Muscles of a different 'color'

Much of our knowledge of the responsiveness of skeletal muscle to neurogenic or myogenic disease is based on information obtained from a limb muscle prototype. Disease states interact with well-characterized skeletal muscle fiber types to produce patterned pathologic changes in muscle biopsies that are reliably used in diagnosis. By contrast, some muscle groups exhibit fundamental departures in both structure and function from the limb muscle prototype. Muscle fiber type differences that have been so clearly defined by a concept as simple as "color" (i.e., red, intermediate, and white) suddenly are not so easily resolved. The atypical fiber types might more appropriately be described as gray in color to denote a rather confusing array of characteristics in these unique muscles. The classically defined neurogenic and myogenic muscle signs either may be accentuated or may not always be present and thus may not be of the same diagnostic value for such muscles. Hoh et al ^[1] coined the term allotype to identify such fundamentally distinct muscle classes. To date, the identified skeletal muscle allotypes include limb/diaphragm, masticatory, and extraocular muscle. Other muscles that have not been fully characterized (e.g., laryngeal muscles and muscles of the middle ear) also may prove to be distinct allotypes. Many of the allotype-specific properties have their origin in muscle precursor cell lineage differences. Although epigenetic factors (e.g., innervation, hormonal influences, local cues) may influence the fiber type characteristics within an allotype, extrinsic regulatory events cannot account for differences in phenotypic options that are available across allotypes. Because disease may differentially involve one allotype while sparing one or more of the others, it is important to understand both the structural and functional differences that exist across allotypes as well as the mechanisms that may protect or predispose a particular allotype to disease. In this review, we focus on distinctions that exist between the extraocular muscle and limb/diaphragm allotypes (Table 1) and speculate as to how such differences may be responsible for either the involvement or sparing of the eye muscle allotype in different diseases (Table 2).

Extraocular muscle does not fit the traditional skeletal muscle fiber type classification schemes.

The visible functional characteristics of skeletal muscle-contraction speed and fatigue resistance-are determined by cellular and molecular properties at the single myofiber level and by the collective properties of all myofibers that comprise a single muscle. The structural characteristics that determine contractile properties are not independently regulated, but, for example, a particular muscle fiber may exhibit inter-related properties that maximize contraction speed but minimize fatigue resistance. Patterned variations in the characteristics of individual skeletal muscle fibers form the basis for muscle fiber type classification schemes. The major classifications ^[2-4] agree on the existence of three to four fiber types in typical skeletal muscle: (1) slow-twitch, fatigue resistant (I, SO, S); (2) fast-twitch, fatigue resistant (IIA, FOG, FR); (3) fast-twitch, fatigable (IIB, FG, FF); and (4) fast-twitch, intermediate (IIC, F (int.)).

The general applicability of the traditional classification schemes breaks down when applied to allotypes other than the classic limb muscle pattern. Extraocular muscle fibers exhibit considerable variation in size and, overall, are considerably smaller in diameter than those of the limb. These features have led to early misinterpretations of their involvement in disease states. Moreover, some extraocular muscle fiber types exhibit either high mitochondrial content or unusual innervation/contractile patterns (see below). The high mitochondrial content of extraocular muscle is associated with elevated oxidative enzyme activity, a highly developed microvascular bed, ^[5] and high blood flow ^[6]; one or more of these characteristics may explain their extraordinarily high fatigue resistance ^[7] and the increased contrast in this muscle group in gadolinium-enhanced magnetic resonance imaging. ^[8] When studies focused on only a few traits (e.g., myofibrillar ATPase), limb muscle fiber type terminology was erroneously applied to the extraocular muscles. If the overall fiber profiles are examined, none of the extraocular muscle fiber types fit the traditional limb/diaphragm muscle classification schemes. To deal with fiber terminology, a classification scheme has been devised for the extraocular allotype that identifies six distinct fiber types that are differentially distributed across orbital and global muscle layers. This classification is based on the features of location within muscle, color, and innervation pattern. The classification includes orbital singly innervated, orbital multiply innervated, global red singly innervated, global intermediate singly innervated, global pale singly innervated, and global multiply innervated fiber types (see Spencer and Porter, ^[5] Spencer and McNeer, ^[9] and Porter et al ^[10]).

Innervation patterns of extraocular muscles set them apart from all other mammalian skeletal muscles.

Maturation of the innervation pattern of individual muscle fibers includes a well-characterized transitory state seen during the perinatal development of all skeletal muscles (reviewed by Hall and Sanes ^[11]). Multiple axons compete for space at a single synaptic site at the midbelly of the muscle fiber, with preservation of the contacts from only one motor neuron per muscle fiber as adult activity patterns are established. Activation at the single synaptic site produces a propagated action potential, leading to an all-or-none twitch in all fiber types of the limb/diaphragm and masticatory allotypes. Although [approximately]80% of fibers in adult extraocular muscles (four of the six fiber types) exhibit this pattern of single innervation and twitch contraction, the remainder of the fibers exhibit an unusual innervation pattern with multiple neuromuscular junctions distributed along the length of single fibers. The global multiply innervated fiber type uses a nontwitch, or tonic, mode of contraction and is activated focally at each synaptic site in the absence of propagated action potentials. ^[12] By contrast, the orbital multiply innervated fiber type has a twitch mode of contraction at midbelly and nontwitch mode at its proximal and distal ends. ^[13] Some avian limb muscle fibers also have the unusual combination of both multiple innervation and a twitch contraction mode, but these fibers do not exhibit the complex longitudinal variations seen in the related extraocular muscle fiber type.

Ocular motor motor units share similarities with and differences from the spinal motor neuron-limb muscle motor unit pattern.

The motor unit is the basic functional subdivision for all skeletal muscles (see Burke ^[3] for review), including the extraocular muscles. Typical human motor units range in size from 10 to 2000 muscle fibers per motor neuron. Ocular motor motor units are among the smallest seen in any skeletal muscle ([approximately]10 muscle fibers/motor neuron), consistent with their capacity to augment (or decrement) force in small increments. Ocular motor motor units composed of singly innervated muscle fibers have exceedingly fast contraction times and high fusion frequencies, but low tetanic force development, when compared with spinal motor units. ^[14-19] In particular, the observation that force development per unit area is about half that of limb muscle ^[15] suggests that basic differences exist in the excitation-contraction coupling apparatus in the extraocular muscle allotype. The global multiply innervated fibers may receive polyneuronal innervation and exhibit slow contractions and low force levels. Although the function that this motor unit type serves in eye movements is unclear, the characterizations that have been made on the basis of single unit studies may be misleading because coactivation of multiple motor neurons might be required to maximally activate single units composed of this fiber type.

The properties of all muscle fibers of a given motor unit are the same, supporting the notion that motor neuron discharge properties play an important role in regulating muscle fiber phenotype. It is then likely that ocular motor motor neuron discharge rates that are an order of magnitude higher than those of spinal cord motor neurons ^[20] contribute toward the differences in the limb and extraocular muscle allotypes. Limb motor neurons use both supranuclear inputs and proprioceptive feedback mechanisms to augment (decrement) force levels by two distinct mechanisms: recruitment of motor neurons according to a size principal and frequency modulation of already active units. At spinal levels, motor neuron soma size-based recruitment appears to be the major means of force modulation. The smallest motor neurons, with the lowest force levels, are recruited first in a process that leads to smooth incrementation of force and serves the task of load compensation well. By contrast, 70% of ocular motor motor neurons are already at threshold at primary position of gaze, with a substantial number active at all gaze positions. ^[20] The importance of soma size in the recruitment of ocular motor motor units is unclear. Frequency modulation is the major means of executing eye movements, ^[19] as motor neuron discharge and the resultant force levels are exquisitely tuned so that the correlation between tonic motor neuron firing rates and eye position is high. Ocular motor motor neuron recruitment occurs via the convergence of supranuclear inputs with no need for external load compensation. Paradoxically, the identified feedback mechanisms that might modulate ocular motor motor neuron discharge, vision and proprioception, are indirect and slow when compared with the neuromuscular spindle afferents of limb muscles.

Developmental factors contribute toward differences in muscle phenotype.

Differences in the three skeletal muscle allotypes likely reflect, at least in part, the properties of the myoblast lineages that arise from different sites in the embryo. Although the origin of limb and body wall musculature from the somites and associated paraxial mesoderm is well documented, craniofacial muscles arise from rostral condensations of mesoderm known as somitomeres ^[21,22] that may be regulated by segment-specific genes. For example, the engrailed-2 (En-2) transcription factor is preferentially expressed in masticatory myoblasts at a time before the expression of any of the more ubiquitous myogenic regulatory genes (e.g., MyoD, myogenin). ^[23] Masticatory myoblasts also have been shown to retain the expression of at least some allotype-specific features, even when transplanted into a limb muscle bed where they encounter limb muscle innervation and local factors. ^[24] We suggest that En-2 may regulate transcription of masticatory allotype-specific genes, leading to the unique identity of this pool of muscle precursor cells, and that an alternative transcription factor functions to commit myoblasts specifically to the extraocular muscle allotype. Cell-cell communication also must be considered as a potential factor in allotype specification because the connective tissues, which are intimately associated with developing extraocular musculature, are derived from neural crest, whereas those associated with limb muscles are from mesoderm. ^[21,22]

A major developmentally regulated protein that serves as an important determinant of both fiber type and allotype is myosin. Myosin heavy chain expression is controlled by a multigene family and the myosin isoform defines fiber traits by specifying contractile velocity. Normally, there is sequential expression of embryonic, neonatal, and adult myosin heavy chain genes during myogenesis, resulting in overlap of multiple isoforms during development but ending in a one-to-one correspondence of adult myosin isoform to fiber type (i.e., type I fibers exhibit slow myosin, type IIA fibers have IIA myosin, and so on). ^[25] The extraocular muscles are an exception to this pattern in at least three respects. First, the temporal regulation of myosin expression is altered in both of the orbital layer muscle fiber types as they continue to express the embryonic myosin isoform in the adult. ^[13] Sarcomeres composed of embryonic myosin in series with others with fast myosin might serve as a damping mechanism (i.e., to reduce net force produced) to increase resolution for precise eye movement control. ^[13] Second, individual orbital layer fibers display spatial differences in myosin expression. That is, longitudinal variations in the ultrastructure and physiology of single orbital layer fibers of both types are in register with longitudinal variations in myosin isoform expression. ^[13] For example, the orbital singly innervated fiber expresses fast myosin along its entire length in addition to the expression of embryonic myosin in the proximal and distal segments of the same fibers. Third, the masticatory and extraocular allotypes are further defined by their expression of one or more myosin heavy chain genes that are not seen in other skeletal muscles. ^[1,26,27] Finally, although extraocular muscles express the same troponin variants as in other skeletal muscles, the major troponin forms in this muscle group are minor forms in other skeletal muscles. ^[28] Of interest is that the major troponin isoform in extraocular muscle, TnT_3f, has a more graded response to calcium. ^[29] The low ratio of twitch to tetanic tension production in eye muscle versus other skeletal muscle may be related to the dynamics of the interaction of calcium with TnT_3f.

Developmental differences in the motor neuron pools also may contribute to the diversity of muscle allotypes and, in turn, to their differential involvement in disease states. Segment-specific genes, such as the homeobox (Hox) family, regulate the development of more or less restricted rostral-caudal regions of the neuraxis. The positional identity conferred by Hox genes not only specifies segment-specific properties but the converse is also true in that Hox gene mutations disrupt development in specific segments of the nervous system. In congenital disorders such as Duane retraction syndrome, a motor neuron population may fail to develop even though the target muscle primordium is in place. Although the basis for the loss of abducens motor neurons in Duane syndrome has not been established, transgenic mouse models may provide insight into this disorder. The disruption of a segment-specific transcription factor, Wnt-1, leads to dysgenesis of the midbrain early in development, ^[30] with loss of oculomotor and trochlear motor neurons and a concomitant aberrant innervation of extraocular muscles by the intact abducens motor neurons (arguably a "reverse" Duane syndrome; Porter JD, Baker RS, unpublished data). The motor neuron pool-specific factors that are regulated by homeobox genes have scarcely been studied. By modulating traits such as neurotropin receptors or voltage-gated channels, these genes may produce differences in the motor neurons that innervate particular muscle groups. Such motor neuron type-specific properties may, in turn, contribute to their differential sensitivity in neuromuscular diseases, such as amyotrophic lateral sclerosis, that typically involve eye movements late, if at all.

The developing motor neuron and muscle fiber are dependent on one another for survival and maturation. Muscle- and motor neuron-specific genes may place restrictions on the particular tissue interactions that can act as positive or negative growth regulators for the motor unit components. In the ocular motor system, this interaction between motor neuron and muscle primordium is highly specific. Coculture of either the correct ocular motor motor neurons or the incorrect spinal motor neurons with explants of neonatal extraocular muscle shows that this muscle group is dependent on innervation by the correct motor neurons for survival, much less for the maturation of allotype-specific characteristics. ^[31] Although the extraocular muscle explants initially developed basic muscle characteristics, spinal motor neurons did not provide adequate trophic support and muscle fibers failed to survive beyond 2 weeks in culture if not provided with innervation by ocular motor motor neurons. Whether the specificity of this tissue interaction is based on activity patterns, trophic factors, or both is unclear at this time.

Proprioceptive mechanisms play a unique role in extraocular muscle.

Skeletal muscle uses specialized stretch receptors, neuromuscular spindles and Golgi tendon organs, to provide monosynaptic feedback of muscle length and tension information in the regulation of motor neuron discharge rate. By contrast, the extraocular muscles have no requirements for external load compensation and thus exhibit dramatic departures from this scheme. Neuromuscular spindles and Golgi tendon organs do not appear to be the major sensory receptors in extraocular muscles nor are such encapsulated receptors even present in many mammalian species. The palisade ending ^[32] appears to be the predominant sensory end organ and is specifically associated with an extrafusal muscle fiber, the global multiply innervated fiber type. This association with a nontwitch muscle fiber type suggests that the information that arises from this receptor does not mediate rapid adjustments in muscle force but rather that eye muscle proprioception may contribute to long-term recalibration of the motor system. Consistent with this hypothesis, extraocular muscle afferent information is conveyed by the trigeminal nerve and targets principal sensory nuclei rather than the ocular motor nuclei. ^[33,34] Afferent signals have been recorded in a wide variety of brainstem and cortical regions, with suggested functions ranging from the modulation of visual development to adaptive gain control (see Porter et al ^[10] for review). Recent data ^[35] suggest that proprioception may provide an orbital eye position signal for use in long-term regulation of ocular alignment and eye movement conjugacy. Deafferentation (e.g., for facial pain) or disease that specifically targets sensory pathways (e.g., herpes zoster) then may have consequences for eye muscle function that are much different from those previously described for limb muscle.

Extraocular muscle does not exhibit the massive denervation atrophy that characterizes other skeletal muscles.

In contrast to most skeletal muscles, denervation of extraocular muscles either by axotomy or pharmacologic neurotransmission blockers is accompanied by a relatively mild response. Axotomy does not produce the severe atrophy and fiber loss nor the muscle fiber type grouping that typically accompanies reinnervation. After transection of the oculomotor nerve, muscle fiber size changes were limited to mild atrophy or hypertrophy with minimal disruption of myofiber architecture. ^[36,37] Reduction of the mitochondrial content of the orbital singly innervated fiber type, and an ensuing reduction in fatigue resistance of the eye muscles, represented the major structural-functional consequences of axotomy. The lack of fiber type grouping in extraocular muscle, when it is so characteristic of denervated limb muscle, suggests either that eye muscle fibers are preferentially reinnervated by motor neurons with appropriate discharge properties, thereby precluding type respecification; adult extraocular muscle fibers are reinnervated by motor neurons at random but are resistant to neural activity-based alterations in their fiber type defining characteristics; or, once specified, the extraocular muscle fiber types may persist, whether or not they are reinnervated.

Likewise, paralysis of the extraocular muscles with botulinum toxin type A produced fiber type-specific changes that were consistent with mild disuse atrophy rather than with loss of function. ^[38] The orbital singly innervated fiber type again was targeted by botulinum denervation, as its characteristic mitochondrial clusters were disrupted by intramuscular toxin injections. In addition, there was a reduction in the microvascular network associated with this fiber type. Neither of these changes were completely reversed after postbotulinum recovery of function. Toxin denervation of the orbicularis oculi, which has fiber types resembling the limb muscle allotype, caused a much more severe atrophic response. ^[39] Denervation atrophy, however, appeared to be completely reversible in this muscle group.

In summary, the response of extraocular muscles to denervation can be best described as a mild disuse response. In general, when limb muscles are denervated, it is the more active muscles that are severely affected. Although the extraocular muscles are highly active and thus would have been predicted to be very sensitive to loss of innervation, they are spared for reasons that are not apparent.

An additional consequence of the different innervation patterns of muscle allotypes is their differential response to neuromuscular blocking agents. Although succinylcholine functions as a depolarizing blocker in typical mammalian skeletal muscle fiber types, it selectively activates the extraocular muscle multiply innervated fibers. Likewise, the organic calcium blocker, diltiazem, produces a reduction in tension development in extraocular muscle multiply innervated muscle fibers in the same way that it alters tension development in cardiac and smooth, but not skeletal, muscle. ^[40] These authors interpreted this as evidence for the hypothesis that the multiply innervated fiber type is dependent on extracellular calcium for excitation/contraction coupling.

Myasthenia gravis may preferentially involve the extraocular muscles because of their acetylcholine receptor isoforms.

Investigators attributed the common, and sometimes exclusive, involvement of the extraocular muscles in myasthenia gravis to differences in the acetylcholine receptor types that are expressed in extraocular versus typical skeletal muscles. ^[41-45] The two identified acetylcholine receptor isoforms are developmentally regulated and are comprised of alpha-, beta-, and delta-subunits plus either gamma- or epsilon-subunits. Rather than downregulating the gamma- (embryonic) subunit in favor of the epsilon- (adult) subunit, as in other skeletal muscles, adult extraocular muscle exhibits expression of both acetylcholine receptor isoforms. ^[45] Each isoform can be expressed at neuromuscular junctions of singly innervated and multiply innervated fiber types. In theory, the eye muscles might be selectively compromised when the embryonic isoform serves as the antigen for antibody formation. The problem with this interpretation is that ptosis is frequent in ocular myasthenia, but the levator palpebrae superioris muscle does not express the embryonic acetylcholine receptor isoform. ^[46]

Extraocular muscle is spared the degeneration/regeneration cycle that affects every other skeletal muscle in Duchenne muscular dystrophy.

Despite the widespread and progressive degeneration of cardiac, smooth, and skeletal muscle tissue, the extraocular muscles are spared in Duchenne muscular dystrophy. Ocular motility examinations of late-stage Duchenne patients indicated that saccade metrics were normal and that nystagmus and significant eye movement abnormalities were absent ^[47] (however, a recent case report ^[48] showed ocular motor involvement in Becker muscular dystrophy). Although the protein that has been linked to the pathogenesis of Duchenne dystrophy, dystrophin, also is absent from the extraocular muscles, morphologic studies in appropriate animal models reported an absence of pathology in this muscle group. ^[49] Our studies in the mdx mouse model have extended these findings by demonstrating that neither significant fiber turnover nor fiber type-specific loss occurred in the extraocular muscles despite the dystrophin gene defect. ^[50] Consequently, in extraocular muscle, either normal physiologic properties or adaptive compensatory mechanisms are capable of overcoming the cascade of pathogenic events that leads to fiber disruption in virtually every other skeletal muscle. Extraocular muscle might either be able to manage the massive calcium influx that accompanies the dystrophin defect, thereby preventing the calpain-driven proteolytic events, or to have more highly developed free radical scavenging systems to mitigate the protein and lipid oxidation that is part of Duchenne muscle pathogenesis. In recent studies, we explored the hypothesis that the high metabolic activity of the extraocular muscles requires that they be more efficient in removing free radicals and that these normal operating properties then serve to protect them from the pathology of Duchenne muscular dystrophy. ^[50] Spectrophotometric assay of Cu/Zn superoxide dismutase in extraocular muscle indicated that the specific activity of this enzyme is approximately twice that of the limb muscles. Thus, morphologic evidence of degeneration and regeneration in the mdx mouse correlates with the capacity of free radical scavenging systems. Beyond muscular dystrophy, the finding of highly developed systems to control free radical damage also may explain the resistance of extraocular muscle to denervation and to other neuromuscular diseases.

Muscle allotypes and motor neuron diversity may provide a framework for differential involvement of skeletal muscles in a variety of other neuromuscular disorders.

Ocular motility disorders that occur in isolation or in conjunction with those of other skeletal muscles may have their basis in allotype-dependent properties. However, thorough analysis of the extraocular muscle consequences of many neuromuscular diseases frequently is lacking. Most primary muscle diseases are not restricted to the extraocular muscles. Congenital fibrosis ^[51] and the inherited mitochondrial cytopathy seen in chronic progressive external ophthalmoplegia ^[52] may affect the extraocular muscles either exclusively or early in the disease course. The mitochondrially inherited diseases may preferentially target eye muscle, perhaps due to the high dependence of this muscle group on oxidative energy metabolism (cf., Chang et al ^[53]). Acquired mitochondrial disease also may frequently involve ocular motility because of the potential for free radical-mediated enzyme and mitochondrial DNA damage in muscles with high metabolic activity. ^[54] Consistent with this view, cytochrome c oxidase exhibits an aging-dependent defect density that is between five and six times higher in eye muscle than in limb, diaphragm, or heart. ^[55] Although secondary muscle diseases, such as polymyositis, may have extensive extraocular involvement, the cellular response appears very similar to that in other skeletal muscles. Among neuropathies, amyotrophic lateral sclerosis, poliomyelitis, and spinal muscular atrophy spare craniofacial muscle (except in their more severe forms) for reasons that are not clear. Differential expression of voltage-gated ion channels or neurotropin receptor types represents but one means of disease targeting of particular motor neuron pools.

Local anesthetic myotoxicity is mitigated in the extraocular muscles.

The aminoacyl class of local anesthetics is widely used in ophthalmology. However, the myotoxic activity of the aminoacyl anesthetics is of such significance that they have been used as a model system for studies of muscle degeneration and regeneration. ^[56] The pathogenic events subsequent to local anesthetic injections are thought to be mediated by sarcolemmal disruption and displacement of calcium from intracellular stores. Supraphysiologic intracellular free calcium levels then trigger calcium activated proteases, resulting in myofilament breakdown and fiber necrosis.

Although the aminoacyl anesthetics occasionally produce ptosis and diplopia when used in ophthalmic procedures, given their demonstrated myotoxicity, it is somewhat surprising that more patients have not experienced complications of ocular motility. Instead, retrobulbar applications of bupivacaine in a monkey model caused a severe response only in the global pale singly innervated fiber type. ^[57] Because mitochondria can serve as a calcium sink, perhaps this one fiber type is involved and all other extraocular muscle fiber types are spared because of their mitochondrial content. Dose-related limits on this protection may explain the finding that direct injections into individual extraocular muscle do produce more significant pathology. As yet unidentified mechanisms must operate to give a significant degree of protection to primate extraocular muscles such that they do not exhibit the pathology seen in other skeletal muscles. Because local anesthetic toxicity invokes calcium-mediated cell damage, the relative sparing seen with these agents may involve similar mechanisms to those responsible for the lack of response in extraocular muscle of Duchenne muscular dystrophy patients.

Thyroid disease may preferentially involve the extraocular muscles.

The involvement of extraocular muscle, and sparing of other skeletal muscles, in Graves ophthalmopathy does not appear to be a muscle allotype issue per se but rather can largely be ascribed to a disease process that specifically targets orbital fibroblasts. The extraocular muscles enlarge because of an abnormal accumulation of glycosaminoglycans in the connective tissues, thereby compressing all orbital tissues. The pathogenic mechanism underlying Graves ophthalmopathy most likely involves circulating T cells, directed against an antigen on thyroid follicular cells, recognizing this same antigen on orbital fibroblasts. Interferon gamma, interleukin 1 alpha, and tumor necrosis factor are present in the orbital connective tissue of patients with ophthalmopathy and act preferentially on orbital fibroblasts to stimulate their proliferation as well as glycosaminoglycan production. ^[58] The preferential involvement of orbital fibroblasts in Graves ophthalmopathy may reflect their unusual neural crest origin. Observations that the sarcomeric organization of the muscle fibers remains intact are consistent with the hypothesis that this disease principally targets fibroblasts. ^[59] Other reports suggest that extraocular muscle fibers proper may be targeted by cell-mediated cytotoxicity. ^[60,61]

Summary.

The rules that govern many aspects of skeletal muscle structure and function are very different for the extraocular muscle allotype. The myoblast lineages present in the extraocular muscle primordia are permissive for generation of an unusually wide range of fiber types. The balance that is struck between genetic specification and activity dependent factors in shaping fiber phenotype to suit the demands of complex visuomotor systems is not yet well defined. Because skeletal muscle has high energy demands, diversity in fiber types is needed to maximize efficiency; greater diversity in fiber composition then indicates a more diverse functional repertoire. Together, the characteristics of small motor unit size, precise dependence of muscle force upon motor neuron discharge rate, high contractile speed but low tension development, and contractile protein heterogeneity contribute toward the high precision and diversity that is required for eye movements. Finally, the structural and functional characteristics and plasticity of the individual extraocular muscle fiber types play an important role in determining their response to disease or manipulation. The lack of uniform responses across the muscle allotypes in disease, or in response to pharmaceutical or surgical interventions, requires that we obtain a better understanding of the fundamental differences that exist between muscle groups.

Acknowledgments

The technical assistance of Olga Itkis and Mary Gail Engle is much appreciated. Discussions with and data from laboratory members Jennifer Brueckner and Robert Ragusa yielded valuable insights that are incorporated into this manuscript.