Journal of Cell Science 111, 1801-1811 (1998) 1801 Printed in Great Britain © The Company of Biologists Limited 1998 JCS3772

The sparing of extraocular muscle in dystrophinopathy is lost in mice lacking utrophin and

J. D. Porter1,*, J. A. Rafael2, R. J. Ragusa3, J. K. Brueckner3, J. I. Trickett2 and K. E. Davies2 1Department of Ophthalmology, Case Western Reserve University, 11100 Euclid Avenue, Cleveland, OH 44106-5068, USA 2Genetics Unit, Department of Biochemistry, University of Oxford, South Parks Road, Oxford, OX1 3QU, UK 3Department of Anatomy and Neurobiology and Department of Ophthalmology, University of Kentucky Medical Center, Lexington, Kentucky, 40536-0084, USA *Author for correspondence

Accepted 20 April; published on WWW 15 June 1998

SUMMARY

The extraocular muscles are one of few skeletal muscles Mice lacking expression of utrophin alone, like the that are structurally and functionally intact in Duchenne dystrophin-deficient mdx mouse, showed no pathological muscular dystrophy. Little is known about the mechanisms alterations in extraocular muscle. However, mice deficient responsible for differential sparing or targeting of muscle in both utrophin and dystrophin exhibited severe changes groups in neuromuscular disease. One hypothesis is that in both the accessory and principal extraocular muscles, constitutive or adaptive properties of the unique with the eye muscles affected more adversely than other extraocular muscle phenotype may underlie their skeletal muscles. Selected extraocular muscle fiber types protection in dystrophinopathy. We assessed the status of still remained spared, suggesting the operation of an extraocular muscles in the mdx mouse model of muscular alternative mechanism for muscle sparing in these fiber dystrophy. Mice showed mild pathology in accessory types. We propose that an endogenous upregulation of extraocular muscles, but no signs of pathology were evident utrophin is mechanistic in protecting extraocular muscle in in the principal extraocular muscles at any age. By dystrophinopathy. Moreover, data lend support to the immunoblotting, the extraocular muscles of mdx mice hypothesis that interventions designed to increase utrophin exhibited increased levels of a dystrophin analog, levels may ameliorate the pathology in other skeletal dystrophin-related or utrophin. These data suggest, muscles in Duchenne muscular dystrophy. but do not provide mechanistic evidence, that utrophin mediates eye muscle protection. To examine a potential causal relationship, knockout mouse models were used to Key words: Muscular dystrophy, mdx mice, Extraocular muscle, determine whether eye muscle sparing could be reversed. Dystrophin, Utrophin

INTRODUCTION been established between muscle or motoneuron group- specific properties and the propensity to disease. Skeletal muscles are not all created equal in their response to We have previously suggested that the selective involvement neuromuscular disease. While the targeting or sparing of or sparing of the extraocular muscles (EOMs) in a variety of particular muscles or muscle groups is not an infrequent neuromuscular diseases may be causally linked to fundamental occurrence, there often are no overt phenotypic or functional differences in the molecular, cellular, structural and functional differences between the affected and unaffected muscles, or properties of these muscles (Porter and Baker, 1996; Porter et their motoneurons, that might provide clues to the underlying al., 1997). EOM has been characterized as one of at least three mechanisms for disease selectivity. Distal myopathies spare allotypes, or classes (Hoh et al., 1989; Porter seemingly identical muscles that are in close proximity to the and Baker, 1996). The currently recognized allotypes, affected musculature. Likewise, amyotrophic lateral sclerosis limb/diaphragm, masticatory and extraocular, certainly share has devastating consequences for spinal motoneurons, but many of the essential traits that typify skeletal muscle tissue, either spares or leaves cranial motoneurons intact until late in but also diverge from one another in several respects that may the disease course. Although the identification of mechanisms relate to their unique functions. For example, EOMs differ that target or spare particular muscle groups clearly would aid from other muscles in expression, innervation pattern, understanding of disease pathogenesis and treatment, we must contractile properties, hormone receptors and cell surface first identify the key differences between muscles that translate markers, perhaps in response to the diverse functional demands into disease targeting patterns. Unfortunately, for most placed upon this muscle group (reviewed in Porter et al., 1995, neuromuscular diseases, such direct relationships have not yet 1997). Even a cursory examination of muscle properties 1802 J. D. Porter and others demonstrates that the traditional muscle fiber classification autosomal product utrophin (Love et al., 1989; Tinsley et al., schemes cannot be used to characterize EOM fiber types. 1992; Winder et al., 1995; Guo et al., 1996). However, the Taken together, it is not difficult to speculate that allotype- developmental regulation of utrophin (Khurana et al., 1991; specific properties might, at least in part, provide the basis for Clerk et al., 1993) and its normal adult pattern of restriction to the selectivity of neuromuscular disorders. a small percentage of the , the neuromuscular and The EOMs are responsible for voluntary movements of the myotendinous junctions (Ohlendieck et al., 1991; Gramolini et eyes. Their structure and function are conspicuously spared in al., 1997), most likely limit the compensatory role that utrophin Duchenne muscular dystrophy and in all studied animal models might play in dystrophinopathy. There is, however, compelling of dystrophinopathy (Karpati and Carpenter, 1986; Karpati et evidence that interventions that increase utrophin to a level al., 1988; Kaminski et al., 1992; Khurana et al., 1995; Ragusa sufficient to saturate β-, and to preserve the et al., 1996). The pathogenic alterations characteristic of transmembrane glycoprotein complex, can structurally and dystrophinopathy have not been found in eye muscle. Since functionally rescue mdx limb and diaphragm musculature diplopia (double vision) is not a clinical finding in Duchenne (Tinsley et al., 1996; Deconinck et al., 1997c). Might an EOM muscular dystrophy, the fine muscular control needed for the capacity to endogenously increase utrophin explain the stabilization and interocular coordination of the eyes is not observed muscle sparing pattern? We have characterized the disturbed. Similarly, the amplitude/peak velocity relationship sparing of EOM in the mdx and, using utrophin- and of exceedingly rapid saccadic eye movements is not altered by dystrophin/utrophin-deficient mice, provided support for the dystrophin deficiency (Kaminski et al., 1992). In limb muscles hypothesis that utrophin protects eye muscle in it is the fast-twitch muscle fiber types that appear to be acutely dystrophinopathy. sensitive to the absence of dystrophin (Webster et al., 1988; Minetti et al., 1991). It is thought that the daily mechanical stresses to which fast-twitch fibers are subjected contribute MATERIALS AND METHODS toward their particular susceptibility in dystrophin-deficient sarcolemmal destabilization. Thus, it is counterintuitive that Mice The C57BL/10 and dystrophin-deficient mdx (C57BL/10ScSn- the EOMs, certainly among the fastest of mammalian skeletal mdx muscles, would be spared in and animal models of Dmd /J) mice are commercially available (Harwell Labs; Jackson Labs, Bar Harbor, ME). To examine the potential compensatory role Duchenne muscular dystrophy. of utrophin in EOMs of dystrophic mice, two transgenic strains were The differential protection of EOM in Duchenne muscular evaluated. Utrophin-knockout mice were generated by homologous dystrophy may relate to muscle group-specific properties that recombination of a NeoR-containing cassette into exon 7 of mouse either prevent or adapt for the cascade of events that lead to utrophin, as described in Deconinck et al. (1997a). Double-knockout myofiber degeneration. As in other muscles, dystrophin is mice (dko), lacking expression of both utrophin and dystrophin, were localized to the extraocular sarcolemma in and all obtained by crossing the utrophin-knockout mice with dystrophin- animal models that have been examined and is absent in deficient mdx mice, as described in Deconinck et al. (1997b). All dystrophinopathy (Matsumura et al., 1992; Khurana et al., animal-use procedures conformed to the appropriate regulations in the 1995). Potential protective or compensatory mechanisms that respective laboratories (USA, National Institutes of Health; UK, may be operative in dystrophin-deficient EOMs have been Medical Research Council). reviewed in Porter (1998). The identification of the definitive Morphology mechanism(s) for eye muscle protection is, in part, limited, We evaluated muscle ultrastructure in C57BL/10, mdx, utrophin- because the precise sequence of events that is triggered by deficient and utrophin/dystrophin-deficient mice. Wild-type and mdx dystrophinopathy is not fully resolved. Most theories of mice were studied at multiple ages; utrophin-deficient and pathogenesis implicate loss of sarcolemmal integrity with utrophin/dystrophin-deficient mice were studied at 8 weeks of age, an consequent increase in intracellular free calcium levels and advanced pathological stage in the dko strain (Deconinck et al., protease activation (see Campbell, 1995). Thus, one option for 1997b). Mice were either perfusion-fixed, or orbital and hindlimb muscle protection is that dystrophinopathy might not exert a tissues were dissected and rapidly immersed in dialdehyde fixative. destabilizing influence upon the EOM sarcolemma. Perhaps Tissue sections were postfixed in 1% osmium tetroxide, stained en the mechanical stresses seen by this muscle group are less bloc in uranyl acetate, dehydrated in a methanol series and propylene oxide, and embedded in epoxy resin. 1 µm sections were cut with a (Karpati and Carpenter, 1986; Karpati et al., 1988), or the Reichert ultramicrotome, stained with Toluidine Blue, and examined extraocular sarcolemma is stabilized by alternate using brightfield optics. Ultrathin (90 nm) sections were stained with transmembrane protein complexes (Matsumura et al., 1992; uranyl acetate and lead citrate and examined with an Hitachi H-7000 Law et al., 1994; Hodges et al., 1997). The localization of electron microscope. To estimate the extent of myofiber regeneration, utrophin to both and non-junctional counts of centrally nucleated fibers were made from the 1 µm epoxy sarcolemma in mdx EOM (Matsumura et al., 1992) lends resin sections. While this method underestimates the actual number support to the alternate protein theory. of myotubes and regenerated myofibers, it provides an effective In this study, we evaluated the hypothesis that a protein cumulative index for regeneration (Karpati et al., 1988). homologous to dystrophin mediates the protection of EOM in Western blots the mdx mouse. A potential substitute for dystrophin is Individual muscles were dissected from 8-week old C57BL/10 and dystrophin-related protein, or utrophin. Molecular models of mdx mice for evaluation of utrophin protein levels by immunoblotting. the protein complex linking to extracellular For each repetition, muscles were pooled from five mice per strain. matrix (see Campbell, 1995; Blake et al., 1996) suggest that The EOM samples included only the rectus and oblique muscles (not the key domains of dystrophin are the F- and β- the retractor bulbi and levator palpebrae superioris) in order to dystroglycan binding sites. Identical sites are present on the evaluate utrophin content only in those muscles with no evidence of Extraocular muscle and muscular dystrophy 1803 dystrophy. 50 µg of total protein homogenate from C57BL/10 and describe the EOMs, we used a system described in Spencer and Porter mdx EOM and quadriceps was loaded on a 6% SDS-PAGE gel, run (1988) and Porter et al. (1995). The rectus and oblique EOMs consist at 80 V for 3-4 hours on a Mini Protean II system (Bio-Rad, Hercules, of six fiber types based upon: (1) distribution into orbital and global CA), and transferred onto nitrocellulose at 80 V for 2 hours in a wet layers, (2) innervation status; single versus multiple nerve transfer apparatus. The blot was probed with an antibody to utrophin contacts/fiber and (3) mitochondrial/oxidative enzyme content. This (Mancho3; kindly provided by Glen Morris, MRIC Biochemistry scheme identifies orbital singly, orbital multiply, global red singly, Group, Wrexham, UK) at 1:100 dilution and developed using the ECL global intermediate singly, global pale singly and global multiply detection system (Boehringer-Mannheim). Western blots were run in innervated fiber types. The accessory EOMs are identified by their triplicate and changes in relative utrophin levels were determined by location relative to rectus and oblique muscles. The levator palpebrae densitometry. superioris lacks the layered organization (i.e. the orbital layer is absent) and exhibits some additional differences in fiber type Myosin immunocytochemistry and in situ hybridization composition (Porter et al., 1989). The fiber type composition of the In order to better characterize some EOM fiber types that were retractor bulbi muscle is closer to that of other skeletal muscles apparently spared in the dko mice, the expression of selected myosin (Pachter et al., 1976; Spencer and Porter, 1988). heavy chain isoforms was studied, as previously described in Brueckner et al. (1996). Immunocytochemistry was used to identify µ global multiply innervated fibers in unfixed 12 m cryosections. RESULTS Sections were processed with a monoclonal antibody to slow myosin (A4.840; 1:5; Developmental Hybridoma Bank, Iowa City, IA) and immunoreactivity was visualized using biotinylated secondary Extraocular muscle is protected in dystrophinopathy antibodies (Vector Labs, Burlingame, CA), followed by Vectastain Although prior reports (Karpati and Carpenter, 1986; Karpati ABC reagent (Vector Labs, Burlingame, CA) and development with et al., 1988; Kaminski et al., 1992; Khurana et al., 1995; nickel-enhanced 3,3-diaminobenzidine (Sigma, St Louis, MO). In situ Ragusa et al., 1996) have established that EOM does not hybridization was used to localize mRNA for the EOM-specific exhibit the frank degeneration and regeneration that typifies 35 myosin isoform in muscle cryosections, with an S-labeled other skeletal muscles, a thorough analysis of the EOM oligonucleotide probe (amino acids 1705-1719; Wieczorek et al., phenotype in dystrophinopathy has not yet been done. While 1985). extensive degeneration was noted in mdx hindlimb musculature Extraocular muscle organization and fiber classification by postnatal day (P) 20, with widespread regeneration between The principal EOMs consist of four rectus and two oblique muscles P35 and P90, qualitative signs of degeneration and arranged radially around the globe and optic nerve. Since traditional regeneration were not observed in the rectus and oblique skeletal muscle fiber-type classification systems do not accurately EOMs at any of the ages evaluated (Fig. 1A,B). The

Fig. 1. Light photomicrographs of extraocular (A,B) and accessory extraocular (C,D) muscles in 8-week mdx mice. Orbital (A) and global (B) muscle layers show the unique morphological fiber types characteristic of wild-type EOM, but exhibit no signs of myofiber degeneration/ regeneration. By contrast, retractor bulbi (C) and levator palpebrae superioris (D) muscles include regenerated myofibers, with central nuclei (arrows). Bar, 25 µm. 1804 J. D. Porter and others

Table 1. Central nucleated muscle fibers in control and mdx EOM and leg muscles P35 P60 P120 >P450 Muscle Control mdx Control mdx Control mdx Control mdx EOM 0.09±0.09 0.12±0.07 0.16±0.08 0.04±0.04 0.04±0.04 0.38±0.25 0.11±0.07 1.18±0.35 RB 0.40±0.12 2.76±1.89 0.15±0.15 16.70±4.24 0 27.43±4.88 0 22.64±3.72 Leg 0.13±0.13 31.72±1.67 0.10±0.10 36.28±4.36 0 49.37±4.18 n.d. n.d.

Values represent the mean percentages of centrally nucleated fibers ± s.e.m. in counts from 1 µm epoxy resin sections of control and mdx muscles at the noted postnatal ages. These serve as an index of level of degeneration/regeneration in extraocular (EOM), retractor bulbi (RB) and gastrocnemius (Leg) muscles of dystrophin-deficient mice. n = 3 mice/age/strain. Analyses were not done (n.d.) for the old mdx hindlimb muscle, but central nuclei reportedly increase in limb muscles of elderly dystrophin-deficient mice (Pastoret and Sebille, 1995). distribution of the six distinct fiber types into orbital and global dystrophin-related protein (DRP) or utrophin, which normally muscle layers was normal in the EOMs of dystrophin-deficient is restricted to the specialized sarcolemma at the mice. Moreover, at the ultrastructural level, the principal neuromuscular junction and myotendinous junction EOMs did not exhibit any of the signs typically associated with (Ohlendieck et al., 1991). Indeed, there is prior experimental dystrophic muscle, including sarcolemmal breaks, myofibril support for sarcolemmal localization of utrophin in mdx EOM disruption, swollen mitochondria and sarcotubular (Matsumura et al., 1992). Key structural similarities might aggregations. In addition, no obvious increase in endomysial allow utrophin to fill the missing link between muscle connective tissue or adipose tissue, as previously described for cytoskeleton and extracellular matrix and functionally the mdx diaphragm (Stedman et al., 1991), was observed. By substitute for dystrophin. To study whether a redundancy contrast, the two accessory EOMs exhibited a delayed between utrophin and dystrophin might mediate the muscle pathologic response to dystrophin deficiency, with onset of group-specific sparing in dystrophinopathy, utrophin levels degeneration by P35 and obvious regeneration between P60 were examined in adult mdx quadriceps and EOM by western and P120 (Fig. 1C,D). Since limb muscle pathology is reported blot analysis. The Mancho3 antibody used to detect utrophin to increase at an advanced age in mdx mice (Pastoret and identified a single band of similar molecular mass in both Sebille, 1995), we also examined EOM in mdx mice older than muscle groups (Fig. 2). Utrophin levels were elevated by 450 days. Again, there was no evidence of pathology in EOM approximately fivefold in dystrophin-deficient quadriceps of old mdx mice, suggesting that the muscle sparing muscles compared to those of age-matched C57BL/10 mice. mechanism is operative throughout the lifespan of the animal. Some of this increase can be attributed to the presence of During myogenesis, the centrally placed nuclei present in regenerative myotubes in the dystrophic hindlimb, which myotubes migrate to assume a subsarcolemmal location in normally show transient expression of utrophin (Clerk et al., adult myofibers. By contrast, the regenerated muscle fibers in the mdx mouse characteristically retain centrally located nuclei (Karpati et al., 1988), providing an effective index of muscle regeneration. To establish the level of myofiber regeneration in extraocular, accessory extraocular and hindlimb muscles, the percentage of centrally nucleated myofibers was evaluated in mdx and control strain mice, at ages from P35 to >P450. Overall, our fiber counts provided slightly lower values for hindlimb muscles than in other studies (Karpati et al., 1988), most likely because 1 µm epoxy resin sections were used instead of thicker paraffin sections. Consistent with prior observations, the percentage of centrally nucleated fibers in mdx gastrocnemius was already elevated by P35 and increased to roughly 50% by P120, versus <1% at all ages in control mice (Table 1). Central nuclei were, however, rare in the principal EOMs of mdx mice, with slightly more than 1% found only in the aged mdx group. Together with the fine structural findings, these data establish that there apparently are no abnormalities in the EOMs of mdx mice. The fiber count analysis also Fig. 2. Western blot analysis of C57BL/10 and mdx quadriceps confirmed the qualitative observations of the relatively mild (quad) and extraocular (EOM) muscles with an antibody raised and temporally delayed pathology in accessory EOMs, as the against the mouse utrophin protein (Mancho3). Only rectus and percentage of centrally nucleated myofibers in the mdx oblique EOMs were used in this analysis. Because of the small size retractor bulbi lay roughly intermediate between the of the EOMs, each lane includes muscle pooled from five mice. gastrocnemius and EOM values from P60 onward. Analyses were done in triplicate. Utrophin levels were elevated in dystrophin-deficient quadriceps (lane 2) muscles compared with mdx extraocular muscle exhibits utrophin C57BL/10 mice (lane 1). Some of this increase can be attributed to the presence of regenerating muscle fibers, which normally show a upregulation transient expression of utrophin. An increase in utrophin level was One potential mechanism for the sparing of the EOM group in also observed in mdx EOM (lane 4) compared with C57BL/10 (lane dystrophinopathy is the upregulation of a homologous protein, 3). None of the EOM increase can be attributed to regenerated fibers. Extraocular muscle and muscular dystrophy 1805

neuromuscular junctions of utrophin-knockout mice (Deconinck et al., 1997a; Grady et al., 1997a), including a decrement in post-junctional folding. Since the neuromuscular junctions associated with most EOM fiber types normally possess few post-junctional folds (Spencer and Porter, 1988), it is difficult to assess alteration in this feature in eye muscles of these mice. Collectively, the mdx and utrophin-knockout data show that neither dystrophin nor utrophin are essential in maintaining the structural and functional integrity of EOM. Mice deficient in both dystrophin and utrophin exhibit severe abnormalities in the extraocular muscles In most skeletal muscles, the endogenous increase in utrophin that accompanies dystrophinopathy (present data; Matsumura Fig. 3. Light photomicrograph from a rectus EOM of an 8-week-old utrophin knockout mouse. Neither orbital (orb) or global (glob) et al., 1992; Helliwell et al., 1992; Karpati et al., 1993) muscle layers showed pathologic alterations. Bar, 25 µm. apparently is insufficient to stabilize the sarcolemma and protect against myofiber degeneration (Straub et al., 1997). By contrast, the correlation between the increase in utrophin levels 1993). Utrophin levels increased approximately threefold in and the absence of pathology in the EOMs of mdx mice eye muscles from dystrophic animals. The western analyses suggests that utrophin actively compensates for dystrophin used only the rectus and oblique EOMs, which lack evidence deficiency and selectively maintains sarcolemmal integrity in of pathology or regeneration in the mdx model, so none of the this muscle group. increase in eye muscle utrophin could be attributed to To further assess this hypothesis, the EOM phenotype was immature or regenerative myofibers. This finding supports the examined in mice lacking both utrophin and dystrophin notion that an endogenous mechanism mediates an increase in (Deconinck et al., 1997b). The accessory EOMs, which are utrophin in a manner that may serve to selectively protect the only mildly sensitive to dystrophin deficiency (see above), EOMs in dystrophinopathy. were severely affected by the absence of both dystrophin and utrophin. Regenerated myofibers, assessed as percentage of Mice deficient in utrophin have morphologically internal nuclei, accounted for 35-45% of dko levator palpebrae normal extraocular muscles superioris and retractor bulbi muscles, compared to 10-16% in The high constitutive level in wild-type mice and the age-matched mdx mice (Table 2). The rodent retractor bulbi substantive increase in utrophin in the EOMs of mdx mice consists of multiple muscle slips and some variability was suggested that utrophin may normally play an important role detected within the muscles of individual animals. In some in either the maturation or maintenance of extraocular regions of the retractor bulbi, virtually all fibers were either myofibers. In order to assess this notion further, we evaluated pathologic or regenerative fibers (Fig. 4C). In these regions, the morphology of EOMs from a knockout mouse model in small diameter, centrally nucleated myotubes comprised the which utrophin was undetectable in skeletal muscle bulk of the muscle, as if the regenerative process was arrested (Deconinck et al., 1997a). In spite of the complete absence of at an early stage. Other fibers were basophilic in appearance at utrophin, no morphologic abnormalities were observed in the light microscopic level, with a corresponding myofibril either EOMs (Fig. 3) or accessory EOMs. As previously noted for hindlimb musculature (Deconinck et al., 1997a; Grady et al., 1997a), myofiber regeneration apparently is negligible in Table 2. Central nucleated fibers in mdx and double- all EOMs, as centrally nucleated fibers were rare (<0.5% in knockout mouse EOMs rectus, levator palpebrae superioris and retractor bulbi muscles; n=3 mice) in utrophin-deficient mice. Light and electron Muscle mdx* dko* P microscopy did not reveal any morphological alterations, even LPSa 10.4±2.5 41.6±2.4 <0.01 RBb 16.2±3.2 39.8±0.9 <0.05 in the retractor bulbi and levator palpebrae superioris muscles c EOMorb 0.5±0.1 1.6±0.6 n.s. d that are sensitive to dystrophin deficiency (see above). EOMglob 0.5±0.1 23.5±1.6 <0.01 Utrophin deficiency produces mild defects in neuromuscular transmission in hindlimb musculature (Deconinck et al., *Values are mean percentage of central nucleated fibers ± s.e.m. from age- 1997a; Grady et al., 1997a). We anticipated that mice lacking matched 8-week-old mdx and double-knockout (dko) mice, with n = 3 mice for mdx and n = 5 for dko. P values from two-tailed Student’s t test; n.s., not utrophin would exhibit selective alterations in the EOM orbital significant. layer fiber types, since these are highly oxidative, continuously aLevator palpebrae superioris (LPS) data based upon counts of 150-225 active in the maintenance of eye position, and particularly muscle fibers/mouse. sensitive to denervation (reviewed in Porter et al., 1995; Porter bRetractor bulbi (RB) data based upon counts of 200-440 muscle fibers/mouse. and Baker, 1996). There were, however, no detectable c EOM rectus muscle, orbital layer (EOMorb) data based upon counts of morphological alterations in the phenotype of the fibers in 290-805 muscle fibers/mouse. d either layer of the rectus and oblique muscles. Prior reports EOM rectus muscle, global layer (EOMglob) data based upon counts of have noted mild structural and functional alterations at the 240-575 muscle fibers/mouse. 1806 J. D. Porter and others

Fig. 4. Light photomicrographs of extraocular (A,B) and accessory extraocular (C,D) muscles in an 8-week-old dko mouse. The orbital layer (A) of rectus muscles showed no signs of pathologic alterations. The rectus global layer (B) exhibited severe pathology, with many abnormally basophilic fibers and small, centrally nucleated myotubes. Large areas of the retractor bulbi (C) and levator palpebrae superioris (D) muscles also were composed of small myotubes. Bar, 25 µm. disruption usually associated with the hypercontracted more prevalent than in the global layer of either C57BL/10 or myofibers that characterize dystrophic muscles. By contrast, utrophin-deficient EOMs. In other rectus and oblique muscles, regenerated (central nuclei), larger diameter myofibers most global layer fibers had matured beyond the myotube predominated in other regions of the retractor bulbi. Similarly, stage, but many exhibited central nuclei indicating that they dko mice exhibited heterogeneity in the levator palpebrae were regenerated myofibers. superioris, with some regions comprising small myotubes (Fig. 4D) and other regions exhibiting larger myofibers, often with Sparing of some fiber types is still apparent in mice central nuclei. This pattern of pathology is often associated deficient in both utrophin and dystrophin with functional ocular motility deficits, including ptosis (i.e. Despite the overall high level of pathology in rectus and drooping of the eyelid due to compromise of the levator oblique EOMs of mice deficient in both utrophin and palpebrae superioris) or exophthalmos (i.e. outward bulging of dystrophin, some muscle fibers remained spared from the the eye, resulting from decreased retractor bulbi function) (e.g. dystrophic process. Myofibers lacking morphopathological as in merosin-deficient 129/ReJ-Lama2dy mice; J. D. Porter, changes or central nuclei were identified as belonging to three unpublished). Neither sign of eye muscle dysfunction was of the six EOM fiber types. Immunocytochemical evaluations observed in the utrophin/dystrophin-deficient mice, perhaps with a slow myosin heavy chain-specific antibody showed that because (1) the presence of hypercontracted muscle fibers in many of the intact muscle fibers in the global layer of severely both muscles maintained normal eyelid and globe position or affected EOMs were of the global multiply innervated fiber (2) muscle residual capacity remained adequate for demands. type (Fig. 6A). Fine structural analysis (Fig. 6B) confirmed this In utrophin/dystrophin-deficient mice, the rectus and oblique finding, as many of the intact fibers exhibited the large EOMs also exhibited severe dystrophic alterations. In myofibrils and poorly defined internal membrane system that particular, morphological evaluation of the global layer of the characterize this muscle fiber type (Spencer and Porter, 1988). EOMs showed extensive signs of degeneration and Motor nerve terminals and the cytoarchitecture of these fibers regeneration (Fig. 4B; Table 2). In some muscles, the global were indistinguishable from those of C57BL/10 mice. Perhaps layer pathology was as severe as was noted in the retractor important to their sparing in the dko mice, these are non-twitch bulbi and levator palpebrae superioris. In such regions, the muscle fibers rarely found in mammals, that achieve slow, muscle was predominately composed of small, centrally graded contractions only focally at neuromuscular junction nucleated myotubes, often exhibiting primitive neuromuscular sites distributed along their entire length (Porter et al., 1995). junctions (Fig. 5). Consistent with the massive regenerative The cytoarchitecture of both orbital singly and orbital response, myelinated and unmyelinated axon profiles were multiply innervated muscle fiber types was also preserved in Extraocular muscle and muscular dystrophy 1807

Fig. 5. Representative electron photomicrographs illustrating morphology of the global layer of rectus muscles from dko mice. Many of the muscle cells present in this area appeared to be arrested at an early myotube stage (A), given the small amount of cytoplasm surrounding centrally placed nuclei (n, muscle nuclei; a, a myelinated axon). Neuromuscular junction morphology on the myotubes (B) also remained primitive, without specializations associated with the presynaptic nerve terminals (nt) or postsynaptic sarcolemma. Bars, 3 µm (A), 1 µm (B).

the utrophin/dystrophin-deficient mice (Figs 4A and 7). Central nuclei did not exceed control levels in this muscle layer (Table 2) and sarcomeric organization, mitochondrial morphology and distribution, internal membrane system content and neuromuscular junction morphology were normal. The EOM- specific myosin (Wieczorek et al., 1985) normally localizes to the orbital layer (Brueckner et al., 1996). In situ hybridization showed no qualitative changes from the wild-type pattern of EOM-specific myosin mRNA in mdx or utrophin-knockout mice. However, the mRNA levels, as indicated by silver grain density, were reduced in the dko mice (Fig. 8). Therefore, with the exception of a muscle-specific myosin, the orbital layer fiber types either mature and function independent of both utrophin and dystrophin or can functionally replace both .

DISCUSSION

Prior studies have shown that EOM is uniquely protected in dystrophinopathy. We extended these studies by demonstrating that there are no apparent signs of either generalized or fiber type-specific alterations in EOM of mdx mice of any age, and only mild damage in the two accessory EOMs. We also focused upon the idea that the EOMs might naturally compensate for a dystrophin deficiency by over-expressing a synergistic protein, utrophin. Utrophin levels increase in the EOMs of mdx mice and the absence of both dystrophin and utrophin eliminates the protection of most EOM fibers. Together, these Fig. 6. Differential interference contrast (A) and electron (B) data provide strong support for the hypothesis that an adaptive photomicrographs depicting intact global layer multiply innervated increase in utrophin mediates the rescue of eye muscle in fibers in dko mice. The global multiply innervated fibers, identified muscular dystrophy. by their characteristic staining with a slow myosin monoclonal antibody (A), represented many of the intact fibers in the global layer Selective sparing of the EOMs in dystrophinopathy in of rectus muscles. These fibers retained their fine structural several species, and partial involvement of accessory EOM in characteristics (B), including small, simple neuromuscular junctions the mdx mouse, have been previously demonstrated (Karpati (nt), large myofibrils in both the A and I bands (A, I), and sparse, and Carpenter, 1986; Karpati et al., 1988; Kaminski et al., small mitochondria (m). Bars, 20 µm (A), 1 µm (B). 1992; Khurana et al., 1995; Ragusa et al., 1996). Because of 1808 J. D. Porter and others

Fig. 8. Darkfield photomicrographs of autoradiograms showing in situ hybridization with an EOM-specific myosin heavy chain oligonucleotide probe in wild-type (A) and dko (B) mice. This myosin isoform is normally abundant in the c-shaped orbital layer of rectus and oblique muscles (A), as previously reported (Brueckner et al., 1996; Brueckner and Porter, 1998). Double-knockout mice Fig. 7. Electron photomicrographs of an orbital singly innervated exhibited a dramatic reduction in silver grain density (B), indicating muscle fiber from a dko mouse. Fiber exhibits the same general fine a reduction in EOM-specific myosin expression. structural organization as in wild type mice, including central and subsarcolemmal mitochondrial (m) aggregates (A) and small myofibrils that are well defined by internal membrane system elements (B). Bars, 3 µm (A), 1 µm (B). that transcription remains restricted to junctional myonuclei, it locates to non-junctional sarcolemma where it could substitute for dystrophin (Matsumura et al., 1992; Helliwell et al., 1992; the complexity of the fiber type composition of EOM (Spencer Karpati et al., 1993). This finding, plus the substantial amino and Porter, 1988; Porter and Baker, 1996), the involvement of acid identity of utrophin with dystrophin, including the key F- this muscle group in neuromuscular disease has often proved actin and β-dystroglycan binding sites (Love et al., 1989; difficult to interpret. We have extended prior studies to Tinsley et al., 1992; Winder et al., 1995; Guo et al., 1996), is demonstrate that there are no generalized or fiber type-specific consistent with a proposed adaptive role for utrophin in alterations in EOM of the mdx mouse. These data also show muscular dystrophy (Karpati et al., 1993; Tinsley and Davies, that not only the retractor bulbi, but also the levator palpebrae 1993). The most compelling argument for this view is that superioris, exhibit a level of pathology intermediate between targeted overexpression of utrophin restores both the structural extraocular and limb musculature. Taken together, the mdx and functional properties of limb musculature in the mdx orbital musculature includes both fully protected EOM and mouse (Tinsley et al., 1996; Deconinck et al., 1997c). partially susceptible accessory EOMs that provide an excellent On the basis of immunoblotting, utrophin levels appear to model for understanding the mechanisms that spare some be constitutively higher in extraocular than hindlimb muscle. muscle groups in dystrophin deficiency. This finding may relate to myofiber size in EOM; the smaller We have examined the question: can an analog to dystrophin muscle fibers have a higher surface to volume ratio and, be responsible for the protection of EOM in Duchenne therefore, a higher proportion of total muscle protein would be muscular dystrophy models? A likely candidate, represented by that at neuromuscular junctions. In addition, the 6-encoded utrophin (Love et al., 1989; Khurana et al., 1990), presence of multiply innervated fibers and the elongated nerve- is localized throughout the sarcolemma during myogenesis, but muscle contact zones of orbital singly innervated fibers subsequently becomes restricted to neuromuscular junction (Spencer and Porter, 1988) would increase the sarcolemmal and myotendinous junction sites (Takemitsu et al., 1991; Clerk surface area that is normally occupied by utrophin. Our et al., 1993). Utrophin normally is transcribed only by immunoblot data show that mdx mouse EOM exhibits a junctional myonuclei and occupies the cytoplasmic surface of threefold elevation in utrophin. A similar augmentation of the sarcolemma at neuromuscular junctions complexed with β- utrophin levels in dystrophic eye muscle was not observed in dystroglycan, F-actin and syntrophin. Although utrophin a prior immunoblot analysis (Khurana et al., 1995). However, mRNA apparently does not increase in dystrophic skeletal by immunocytochemistry, Matsumura et al. (1992) muscle (Karpati et al., 1993; Schofield et al., 1995), suggesting demonstrated that utrophin spreads from the neuromuscular Extraocular muscle and muscular dystrophy 1809 and myotendinous junction loci to be expressed throughout the Therefore, the high activation pattern of EOM probably is not sarcolemma in mdx mouse EOM. A dystrophy-associated a factor in its ability to modulate utrophin levels. It will be increase in utrophin is not altogether unusual, at least in small important to understand the molecular mechanisms that are caliber, presumptive regenerative, fibers. Previous studies have operative in expression of sufficient quantity of utrophin to shown both increased utrophin levels and apparent expansion rescue the EOM group. of cellular localization to sites formerly occupied by dystrophin Many of the unique properties of EOM directly relate to their in limb and diaphragm muscles of human and animal distinct functional roles in eye movement. In turn, some of Duchenne muscular dystrophy models (Khurana et al., 1991; these properties may provide fortuitous protection or targeting Takemitsu et al., 1991; Tanaka et al., 1991; Clerk et al., 1993; in neuromuscular disease (Porter and Baker, 1996; Porter et al., Karpati et al., 1993). However, the level of utrophin 1997). mdx EOMs either maintain the embryonic utrophin upregulation might not be sufficient to fully compensate for the expression pattern into adulthood, which does not occur in absence of dystrophin, even in the mdx mouse which shows other skeletal muscles in dystrophy (Pons et al., 1994; Taylor mild signs of limb dystrophy (Matsumura and Campbell, et al., 1997), or must increase utrophin levels in mature 1994). Consistent with this view, transgenic models provide myofibers in response to the disease process. The unusual support for the ‘dose dependency’ of utrophin in rescue of developmental profile of EOM, particularly their relatively late diaphragm and hindlimb muscles (Tinsley et al., 1996; maturation compared to other muscles (Porter and Baker, 1992; Deconinck et al., 1997c). Studies of truncated utrophin Brueckner et al., 1996; Porter et al., 1997), makes it difficult construct insertion into the mdx genome show that there is a to exclude either possibility. These data strongly support the threshold utrophin level for myofiber rescue (Deconinck et al., concept that utrophin is sufficient to serve as a key element in 1997c). What is unique about utrophin expression in the sparing of EOM in mdx mice, but do not necessarily dystrophic mouse EOM is that none of the increase can be exclude the involvement of other mechanisms (Porter, 1998). attributed to regenerative fibers, since mdx rectus and oblique The full answer as to why EOM is protected in muscular muscles are completely spared. The elevation in eye muscle dystrophy is limited at this time, until the precise mechanisms utrophin occurs in mature myofibers, without the need for a for myofiber degeneration in muscular dystrophy are fully degeneration/regeneration cycle in which an embryonic understood. Certainly, the hypothesis of sarcolemmal utrophin expression pattern is first restored and then adaptively destabilization leading to loss of calcium homeostasis retained. Thus, the critical factor in EOM sparing in dystrophic (reviewed by Campbell, 1995) has much support; calcium mice might be spatial or temporal factors underlying the scavenging capacity may indeed play at least a secondary role increase in utrophin in mature, fully functional myofibers. in EOM protection (Khurana et al., 1995; Porter, 1998). Resolution of the utrophin expression pattern in EOM of mdx However, alternative mechanisms, including muscle group- mice at different developmental stages will yield further insight specific signaling pathways, might be significant in the sparing into this question. of the eye muscles in Duchenne and other muscular In isolation, the increased utrophin in mdx EOM only dystrophies (Vachon et al., 1997; Straub et al., 1997). provides correlative, rather than causal, evidence for a Our findings also establish that, for some EOM fiber types, homologous protein substituting for dystrophin. We chose to utrophin cannot be the key protective mechanism. Despite the address the potential role of utrophin in mediating muscle severe myopathic changes induced in EOMs of dko mice, protection by using a mouse model deficient in both utrophin specific fiber types still retain their specialized morphologic and dystrophin (Deconinck et al., 1997b). Our data show that characteristics. We did observe a decrement in EOM-specific the loss of the protected status of EOM represents an important myosin expression in the otherwise spared orbital layer fibers; difference between the mdx and dko mice. By comparing the but, this trait usually appears late in orbital layer maturation, severity of pathologic findings in EOM (present data) with and is activity dependent (Brueckner et al., 1996; Brueckner other skeletal (Deconinck et al., 1997b; Grady et al., 1997b) and Porter, 1998), so alterations in this myosin may relate to a muscles, it is apparent that the consequences of the dko for the general reduction in oculomotor behavior that accompanies the EOMs are more deleterious than for the hindlimb or severe global layer pathology. While all fiber types in EOM are diaphragm. Regenerated fibers in the mdx hindlimb attain unique, those types that remain spared in mice deficient in both approximately 80% of wild-type myofiber diameter (Karpati et dystrophin and utrophin, the orbital singly and multiply al., 1988) and a similar level of recovery may occur in the dko innervated and global multiply innervated fiber types, exhibit mouse. By contrast, the eye muscles that are protected (EOM) particularly unusual properties (see Porter and Baker, 1996; or only mildly affected (accessory EOM) in the mdx, exhibit Porter et al., 1997). The constitutive or adaptive means by regeneration that appears to be arrested, often at the myotube which such apparently dystrophin- and utrophin-independent stage, in dko mice. In such regions, myonuclei are central, myofibers are spared is unknown. occupy a high percentage of the cross-sectional area, and In conclusion, these data provide compelling support for the neuromuscular junctions are immature. It is unlikely that such idea that utrophin actively participates in the rescue of EOM fibers contribute substantively in the delicate task of in the mdx mouse. First, an atypical increase in utrophin in coordinated eye movements. We conclude that the increase in mature myofibers correlates with eye muscle sparing in utrophin directly participates in rescue of EOM in otherwise dystrophic mice. Second, adding the deletion of dystrophinopathy. Although utrophin may be localized to the utrophin to the dystrophin deficiency of mdx mice introduces neuromuscular junction by means of an agrin-responsive a dramatic reversal in this protected phenotype, leading to promoter element (Dennis et al., 1996; Gramolini et al., 1997; severe myopathic changes in all EOMs. We propose that the Meier et al., 1997), synaptic activity patterns apparently are not increase in utrophin by an as yet unidentified, but endogenous, the key factor in upregulation of utrophin (Jasmin et al., 1995). means is mechanistically responsible for the sparing of EOM 1810 J. D. Porter and others in dystrophinopathy. Perhaps more importantly, this apparent Cellular and Molecular Biology of Muscle Development (ed. L. H. Kedes rescue of EOMs in dystrophic mice helps to validate the and F. E. Stockdale), pp. 12-26. Alan R. Liss, Inc., New York. hypothesis that the manipulation of utrophin expression, Jasmin, B. J., Alameddine, H., Lunde, J. A., Stetzkowski-Marden, F., Collin, H., Tinsley, J. M., Davies, K. E., Tomé, F. M. S., Parry, D. J. and through either therapy or pharmaceutical agents directed Cartaud, J. (1995). Expression of utrophin and its mRNA in denervated at the utrophin promoter (Dennis at al., 1996), may be used to mdx mouse muscle. FEBS Lett. 374, 393-398. ameliorate the consequences of Duchenne muscular dystrophy Kaminski, H. J., Al-Hakim, M., Leigh, R. J., Katiriji, M. B. and Ruff, R. in other skeletal muscles. L. (1992). Extraocular muscles are spared in advanced Duchenne dystrophy. Ann. Neurol. 32, 586-588. We thank Mary Gail Engle, Paraskevi Karathanasis, William J. Karpati, G. and Carpenter, S. (1986). Small-caliber skeletal muscle fibers do not suffer deleterious consequences of dystrophic . Am. Miller and Allyson C. Potter for technical assistance and Dr Francisco J. Med. Genet. 25, 653-658. Andrade for comments on the manuscript. These studies were Karpati, G., Carpenter, S., Morris, G. E., Davies, K. E., Guerin, C. and generously supported by grants from the National Eye Institute (J. D. Holland, P. (1993). Localization and quantitation of the - P.), the Muscular Dystrophy Association of America (J. D. P. and K. encoded dystrophin-related protein in normal and pathological human E. D.), Research to Prevent Blindness (J. D. P.), the Medical Research muscle. J. Neuropathol. Exp. Neurol. 52, 119-128. Council (K. E. D.), the Muscular Dystrophy Association of Great Karpati, G., Carpenter, S. and Prescott, S. (1988). Small-caliber skeletal Britain and Northern Ireland (K. E. D.) and the Burroughs Welcome muscle fibers do not suffer necrosis in mdx mouse dystrophy. Muscle Nerve Foundation (J. A. R.). 11, 795-803. Khurana, T. S., Hoffman, E. P. and Kunkel, L. M. (1990). Identification of a chromosome 6-encoded dystrophin-related protein. J. Biol. Chem. 265, 16717-16720. REFERENCES Khurana, T. S., Prendergast, P. A., Alameddine, H., Tomé, F. M. S., Fardeau, M., Arahata, K., Sugita, H. and Kunkel, L. M. (1995). Absence Blake, D. J., Tinsley, J. M. and Davies, K. E. (1996). Utrophin: a structural of extraocular muscle pathology in Duchenne muscular dystrophy: role for and functional comparison to dystrophin. Pathol. 6, 37-47. calcium homeostasis in extraocular muscle sparing. J. Exp. Med. 182, 467- Brueckner, J. K., Itkis, O. and Porter, J. D. (1996). Spatial and temporal 475. patterns of myosin heavy chain expression in developing rat extraocular Khurana, T. S., Watkins, S. C., Chafey, P., Chelly, J., Tomé, F. M. S., muscle. J. Mus. Res. Cell Motil. 17, 297-312. Fardeau, M., Kaplan, J.-C. and Kunkel, L. M. (1991). Brueckner, J. K. and Porter, J. D. (1998). Visual system maldevelopment Immunolocalization and developmental expression of dystrophin related disrupts extraocular muscle-specific myosin expression. J. Appl. Physiol. (in protein in skeletal muscle. Neuromuscul. Disord. 1, 185-194. press). Law, D. J., Allen, D. L. and Tidball, J. G. (1994). , and DRP Campbell, K. P. (1995). Three muscular dystrophies: loss of cytoskeleton- (utrophin) concentrations are increased at mdx myotendinous junctions extracellular matrix linkage. Cell 80, 675-679. following onset of necrosis. J. Cell Sci. 107, 1477-1483. Clerk, A., Morris, G. E., Dubowitz, V., Davies, K. E. and Sewry, C. A. Love, D. R., Hill, D. F., Dickson, G., Spurr, N. K., Byth, B. C., Marsden, (1993). Dystrophin-related protein, utrophin, in normal and dystrophic R. F., Walsh, F. S., Edwards, Y. H. and Davies, K. E. (1989). An human fetal skeletal muscle. Histochem. J. 25, 554-561. autosomal transcript in skeletal muscle with to dystrophin. Nature Deconinck, A. E., Potter, A. C., Tinsley, J. M., Wood, S. J., Vater, R., 339, 55-58. Young, C., Metzinger, L., Vincent, A., Slater, C. R. and Davies, K. E. Matsumura, K. and Campbell, K. P. (1994). Dystrophin-glycoprotein (1997a). Postsynaptic abnormalities at the neuromuscular junctions of complex: its role in the molecular pathogenesis of muscular dystrophies. utrophin-deficient mice. J. Cell Biol. 136, 883-894. Muscle Nerve 17, 2-15. Deconinck, A. E., Rafael, J. A., Skinner, J. A., Brown, S. C., Potter, A. C., Matsumura, K., Ervasti, J. M., Ohlendieck, K., Kahl, S. D. and Campbell, Metzinger, L., Watt, D. J., Dickson, J. G., Tinsley, J. M. and Davies, K. K. P. (1992). Association of dystrophin-related protein with dystrophin E. (1997b). Utrophin-dystrophin-deficient mice as a model for Duchenne associated proteins in mdx mouse muscle. Nature 360, 588-591. muscular dystrophy. Cell 90, 717-727. Meier, T., Hauser, D. M., Chiquet, M., Landmann, L., Ruegg, M. A. and Deconinck, N., Tinsley J., de Backer, F., Fisher, R., Kahn, D., Phelps, S., Brenner, H. R. (1997). Neural agrin induces ectopic postsynaptic Davies, K. and Gillis, J.-M. (1997c). Expression of truncated utrophin leads specializations in innervated muscle fibers. J. Neurosci. 17, 6534-6544. to major functional improvements in dystrophin-deficient muscles of mice. Minetti, C., Ricci, E. and Bonilla, E. (1991). Progressive depletion of fast Nature Med. 3, 1216-1221. alpha--positive muscle fibers in Duchenne muscular dystrophy. Dennis, D. L., Tinsley, J. M., Deconinck, A. E. and Davies, K. E. (1996). Neurology 41, 1977-1981. Molecular and functional analysis of the utrophin promoter. Nuc. Acids Res. Ohlendieck, K., Ervasti, J. M., Matsumura, K., Kahl, S. D., Leveille, C. J. 24, 1646-1652. and Campbell, K. P. (1991). Dystrophin-related protein is localized to Grady, R. M., Merlie, J. P. and Sanes, J. R. (1997a). Subtle neuromuscular neuromuscular junctions of adult skeletal muscle. Neuron 7, 499-508. defects in utrophin-deficient mice. J. Cell Biol. 136, 871-882. Pachter, B. R., Davidowitz, J. and Breinin, G. M. (1976). Morphological Grady, R. M., Teng, H., Nichol, M. C., Cunningham, J. C., Wilkinson, R. fiber types of retractor bulbi muscle in mouse and rat. Invest. Ophthalmol. S. and Sanes, J. R. (1997b). Skeletal and cardiac myopathies in mice 15, 654-657. lacking utrophin and dystrophin, a model for Duchenne muscular dystrophy. Pastoret, C. and Sebille, A. (1995). Mdx mice show progressive weakness Cell 90, 729-738. and muscle deterioration with age. J. Neurol. Sci. 129, 97-105. Gramolini, A. O., Dennis, C. L., Tinsley, J. M., Robertson, G. S., Cartaud, Pons, F., Robert, A., Marini, J. F. and Léger, J. J. (1994). Does utrophin J., Davies, K. E. and Jasmin, B. J. (1997). Local transcriptional control of expression in muscles of mdx mice during postnatal development utrophin expression at the neuromuscular synapse. J. Biol. Chem. 272, 8117- functionally compensate for dystrophin deficiency? J. Neurol. Sci. 122, 162- 8120. 170. Guo, W.-X. A., Nichol, M. and Merlie, J. P. (1996). Cloning and expression Porter, J. D. (1998). Extraocular muscle sparing in muscular dystrophy: a of full length mouse utrophin: the differential association of utrophin and critical evaluation of potential protective mechanisms. Neuromuscul. dystrophin with AChR clusters. FEBS Lett. 398, 259-264. Disord. (in press). Helliwell, T. R., Man, N. T., Morris, G. E. and Davies, K. E. (1992). The Porter, J. D. and Baker, R. S. (1992). Prenatal morphogenesis of primate dystrophin-related protein, utrophin, is expressed on the sarcolemma of extraocular muscle: neuromuscular junction formation and fiber type regenerating human skeletal muscle fibres in dystrophies and inflammatory differentiation. Invest. Ophthalmol. Vis. Sci. 33, 657-670. myopathies. Neuromuscul. Disord. 2, 177-184. Porter, J. D. and Baker, R. S. (1996). Muscles of a different ‘color’: the Hodges, B. L., Hayashi, Y. K., Nonaka, I., Wang, W., Arahata, K. and unusual properties of the extraocular muscles may predispose or protect Kaufman, S. J. (1997). Altered expression of the α7β1 integrin in human them in neurogenic and myogenic disease. Neurology 46, 30-37. and murine muscular dystrophies. J. Cell Sci. 110, 2873-2881. Porter, J. D., Baker, R. S., Ragusa, R. J. and Brueckner, J. K. (1995). Hoh, J. F. Y., Hughes, S., Hugh, G. and Pozgaj, I. (1989). Three hierarchies Extraocular muscles: basic and clinical aspects of structure and function. in skeletal muscle fibre classification: allotype, isotype and phenotype. In Surv. Ophthalmol. 39, 451-484. Extraocular muscle and muscular dystrophy 1811

Porter, J. D., Burns, L. A. and May, P. J. (1989). Morphological substrate Tanaka, J., Ishiguro, I., Efuchi, C., Saito, K. and Ozawa, E. (1991). for eyelid movements: innervation and structure of primate levator palpebrae Expression of a dystrophin-related protein associated with the skeletal superioris and orbicularis oculi muscles. J. Comp. Neurol. 287, 64-81. membrane. Histochemistry 96, 1-5. Porter, J. D., Karathanasis, P., Bonner, P. H. and Brueckner, J. K. (1997). Taylor, J., Muntoni, F., Dubowitz, V. and Sewry, C. A. (1997). The abnormal The oculomotor periphery: the clinician’s focus is no longer a basic science expression of utrophin in Duchenne and Becker muscular dystrophy is age stepchild. Curr. Opin. Neurobiol. 7, 880-887. related. Neuropathol. Appl. Neurobiol. 23, 399-405. Ragusa, R. J., Chow, C. K., St. Clair, D. K. and Porter, J. D. (1996). Tinsley, J. M., Blake, D. J., Roche, A., Fairbrother, U., Riss, J., Byth, B. Extraocular, limb and diaphragm muscle group-specific antioxidant enzyme C., Knight, A. E., Kendrick-Jones, J., Suthers, G.K., Love, D. R., activity patterns in control and mdx mice. J. Neurol. Sci. 139, 180-186. Edwards, Y. H. and Davies, K. E. (1992). Primary structure of dystrophin- Schofield, J. N., Górecki, D. C., Blake, D. J., Davies, K. and Edwards, Y. related protein. Nature 360, 591-593. H. (1995). Dystroglycan mRNA expression during normal and mdx mouse Tinsley, J. M. and Davies, K. E. (1993). Utrophin: a potential replacement embryogenesis: a comparison with utrophin and the apo-. Dev. for dystrophin? Neuromuscul. Disord. 3, 537-539. Dynam. 204, 178-185. Tinsley, J. M., Potter, A. C., Phelps, S. R., Fisher, R., Trickett, J. I. and Spencer, R. F. and Porter, J. D. (1988). Structural organization of the Davies, K. E. (1996). Amelioration of the dystrophic phenotype of mdx extraocular muscles. In Reviews in Oculomotor Research, Vol. 2, mice using a truncated utrophin transgene. Nature 384, 349-353. Neuroanatomy of the Oculomotor System (ed. J. A. Büttner-Ennever), pp. Vachon, P. H., Xu, H., Liu, L., Loechel, F., Hayashi, Y., Arahata, K., Red, 33-79, Elsevier Science Publishers, New York. J. C., Wewer, U. M. and Engvall, E. (1997). Integrins (α7β1) in muscle Stedman, H. H., Sweeney, H. L., Shrager, J. B., Maguire, H. C., Panettieri, function and survival: disrupted expression in merosin-deficient congenital R. A., Petrof, B., Narusawa, M., Leferovich, J. M., Sladky, J. T. and muscular dystrophy. J. Clin. Invest. 100, 1870-1881. Kelly, A. M. (1991). The mdx diaphragm reproduces the degenerative Webster, C., Silberstein, L., Hays, A. P. and Blau, H. M. (1988). Fast muscle changes of Duchenne muscular dystrophy. Nature 352, 536-539. fibers are preferentially affected in Duchenne muscular dystrophy. Cell 52, Straub, V., Rafael, J. A., Chamberlain, J. S. and Campbell, K. P. (1997). 503-513. Animal models for muscular dystrophy show different patterns of Wieczorek, D. F., Periasamy, M., Butler-Browne, G. S., Whalen, R. G. and sarcolemmal disruption. J. Cell Biol. 139, 375-385. Nadal-Ginard, B. (1985). Co-expression of multiple myosin heavy chain Takemitsu, M., Ishiura, S., Koga, R., Kamakura, K., Arahata, K., Nonaka, , in addition to a tissue-specific one, in extraocular muscle. J. Cell Biol. I. and Sugita, H. (1991). Dystrophin-related protein in the fetal and 101, 618-629. denervated skeletal muscles of normal and mdx mice. Biochem. Biophys. Winder, S. J., Gibson, T. J. and Kendrick-Jones, J. (1995). Dystrophin and Res. Comm. 180, 1179-1186. utrophin: the missing links! FEBS Lett. 369, 27-33.