or two of the histochemical fiber types making ultrastructural correlation easier (Padykula and Gauthier 1967a; Schiaffino, Hanzlikova and Pierobon 1970; Gauthier 1971; Shafiq, Askanas and Milhorat 1971). We have recently had the opportunity to study biopsied muscle from patients with 3 muscle disease states in which there was a relative deficiency of at least one of the 3 fiber types. Our correlated histochemical and electron-microscopic results prompt this report.

CLINICAL SUMMARIES

Control patient~ ( 77o objeetit~e el~idence q] neuromuscular disease) Patient H.J. This 57-year-old w o m a n underwent a vastus lateralis muscle biopsy for evaluation of myalgia

This work was supported in part by a grant from the Muscular Dystrophy Associations of America (to Dr. L. Z. Stern and Dr. R. Gruener), NIH Grant NS-10417, and by a gift from Mr. and Mrs. Edward Watz.

100

C. M. PAYNE, L. Z. STERN, R. G. CURLESS, L. K. HANNAPEI.

of 9 months. Her physical and laboratory examinations were within normal limits except lot a mildl) elevated sedimentation rate of 28 mm/hr. Serum CK. L D H and G O T were normal as were rc,nlls of,qty. tromyography and nerve conduction velocity studies. Patient R.H. This 26-year-old w o m a n without neuromuscular disease underwent a diagnostic thoracotomy for a bronchial adenoma. At the time of thoracotomy a specimen of externa) intercostal muscle was removed for histochemical and electron-microscopic study. Patient M.S. This 59-year-old w o m a n without neuromuscular disease underwent a diagnostic thoracotomy for a s q u a m o u s cell hronchogenic carcinoma. At the time of thoracotomy a specimen of external intercostal muscle ~ , removed for h i~toehemical and electron-microscopic study. Palienl C. ('. Thi~ ~5-,,c~r-old man ~ ith previously reported elevations of serum CK underwent a gastro~nemius muscle biopsy for ex aluation of severe pain in the legs. Serum CK, LDH and G O T wevc normal as were the ESR and results of electromyography and nerve conduction velocity studies. Bi(~F},\ resulls were also normal, and no diagnosis has been established. Patients with evidence o/ a neuromuscular disorder Patient S.T. This 26-year-old w o m a n underwent an external intercostal muscle biopsy 1o~ evaluation of progressive proximal muscle weakness and wasting beginning approximately 15 years before. There was a family history of muscular dystrophy with 2 of her 3 brothers similarly affected. On examination, she had the clinical findings of moderately advanced dystrophy but was still ambulatory. Laboratory evaluation revealed an elevated serum C K of 260 IU (normal 5 7 0 IU). Serum L D H and G O T were normal. Results of electromyography were consistent with the clinical diagnosis of limb-girdle muscular dystrophy. Patient H.B. This 59-year-old woman, who had been followed for several years at the University of Arizona Neurology Clinic for complaints referable to cervical spondylosis, developed proximal muscle weakness in her extremities 6 m o n t h s prior to admission. Evaluation revealed that she had developed acromegaly. She underwent a vastus lateralis muscle biopsy. Serum CK, G O T and L D H had been normal. but results of electromyography were consistent with a myopathic process. Nerve conduction velocity studies were within normal limits. The diagnosis was acromegalic myopathy. Patient M.R. This l I-month-old child was born after an uncomplicated labor, but there was mild hypotonia during the neonatal period. Her motor milestones were delayed. During the first 4 m o n t h s of life, she developed dysphagia which improved by 8 m o n t h s of age. At 11 months, she was sitting up and crawled slowly with considerable difficulty. Her head circumference of 43.1 cm was on the minus 2 standard deviation mark for her age. She had generalized hypotonia which was greatest in the trunk and lower extremities. The tendon reflexes were normal. An E M G was within normal limits. A biopsy of the left vastus lateralis muscle was performed which revealed no morphological abnormalities. The diagnosis was one of probable cerebral hypotonia.

MATERIALS AND METHODS

H istochemistr y

Muscle biopsy specimens were frozen in isopentane cooled in liquid nitrogen and cross-sectioned at 10/~m thickness in a cryostat. Sections were stained with the modified Gomori trichrome (Engel and Cunningham 1963) and with the succinic dehydrogenase (SDH) (Nachlas, Tsou, DeSouza, Chang and Seligman 1957) and nicotinamide adenine dinucleotide tetrazotium reductase (NADH-TR) reactions (Farber, Sternberg and Dunlap 1956). Muscle fiber typing was done using the pH 9.4 myofibrillar adenosine triphosphatase (ATPase) reaction (Padykula and Herman 1955). Histochemical subtyping of the Type II fibers was accomplished using the acid preincubation techniques described by Brooke and Kaiser (1970). Electron microscopy

Specimens for electron microscopy were fixed for 2 hr in ice-cold 30/0 glutaraldehyde buffered with 0.1 M phosphate buffer (pH 7.2). The tissue was then transferred to 0.1 M phosphate buffer (containing sucrose) and post-fixed for 1½ hr in cold 1 ~Jo osmium tetroxide buffered with 0.1 M phosphate. The tissue was dehydrated through a graded

ULTRASTRUCTURAL FIBER TYPING IN HUMAN MUSCLE

101

series of alcohol and embedded in Spurr's low-viscosity epoxy (Spurr 1969). Semithin sections (0.5 #m) were cut and stained with toluidine blue and examined with the light microscope. Ultrathin sections of the selected areas were cut on a Sorvall MT2-B ultra-microtome and mounted on uncoated 200-mesh copper grids. Increased contrast was obtained by staining in 5~o aqueous uranyl acetate followed by lead citrate. The grids were lightly carbon-coated before examination with the Hitachi HU-12 electron microscope. Z- and M-line measurements

All negatives of muscle fibers were taken at 7,000 magnification with the pre-calibrated electron microscope. Then 8 × 10 cm prints were made from each negative and the final magnification determined. Only those fibers with clearly demarcated Z-lines and M-lines were chosen for measurement. The 8 × 10 cm prints were placed on a light box, examined with a calibrated lens and Z-lines and M-lines measured to the nearest 0.1 mm. From each of 4 control patients, approximately 100 Z-lines and M-lines were measured from 3 muscle fibers of each histochemical type. The same measurements were done for each of 4 fibers of each histochemical type from each of the 3 patients with neuromuscular disorders. The mean widths were determined for each muscle fiber measured. The overall mean was then obtained for each of the 3 histochemical fiber types from each of the 7 patients. Mean Z-line and M-line widths for the 3 histochemical fiber types in each patient were then compared using Student's t-test.

RESULTS

Control muscle

Histochemical Type I fibers were identified ultrastructurally in all 4 control muscles as those having the largest and most numerous mitochondria (Fig. IA), Type IIA as having smaller and less numerous mitochondria (Fig. 1B) and Type IIB as having the smallest and least numerous mitochondria (Fig. 1C). Mitochondria were located on either side of the Z-line at the A-I junction, but sometimes extended along the length of the sarcomere thereby separating one myofibril from another. Since the mitochondria were plentiful in both the Type I and IIA fibers, myofibrils were clearly delineated (Fig. 1A, B). On the other hand, mitochondria being relatively sparse in the IIB fibers made it difficult to distinguish the individual myofibrils (Fig. 1C). Lipid droplets were most abundant in the Type I fiber, moderate in the IIA and scarce in the IIB fiber. In most cases, the lipid droplets were found in close proximity to mitochondria. Calculations of the mean Z-line width and mean M-line width from each of the 3 fiber types from 4 control muscles are shown in Table I. The Type I fibers had the widest Z-lines (95 nm), whereas the Type IIA and IIB fibers were both considerably (P < 0.001) narrower (74 and 69 nm, respectively). One can easily distinguish between the I and II fibers on the basis of Z-line width but not between the IIA and IIB fibers. However, comparison of M-line widths shows that the IIB fiber has an appreciably (P < 0.001) narrower M-line (60 nm) than that of the IIA fiber (79 nm). The IIA fiber can therefore be considered as being intermediate between the I and IIB fiber, the

Fig. l. l.Jltrastructural correlates of the 3 major histochcmica[ fiber types in control vastus !a~t:~alis muscle (patient H.J,). ,,I : Type 1 fibers are shown to have tile largest and most n u m e r o u s mitochtmdria, Lipid droplets (L) are most n u m e r o u s in this fiber type : B : Type 11A have smaller and less numerou,~ rail ochondria : C: Type liB have lhe smallest and least n u m e r o u s mitochondria. Myofibrils are not ck'z,1~13 delineated in this fiber type, [ 7ranyl acetate, lead citrate. ~ 14,000 TABII: ( ()MPARAIIVI:

I

~,VII)'I I | S ()1' Z - I I N I S , \ N I t M - I I N I S I R ( ) M ( O N t R ( ) I

Fiber type 1 I IA 1t B

Z-liHe.~ 94.7+_ 1.3 (402} 73,6q::1.1 (437} 68.7 + 1.1 (479)

Ill NI~N NI~S( i I

,~4-1ine,s 89.4+_1.1 (373) 79,1 +_ 1.3 (430) 59.5 +_ 1.3 (440)

Values are means in nm ± SEM. N u m b e r s in parentheses indicate total n u m b e r of lines measured from 12 separate fibers from 4 control patients, P-values are stated in text.

103

ULTRASTRUCTURAL FIBER TYPING IN HUMAN MUSCLE

M-line width being closer to that of the Type I fiber but the Z-line width being closer to that of the IIB fiber. Another characteristic which distinguishes the liB fiber from the other two fiber types is the sarcomere length. In all 4 control muscles, the Type liB fibers consistently had the shortest sarcomere length, differing from the other two fiber types by 100-300 rim.

No consistent difference in the content and distribution of the sarcoplasmic reticulum among the 1, IIA and IIB fibers was seen in the control muscle examined. The glycogen content of the 3 fiber types was so highly variable as to make any apparent differences between individual fibers insignificant.

Diseased muscle Mitochondrial content cannot be used to identify fiber types in diseased muscle since mitochondria are among the first organelles to be affected in many pathological states. Our results with control muscle show that it is possible to identify histochemical Type I, IIA and IIB fibers ultrastructurally on the basis of their Z-line and M-line widths. If this system is to have practical application, one should be able to identify fiber types in diseased muscle, assuming that the Z-lines and M-lines are intact (i.e. no Z-line smearing, etc.). We chose 3 diseased muscles, each of which showed a deficiency in at least one of the major histochemical fiber types (Table 2). Control values are shown for comparison. Patient S.T. with limb-girdle dystrophy had 100% Type I fibers in her biopsy. Patient H.B., with acromegaly, had a marked deficiency of IIB fibers. Patient M.R., with hypotonia0 had a marked deficiency of IIA fibers. The frequency distributions of the latter two diseased muscles were similar to the frequency distribution reported in some normal animal muscles (Gauthier 1971). In the diseased muscles, the fiber types were identified ultrastructurally not by their mitochondrial

TABLE 2 P E R C E N T A G E S OF THREE MAJOR H I S T O C H E M I C A L FIBER TYPES FROM C O N T R O L A N D DISEASED MUSCLE

Type 1

Type IIA

Type liB

68

17

15

28

19

53

100

0

0

Vastus lateralis (acromegalic myopathy)

65

27

8

Vastus lateralis (congenital hypotonia)

62

8

30

Control muscle External intercostal (Patients R.H. and M.S.) Vastus lateralis (Patient H.J.) Diseased muscle External intercostal (limb-girdle muscular dystrophy)

104

('. M. PAYNE, L. Z. STERN, R. (}. CURLESS. L. K. IIANNAPEI.

content but by their relative Z-line and M-line widths. The frequency distribution of each fiber type thus identified was then correlated with the frequency distribution determined histochemically. Histochemical analysis of the muscle from the patient with limb-girdle muscular dystrophy showed this intercostal biopsy to contain 100°,~; Type 1 fibers (Table 2) and at the ultrastructural level, all fibers examined contained wide Z-lines and wide M-lines. Measurements taken from fibers with relatively intact Z-lines revealed the I'A FI1 ! ( ' O M I ' A R A I I V I kk'll) f t t S Ol 7-1 I N I S AN1) M - l I N E S I R O M DISEASED M u g ( I I a

Fiber type 1

IIA

liB

I>~d<'m

Z-lim'~

U-/me.,

115.5±2.1 (110)

91.0±1.2 (i14)

H. B. (vastus lateratis) (acromegalic myopathy)

76.7+ 1.0 (146)

82.1 :L 1.2 (135)

M. R. (vastus tateralis) (congenital hypotonia)

69.1 +-1.0 174)

56.9 + I.(} (162)

S. T. (intercostal) (limb girdle muscular dystrophy')

" Values are means in n m + SEM. N u m b e r s in parentheses indicate total number of lines measured from 4 separate fibers in each case. P-values are stated in text.

1EEID 2El 3D

I

[] []

4D]

E3 5153

E~

2EB 3EL3 [] 4E3C3 6D [ ]

IIA

[Zq []

2D lib [

3~

75

4~

[]

7D t 50

i

60

70

80

90

100

110

120

WIDTH OF Z-LINES AND M-UNES (nm) Z-lines (mean ± SEM) M-lines (mean _+SEM)

Fig. 2. Numbers 1 4 are from controls and 5 7 from diseased muscles. Each number is in line with 2 sets of measurements (Z-line and M-line widths). Three different groups of fibers are clearly distinguished. Type I fibers in both control and diseased muscle can be distinguished from Type II fibers on the basis of nonoverlapping Z-line widths. IIA fibers can be distinguished from IIB fibers on the basis of non-overlapping M-line widths. Two of the 3 diseased muscles (numbers 6 and 7) have Z-line and M-line widths comparable to those of the controls. Muscles from the patient with limb girdle muscular dystrophy (number 5) had significantly wider Z-lines than the controls. (number 1 = p a t i e n t H.R.: 2 = p a t i e n t R.H. ; 3 = p a t i e n t M.S, : 4 = p a t i e n t C.C. ; 5 = p a t i e n t S.T. ( l i m b girdle muscular dystrophy): 6 = patient H.B. (acromegalic myopathy); 7 = patient M.R. (congenital hypotonia).

ULTRASTRUCTURAL FIBER TYPING IN HUMAN MUSCLE

105

mean Z-line and M-line width to be 116 and 91 nm, respectively (Table 3). While the mean M-line widths are comparable, the mean Z-line width is much larger than that of Type I fibers from the controls (Table 1). Histochemical analysis of the vastus lateralis biopsy obtained from the patient with acromegaly shows that of the Type II fibers, the IIA occurs with greatest frequency (Table 2). Ultrastructural examination of this muscle revealed that of the Type II fibers, the fiber that occurs with greatest frequency has narrow Z-lines and wide M-lines. The mean Z-line and M-line widths are 77 and 82 nm, respectively (Table 3). These widths are both comparable to that of the IIA fibers from control muscle (Table 1). Histochemical analysis of the vastus lateralis biopsy obtained from the infant with cerebral hypotonia shows that of the Type II fibers, the IIB occurs with greatest frequency (Table 2). Ultrastructural examination of this muscle revealed that of the Type II fibers, the fiber that occurs with greatest frequency has narrow Z-lines and narrow M-lines. The mean Z-line and M-line widths are 69 and 57 nm, respectively (Table 3). These widths are both comparable to that of the IIB fibers from control muscle (Table 1). The comparative dimensions of Z-lines and M-lines from both control and diseased muscle are graphically shown in Fig. 2. It can be readily observed that all muscle fibers can be placed into 1 of 3 groups if both the mean Z-line and M-line widths are taken into consideration. Two of the 3 diseased muscles examined have Z-line and M-line widths comparable to that of the controls (numbers 6 and 7, Fig. 2). In both of these muscles, all 3 fiber types are represented. The diseased muscle which contained 100~o Type I fibers (number 5, Fig. 2) showed significantly wider Z-lines than any of the 4 control muscles examined.

DISCUSSION

The identification of the 3 major histochemical muscle fiber types under the electron microscope may facilitate cross-correlation between pathological changes occurring at the light microscope and ultrastructural levels. This may also permit one to determine if certain pathological changes which are only observed under the electron microscope are limited to just one fiber type. This information should enable us to understand more about the role that each fiber type plays in both normal and pathological conditions and to define better various neuromuscular diseases. Several investigators continue to use the terminology red, intermediate and white to describe individual muscle fibers when color is an attribute of whole muscle. We have chosen to use the terminology I, IIA and IIB which is based on the myofibrillar ATPase activity of the individual muscle fibers. This enzymatic activity can be crosscorrelated with oxidative activity which can be assessed under the electron microscope by thc relative content of mitochondria. There are two previous studies on human muscle which attempt to define fiber types at the ultrastructural level. Since both of these studies use the terms red, intermediate and white we must relate these terms to the current classification. The red fiber has a strong oxidative activity, the white fiber has a weak oxidative activity, and the intermediate fiber has an oxidative activity

106

(:. M. PAYNE, L. Z. STERN, R. G. CURLESS, L. K. HANNAPE1

which is between that of the red and white fibers (Stein and Padykula 1962). In our current classification, red =Type I, intermediate =Type IIA, and white =Type IIB. Shafiq, Gorycki, Goldstone and Milhorat (1966) studied the fine structure of fiber types in human vastus lateralis muscle and reported that red fibers had a greater content ofmitochondria than white fibers. The intermediate fiber was not clearly separated. Ogata and Murata (1969) studied the cytological differences among the red, intermediate and white fibers in normal human intercostal muscle. Similarly, they found that in the red fiber the mitochondria were plump and multitbrm, and in the intermediate and white fibers they were slender and elongated. They distinguished between the red, intermediate and white fibers on the basis of the extent of the mitochondrial chains among the myofibrillar spaces and the deposits of mitochondria beneath the sarcolemma, both of these being greatest in the red fiber. Although we have occasionally found sub-sarcolemmal aggregates and mitochondrial chains in principally Type I fibers, they were by no means extensive. This difference is most probably a reflection of the age of our respective controls. Ogata and Murata (1969) examined muscle from an 8-year-old control, the mitochondrial content of which might well be higher than ours at ages 26, 35, 57 and 59 years, respectively,. Our results on the relative abundance of lipid droplets comparing tile 3 fiber types are consistent with those of Ogata and Murata (1969). They found the most abundant lipid droplets in the red fiber, a medium abundance in the intermediate fiber and a scarce amount in the white fiber. They also found the sarcoplasmic reticulum to be poorly developed and the glycogen particles to be numerous in all 3 fiber types, Ogata and Murata (1969) state that "the demonstration of the structural features in three fiber types is indispensable for studying pathological alterations in individual fibers of human muscle". Since their distinction between the 3 fiber types at the ultrastructural level is mostly based on mitochondrial content and distribution, any pathological reaction which alters the content or distribution of mitochondria would make fiber typing difficult. By taking into consideration Z-line and M-line widths, we have added a new component to fiber typing at the ultrastructural level, a component which should be minimally altered by varying physiotogic~il and pathological conditions. In two of the diseased muscles examined in this study, measurements of Zline and M-line width were comparable to control values. In the case of limb-girdle dystrophy where all the fibers were Type I, the Z-lines were considerably larger than normal. Whether or not this represents a compensation for the 2 missing fiber types cannot be ascertained at present. Three fiber types in human muscle can be distinguished on the basis of mitochondrial content, as in animal muscle (Padykula and Gauthier 1967b). However, subsarcolemmal aggregates and interfibrillar chains of mitochondria are more prevalent in animal muscle than in human muscle, making it easier to distinguish fiber types at the ultrastructural level. The sarcoplasmic reticulum is also more highly developed in animal muscle and differences in the content and distribution of this membrane system between fiber types (Padykula and Gauthier 1963, 1967a and b; Ogata 1964: Gauthier and Padykula 1966) and between fast-twitch and slow-twitch muscles(Schiaffino et al. 1970) have been described. Our measurements of Z-line widths in human muscle has revealed a smaller dif-

ULTRASTRUCTURAL FIBER TYPING 1N HUMAN MUSCLE

107

ference between the I and IIB fibers and between the IIA and IIB fibers than has been reported for animal muscle. Considering averages for all 4 human control muscles, the Z-line widths of the Type I, IIA and lIB fibers are 95, 74 and 69 nm, respectively, compared to 104, 80 and 55 for chicken sartorius (Shafiq et al. 1971) and 100, intermediate values and 55 for rat extensor digitorum longus muscles (Schiaffino et al. 1970). The difference in Z-line width between the Type I and IIB fiber in human muscle is approximately 26 nm, whereas in animal muscle it is 45-49 nm. The Z-line width in rat diaphragm muscle measures 63, 43 and 34 nm in the red, intermediate and white fibers, respectively (Gauthier 1970). The actual width of the Z-lines is much smaller than in human muscle and other animal muscles, but the difference between the fiber types is similar to that seen in human muscle. The difference in Z-line width between the red and intermediate fibers in the rat diaphragm is 20 nm and that between the I and IIA fibers in human muscle is 21 nm. The difference between the intermediate and white fiber in the rat diaphragm is 9 nm and between the IIA and IIB fibers in human muscle is 5 nm. To our knowledge there are no reported measurements of M-line widths in either human or animal studies. Schiaffino et al. (1970) state that the M-lines in the small mitochondria-rich fibers are wider than in the other fiber types and the sarcomere length is slightly longer. Although we found that the M-lines in human Type I fibers tended to be wider (89 nm) than the IIA (79 nm) and liB (60 nm) fibers, there was some overlap between the I and the IIA. However, the IIB fiber had the narrowest M-line and consistently had the shortest sarcomere length. Whether there is any relationship between M-line width and sarcomere length is unknown at the present time. More measurements of Z-lines and M-lines need to be obtained from control human muscle from both sexes in different age groups. Once fiber typing can readily be accomplished consistently at the ultrastructural level, our understanding of pathological changes in skeletal muscle will be aided. ACKNOWLED(;Ii~n~YTS The authors wish to thank Ms. Rita Redman for excellent technical assistance and Ms. Catherine McVeigh for the typing of the manuscript.

SUMMARY

The differential staining of the 3 fiber types for oxidative enzyme activity at the histochemical level provides the basis of their identification at the ultrastructural level. Type I fibers have the largest and most numerous mitochondria, the Type IIA smaller and less numerous mitochondria, and Type IIB have the smallest and least numerous mitochondria as studied in 4 patients without neuromuscular disease. Type I fibers could be distinguished from Type II fibers on the basis of mean Z-line width and IIA fibers could be distinguished from IIB fibers on the basis of mean Mline width. Type I fibers had wide Z-lines (95 nm) and wide M-lines (89 nm), Type IIA had narrow Z-lines (74 nm) and wide M-lines (79 nm) snd Type IIB fibers had narrow Z-lines (69 nm) and narrow M-lines (60 nm). This system of fiber typing based on

108

('. M. PAYNE, L. Z. STERN, R. G. CURLESS. L. K. HANNAPI.I

relative Z-line and M-line widths was applied to several abnormal muscle biopsies each of which showed a deficiency in at least 1 of the major histochemical fiber types. In each case, the same deficiency, was revealed at the ultrastructural level by measu ring the relative Z-line and M-line widths of the remaining fiber types.

REFERENCES BROOKE. M. H. AND K. K. KAISER (1970) Muscle fiber types. How many and what kind?, 4rch Neuro/. :Chic.). 2 3 : 3 6 9 37¢). ENGEL, A. G. (1966) Electron microscopic observations in thyrotoxic and corticosteroid-induced myopathies, Mayo Clin. Proc., 4l : 785--796. ENGEL, W. K. (1962) The essentiality of histo- and cytochemical studies of skeletal muscle in the investigation of neuromuscular disease, Neuroh)qy (Minne~q~.), 12:778 784. t!NGEL. W. K. AND G. G. CUNN1NGHAM(1963) Rapid examination of muscle tissue. An improved trichrome method for flesh-frozen biopsy sections. Neurolo~dy : Minneap.), 13:919 923. t .XRBER, E., W. H. STt:I~.NBERG AND C. D. DUNLAP (1956) Histochemical localization of specific oxidative enzymes, J. Histochem. C.~tochem.. 4 : 2 5 4 263. GAt~TmER, G. F. (1970) The ultrastructure of three fiber types in m a m m a l i a n skeletal muscle. In: 17Z ,I. BRISKEY, R. G. CASSENSAND B. B. MARSH (Eds.). The Physiolooy and Biochemistry ol ,141t~'ch,,s a F})od, Vol. 2, University of Wisconsin Press, Madison, Wisc., p. 103. G,~UTmER, G. F. (1971) The structural and cytochemical heterogeneity of m a m m a l i a n skeletal muscle fibers. In: R. J. PODOLSKV fEd. J, Contractility of Muscle Cells" and Related Processe.~. Prentice-Hall, Englewood Cliffs. N. J,, p. 131. G:~t:THIER, G. F. AND H. A. PADYKULA (1966) Cytological studies of fiber types in skeletal muscle. A comparatixe study of the m a m m a l i a n diaphragm, J. Cell Biol,, 28:333 354. NA< HLAS. M. M., K. C. TSOU, E. Dt;,SouzA, C. S. CHANG aND A. M. SEHGMAN (1957} Cbtochemical demonstration of succinic dehydrogenase by the u ~ of a new p-nitrophenyl substituted ditetrazole, d. Histoehem. Cytochem., 5: 420-436. {)~;.~-r,~. T. (1964) An electron microscopic study on the red, white and intermediate muscle fibers of mouse. Acta reed. Okayama, 18 : 271 280. O{;ATA, T, AND F. Mt;RMA (1969) Cytological features of three fiber types in h m n a n striated muscle, Tohoku J. exp. Med., 99: 225-245. PAD'CKULa, H. A. AND G. F. GAUTHW~ (1963) Cytochemical studies of adenosine triphosphatases in skeletal muscle fibers. J. Cell Biol., 18: 8%107. PADYKULA, H. A. AND G. F. GAU'EHIER (I 967a) Morphological and cytochemical characteristics of tibet Dpes in normal malnmaliall skeletal muscle. In: A. 1-. MII IIORAF fed.), Exldoratorl ('omcf~t~ ,~ Ult,~u/u: Dystrophy and Reluted Disor&'rs (Proceedings of the International Conference convened by the Muscular Dystrophy Associations of America. Ne\~ York, 1966), (International Congress Series. No. 147). Excerpta Medica, Amsterdam, p. 117 PAD','KUI ',. H. A. AND G. F. GAU1HIER (1967b) UItrastructural features of three fiber types in the rat diaphragm, .Iraqi. Rcc.. 157:296 297 PADYKt. LA, 17t.A. AND E. HERMAN (1955} The specificity of the histochemical method of adenosine triphosphatase. J. Histochem., 3 : 170- 195. RO~aANUL, F. C. A. (1964) Enzymes m muscle. Part I (Histochemical studies of enzymes in individual muscle fibers), Arch. Neurol. :Chic. ), I I : 355-368. N('HIAEFINO, S., V. HANZLIKOVA AND S. PIEROBON (1970) Relations between structure and function in rat skeletal muscle fibers, J. Cell Biol., 47: 107-119. SHAFIO. S A., V. ASKANAS AND A. T. Mn.UOR:,T (1971) Fiber types and preclinical changes in chicken muscular dystrophy, .4rch. Neurol. (('hie.), 25 : 560- 571. SHAFIQ, S, A., M. GORYCKI, L. GOLDSTONE AND A. T. MILHORAT (1966) Fine structure of fiber types in normal h u m a n muscle, Anat. Rec., 156: 283-301. SPURR, A, R. (1969) A low viscosity epoxy resin embedding medium for electron microscopy, J. Ultrastruct. Res., 26: 31-43. STEIN, J. M. AND H. A. PADYKUL~ (1962) Histochemical classification of individual skeletal muscle fibers of the rat. Amer. J. Anat., I10: 103-123. STERN', L. Z., C. M. PARSE AND L. K. HANNAPEL 0974) Acromegaly - Histochemical and electron microscopic changes in deltoid and intercostal muscle, Neurology (Minneap.), 24: 589-593.