Muscle Fiber Types of Human Extraocular Muscles: A Histochemical and Immunohistochemical Study

Abstract

purpose. To classify muscle fibers of human extraocular muscle (hEOM) and to compare them to previous studies on hEOM, as well as to nonhuman EOM classification schemes and skeletal muscle fiber types.

methods. Muscle fibers cut in different muscle planes were followed on consecutive cross sections and typed with regard to their oxidative profile in combination with their myosin–immunohistochemical characteristics.

results. Three zones were observed. In the global layer three muscle fiber types were observed: global layer singly innervated granular fibers, 79.4 ± 8.1 μm (perimeter [values at midmuscle region] ± SD); 59%; global layer singly innervated coarse fibers (80.3 ± 10.8μ m; 21%); and global layer multiply innervated muscle fibers (4.1 ± 9.7 μm; 21%). Two muscle fiber types were detected in the orbital layer: orbital layer singly innervated muscle fibers (54.1 ± 8.5 μm; 83%) and orbital layer multiply innervated muscle fibers (53.5 ± 7.6 μm; 17%). Three muscle fiber types were differed in the marginal zone: marginal zone singly innervated muscle fibers (83.1 ± 15.8 μm; 56%), marginal zone multiply innervated low oxidative muscle fibers (84.4 ± 23.3 μm; 7%), and marginal zone multiply innervated high oxidative muscle fibers (88.4 ± 14.5 μm; 37%). Coexpressions of developmental myosin heavy chain isoforms and fast myosin heavy chain isoform were detected mainly in the marginal zone.

conclusions. hEOMs resemble mammalian EOM with regard to their organization. However, in addition to an inner global layer and an orbital layer an external marginal zone was described for the first time in hEOM in the present study.

There are many differences between mammalian extraocular muscle (EOM) and skeletal muscle. EOM have an inner global layer (GL), bordering the surface adjacent to the globe and containing mainly large muscle fibers, and a surrounding orbital layer (OL) composed of small muscle fibers. ^{1 2} This is most clearly seen in the recti muscles. The innervation pattern of EOM differs from that in skeletal muscle. In EOM, both singly innervated muscle fibers (SIFs) and multiply innervated fibers (MIFs) are present, ^{3 4} whereas skeletal muscle contains singly innervated muscle fibers exclusively. 

The myosin heavy chain pattern of adult EOM is different from that in adult skeletal muscle. In adult EOM single muscle fibers coexpress developmental (neonatal and embryonic) myosin heavy chain (MHC) isoforms ^{5 6} in combination with adult MHC isoforms. Furthermore, in adult EOM, variations in MHC characteristics are observed along single muscle fibers ⁷ and an extraocular specific MHC isoform is detected. ⁵ Normal adult extrafusal skeletal muscle fibers contain only an adult MHC pattern. ⁸  

Extraocular muscle fibers can be classified in several different ways. They were initially distinguished according to their histologic appearance into “Felderstruktur” and “Fibrillenstruktur” fibers. ⁹ Later, based on the amount and distribution of mitochondria, Durston ¹⁰ differentiated between“ coarse,” “fine,” and “granular” EOM muscle fibers. Mayr ¹¹ was the first who described (in the rat) six fiber types, taking into account location (orbital layer versus global layer), muscle fiber diameter, innervation pattern (singly innervated fibers versus multiply innervated fibers), histochemical features, and ultrastructure. Subsequent investigations of EOM in different mammals confirmed the concept of six EOM fiber types. ²  

In human EOM (hEOM) fiber classifications were carried out using the mitochondrial pattern as a differentiation criterion. ^{12 13} Staining characteristics for actomyosin ATPase in combination with glycolytic and oxidative enzymes were used by Hoogenraad et al. ¹⁴ Immunohistochemically, anti-slow twitch and anti-slow tonic MHC antibodies served to classify hEOM fibers. ¹⁵  

In each of these investigations, only one method was used to classify the fiber types, leading to different and non-corresponding hEOM fiber classifications. Although Hoogenraad et al. ¹⁴ and Fujii et al. ¹⁵ describe four hEOM fiber types, only three were found by Carry et al. ¹² Furthermore, the muscle fiber types 1, 2, and 3 of Fujii et al. ¹⁵ seem to all correspond to only muscle fiber type 1 of Hoogenraad et al. ¹⁴ Moreover, the location of the different fiber types was not adequately considered in both above-mentioned papers. 

The aim of the present study was to classify hEOM fiber types by combining for the first time histochemical and immunohistochemical techniques. This combined with the location and fiber diameters is used to develop a general hEOM fiber classification scheme, allowing comparisons to EOM fiber classifications of other mammalian species as well as to skeletal muscle fiber types. 

In addition to the global and the orbital layers, a new third layer is described here for the first time. It is called “marginal zone” and covers parts of the outer surface of the orbital layer. 

Methods

Sample Collection

Four hEOMs, rectus medialis (MR), rectus superior (SR), rectus inferior (IR), and obliquus superior (SO) muscles from a 52-year-old woman and a distal part of SR from a 60-year-old man, who did not suffer from neuromuscular disease, were harvested postmortally according to the Austrian federal law of transplantation. Another distal part of the SR was obtained from a 77-year-old man, who underwent a surgical enucleation after carcinoma. Methods for securing human tissue were humane and complied with the tenets of the Declaration of Helsinki. Immediately after excision from the orbit, the four muscles were transferred into a 20% sucrose solution at pH = 7.4 for up to 20 minutes, followed by freezing in liquid nitrogen. Slides were kept at −80°C until used. In three muscles (SR, IR, and SO) consecutive transversal sections (5 μm) from midmuscle regions were cut on a Leitz KryoCut 3000 cryostat microtome. The MR was divided into six parts of equal length, the two distal parts from SR were divided into two parts each. From each part consecutive sections were performed to investigate variations of muscle fiber composition along the muscle’s length. 

Histochemistry

Sections were stained for succinic dehydrogenase (SDH), ¹⁶ cytochrome C oxidase (Cyt-c-ox), ¹⁷ and nicotinamide tetrazolium reductase (NADH–TR). ¹⁶  

Staining for myofibrillar actomyosin adenosine triphosphatase (mATPase) was performed after alkaline (pH = 10.4) and after acid (pH = 4.4) preincubations according to Guth and Samaha. ¹⁸  

Immunohistochemistry

For immunohistochemical detection of MHC isoforms unfixed sections were incubated with primary monoclonal antibodies (mouse monoclonal; Novocastra Laboratories, Newcastle, UK) against fast MHC (NCl-MHCf), slow MHC (NCl-MHCs), developmental MHC (NCl-MHCd), and (9) neonatal myosin heavy chain (NCl-MHCn) for 1 hour at 25°C. After washing in phosphate-buffered saline three times for 10 minutes (PBS, pH = 7.4), sections were incubated with the secondary antibody (goat anti-mouse peroxidase-conjugated immunoglobulin (NCl G-AMP, polyclonal; Novocastra Laboratories, Newcastle, UK) for 1 hour at 25°C followed by washing in PBS 3 times for 10 minutes. DAB (diaminobenzidine) was used as chromagen. 

Course of Investigation

Serial sections stained for SDH, Cyt-c-ox, NADH–TR, mATPase (pH = 10.4), mATPase (pH = 4.4), anti–slow-, anti–fast-, anti–developmental-, and anti–neonatal MHC were examined in the light microscope. Muscle fibers were investigated in consecutive sections, and their staining characteristics were assessed. Photodocumentation and immunohistomorphometrics were done with a Zeiss Axioskop supplied with a semiautomatic Kontron Imaging System (KS-300). 

Different fiber types were characterized with regard to their oxidative profiles in combination with their MHC characteristics. 

Statistics

SPSS for Windows (version 7.5.2G) was used for statistical analysis. Values of fiber type diameters are given as mean ± SD (n = 120–180). If applicable an unpaired t-test was performed. Probability values are the results of two-tailed tests. 

Terminology

The anti-developmental myosin antibody used in this study has been reported to show results identical to those of BF–G6, ⁷ an anti-embryonic myosin antibody used and raised by Schiaffino et al. ¹⁹ To avoid confusion we use the term “embryonic” instead of “developmental” in this study. “Embryonic” and“ neonatal” myosin together will be called “developmental.” 

In the present study the terms “fast” and “slow” fibers are referred to fibers expressing fast or slow MHC isoform, but do not describe their physiological contraction characteristics. 

From other species it is well known that EOM fibers positive for alkaline stabile/acid labile mATPase and showing fast MHC characteristics are SIFs and that MIFs contain slow MHC and stain positive for alkaline labile/acid stabile mATPase. ² In our study we use the terms “singly innervated fibers” (SIF) and“ multiply innervated fibers” (MIF) for muscle fibers displaying a certain mATPase pattern and MHC expression, although we did not investigate the innervation pattern of single muscle fibers in this study. This will be recorded in a future study. 

Results

Organization

In cross sections of all hEOM investigated, three muscle layers could be distinguished. The main and central portion of the muscles, in all rectus muscles adjacent to the ocular globe, the GL displayed larger fibers of varying diameters. At its orbital surface, the OL, containing smaller muscle fibers less variant in size, enclosed the GL in rectus muscles in a C-shaped manner. These two “classic” layers typical for mammalian EOM ¹ were separated by an ill-defined boundary, consisting of perimysium internum and an“ intermediate zone.” ² The muscle fiber bundles of the intermediate zone, which are made up of smaller fibers of the OL, intermingled between the bigger ones of the GL. 

In addition to the GL and the OL a third muscle layer was found (Fig. 1) . This layer lined the OL at its outer surface and consisted mainly of larger muscle fibers. This layer, designated as marginal zone (MZ), was observed to be better developed near the muscles’ edges, where the plane global surface meets the convex orbital one. Apart from those edges the MZ sometimes becomes discontinuous in its width. The amount of connective tissue between muscle fiber bundles increased from the global to the orbital surface. 

This organization into GL, OL, and MZ, observed in all muscles investigated, becomes unclear in the very proximal and distal parts of the muscle. 

Variations of the fiber composition along the MR muscle’s length are given in Table 1 . Variations of the fiber composition along the distal parts of the SR muscle’s length are given in Tables 2 and 3 . No major interindividual differences were observed in the muscle fiber distribution. 

The following descriptions are given for a transversal section through the midmuscle region of the MR. 

Histochemical Findings

Muscle fibers positive for mATPase after acid preincubation were found in both the OL and the GL and distributed evenly, with increasing percentage proportion toward the distal muscle part. In the MZ the number of these fibers increased up to 49.2% (Table 1 ; MZ MIF low oxidative; MZ MIF high oxidative). The majority of muscle fibers in all three muscle layers exhibited a positive mATPase activity after alkaline pretreatment. 

In the GL, some muscle fibers were stable for mATPase after acid as well as after alkaline preincubation. 

Oxidative enzymes, NADH–TR and SDH, showed evenly high activity in fibers of the OL. Variable staining intensity resulting in a checkerboard appearance was observed in muscle fibers of the GL as well as the MZ. 

Staining for Cyt-c-ox showed mainly a pattern similar to that of the oxidative enzymes. Generally, a higher concentration of this enzyme was observed in muscle fibers of the OL, whereas muscle fibers of the GL and the MZ displayed lower staining intensity. Single muscle fibers, throughout the transverse section, were negative for Cyt-c-ox but positive for NADH–TR and SDH, which is consistent with an observation of Müller–Höcker et al. ²⁰ They found this peculiar enzyme pattern based on a defect of Cyt-c-ox protein as a result of a degenerative alteration, age dependent and more progressive in EOM than in other skeletal muscles. 

Immunohistochemical Findings

Results of treatment with anti–fast MHC corresponded to the outcome of staining for ATPase after alkaline preincubation. Anti–slow MHC distribution was similar to the distribution of ATPase positive muscle fibers succeeding acid preincubation. Those fibers exhibiting positive reactions for ATPase after alkaline and acid preincubations turned out to be positive for anti–fast MHC antibodies and negative after treatment with anti–slow MHC antibodies (Fig. 2) . 

Staining for developmental MHC isoforms was found predominantly in the MZ and the superficial cell layers of the OL. Sparse muscle fibers in the inner OL were positive for anti–neonatal or anti–embryonic MHC antibodies (Fig. 3) . Muscle fibers containing neonatal MHC or embryonic MHC coexpressed adult fast MHC. The difference in the mean perimeter of muscle fibers expressing neonatal MHC (41.1 ± 13 μm; 189/15,000 muscle fibers) and those expressing embryonic MHC (68.2 ± 15.5 μm; 108/15,000 muscle fibers) was statistically significant (t-test; P < 0.001). Sparse muscle fibers were positive for neonatal and embryonic MHCs in addition to fast MHC. 

Muscle fibers expressing neonatal MHC isoform could be followed for only a few consecutive cross sections. Thus, it was not possible to characterize those fibers according to all characteristics as summarized in Table 4 . Therefore these muscle fibers are not taken into account in the muscle fiber classification adopted in this investigation (Table 4) . 

Muscle Fiber Types

Global Layer.

In the GL three muscle fiber types were distinguished. Positivity for mATPase after alkaline preincubation was found in the majority of fibers (Tables 1 2 3 ; GL SIF granular and GL SIF coarse muscle fibers). In the proximal muscle portion the lower oxidative “GL SIF granular” muscle fibers turned out to comprise only a small fiber population strongly increasing toward the muscles’ distal part. Inversely, the amount of the higher oxidative “GL SIF coarse” muscle fibers decreased toward the distal muscle portion. “GL MIFs” were positive for mATPase succeeding acid preincubation. Oxidative enzyme activity was found to be high in these fibers. GL SIF granular and GL SIF coarse muscle fibers exhibited smaller perimeters than GL MIF muscle fibers. Among GL SIF granular and GL SIF coarse muscle fibers, approximately 50% of the fibers appeared stable for both alkaline and acid mATPase. This histochemical double staining of fibers was not confirmed by the means of myosin immunohistochemistry, showing those fibers to be positive for fast myosin but negative for slow myosin (Fig. 2) . Thus, it seems possible, that the mATPase reaction of alkali stabile and acid labile muscle fibers was not inhibited satisfactorily in the pH 4.4 environment used. However, a pH = 4.3 environment will be somehow “optimal” for inhibiting alkali labile mATPase in hEOM (authors’ unpublished observations). A slightly more alkaline pH leads to an alkaline and acid stabile mATPase pattern, which cannot be confirmed by the means of myosin immunohistochemistry (authors’ unpublished observation). 

Orbital Layer.

In the OL two muscle fiber types were found to differ in their mATPase pattern after alkaline and acid pretreatment. Muscle fibers positive for alkali stable and acid labile ATPase (OL SIF) displayed the main fiber type in the OL. This muscle fiber type contained high amounts of oxidative enzymes. The second muscle fiber type (OL MIF) stained darkly for alkali labile and acid stable mATPase, whereas only a slightly less intense staining for oxidative enzymes was detected (Fig. 4) . Immunohistochemical staining for adult MHC isoforms correlated strictly with mATPase pattern. Fast MHC pattern corresponded to alkali stable/acid labile mATPase, whereas slow MHC distribution corresponded to the alkali labile and acid stable mATPase positivity. OL muscle fiber types did not differ significantly in their perimeter size. 

Marginal Zone.

One type of MZ muscle fibers displayed alkali stable and acid labile mATPase (MZ SIF). Two types of muscle fibers exhibited alkali labile and acid stable ATPases. A very small amount of these fibers showed a weak staining for oxidative enzymes (MZ MIF low oxidative). In contrast, “MZ MIF high oxidative” muscle fibers exhibited a more intense staining for SDH and NADH–TR (Fig. 5) . 

The total amount of MZ muscle fibers with alkali labile and acid stable ATPase increased toward the distal muscle portion up to nearly half of the muscle fibers of this layer. 

All three MZ muscle fiber types were larger than GL and OL muscle fibers. 

Alterations of perimeter and percentages of muscle fiber types along the muscle’s length were most accentuated in the MZ (Tables 1 2 3) . 

Mitochondrial Pattern

The mitochondrial pattern of EOM fibers is more complex than that of skeletal muscle fibers. Therefore, any pathologic alterations concerning the arrangement of mitochondria are difficult to interpret. ^{21 22} In contrast to the MZ and the OL, where no correlation of muscle fiber type and mitochondrial pattern was detected, muscle fibers of the GL showed differences in the mitochondrial pattern, most conspicuous after staining for NADH–TR. Dependent on the size and density of mitochondria, muscle fibers containing dispersed small mitochondrial profiles could be classified as “granular” fibers; those exhibiting larger mitochondrial profiles or clusters of mitochondria as “coarse” fibers. Small but densely packed NADH—TR–positive profiles were classified as“ fine” fibers ^{10 12} (Fig. 6) . A clear correlation of “fine” NADH–TR appearance with fibers positive for anti–slow MHC antibodies was found in GL MIF muscle fibers. One type of GL SIF muscle fibers was granular (GL SIF granular muscle fibers), whereas the second one showed a “coarse,” and some a “fine” appearance (GL SIF coarse muscle fibers). GL MIF“ fine” fibers were clearly distinguishable from GL SIF coarse MF displaying a “fine” appearance, because of a slightly lesser density in their mitochondrial pattern but more intense staining for NADH–TR. Furthermore, in the midbelly section (Table 1) very large muscle fibers (119.1 ± 15.4) exhibited a dense granular appearance with coarse elements. These muscle fibers were counted as“ coarse, despite their statistically significant (t-test; P < 0.001) larger size compared with GL SIF coarse (87.7 ± 17.7) at other levels. 

Discussion

This study confirms observations that recti and obliqui EOM are organized into a GL, adjacent to the eye globe, surrounded by the OL. In this article we describe an additional MZ at the muscle’s periphery for the first time in hEOM. This MZ is comparable to the “peripheral patch layer” detected in sheep EOM ²³ in three ways: (1) by its location at the outside of the OL, (2) it is composed by much larger fibers than the OL muscle fibers, and (3) the higher amount of MZ MIF type fibers in hEOM or intermediate-G fibers in sheep EOM, respectively. Harker ²³ found the loose “peripheral patch layer” covering only the proximal and distal end of sheep SR muscle, and described it as containing two fiber types. He observed a high number of MIFs, which he called intermediate G (grape-like) fibers in sheep EOM. Besides intermediate G fibers, Harker ²³ described “small” C fibers with SIF-like characteristics in the peripheral patch layer. 

In the present study, the existence of a MZ in hEOM is described for the first time. In contrast to the “peripheral patch layer” in sheep EOM, the MZ of hEOM was found to cover the whole muscle length except the very proximal and distal muscle portion. 

Although innervation was not studied by direct evidence of myoneural synapses (acetylcholine-esterase, choline-acetyl-transferase,α -bungarotoxin), we found two types of MZ MIFs (MZ MIF high oxidative and low oxidative). But only the human MZ MIF low oxidative muscle fibers were comparable to the sheep’s intermediate G fiber with respect to a sparse staining profile for SDH. The amount of MZ MIFs increased toward the distal muscle portion. 

The MZ SIF type may resemble “small” C fibers in sheep EOM. 

The conspicuous increase of the number of slow MHC–positive muscle fibers toward the distal MZ might be explained by a higher degree of stretch and isometric contraction forces to the MZ than to other regions of the hEOM. Goldspink ²⁴ has suggested that the expression of slow MHC isoform in muscle fibers might be dependent on stretching forces or on isometric contractions. 

In contrast, to the MZ in the GL and the OL, we found only one MIF type, but two SIF types in the GL and one in the OL. 

In summary, the studies on histochemical/ultrastructural classifications in mammalian EOMs are in general agreement that there are 2 to 3 GL SIF types and 1 GL MIF type, and 1 to 2 OL SIF types and one OL MIF type. ²  

In previous hEOM classifications ^{12 13} the GL SIF granular fiber type turned out to be the main fiber type. However, our investigation indicated varying amounts of GL SIF type fibers, with the GL SIF granular fiber type more concentrated in the distal parts of the EOM. Histochemically, II_B skeletal muscle fibers resembled our GL SIF granular muscle fiber. The other SIF type muscle fiber, GL SIF coarse type, showed skeletal muscle fiber type II_A features. 

In this article the terms “I-like,”“ II_A-like,”“ II_B-like,” and“ II_C-like” are used with regard to a combination of demonstration of mATPase with the histochemical demonstration of oxidative enzyme profiles and the MHC protein characteristics to ensure compatibility with previous investigations of human EOM fibers. ^{14 15} However, human type II_B MHC isoform corresponds to II_X MHC isoform in the rat rather than to the faster rodent II_B MHC isoform. ²⁵ When SIF types as classified in nonhuman EOM are compared with our results, the low fatigue resistant “global pale SIF” ²⁶ most likely corresponds to our GL SIF granular fiber. The “global red SIF,” ²⁶ which is suggested to be highly fatigue resistant, corresponds to our GL SIF coarse fiber. The fiber type corresponding to the “global intermediate SIF” ²⁶ was probably included among GL SIF coarse fibers. 

We found the proportion of GL MIFs to increase from 11% to 22% from the proximal to the distal muscle section. This variation in the number of MIFs along the muscle’s length might be an explanation for slightly differing numbers of MIFs (12%–16%) in previous studies on hEOM. ^{12 13 15} In nonhuman mammalian EOM, approximately 10% of the GL muscle fibers were reported to be GL MIFs. ^{2 11} Ringel et al. ¹³ have described the human GL MIF as “fine,” and it was also reported to resemble skeletal muscle fiber type I ¹⁴ or“ slow.” ¹⁵  

Nonhuman GL MIFs were reported to resemble slow tonic fibers in amphibian skeletal muscle; however, these show a weak staining profile for NADH–TR, ^{2 4} in contrast to our finding of an intensive staining for NADH–TR. Consistent with slow tonic behavior of GL MIFs, physiological studies demonstrated the presence of non-twitch motor units in the GL of the rat and cat. ^{27 28 29} Although hEOM GL MIFs do not resemble slow tonic fibers in amphibians in their oxidative enzyme pattern, a high amount of human GL MIFs coexpresses slow tonic and slow twitch MHC isoforms. ¹⁵  

In previous studies muscle fiber types displaying a “fine” intermyofibrillar pattern were misinterpreted as SIFs. ^{10 13 30} In subsequent studies this muscle fiber type was found to be multiply innervated. ^{2 4}  

In agreement with findings in other species, ² the hEOM OL was composed of small oxidative fibers, one OL MIF type and one OL SIF type. However, in sheep EOM ²³ and in rat EOM ³¹ two types of SIFs were found. Hoogenraad et al. ¹⁴ classified two types of hEOM OL SIFs, type II_C muscle fibers (78%) and type II_A muscle fibers (7%). In contrast, we observed all human OL SIFs to display instead a pattern similar to type II_A (alkali stabile and acid labile ATPase, fast MHC expression in combination with high activity of oxidative enzymes). Consistent with the histochemical findings the OL SIFs failed to display coexistence of fast and slow MHC characteristics, as has been described for type II_C skeletal muscle fibers. ^{32 33 34} However, there is some evidence of the possibility that II_A muscle fibers get II_C characteristics along their length, ³⁵ a finding that we cannot confirm in hEOM by immunohistochemical means. 

The human OL MIFs expressed only slow MHC and showed corresponding to this alkaline labile and acid stabile mATPase activity. In combination with a dark blue stain for NADH–TR, indicating high activity of oxidative enzymes, the OL MIF had histochemical and immunohistochemical features similar to those of skeletal muscle fiber type I, although in contrast to the focal innervation of skeletal muscle slow twitch muscle fibers OL MIFs are supposed to have multiple innervation and might be unable to conduct propagated action potentials. This finding is consistent with the description of Hoogenraad et al. ¹⁴  

A number of nonhuman OL MIFs were reported to show alkaline stabile as well as acid stabile mATPase activity, like certain intrafusal muscle fibers in the skeletal muscle. ^{2 4 36} Alkali stabile and acid stabile mATPase was reported to occur in the midfiber portion only, whereas distal and proximal endings exhibited only alkali labile and acid stabile mATPase. ⁴ In contrast to this, OL MIFs with dual mATPase activity could not be observed in this study. This discrepancy with previous studies and the present study could be due to differing pH values (see the Results section), to differing methods, or both. For demonstration of alkaline stabile/acid labile mATPase the method of Guth and Samaha ¹⁸ was used in the present study, which is based on the sensitivity of mATPase to formaldehyde. Brooke and Kaiser, ³² referring to a similar method for demonstration of mATPase based on the sensitivity toward pH, introduced the terms I, II_A, II_B, and II_C. Slow twitch fiber typing based on those two methods ^{18 32} turned out to be compatible, but fast fiber subpopulations have been found to correspond to a lesser degree, varying in various species, if classified with both methods. ³⁷  

Binding of anti–slow tonic MHC antibodies showed positive fibers to be concentrated in human OL ^{15 38} and rat, rabbit, and guinea pig OL. ³⁸ However, some EOM fibers expressing slow twitch MHC own slow tonic MHC, too, ¹⁵ but these muscle fibers were not detectable by means of traditional histochemistry. On the other hand, in the OL of cat lateral rectus muscle only twitch, and slow fatigable and fast fatigue-resistant, motor units were found. ³³  

In MR, midbelly region, expression of developmental MHC isoforms was observed in fast fibers in the muscle’s periphery only. Sparse fibers were found to be positive for neonatal, embryonic, and fast MHC isotypes. Fibers expressing neonatal MHC were much smaller than fibers expressing embryonic MHC. Developmental MHC isoforms, normally suppressed in adult skeletal muscles, ⁸ are (re-)expressed in intrafusal muscle fibers ³⁹ or in regenerating muscle fibers, or in muscle dystrophy. ^{40 41} In EOMs mRNA transcripts and protein products of developmental MHC isoforms were detected even in adult stages. ^{5 42}  

The surprising variability of fiber composition at different muscle levels might be accounted for by the change in the total number of muscle fibers along the length of the EOM, increasing strongly toward the midbelly region. ⁴³ The conspicuous increase of fiber perimeters in the midbelly region (Table 1) may be due to the presence of myomyos junctions, as demonstrated in cat EOM. ⁴⁴  

Extraocular muscles are among the fastest and most fatigue-resistant skeletal muscles. ²⁶ Their highly specialized function, to move a sensory organ, the eyeball, is reflected in their specific MHC content and the multiplicity of fiber types. In addition to the described developmental MHC isoforms and adult MHC isoforms (including slow tonic MHC), a very fast tissue-specific MHC isoform called EOM-specific MHC has been detected at protein and mRNA levels. ^{5 6 7 45 46} Extraocular muscles also express cardiac MHC isoforms. ⁴⁷ However, the EOM function cannot be designated to any one particular layer. Furthermore, it would be unwise to predict the function of any muscle part without first establishing the mechanical connections between the fibers. 

In addition to the classic histochemical and MHC immunohistochemical patterns of the different fiber types, the single or multiple innervation also plays an as yet unknown role in the physiological behavior. Beside this the function of the MZ remains unclear. 

 Submitted for publication March 16, 1999; revised October 26, 1999; accepted November 8, 1999.

 Commercial relationships policy: N.

 

Figure 1.

 

Cross section of human MR muscle. mz, marginal zone; ol, orbital layer; gl, global layer. (A) Staining for SDH, (B) anti–slow MHC antibody, and (C) anti–fast MHC antibody. Magnification, ×100.

Table 1.

 

Muscle Fiber Distribution (Classes) in Different Sections of a Human MR Muscle from a 52-Year-Old Woman

Muscle Section	GL							MZ
		GL SIF			GL MIF	OL			MZ SIF	MZ MIF
		Granular	Coarse			OL SIF	OL MIF			Low Oxidative	High Oxidative
Proximal	74.3 ± 10.9, 7.8%	74.1 ± 13.2, 81.7%	75.3 ± 13, 10.5%	60 ± 6.6, 94%	29.7 ± 20.4, 6%	72.1 ± 11.2, 86%	79.4 ± 15.6, 8%	75.1 ± 15.2, 6%
Midbelly	84 ± 12.7, 23.7%	94.8 ± 17.5, 60%	91.9 ± 29.7, 16.3%	62.8 ± 11.9, 92.4%	42.3 ± 27.9, 7.6%	105 ± 17.8, 85.1%	106.3 ± 10.5, 3.7%	117.6 ± 26, 11.2%
Midmuscle	79.4 ± 8.1, 58.5%	80.3 ± 10.8, 20.7%	84.1 ± 9.7, 20.7%	54.1 ± 8.5, 82.6%	53.5 ± 7.6, 17.4%	83.1 ± 15.8, 56.3%	84.4 ± 23.3, 6.8%	88.4 ± 14.5, 36.9%
Distal	72.2 ± 7.3, 41.1%	74.4 ± 9.7, 37%	80.9 ± 17.6, 21.9%	55.6 ± 7.7, 80.6%	55.3 ± 8.3, 19.4%	91.6 ± 13.3, 50.8%	109.3 ± 16.1, 23.3%	90.6 ± 19, 25.9%

 Values are perimeter (in micrometers; mean ± SD) and percentage.

Table 2.

 

Muscle Fiber Distribution (Classes) in the Distal Sections of Human SR Muscle from a 60-Year-Old Man

Muscle Section	GL							MZ
		GL SIF			GL MIF	OL			MZ SIF	MZ MIF
		Granular	Coarse			OL SIF	OL MIF			Low Oxidative	High Oxidative
Midmuscle	83.4 ± 15.3, 42%	78.4 ± 17.1, 39%	74.4 ± 10.2, 19%	60 ± 8, 84%	54 ± 13, 16%	72.9 ± 12.5, 55%	87.9 ± 14.3, 28%	77 ± 19.5, 17%
Distal	87.8 ± 12, 36%	86.7 ± 14.1, 46%	94.6 ± 12.2, 18%	66.2 ± 8.2, 82%	61.7 ± 8, 18%	93.3 ± 20.5, 52%	113 ± 11.4, 17%	123.3 ± 24.7, 31%

 Values are perimeter (in micrometers; mean ± SD) and percentage.

Table 3.

 

Muscle Fiber Distribution (Classes) in the Distal Sections of Human SR Muscle from a 77-Year-Old Man

Muscle Section	GL							MZ
		GL SIF			GL MIF	OL			MZ SIF	MZ MIF
		Granular	Coarse			OL SIF	OL MIF			Low Oxidative	High Oxidative
Midmuscle	92.8 ± 14.2, 33%	95.2 ± 14.9, 56%	79.7 ± 9.9, 11%	56.4 ± 6.5, 85.3%	56.5 ± 9.8, 14.7%	104.8 ± 24.7, 83.5%	92.6 ± 14.1, 6.4%	81.5 ± 13.1, 10.1%
Distal	78.1 ± 9.6, 40.8%	76.7 ± 7.8, 40.8%	79.2 ± 14.4, 18.4%	72.4 ± 8.8, 84%	63.4 ± 4.5, 16%	97.5 ± 16.5, 47%	98 ± 15, 27%	107.6 ± 21.3, 26%

 Values are perimeter (in micrometers; mean ± SD) and percentage.

Figure 2.

 

Global layer. (A) Staining for NADH–TR, (B) slow MHC, (C) mATPase after acid preincubation, and (D) mATPase after alkaline preincubation. 1, GL SIF granular muscle fiber; 2, GL SIF coarse muscle fiber; and 3, GL MIF muscle fiber. GL muscle fiber stable for ATPase after alkaline and acid pretreatment, but negative for anti–slow MHC antibody (arrow). Magnification, ×400.

Figure 3.

 

Marginal zone. Staining for neonatal MHC (A) and embryonic MHC (B). Muscle fibers expressed embryonic MHC (arrowhead). Muscle fibers expressed neonatal MHC (large arrows). Muscle fibers coexpressed neonatal and embryonic MHCs (small arrows). Magnification, ×1000.

Table 4.

 

Comparison of Studies on Muscle Fiber Types in hEOM

	Central					Peripheral
Skeletal muscle–like pattern 14	IIB-like 27%	IIA-like/IIC-like 24.9%/31.6%	I-like 14.3%	IIA-like/IIB-like IIC-like 6.9%/0.1%/77.9%	I-like 15.1%		?
	Global Layer					Orbital Layer
Mitochondrial pattern 12	Granular 55%	Coarse 33%	Fine 12%	Coarse/granular 92%/7%	Fine 1%		?
	Central					Peripheral
Slow immunohistochemical 15	Group 4 83.3%			Group 1;2;3 16%	Group 4 71.3%	Group 1;2;3 27.8%		?
	Global Layer					Orbital Layer			Marginal Zone
Wasicky et al., present study	GL SIF granular	GL SIF coarse	GL MIF	OL SIF	OL MIF	MZ SIF	MZ MIF low oxidative	MZ MIF high oxidative
ATPase pH = 10.4	+	+	−	+	−	+	−	−
ATPase pH = 4.4	−	−	+	−	+	−	+	+
NADH–TR	Intermediate/low	Intermediate/high	Intermediate/high	Very high/high	Very high/intermediate	High/intermediate	Intermediate	High/intermediate
SDH	Intermediate/low	Intermediate/high	Intermediate/high	Intermediate/high	Intermediate/high	High/intermediate	Low	High/intermediate
Cyt-c-ox	Intermediate/low	Intermediate/high	Intermediate/low	High	High	High/intermediate	Intermediate/low	Intermediate
Anti–fast MHC ab	+	+	−	+	−	+	−	−
Anti–slow MHC ab	−	−	+	−	+	−	+	+
Perimeter, μm mean (±SD)	79.4 (±8.1)	80.3 (±10.8)	84.1 (±9.7)	54.1 (±8.5)	53.5 (±7.6)	83.1 (±15.8)	84.4 (±23.3)	88.4 (±14.5)
%	58.5%	20.7%	20.7%	82.6%	17.4%	56.3%	6.8%	36.9%
Probable type of innervation	SIF	SIF	MIF	SIF	MIF	SIF	MIF	MIF

 The terms central and peripheral correspond to the subdivisions orbital and global layers, respectively. The ? indicates that the MZ has no equivalent in other studies.

Figure 4.

 

Orbital layer. (A) Staining for NADH–TR. (B) mATPase after acid preincubation. (C) mATPase after alkaline pretreatment. Note the high NADH–TR activity in all OL muscle fibers. 1, OL MIF muscle fiber type mATPase stained darkly after acid preincubation and negatively after alkaline preincubation. The 2, OL SIF muscle fiber type shows an inverse pattern. Magnification,× 400.

Figure 5.

 

Marginal zone. Staining for NADH–TR (A), mATPase after alkaline pretreatment (B), mATPase after acid preincubation (C), and fast MHC (D). 1, MZ SIF muscle fiber type. MZ MIF low oxidative muscle fiber type is indicated by large arrow. Marginal zone MIF high oxidative muscle fiber type is indicated by small arrows. Magnification,× 400.

Figure 6.

 

Global layer. Staining for alkaline mATPase (A) and NADH–TR (B). Coarse (C), fine (F), and granular (G) muscle fibers. Magnification, ×1000.

The authors thank Jean A. Buettner–Ennever for reading the manuscript and Marietta Lipowec for her valuable technical aid.