Neuroanatomy of the oculomotor system by jean buttner
List of ContributorsN.H. Barmack, Neurological Sciences Institute, Oregon Health and Science University, 505 NW 185th Avenue, Beaverton, OR 97006, USAR.H.I. Blanks, Florida Atlantic University, Charles E. Schmidt College of Science, 777 Glades Road, P.O. Box 3091, Boca Raton, FL 33431-0991, USAR. Blumer, Institute of Anatomy, University of Vienna, Waehringerstrase 13, A-1090 Vienna, AustriaU. Buttner, Department of Neurology, Ludwig-Maxmilian University Munich, Klinikum Grosshadern, ¨ Marchioninistr. 15, D-81377 Munich, GermanyJ.A. Buttner-Ennever, Institute of Anatomy, Ludwig-Maximilian University of Munich, Petten Koferstr. ¨ 11, D-80336 Munich, GermanyP.D.R. Gamlin, Department of Vision Sciences, University of Alabama at Birmingham, 924 South 18th Street, Birmingham, AL 35294-4390, USAR.A. Giolli, Department of Anatomy and Neurobiology, University of California, College of Medicine, Irvine, CA 92697-1275, USAJ.K. Harting, Department of Anatomy, University of Wisconsin Medical School, 1300 University Avenue, Madison, WI 53706, USAY. Hata, Department of Systems Neurophysiology, Graduate School of Medicine, Tokyo Medical and Dental University, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8519, JapanS.M. Highstein, Washington University School of Medicine, Department of Otolaryngology, Box 8115, 4566 Scott Avenue, St. Louis, MO 63110, USAG.R. Holstein, Department of Neurology and Cell Biology, Mount Sinai School of Medicine, Box 1140, One Gustave Levy Place, New York, NY 10029, USAA.K.E. Horn, Institute of Anatomy, Ludwig-Maximilian University of Munich, Pettenkoferstrasse 11, 80336 Munich, GermanyY. Izawa, Department of Systems Neurophysiology, Graduate School of Medicine, Tokyo Medical and Dental University, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8519, JapanK.Z. Konakci, Center of Anatomy and Cell Biology, Integrative Morphology Group, Medical University Vienna, Waehringerstrasse 13, A-1090 Vienna, AustriaF. Lui, Dipartimento di Scienze Biomediche, Sezione di Fisiologia, Universita di Modena e Reggio Emilia, Via Campi 287, 41100 Modena, ItalyJ.C. Lynch, Department of Anatomy, Ophthalmology and Neurology, University of Mississippi Medical Center, 2500 N. State Street, Jackson, MS 39216, USAP.J. May, Department of Anatomy, Ophthalmology and Neurology, University of Mississippi Medical Center, 2500 North State Street, Jackson, MS 39216, USAR.A. McCrea, Department of Neurobiology, Pharmacology and Physiology, University of Chicago, Abbott 09/MC 0926, 947 E. 58th Street, Chicago, IL 60637, USAM. Mock, Department of Anatomy, Visual Sensorimotor Section, Neurological Clinic, University Hospital ¨ Tubingen, D-72076 Tubingen, GermanyR.M. Muri, Perception and Eye Movement Laboratory, Departments of Neurology and Clinical Research, ¨ University of Bern, Inselspital, CH-310 Bern, Switzerland v
viJ.D. Porter, National Institutes of Neurological Disorders and Stroke, 6001 Executive Blvd, NINDS/NSC 2142, Bethesda MD 20892, USAY. Shinoda, Department of Systems Neurophysiology, Graduate School of Medicine, Tokyo Medical and Dental University, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8519, JapanyR.F. Spencer, Departments of Anatomy and Otolaryngology, Medical College of Virginia, Richmond, VA 23298, USA (deceased 2001)Y. Sugiuchi, Department of Systems Neurophysiology, Graduate School of Medicine, Tokyo Medical and Dental University, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8519, JapanP. Thier, Department of Cognitive Neurology, Hertie-Institute of Clinical Brain Research, University of Tubingen, Hoppe-Seyler 3, 72076 Tubingen, GermanyJ.-R. Tian, Jules, Stein Eye Institute, 3-310 DSERC, UCLA Medical Center, 100 Stein Plaza, Los Angeles, CA 90095-7002, USAB.V. Updyke, Department of Anatomy, University of Wisconsin Medical School, 1300 University Avenue, Madison, WI 53706, USAJ. Voogd, Erasmus Medical Center Rotterdam, Department of Neuroscience, Box 1738, 3000 DR Rotterdam, The Netherlands
Preface This book is dedicated to Bernard Cohen who discovered that the paramedian pontine reticular formation (PPRF) was essential for the generation of saccadic eye movements 40 years ago. His work and his unfailing enthusiasm continue to inspire the ﬁeld of oculomotor and vestibular research.The updated and extended version of ‘Neuroanatomy of the Oculomotor System’ is a set of reviews whichfocus on the functional neuroanatomy and connectivity of the brain areas involved in controlling eyemovements. The ﬁrst edition of ‘Neuroanatomy of the Oculomotor System’ was published as volume 2 of‘Reviews in Oculomotor Research’. This series outlived its commercial life and has been discontinued. Butwe are delighted to be able to continue the spirit of these reviews in a volume of ‘Progress in BrainResearch’. We chose to publish this updated and extended version as part of this series because it ﬁts wellwith the character of ‘Progress in Brain Research’, and because this series is available in most universitylibraries. The ﬁrst chapter is written as an introduction to the oculomotor system: it discusses the differenttypes of eye movements, the structures involved in their generation and some clinical aspects; it deals withsaccades, the vestibulo-ocular reﬂex, optokinetic responses, vergence, smooth pursuit and gaze-holding.Chapter 1 also introduces current concepts such as ‘pulleys’ in the orbit (i.e. the functional consequences ofthe Tenon’s capsule), and integrators for gaze-holding. Each of the various topics is picked up in a laterchapter and the neuroanatomy dealt with in more detail. The subsequent chapters are arranged in a‘bottom –up’ approach; they review the structure and control of eye muscles in the periphery, the nextchapters are on the oculomotor nuclei in the brainstem, then the reticular formation, the vestibular nucleiand cerebellum. The following chapters move on to more rostral structures, the tectum, the pretectum,basal ganglia, thalamus and cerebral cortex. Many new networks inﬂuencing eye movements have been discovered, and many new hypotheses havebeen proposed, over the 17 years separating the two editions of this book; and as a consequence six newchapters have been added to the original version. The most provocative of these is Chapter 3, which is areview of eye muscle proprioceptors and their relationship to the control of eye movements. Here we havemade an attempt to integrate the slightly unpopular ﬁeld of ‘extraocular proprioception’ into the currentconcepts of the oculomotor system, although the evidence for these hypotheses is incomplete. Perhaps it istoo early to come to conclusions on the role of extraocular proprioception, but we have tried to show thatthe established facts can be re-interpreted in the light of recent discoveries in ﬁelds such as neuraldevelopment, genetics and neurotrophins etc., which reveal the factors inﬂuencing the development ofmuscle spindles, Golgi tendon organs and their neural circuitry. The ﬁeld of proprioception is fraught withcontroversy, and this is reﬂected in Chapter 3 by the differing views of the authors on the function of aneural structure unique to eye muscles – the palisade ending. One camp supports the view that they aremotor, the others provide evidence for their sensory nature. Nevertheless, we are convinced that ourdifferences will be resolved in the future by collaborating with each other; and hence the combinedauthorship of Chapter 3. The other ﬁve new chapters in this updated and extended version are devoted to the following topics: theinferior olive (Chapter 9), which shows how the olivary climbing ﬁbers impose a topography onto the vii
viiicerebellum: the pontine nuclei and nucleus reticularis tegmenti pontis (Chapter10), which likewisedetermine the organization of cerebellar afferents but of the mossy ﬁber type: the accessory optic nuclei(Chapter 13), which provide optokinetic signals to the brainstem, but whose clinical relevance is completelyunknown up to now: and the basal ganglia (Chapter 14), where functional oculomotor networks can nowbe followed within the circuitry of the forebrain. Finally, a new review of functional magnetic resonanceimaging (fMRI) studies of oculomotor-related structures has also been introduced (Chapter 16). The elevenoriginal chapters have been re-written and updated. In almost all cases they have completely altered theircharacter, depending on whether or not a new scientist, or group of scientists, have taken on the authorship:this holds for the chapters on the oculomotor nuclei (Chapter 4), reticular formation (Chapter 5), thevestibular nuclei (Chapter 6), prepositus hypoglossi (Chapter 7), cerebellum (Chapter 8), the superiorcolliculus (Chapter 11), pretectum (Chapter 12), cerebral cortex (Chapter 15), and spinal cord (Chapter 17).In this respect, the old edition is by no means replaced by the new updated version: the chapters of the oldedition will remain useful in their own right because the new authors review different aspects of thestructure. The old Chapter 2 is a masterly review of the properties of eye muscles: Bob Spencer told me in1987 that he was slow writing it because he had to do a lot of new experiments in order to write it properly.It has now been thoroughly updated but the authorship of the new Chapter 2 was left in its originalconstellation in respect to Robert F. Spencer (1950 – 2001), a great scientist. The idea of this new and extended version was initiated several years ago by Volker Henn (1943–1997),who we still sorely miss. Its production has only been possible with the enormous patience and hardwork ofthe authors, each of which were chosen for their scientiﬁc expertise. I have been very fortunate to have hadthe support of Maureen Twaig at Elsevier, as well as the continual encouragement and assistance fromAhmed Messoudi and Rita Buttner in Munich: I am very grateful to all of them. ¨ Jean A. Buttner-Ennever ¨ Munich, April 2005
x12. The pretectum: connections and oculomotor-related roles P.D.R. Gamlin (Birmingham, AL, USA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37913. The accessory optic system: basic organization with an update on connectivity, neurochemistry and function R.A. Giolli, R.H.I. Blanks and F. Lui (Irvine, CA and Boca Raton, FL, USA and Modena, Italy) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40714. Oculomotor-related pathways of the basal ganglia J.K. Harting and B.V. Updyke (Madison, WI, USA) . . . . . . . . . . . . . . . . . . . . . . . . 44115. Cortico-cortical networks and cortico-subcortical loops for the higher control of eye movements J.C. Lynch and J.-R. Tian (Jackson, MS, USA and Los Angeles, CA, USA) . . . . . . . 46116. MRI and fMRI analysis of oculomotor function R.M. Muri (Bern, Switzerland) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ¨ 50317. Long descending motor tract axons and their control of neck and axial muscles Y. Shinoda, Y. Sugiuchi, Y. Izawa and Y. Hata (Tokyo, Japan). . . . . . . . . . . . . . . . 527Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 565
Progress in Brain Research, Vol. 151ISSN 0079-6123Copyright r 2006 Elsevier B.V. All rights reserved CHAPTER 2 Biological organization of the extraocular muscles Robert F. Spencer1 and John D. Porter2,Ã 1 Departments of Anatomy and Otolaryngology, Medical College of Virginia, Richmond, VA 23298, USA2 Departments of Neurology and Neurosciences, Case Western Reserve University and University Hospitals of Cleveland, Cleveland, OH 44106, USAAbstract: Extraocular muscle is fundamentally distinct from other skeletal muscles. Here, we review thebiological organization of the extraocular muscles with the intent of understanding this novel muscle groupin the context of oculomotor system function. The specific objectives of this review are threefold. The ﬁrstobjective is to understand the anatomic arrangement of the extraocular muscles and their compartmental orlayered organization in the context of a new concept of orbital mechanics, the active pulley hypothesis. Thesecond objective is to present an integrated view of the morphologic, cellular, and molecular differencesbetween extraocular and the more traditional skeletal muscles. The third objective is to relate recent datafrom functional and molecular biology studies to the established extraocular muscle ﬁber types.Developmental mechanisms that may be responsible for the divergence of the eye muscles from a skeletalmuscle prototype also are considered. Taken together, a multidisciplinary understanding of extraocularmuscle biology in health and disease provides insights into oculomotor system function and malfunction.Moreover, because the eye muscles are selectively involved or spared in a variety of neuromuscular diseases,knowledge of their biology may improve current pathogenic models of and treatments for devastatingsystemic diseases.1. Introduction eye movements. The complexity and precision of eye movements is reﬂected not only in theThe extraocular muscles (EOMs) are the effector organization of the central oculomotor systemsorgan for voluntary and reﬂexive movements of described elsewhere in this volume, but also in thethe eyes. Because the area of high acuity vision, the very biology of the EOMs. Since skeletal muscle isfovea, subtends a very small angle of visual space, a highly plastic tissue, readily adapting to usagethe task of gaze control must be accomplished with patterns, one can hypothesize that properties ofhigh precision through the coordinated activity of the novel EOM phenotype exist to meet a complexthe six EOMs. EOM innervation is by motoneu- ‘‘job description’’ of stabilizing and reorienting eyerons in the oculomotor, trochlear, and abducens position for clear vision. The biological organiza-nuclei, which represent the ﬁnal common pathway tion of EOM is then a consequence of the structureupon which signals from a variety of supranuclear and function of oculomotor systems and, in turn,areas converge to produce ﬁve distinct classes of careful analysis of EOM properties can provide fundamental insights into the status of these neuralÃNational Institutes of Neurological Disorders and Stroke, control systems in health and disease. The layered or compartmentalized organization of the EOMs,6001 Executive Blvd, NINDS/NSC 2142, Bethesda MD 20892.Tel.: +1 301 496 1917; Fax: +1 301 402 1501; into distinctive orbital and global layers, also hasE-mail: firstname.lastname@example.org important connotations for the function of thisDOI: 10.1016/S0079-6123(05)51002-1 43
44novel muscle group. In this review, we seek an was necessary to interpret later cell and molecularunderstanding of EOM biology in the context of studies and to subsequently develop an overalloculomotor system function. model of EOM myoﬁber function. In spite of recent Knowledge of skeletal muscle biology does not progress, we only now are beginning to appreciatemean that one understands EOM. As we show the full breadth of adaptations of EOM myoﬁberhere, many of the ‘‘rules’’ that govern skeletal types to their novel role in eye movement control.muscle biology do not apply to EOM. The EOMs A second objective of this review is to integrateof some ﬁsh exhibit the most remarkable examples new, multidisciplinary data with establishedof plasticity found in any skeletal muscle. EOM morphologic proﬁles to begin to construct anprecursor cells take alternative developmental paths overall model of the biology of the diverse EOMto form a weakly electric organ, for navigation, in ﬁber types.the stargazer (Astroscopus sp.) or a heater organ, Although EOM compartmentalization into thethat keeps the eye and brain warm during deep two distinctive, orbital and global, layers is a long-dives, in billﬁsh (Scombroidei order) (Bennett and recognized and highly conserved feature, itsPappas, 1983; Block and Franzini-Armstrong, functional significance has only recently become1988; Block, 1991). While the EOMs are among clear. Discovery of the EOM pulleys, and thethe fastest muscles in mammals, they also possess unique relationship of the muscle layers to theslow, non-twitch muscle ﬁbers that are character- pulleys and globe, has created a new concept foristic of phylogenetically older avian and amphibian the division of labor in EOM (Demer et al., 2000).muscles and other traits more typically associated A third objective of this review is to relate an-with cardiac muscle or embryonic skeletal muscle. atomic and molecular properties of the two EOMIt is perhaps because of this paradoxical complex- layers to this new hypothesis of orbital function.ity in their structural organization that a funda- Here, we address these objectives by buildingmental enigma remains in regard to EOM function upon the anatomical framework established in ourin normal eye movements and ocular motility dis- prior reviews of EOM (Spencer and Porter, 1988;orders. One objective of this review is to convey an Porter et al., 1995, 1997; Porter and Baker, 1996;integrated view of the morphologic, cellular, and Porter, 2002). Knowledge of the compartmentalmolecular divergence of EOM from prototypical and myoﬁber type organization of mammalianskeletal muscle. EOM and their relationships to oculomotor system The multinucleate muscle ﬁber or myoﬁber is the development, function, and dysfunction representsautonomous structural and functional unit of skel- an essential framework for future studies.etal muscle, but all myoﬁbers are not created equal(Ranvier, 1874). Muscle-to-muscle variability infunction has been ascribed to the relative percent- EOM and orbital gross anatomyage composition of four highly conserved muscleﬁber types (types I, IIA, IIX, and IIB) (Brooke and The EOMs exhibit remarkable variation in number,Kaiser, 1970; Burke et al., 1971; Peter et al., 1972; arrangement (origin and insertion), and innervationSchiafﬁno et al., 1989). By contrast, the myoﬁbers throughout phylogeny. From an early prototype ofcomprising EOM are singularly unique because four EOMs, a pattern of six ‘‘primitive’’ EOMs hasthey do not respect any of the traditional skeletal emerged by an evolutionary process of differentiationmuscle ﬁber type classiﬁcation schemes. Several or degeneration. The presence of these six muscles,reviews have described the anatomic organization the four recti (superior, inferior, medial, andof EOM ﬁber types (Peachey, 1971; Mayr, 1978; lateral) and two obliques (superior and inferior),Asmussen, 1979; Chiarandini and Davidowitz, is rather constant across the vertebrate classes1979; Spencer and Porter, 1988; Ruff et al., 1989; from cyclostomes to avians, despite variations inPorter and Hauser, 1993a; Porter et al., 1995, 1997; arrangement and innervation (Isomura, 1981). ThePorter and Baker, 1996). The emergence of a last principal EOM, the levator palpebrae superi-consensus EOM ﬁber type classiﬁcation scheme oris, did not make its appearance in phylogeny
45until mammals. These seven EOMs are relatively orbits point outward at approximately 231. Thisconsistent across mammalian species in their gen- relationship is important to understanding of theeral location and innervation pattern, although actions of the EOMs because the origin of theindividual muscle actions show interspecies rectus muscles is at the orbital apex and they insertvariation, particularly apparent in frontal-eyed in a spiral around the ocular limbus in such a(e.g., cat, monkey) versus lateral-eyed (e.g., mouse, fashion that the superior and the inferior rectirat, rabbit) animals. These variations are coincident form an angle of 231 with the anterior-posteriorwith species differences in the forward extension of visual axis in the straight-ahead position. Thethe maxillary process (Fink, 1953) and the relative gross anatomy and general functions of the sixangles of the visual axis and the semicircular canals EOMs are reviewed here.(Simpson and Graf, 1981; Ezure and Graf, 1984). The eye sits within the bony orbit surrounded by Medial rectus and lateral rectus musclesthe EOMs, connective tissue, and orbital fat. Thepositions of the six rectus and oblique EOMs in cat The four rectus muscles have a tenon-and-mortise-and monkey are shown in Fig. 1. Inﬂections in the like origin from a tendinous ring (annulus of Zinn)muscle paths due to orbital connective tissue which surrounds the optic foramen and a portionorganization impact EOM actions in primary of the superior orbital ﬁssure (Sevel, 1986). Theand secondary gaze positions (see section ‘‘EOM medial rectus muscle exhibits a single head ofpulleys’’). Although the reference visual axes are origin from both the tendinous ring and the duraparallel and directed straight ahead, the bony that surrounds the optic nerve, and lies medial toFig. 1. Drawings illustrating superior and lateral views of monkey and cat orbits, with positions of EOMs and accessory EOMs.Superior oblique tendon and select other muscles are cut away for clarity of drawings. Note similar localization and insertion ofmonkey ALR and cat RBsl. Arrangement of EOMs and accessory EOMs in mouse and rat orbits is similar to that of cat, but theEOMs are surrounded by the Hardarian gland in rodent. ALR, accessory lateral rectus; IO, inferior oblique; IR, inferior rectus; LR,lateral rectus, MR, medial rectus, RB, retractor bulbi (sl, il, sm, and im denote superior lateral, inferior lateral, superior medial, andinferior medial slips, respectively); SO, superior oblique; SR, superior rectus. Drawings by Alex Meredith, PhD.
46the globe as it courses forward to insert just pos- the oculomotor nerve respectively, enter the globalterior to the corneoscleral junction. The lateral re- surface of the proximal portion of each muscle.ctus, arising from the tendinous ring as two distinct Although elevation and depression of the globe areslips, passes lateral to the globe to insert into the the primary actions of the vertical recti, theirsclera via a long, broad tendinous expansion. The origins lie medial to the globe such that they exhibitinnervation of the medial rectus is provided by secondary roles in the horizontal and torsionalthe inferior division of the oculomotor nerve, while planes. With the eyes in the primary position, thethe abducens nerve innervates the lateral rectus primate vertical recti intersect the globe at an anglemuscle. The nerves to both muscles enter proxi- of 231 lateral to the visual axis. Thus the superiormally on their global surfaces. Since the insertions rectus has a secondary role in adduction andof these muscles are symmetrically distributed intorsion, while the inferior rectus assists in ad-around the horizontal meridian on opposite sides duction and extortion. Though the primary actionsof the globe, the medial and lateral recti are of the vertical recti are the same in lateral-eyedfunctional antagonists that serve as the principal mammals, the angle of force of these muscles shiftsadductor and abductor of the eye, respectively. No 201 medial to the visual axis thereby altering theirsecondary actions of these muscles are expressed secondary actions. As a result, in lateral-eyedduring movements initiated from the primary mammals the superior rectus secondarily extortsposition. Slight vertical and torsional components and abducts, while the inferior rectus has a supple-induced in extreme positions of gaze are attribut- mental role in intorsion and abduction. A modiﬁedable to the actions of the other rectus and oblique distal segment of the superior rectus, the lateralmuscles. The arrangements and actions of the arm, is evident in rabbit and is primarily com-medial and lateral recti are basically identical in prised of orbital layer multiply innervated musclelateral- and frontal-eyed mammals. Expansions of ﬁber types (Briggs et al., 1988) (see sectiontheir tendons of insertion also have attachments ‘‘Detailed organization of EOM ﬁber types’’). Ato lacrimal and zygomatic bones (i.e., check similar structure may exist in humans (Kono et al.,ligaments of Lockwood), which were thought to 2002). Unlike the horizontal recti, the superior andserve to restrict extreme movements in the horizontal inferior recti do not have the enthesis connectionsplane. More recently, these ‘‘check ligaments’’ to the orbital walls (Kono et al., 2002).have been incorporated into an overall scheme oforbital connective tissue organization (see section‘‘EOM pulleys’’) and have been termed entheses Superior oblique muscle(Kono et al., 2002). The muscle planes of these andall other EOMs are ﬁxed within the orbit in all The superior oblique muscle, like the four recti,gaze positions (e.g., the horizontal recti do not arises from the tendinous annulus at the apex of thesideslip during up and down gaze) (Miller and orbit. A small fascicle of muscle ﬁbers on the medialRobins, 1987; Miller, 1989). surface of the muscle originates from the medial bony wall of the orbit. Coursing forward from an origin which lies dorsomedial to that of the medialSuperior rectus and inferior rectus muscles rectus, the superior oblique passes through a ﬁb- rocartilaginous ring, the trochlea, and turns laterallyLike the horizontal recti, the vertical (superior and to insert on the superior aspect of the globe. Theinferior) recti originate from the tendinous ring at insertion of this muscle falls posterolateral to thethe apex of the orbit and course forward to insert central point of the globe in frontal-eyed mammals,anterior to the equator of the globe. The superior but anterolateral in lateral-eyed mammals. Therectus, like the medial rectus, muscle has an addi- trochlear nerve, upon entering the superior orbitaltional origin from the dura of the optic nerve ﬁssure, courses medially to enter the superior(Sevel, 1986). The nerves to the superior and portion of the orbital surface of the muscle. Frominferior recti, the superior and inferior divisions of the primary position, the predominant action of
47this muscle in both lateral- and frontal-eyed onto both the skin of the upper eyelid and theanimals is intorsion. Differences in the point of superior tarsal plate. Since a scleral insertion is ab-insertion of the superior oblique in the primate sent, this muscle exerts no direct inﬂuence upon theversus the rabbit lead to clear differences in its globe, although an indirect inﬂuence is mediated bysecondary actions. The primate superior oblique a partial blending of the levator aponeurosis withsecondarily depresses and abducts the globe, while the tendon of the superior rectus muscle. Innerva-that of the rabbit secondarily elevates and adducts. tion is provided by a branch of the superior divisionAn anomalous muscle, the gracillimus orbitis (of of the oculomotor nerve that passes to the proximalBochdalek) or comes obliqui superioris (of Albin), portion of the muscle either through or lateral towhen present, originates from the proximal dorsal the superior rectus. The levator palpebrae functionssurface of the superior oblique muscle, inserts on in elevation of the upper eyelid. Occasionally,the trochlea and/or its surrounding connective though perhaps not infrequently, two anomaloustissue, and is innervated by a branch of the fourth muscles are associated with the levator palpebrae,nerve (Whitnall, 1921). and, like the latter, are innervated by branches from the superior division of the 3rd nerve (Whitnall, 1921; Isomura, 1977; Sacks, 1985). The tensorInferior oblique muscle trochleae (of Budge) arises from the medial border of the levator muscle and inserts onto the trochleaIn contrast to the origin of the other principal of the superior oblique muscle and/or its surround-EOMs from the annulus of Zinn, the inferior ing connective tissue. A muscle of similar name andoblique muscle arises from the maxillary bone in insertion, but originating from the ventral rim of thethe medial wall of the orbit. The origin of the optic foramen in proximity to the origin of the su-inferior oblique muscle furthermore may display perior rectus muscle and innervated by the fourthconsiderable variation in its anatomical relation- nerve, has been described in the rabbit (Murphyship to the nasolacrimal canal (Whitnall, 1921). et al., 1986). The transversus orbitus attachesThe muscle passes ventral to the tendon of the between the medial and lateral walls of the orbitinferior rectus and inserts on the lateral aspect of connecting with the levator muscle en route.the globe medial to the tendon of the lateral rectus.The insertion of the inferior oblique, like that ofthe superior oblique, is posterior to the equator in Accessory EOMsthe primate and anterior to the equator in therabbit, thereby resulting in the same primary ac- In addition to the seven principal EOMs, manytion, extorsion, but different secondary actions. vertebrates possess accessory EOMs. AccessoryThe inferior oblique of lateral-eyed animals second- EOMs in cat and monkey are shown in Fig. 1. Inarily depresses and adducts, while that of frontal- most species, the accessory muscle takes the formeyed animals elevates and abducts. Innervation is of the retractor bulbi (Hopkins, 1916; Cords, 1924;provided by a branch of the inferior division of the Isomura, 1981). The retractor bulbi is correlatedoculomotor nerve that enters the muscle near its with the presence of a nictitating membrane, andposterior border. these structures are synergistic in reﬂex retraction of the globe in response to corneal stimulation.Levator palpebrae superioris muscle The retractor bulbi ﬁrst appears in phylogeny as a continuous sheath that surrounds, or two slipsThe levator palpebrae superioris has a narrow lying dorsal and ventral to, the optic nerve inorigin from the orbital surface of the lesser wing of amphibians. In both amphibians and reptiles, thethe sphenoid bone, just above the optic foramen retractor bulbi is paired with the membranae nic-and the origin of the superior rectus. In its distal titans muscle. In avians, these two muscles are re-course, this muscle crosses the superior aspect of the placed by quadratus membranae nictitans andglobe and fans out to insert via broad aponeuroses pyramidalis muscles. The mammalian retractor
48bulbi muscle variably has two (mouse), three which movements of pulley (by the orbital layer)(dog), or four (rat, rabbit, cat) slips. Innervation and globe (by the global layer) are coordinated butof the retractor bulbi exhibits species-specific not necessarily coincident (Demer et al., 2000). Thepatterns from branches of the oculomotor and/or active pulley system uses orbital layer motor unitsabducens nerves (Spencer and Sterling, 1977; to alter pulley positions and thereby adjust EOMGrant et al., 1979, 1981; Crandall et al., 1981; vector forces in different gaze positions, greatlyMeredith et al., 1981; Evinger et al., 1987). With simplifying the task of central oculomotor controlthe regression of the nictitating membrane to a systems by making commands independent of in-vestigial plica semilunaris in primates, the retrac- itial eye position (Clark et al., 2000). Any role thattor bulbi is reduced to a single homologous slip in the smooth muscle tissues, and their specialized in-the monkey, the accessory lateral rectus muscle, nervation (Demer et al., 1997), might play in pulleywhich is innervated by the abducens nerve (Spencer positional control is poorly understood at this time.and Porter, 1981; Schnyder, 1984). An accessorylateral rectus muscle may render monkeys resistantto esotropia (Boothe et al., 1990) and has been re- The functional context of the EOMsported in humans only in one case of congenitaloculomotor nerve palsy (Park and Oh, 2003). An understanding of the novel biology of the EOMs is incomplete without an appreciation for the demands of ocular motility (for a thoroughEOM pulleys review, see other chapters of this volume and Leigh and Zee, 1999). The reﬂexive oculomotor controlThe recent discovery of EOM pulleys, and their systems that stabilize images on the retina, therebyinterrelationship with the compartmentalized struc- preventing blur during head/body movement, areture of the EOMs (see section ‘‘Compartmental the phylogenetically oldest and form a base uponorganization of EOM’’), represents a paradigm which the other eye movement systems operate.shift in oculomotor function. Evidence that rectus Thus, the vestibulo-ocular and optokinetic reﬂexesmuscle bellies remain relatively ﬁxed in the orbit are found in all vertebrates, but visual targetingdespite surgical transposition of their insertions movements, such as saccades and smooth pursuit,provided the ﬁrst suggestion that EOM muscle appear later in phylogeny and vergence move-paths were ﬁxed relative to the orbit (Miller et al., ments are associated only with the evolution of1993). Subsequent anatomical and imaging studies frontally placed eyes and high acuity specializa-characterized ﬁbroelastic sleeves, or pulleys, repre- tions of the retina (e.g., area centralis, fovea).senting specializations in Tenon’s capsule (Fig. 2). Elaboration of the more sophisticated oculomotorEOM pulleys are located approximately at the control systems correlates with specific, patternedequator of the globe and suspended from the bony changes in EOM biology (see section ‘‘Differencesorbit by collagen/elastin/smooth muscle struts or in EOM ﬁber types in the same and differententheses (Demer et al., 1995; Porter et al., 1996; species’’). Accessory EOMs are typically restrictedClark et al., 1997; Kono et al., 2002). Adjacent to species with incomplete bony orbits, where re-muscle pulleys are intercoupled by connective ﬂex retraction is required to protect the eye.tissue bands. The pulleys provide inﬂection points While the oculomotor system is arguably thein EOM paths, thereby serving as functional or- best understood of skeletomotor control systems,igins for each muscle. Species differences in pulley it also is among the most complex. Unlike mostmorphology correlate with known differences in skeletal muscles, which often are tightly role-specific,EOM biology and visuomotor function in rat ver- individual EOMs serve very diverse functionalsus humans (Khanna and Porter, 2001). The recent repertoires and execute eye movements almostﬁnding that the two distinct EOM compartments continuously throughout waking hours. Binocularor layers have separate insertion points led to alignment and maintenance of steady ﬁxation uponformulation of the active pulley hypothesis, in targets are essential for clear vision and must be
49Fig. 2. Diagrammatic representation of orbital connective tissue relationships to the EOMs and eye, including the specializations ofTenon’s capsule, the rectus muscle pulleys. The connective tissues of the orbit are thickened to form pulleys for the four rectus musclesand inferior oblique. Interconnections between, and anterior and posterior to, the pulleys are the pulley sling. Differential distributionof orbital smooth muscle, collagen, and elastin components of pulleys and associated tissues is indicated. The three coronal views arerepresented at the levels indicated by arrows in the horizontal section. Separate insertions of orbital and global layers upon pulley andglobe, respectively, also are indicated. IO, inferior oblique; IR, inferior rectus; LPS, levator palpebral superioris; LR, lateral rectus;MR, medial rectus; SO, superior oblique; SR, superior rectus (ﬁgure courtesy of J.L. Demer and J.M. Miller; see Demer, 2000).accomplished within very ﬁne tolerances or else blur stereotyped discharge patterns (Robinson, 1970),and diplopia (double vision) result. On one hand, including: (a) tight linkage between sustainedEOM responds to polymodal sensory signals to activity and eye position, (b) rapid and large puls-produce slow, smooth changes in eye position in es of motoneuron discharge associated with sac-vestibulo-ocular, optokinetic, vergence, and pursuit cadic eye movements, and (c) an overall high levelmovements that stabilize and/or track visual targets. of motoneuron activity, exceeding that of spinalOn the other hand, in acquiring novel visual targets motoneurons by an order of magnitude. EOMthe EOMs must execute saccadic eye movements ﬁbers then must be responsive over an unprece-that can exceed 6001/s. dented dynamic range that requires adaptations for Skeletal muscle characteristics are directly inﬂu- contraction speed and fatigue resistance well be-enced by the patterned activity of the motoneurons yond that experienced by the more typical skeletalthat innervate them (Pette, 2002). Oculomotor muscles. To this end, EOM utilizes the full rangemotoneurons represent the common output of the of phenotypic options available to adult skeletalcontrol systems described above and have highly muscle plus traits strategically borrowed from
50phylogenetically primitive skeletal muscle, imma- orbital bone and an inner global layer close to theture skeletal muscle, and cardiac muscle. There optic nerve and eye. In some species, a transitionallikely is a causal relationship between the wide zone (e.g., monkey), containing an admixture ofdynamic range of oculomotor control systems and muscle ﬁber types from either layer, or a connec-the complexity and diversity of EOM. tive tissue band (e.g., rabbit) may be evident between the orbital and global layers. A thin muscle ﬁberCompartmental organization of EOM layer external to the orbital layer, termed the mar- ginal zone (Wasicky et al., 2000) or peripheralSkeletal muscles are generally heterogeneous in patch layer (Harker, 1972), has been documentedcross-sectional appearance and compartmentalized in some species. In the rectus muscles, the orbitalor layered patterns may be evident. Various func- layer is comprised of smaller diameter ﬁbers andtional advantages of compartmentalization in tra- typically is c-shaped, encompassing the globalditional skeletal muscles have been previously layer except for a gap left adjacent to the opticaddressed (English and Letbetter, 1982; Eason nerve or globe. In the oblique muscles, the orbitalet al., 2000). Likewise, the rectus and oblique EOMs layer often completely encircles the global layer.are characterized by a distinctive compartmentali- The global layer extends the full muscle length,zed organization (Kato, 1938) (Fig. 3A). Each has inserting into the sclera via a well-deﬁned tendon,an outer orbital layer adjacent to the periorbita and while the orbital layer ends before the muscle be- comes tendinous. Recent studies have shown that this early termination of the orbital layer is a con- sequence of its insertion into the muscle pulley, at approximately the equator of the globe (Demer et al., 2000) (Fig. 2). By contrast, neither the le- vator palpebrae superioris nor the accessory EOMs have an orbital layer compartment, a ﬁnd- ing that correlates with their lack of muscle pulleys. In addition to the clear differences in myoﬁber diameter, the two EOM layers are distinguished by substantial morphologic and immunocytochemical differences. First, the interrelated features of mitochondrial content, oxidative enzyme activities (e.g., SDH, NADH-TR), and microvascular net- work all are more developed in the orbital layer. Collectively, these traits correlate with the high fatigue resistance and continuous activation of the orbital layer. Second, the orbital layer expresses traits usually associated with developing skeletal muscle. While traditional skeletal myoﬁbers exhibit a developmental transition in expression of embry- onic to neonatal to adult myosin heavy chain iso-Fig. 3. Histological proﬁles of the EOM layers (A) and ﬁbertypes (B, C) in the monkey lateral rectus muscle. Note general forms, adult orbital layer myoﬁbers retain theﬁber type size differences, with the c-shaped orbital layer con- embryonic myosin heavy chain (Myh3) (Wieczorektaining smaller diameter ﬁbers. Proﬁles of the SIFs (1, 3–5) and et al., 1985; Jacoby et al., 1990; Brueckner et al.,MIFs (2, 6) in the orbital (B) and global (C) layers are indi- 1996). Neural cell adhesion molecule (NCAM), acated. Phase contrast light photomicrographs of semithin cell surface molecule normally downregulated dur-(1 mm) sections highlight differences in mitochondrial contentof different muscle ﬁber types. 1, orbital SIF; 2, orbital MIF; 3, ing myogenesis, also persists on virtually all orbital,global red SIF; 4, global intermediate SIF; 5, global white SIF; but only some global, layer ﬁbers (McLoon and6, global MIF. Wirtschafter, 1996). A similar pattern is apparent
51for the embryonic (g) acetylcholine receptor (AChR) independently regulated. Instead, myoﬁber prop-subunit, as it is present at all neuromuscular junc- erties that determine speed and fatigability aretions of orbital layer myoﬁbers, but only at those of co-expressed in specific patterns that led to thesome global layer myoﬁbers (Kaminski et al., 1996). recognition of discrete muscle ﬁber types. The Because few investigators work on the cell and major myoﬁber classiﬁcation schemes (Brookemolecular biology of EOM, observations such as and Kaiser, 1970; Peter et al., 1972; Burke et al.,the orbital layer retention of embryonic traits are 1973; Gauthier and Lowey, 1979; Schiafﬁno et al.,sparse. To more efﬁciently identify such orbital and 1989) agree on three to four ﬁber types in typicalglobal layer specializations, we used laser capture skeletal muscle: (a) slow-twitch, fatigue resistantmicrodissection to isolate EOM layer-specific sam- (red or type I), (b) fast-twitch, fatigue resistantples and then determined their gene expression sig- (intermediate or type IIA), (c) fast-twitch, inter-natures by high-throughput DNA microarray mediate (type IIX), and (d) fast-twitch, fatigableanalyses. Differential expression proﬁling identiﬁed (white or type IIB). Structural and functional181 transcripts with preferential expression in the properties of these four traditional ﬁber types areorbital or global layer, encompassing genes with a summarized in Table 1. Muscle ﬁber types havewide range of functions (see Khanna et al., 2004 and distinct functional identities (Close, 1972; Burkeaccession number GSE 907 in the National Center et al., 1974), each with a relatively narrow optimalfor Biotechnology Information (NCBI) Gene Ex- working range such that their collective actionspression Ontology (GEO) database). Among these, are required to achieve typical physiologic wholeseveral slow/cardiac muscle markers were preferen- muscle force-velocity proﬁles. These four ﬁbertially expressed in the orbital layer (TNNC1, types are found in various proportions in virtuallyMYH7, MYH6, CSRP3, TNNT2, FHL1, NRXN3, every mammalian skeletal muscle. For example,and NEBL). These data suggest that the orbital may slow fatigue-resistant muscles like soleus are prin-be functionally slower than the global layer and that cipally comprised of types I and IIA, while typeproperties of orbital layer ﬁbers alone may explain IIB ﬁbers predominate in fast, fatigable musclesand extend several prior ﬁndings of cardiac muscle- like gastrocnemius. It is well recognized that thespecific gene or protein expression in EOM. four discrete myoﬁber types may represent pheno- Overall, the orbital and global layers are very types along a continuum in variation of myoﬁberdifferent in their morphologic and gene expression traits. Nonetheless, the ﬁber type classiﬁcationproﬁles, consistent with their respective muscle pul- schemes have been an essential means of under-ley and eye movement roles. Preferential expression standing muscle function and are of considerableof the transcription factor, CSRP3 (and transcripts value in diagnosis and muscle disease modeling, asthat are regulated by CSRP3; e.g., FHL1, MYH3) several neuromuscular diseases preferentially in-(Khanna et al., 2004), by the orbital layer is a par- volve specific muscle ﬁber types.ticularly interesting ﬁnding. CSRP3 responds to Initial gene expression proﬁling studies have sug-muscle stretch by activating transcripts associated gested that differences between muscles that arewith early myogenesis (Knoll et al., 2002). Orbital comprised of predominately type I versus type IIBlayer expression of CSRP3 may mechanistically myoﬁbers are relatively modest (Campbell et al.,link the continuous activity of this layer against 2001). However, more recent data support a greaterelastic elements of muscle pulleys to the orbital degree of divergence among skeletal muscle groupslayer retention of various embryonic traits. than can be explained simply by differences in com- position of stereotypic ﬁber types (Porter et al.,Traditional skeletal muscle ﬁber types 2004). These data suggest that there might be more variability among the traditional muscle groups andMost skeletal muscles are comprised of variable myoﬁber types than is currently known.percentages of four conserved muscle ﬁber types. Despite the value of the ﬁber type concept forThe myoﬁber traits that are responsible for over 130 years (Ranvier, 1874), myoﬁbers presentcontraction speed and fatigue resistance are not in some muscle groups do not appear to respect
52Table 1. Structural and functional proﬁles of the routine skeletal muscle ﬁber typesTerminologiesaBrooke and Kaiser (1970) I IIA IIX IIBPeter et al. (1972) Slow-twitch-oxidative Fast-twitch-oxidative Fast-twitch-glycolytic glycolyticBurke et al. (1973) S FR F (int.) FFGauthier and Lowey (1979) Red (slow) oxidative Red (fast) oxidative Intermediate White glycolytic glycolyticSchiafﬁno et al. (1989) I IIA IIX IIBHistochemical proﬁlesMyosin ATPase (pH 9.4) Low High High HighMyosin ATPase (pH 4.6) High Low Intermediate IntermediateMyosin ATPase (pH 4.3) High Low Intermediate LowSDH (mitochondrial aerobic) High Intermediate–high Intermediate LowNADH-TR (aerobic) High Intermediate–high Intermediate LowLDH (anaerobic) High Intermediate–high Intermediate LowMen-a-GPD (anaerobic) Low High Intermediate HighPAS (glycogen) Low High Intermediate Intermediate–highPhosphorylase Low High High HighOil Red O (lipid) High Low Low LowAlkaline phosphatase High High Low Low(capillaries)Immunocytochemical proﬁlesMyosin heavy chain Myh7 (I/b-cardiac) Myh2 (IIA) Myh1 (IIX) Myh4 (IIB)Ultrastructural proﬁlesbZ-line Wide Wide Narrow NarrowMitochondria Many, small Many, large Moderate, small Few, smallSarcoplasmic reticulum, T- Elaborate, narrow Elaborate, narrow Moderate, small Compact, broad,tubules parallelNeuromuscular junctions Large, widely spaced, Discrete, separate, Long and ﬂat; long, deep folds small, elliptical, shallow, branching, closely sparse folds spaced foldsPhysiological proﬁlescTwitch contraction time (ms) Slow Intermediate Fast FastTwitch tension (g) Very low Low Intermediate HighRelative fatigue resistance Resistant (very) Resistant (moderate) Intermediate Sensitivea Terminology from the major skeletal muscle ﬁber type classiﬁcation schemes. bConsenus morphologic traits from multiple studies(Gauthier, 1969; Padykula and Gauthier, 1970; Schiafﬁno et al., 1970). c Consensus physiological traits derived from multiple studies(Close, 1972; Burke et al., 1973, 1974).traditional classiﬁcation schemes. The allotype Overview of EOM ﬁber typesconcept originated as a framework to accountfor the phenotypic range available to skeletal mus- Early morphologic and physiologic studies recog-cle (Hoh et al., 1988, 1989). Three allotypes were nized that myoﬁbers present in mammalian EOMdeﬁned on the basis of their potential to express were atypical. Siebeck and Kruger (1955) identiﬁedspecialized myosins: masticatory (super fast myo- two basic EOM ﬁber types, one type similar to thesin), EOM (EOM-specific myosin, designated typical twitch ﬁbers of mammalian skeletal musclesMyh13), and limb (no allotype-specific myosins), (now designated singly innervated ﬁbers or SIFs)and their appearance is dependent upon an inter- and the other similar to slow ﬁbers atypical foraction of muscle lineage with appropriate inner- mammalian skeletal muscle (now designated mul-vation patterns. The distinctive ﬁber types tiply innervated ﬁbers or MIFs). The SIFs of rectuscomprising the EOM allotype are discussed here. and oblique EOMs are invariably fast-twitch
53(among EOMs, slow-twitch ﬁbers are found only in the six ﬁber type scheme (McLoon et al., 1999;the levator palpebrae superioris; see Porter et al., Kjellgren et al., 2003a, b). Myosin heavy chain is a1989). MIFs have been found in EOM and a few key determinant of contractile properties; multi-other, highly specialized craniofacial muscles (e.g., ple myosin genes encode proteins differing intensor tympani and laryngeal muscles) (Fernand contraction speed and energetic demands suchand Hess, 1969; Mascarello et al., 1982; Veggetti that an individual skeletal muscle ﬁber typicallyet al., 1982; Han et al., 1999). We suggest that expresses the one myosin isoform that is bestMIFs, while exceptionally rare in skeletal muscle, suited for its workload. EOM is unique in itsmay be more prevalent among craniofacial muscles broad utilization of options from the myosinthan is currently appreciated. Physiologic studies heavy chain family and its frequent heterogeneityidentiﬁed two MIF types in EOM, differing on the in myosin expression within single myoﬁbers.basis of location within the orbital or global layers Specifically, EOM expresses virtually all knownand their physiological ability to propagate action striated muscle isoforms of myosin heavy chain,potentials (Hess and Pilar, 1963; Bach-y-Rita and including traditional adult skeletal (Myh1 or typeIto, 1966; Pilar and Hess, 1966; Pilar, 1967). The IIX, Myh2 or IIA, Myh4 or IIB, and Myh7 or I/b-two types of EOM MIFs resemble the multiply cardiac), developing skeletal (Myh3 or embryonicinnervated ﬁbers that are found in amphibian and Myh8 or perinatal), cardiac-specific (Myh6 or(similar to global layer MIFs) and avian (similar to a-cardiac), and a tissue-specific (Myh13 or EOM-orbital layer MIFs) skeletal muscles (Morgan and specific) isoform (Bormioli et al., 1979; WieczorekProske, 1984). Interestingly, the neuromuscular et al., 1985; Jacoby et al., 1990; Asmussen et al.,junctions associated with SIFs and MIFs appear to 1993; Rushbrook et al., 1994; Brueckner et al.,exhibit very similar molecular organization and 1996; Jung et al., 1998; Winters et al., 1998;both have only modest differences from those of McLoon et al., 1999; Pedrosa-Domellof et al., ¨other skeletal muscles (Khanna et al., 2003b). 2000; Rubinstein and Hoh, 2000; Wasicky et al., Since these early studies, EOM ﬁber typing has 2000; Briggs and Schachat, 2000, 2002; Schachatevolved such that there now is a consensus on a six and Briggs, 2002). If we are to obtain an overallﬁber type classiﬁcation scheme for mammalian understanding of the properties of EOM ﬁberEOM (for a historical review, see Spencer and types, it is essential to relate myosin expressionPorter, 1988). There are also several extensive patterns, identiﬁed by immuncytochemistry and/orreviews of this ﬁber classiﬁcation scheme (Spencer in situ hybridization, to the range of other myo-and Porter, 1988; Porter et al., 1995; Porter and ﬁber traits. Incorporation of much of the recentBaker, 1996). While any single measure (e.g., myo- myosin expression data into the existing EOMﬁbrillar ATPase) might lead one to believe that myoﬁber classiﬁcation scheme, however, is prob-EOM is comprised of traditional skeletal muscle lematic. Possible species differences in myosinﬁber types, broader morphologic/histochemical/ expression patterns, heterogeneity in the batteriesimmunocytochemical proﬁles show that the estab- of myosin antibodies used, failure to consider factorslished skeletal muscle classiﬁcation schemes simply such as the longitudinal variations in the same ﬁber,do not apply to EOM. A reasonable assumption is and the frequent failure to use a ﬁber type identifyingthat the relatively large number of EOM ﬁber types, marker (e.g., trichrome stain) in adjacent sectionssix versus the three to four of typical skeletal muscle, serve to complicate any synthesis of myoﬁber traits.reﬂects the complexity and variety of eye movements. As noted above, it can be misleading to base EOM ﬁber types have been extensively charac- ﬁber classiﬁcation schemes upon any single trait,terized in monkeys, rabbits, rats, and mice and as this may identify mere variations in the samethere is evidence that human EOMs contain sim- ﬁber types. Although four SIF types are describedilar ﬁber types (Wasicky et al., 2000). Recent in mammalian EOM, every ﬁber cannot be ﬁt tostudies, relying upon myosin heavy chain expres- the absolute criteria of a single type. The aggregatesion patterns alone, have suggested that EOM population of EOM SIFs, therefore, may form amay be more complex in ﬁber type content than continuum of fast-twitch ﬁbers that differ in
54contraction speed and fatigability (Nelson et al., isoforms that can be clearly linked to specific ﬁber1986), not unlike the situation for the three fast- types; thus, these likely are incomplete represen-twitch ﬁber types in traditional skeletal muscles. If tations of the actual expression patterns.ﬁber typing is to provide a useful tool for under-standing EOM biology, we argue that furtheradditions to the myoﬁber classiﬁcation schemes The orbital singly innervated ﬁber typemust allow for such variability and base any newtypes upon the identiﬁcation of conserved patterns Orbital SIFs (Figs. 3B and 4A, and Table 2)across a broad range of myoﬁber traits. represent the predominant ﬁber type (80%) in the The six established EOM ﬁber types are desig- orbital layer of rectus and oblique muscles. The or-nated according to their layer distribution (orbital bital SIFs contain small myoﬁbrils, surrounded byor global), innervation type (singly or multiply), abundant sarcoplasmic reticulum, and highand mitochondrial content (red, intermediate, or mitochondrial content (Fig. 5A). At mid-belly, or-white). All EOM SIFs have proﬁles consistent with bital SIF diameter is largest and the ﬁbers taperfast-twitch function, but atypical for skeletal mus- proximally and distally. Mitochondria form char-cle fast-twitch ﬁbers, contain very little glycogen. acteristically large central and subsarcolemmal clus-Fiber type traits are summarized in Table 2 and in ters. Since mitochondria comprise a rather largethe following section. Fiber type repertoires of the volume of the orbital SIFs (20% by volume), myo-levator palpebrae superioris, retractor bulbi, and ﬁbril volume is exceptionally low (60%) in compar-accessory lateral rectus muscles differ from the ison to the range seen in most skeletal musclesscheme presented here and are discussed elsewhere (70–85%) (Hoppeler and Fluck, 2002). This is con-(Alvarado et al., 1967; Pachter et al., 1976; Spencer sistent with the general EOM trait of low force de-and Porter, 1981; Gueritaud et al., 1986; Porter velopment. The histochemical proﬁle of orbital SIFs et al., 1989). It is important to note that ﬁber types suggests that they are fast-twitch and fatigue resist-of the retractor bulbi are more like those of limb ant, but also have capacity for anaerobic metabo-musculature; this may have direct disease conse- lism. Orbital SIFs contain unusually high lipidquences, since the retractor does not exhibit the content. A single neuromuscular junction is presentsparing in muscular dystrophy that is seen for EOM at approximately the middle of each ﬁber, usually(Ragusa et al., 1996; Porter and Karathanasis, encircling the ﬁber; nerve terminals are embedded in1998; Porter et al., 2001b, 2003b). deep depressions of the sarcolemma and exhibit few, irregular postsynaptic folds (Fig. 6A). Myosin expression in orbital SIFs is heterogene-Detailed organization of EOM ﬁber types ous, with expression of a unique myosin gene only seen in EOM and laryngeal muscles (Myh13) and aHere, we present a composite view of each of the developmental myosin isoform (Myh3) (Wieczoreksix recognized EOM myoﬁber types. Two of these et al., 1985; Jacoby et al., 1990; Brueckner et al.,ﬁber types localize to the orbital layer (one SIF 1996). This myosin expression pattern raises twoand one MIF) and four localize to the global layer critical issues: (a) phylogenetic analysis of Myh13(three SIFs and one MIF). A characteristic feature indicates that it diverged early from an ancestralof EOM is the overall small myoﬁber diameter myosin and has substantial structural differencesrelative to most other skeletal muscles. The EOM from other fast isoforms (Briggs and Schachat, 2000;ﬁber types are largely conserved across species; Shrager et al., 2000) and (b) retention of develop-known species differences are addressed in a sub- mental myosin isoforms in adult skeletal muscle issequent section. Morphologic descriptions are rare. Myosin isoforms are specialized to providebased upon rhesus monkey EOM, while histo- specific contractile force/velocity at a specific energychemical and immunocytochemical data are com- cost. The unique myosin expression proﬁle of orbitalpiled from a variety of species. The myosin SIFs is suggestive of a highly specialized role in eyeexpression patterns indicated here are only those movements. Lucas and Hoh (2003) have suggested
Table 2. Ultrastructural and histochemical proﬁles of extraocular muscle ﬁber types Orbital GlobalFiber type: 1 2 3 4 5 6Ultrastructural proﬁlesMyoﬁbrils Extent Small Large Small Small Small Large Size (mm) Rata 0.24 0.20–0.81 0.26 0.34 0.41 0.61 Monkeyb 0.30 0.36–0.58 0.27 0.41 0.51 0.67 Separation Well delineated Moderately delineated Well delineated Well delineated Well delineated Poorly delineated Volume fractionc (%) 60 78 55 65 71 83Sarcoplasmic recticulum Extent Moderately developed Modestly developed Moderately developed Well developed Well developed Poorly developed Location Predominantly I band I band Predominantly I band Predominantly I band I4A band I bandVolume fraction (%) Ratc 9 6 10 14 16 4 Monkeyb 7.7–17.4 7.0–16.2 9.2 18.3 19.5 4.8T-tubules Extent Well developed Poorly developed Well developed Well developed Well developed Poorly developed Location A/I junction Irregular A/I junction A/I junction A/I junction IrregularMitochondria Number Very many Few–many Many Many Few Few Extent Large Small Large Moderate Small Small Size (mm) Monkeyb 0.13–0.28 0.07–0.14 0.22 0.19 0.20 0.06 Disposition Aggregated Single Aggregated Single Single SingleVolume fraction (%) Ratc 20 6 24 13 5 5 Monkeyb 18.1–27.3 7.9–20.6 22.3 13.8 6.8 6.8Z-line Extent Intermediate Wide Intermediate Narrow Narrow WideWidth (mm) Ratc 73 118 76 54 48 100Histochemical proﬁlesd Trichrome Coarse/granular Granular/ﬁne Coarse/granular Granular Granular/ﬁne Fine Mean diameter (mm) 24.873.8 19.373.2 27.274.7 34.574.6 46.776.2 35.774.1Percentage (%) 80 20 33 25 32 10 Myosin ATPase 9.4 +++ +++ +++ +++ +++ +/–Myosin ATPase 4.6 +/- +++ +/- +/– +/– ++++ SDH +++/++++ ++ ++++ +++ ++ + NADH-TR +++ ++ ++++ +++ ++ + LDH ++/+++ ++ ++++ +++ ++ + 55 Men-a-GPD ++/+++ + ++ +++ ++++ +
56Table 2 (continued ) Orbital GlobalFiber type: 1 2 3 4 5 6 Sudan black ++/+++ + +++ ++ ++ + PAS ++/+++ +/– ++ + + +/– Phosphorylase ++/+++ + +++ + + + Oil RedO ++/+++ + +++ ++ + + Alkaline phosphatase ++++ ++ +++ ++ + + AChE Focal, encircle Multiple Focal Focal Focal MultiplePhysiological proﬁles Contraction speed Fast Twitch/tonic Fast Fast Fast Tonic Fatigue resistance High Intermediate High Intermediate Low Lowa Quantitative data from Pachter (1983) in rat superior oblique muscle. Ranges for the orbital MIF (2) indicate proximo-distal variations within single ﬁbers examined inserial sections. bQuantitative data from Pachter (1982) in monkey superior rectus muscle. Ranges for the orbital SIF (1) and MIF (2) types indicate proximo-distalvariations within single ﬁbers examined in serial sections. c Quantitative data from Mayr (1973) in rat extraocular muscle. d Histochemical data from cat EOM: SDH,succinct dehydrogenase; NADH-TR, nicotinamide adenine nucleotide dehydrogenase-tetrazolium reductase; LDH, lactic dehydrogenase; Men-a-GPD, menadione-linkeda-glycerophosphate dehydrogenase; PAS, periodic acid-Schiff; AChE, acetyl-cholinesterase. Level: +/– (very low), +(low), ++ (intermediate), +++ (high), ++++(very high).
57Fig. 4. Ultrastructural proﬁles of the SIF (A) and MIF (B) muscle ﬁbers of the orbital layer, and the red (C), intermediate (D), andwhite (E) SIFs and the MIF (F) of the global layer, of the monkey lateral rectus muscle. Muscle ﬁber types are differentiated on thebasis of the size, number and distribution of mitochondria, the size and delineation of the myoﬁbrils, and the extent of development ofthe internal membrane system (sarcoplasmic reticulum and T-tubules). c, capillary; mn, myonucleus; s, neuromuscular synaptic ending;a, preterminal axon. Scales: A, C, 10 mm; B, D–F, 5 mm.that EOM contains two distinct forms of the em- 2002; Lucas and Hoh, 2003). The fast isoform of thebryonic myosin heavy chain protein, one potentially sarcoplasmic reticulum calcium ATPase (Atp2a1)unique to EOM, that may represent alternative shows a similar pattern of longitudinal variation,splicing of Myh3 or an alternative gene. Myosin dropping out distal to neuromuscular junction sitesisoforms also show variation along the length of (Jacoby and Ko, 1993). A substantial number ofindividual ﬁbers; Myh13 is expressed only in the vi- orbital layer ﬁbers express the neonatal myosincinity of the neuromuscular junction, while Myh3 is heavy chain isoform (Myh8), although it is currentlyexpressed both proximal and distal to this site unclear which ﬁber types these are (Wieczorek et al.,(Rubinstein and Hoh, 2000; Briggs and Schachat, 1985; McLoon et al., 1999). The overall orbital SIF
58Fig. 5. Ultrastructural proﬁles of the mitochondria (m), myoﬁbrillar organization in the A-band (A) and I-band (I), and the de-lineation of the myoﬁbrils by T-tubules (t) and sarcoplasmic reticulum (sr) in the orbital SIF (A) and MIF (B), the global red (C),intermediate (D), and pale (E) SIFs, and the global MIF (F) in the monkey lateral rectus muscle. Scale: A–F, 0.5 mm.proﬁle is consistent with rapid, highly fatigue resist- surpassed only by myocardium (Wooten and Reis,ant muscle contractions. 1972; Wilcox et al., 1981). Consistent with the high mitochondrial andoxidative enzyme content, individual orbital SIFsare ringed by capillaries. The vascular supply of The orbital multiply innervated ﬁber typethe orbital layer and the high oxidative activity ofthe SIFs may account for the high blood ﬂow in Orbital MIFs (Figs. 3B and 4B, and Table 2) ac-EOM, which exceeds that of skeletal muscle and is count for the remainder of ﬁbers (20%) in the
59Fig. 6. Ultrastructural proﬁles of neuromuscular junctions associated with the orbital SIF (A) and MIF (B), the global intermediate(C) and pale (D, E) SIFs, and the global MIF (F) in the monkey lateral rectus muscle visualized by the histochemical localization ofacetylcholinesterase. s, neuromuscular synaptic ending; a, preterminal axon; Sch, Schwann cell; mn, myonucleus. Scales: A, B, F, 2 mm;C–E, 5 mm.orbital layer. Like the orbital SIF, this ﬁber type (fast) and acid (slow) myoﬁbrillar ATPase. Myo-shows considerable structural and biochemical ﬁbrils are larger than those of orbital SIFs andvariation along its length. At mid-belly, the orbit- sarcoplasmic reticulum development is moderateal MIF has traits consistent with twitch contrac- (together, suggestive of slower twitch contractions)tion, exhibiting dual staining with both alkaline (Fig. 5B). By contrast, proximal and distal to the
60ﬁber mid-section orbital MIFs exhibit slow myoﬁ- layer. The histochemical, ultrastructural (Fig. 5C),brillar ATPase and ﬁne structural characteristics of and myosin heavy chain expression proﬁle of thisslowly contracting ﬁbers (large myoﬁbrils and sparse ﬁber type is similar to that of the orbital SIF, ex-sarcoplasmic reticulum). Unlike other adult skeletal cept that it does not exhibit the longitudinal var-muscle ﬁbers, multiple nerve terminals are distrib- iations in ultrastructure and does not co-expressuted along the myoﬁber length. At mid-belly, ne- the developmental myosin isoforms. Instead, globaluromuscular junctions resemble those of the orbital red SIFs express the IIA myosin isoform (Myh2)SIFs (Fig. 6B). By contrast, proximal and distal to (Brueckner et al., 1996; Rubinstein and Hoh, 2000);its center the nerve terminals are small and rest on because of its relationship to the orbital SIF, thisthe sarcolemmal surface or in slight depressions, ﬁber type may be among a population of globalwith no postjunctional folds. ﬁbers that express Myh13 near their neuromuscular Based on enzyme histochemistry, orbital MIFs junctions (Briggs and Schachat, 2002). Like its or-exhibit only modest oxidative and weak glycolytic bital counterpart, the global red SIF has a highcapacity. Myosin heavy chain expression is consi- mitochondrial volume (420%) and very low myo-stent with this proﬁle in that mid-ﬁber regions ﬁbril volume fraction (55%), suggesting that thestain for the slow-twitch isoform (type I or Myh7) considerable fatigue resistance is achieved at theand proximal/distal regions stain for both the em- cost of force reduction. Neuromuscular junctionbryonic myosin (Myh3) and Myh7 (Rubinstein morphology is nearly identical to that of the orbitaland Hoh, 2000; Briggs and Schachat, 2002). SIF. Collectively, these observations suggest simi-Immunoreactivity for an avian slow-tonic myosin larities with the skeletal IIA ﬁber type, but the veryheavy chain also has been linked to orbital MIFs high mitochondrial content and overall histochem-(Pierobon-Bormioli et al., 1980). As noted above, ical proﬁle is very different from typical IIA ﬁbers.orbital layer ﬁbers express the neonatal myosin The global red SIF’s proﬁle suggests that it is fast-heavy chain isoform (Myh8), but it is not clear twitch and highly fatigue resistant.which ﬁber types these are (Wieczorek et al., 1985;McLoon et al., 1999). Orbital MIFs also exhibitatypical myosin light chain patterns; instead of the The global intermediate singly innervated ﬁbertraditional skeletal muscle slow isoform, they ex-press an embryonic skeletal/atrial isoform of myo- Global intermediate SIFs (Figs. 3C and 4D, andsin light chain 1 (Bicer and Reiser, 2004). Table 2) comprise approximately one-fourth of the Physiological studies suggest that orbital MIFs ﬁbers in the global layer, with rather uniformexhibit twitch capability in mid-belly and non- distribution throughout this layer. Myoﬁbrillartwitch contractions in proximal and distal ﬁber ATPase and ultrastructural characteristics indicatesegments (Jacoby et al., 1989). Collectively, the that this is a fast-twitch ﬁber type; myosin isoformheterogeneous features of this ﬁber type are unlike content is likely IIX (Myh1) (Rubinstein and Hoh,any that previously has been described for skeletal 2000). Moderate levels of oxidative enzymes andmuscle, with parallels only to intrafusal (neuro- anaerobic enzymes are apparent. Numerousmuscular spindle) ﬁbers, and it is difﬁcult to draw medium-sized mitochondria are distributed singlyconclusions regarding its function. or in small clusters. Myoﬁbrillar size and sarcoplas- mic reticulum content are intermediate between the other two types of global SIFs (Fig. 5D). Neuro-The global red singly innervated ﬁber muscular junctions include clusters of large nerve endings that are located in synaptic depres-Global red SIFs (Figs. 3C and 4C, and Table 2) sions that include regularly spaced postjunctionalrepresent about one-third of the muscle ﬁbers in the folds (Fig. 6C). Overall, this proﬁle ﬁts that of aglobal layer, predominating in the intermediate fast-twitch ﬁber with an intermediate contractionzone between orbital and global layers and declin- speed and level of fatigue resistance, probablying in frequency with progression into the orbital lying between global red and white SIFs.
61The global white singly innervated ﬁber There are numerous small superﬁcial grape-like endings distributed along the longitudinal extentGlobal white SIFs (Figs. 3C and 4E, and Table 2) of individual ﬁbers of global MIFs (Fig. 6F).comprise about one-third of the global layer. Global A novel type of sensory nerve terminal, the myo-white SIFs exhibit modest levels of oxidative en- tendinous cylinder or palisade ending, is associatedzymes, high anaerobic metabolic capacity, and a with the myotendinous junction of this ﬁber type.fast type ATPase proﬁle. This ﬁber type likely ex- Like amphibian muscles, the global MIF typepresses type IIB myosin heavy chain (Myh4) exhibits a slow graded, non-propagated response(Rubinstein and Hoh, 2000). There are few, small following either neural or pharmacologic activa-mitochondria that are singly arranged between the tion (Chiarandini and Stefani, 1979). The ﬁndingmyoﬁbrils (Fig. 5E). Neuromuscular junctions are of a phylogenetically primitive muscle ﬁber type inthe most elaborate of any of the six EOM ﬁber one of the fastest skeletal muscles is difﬁcult totypes. Multiple axon terminals are clustered togeth- reconcile, unless one considers a potential role iner in deep depressions of the sarcolemma; post- either very ﬁne foveating movements of the eye orjunctional folds are regular, numerous, and deep as part of a specialized proprioceptive apparatus(Fig. 6D, E). The overall ﬁber proﬁle is consistent (Ruskell, 1978) (cf. Chapter 3).with a fast-twitch type that is used only sporadicallybecause of low fatigue resistance. Differences in EOM ﬁber types in the same and different speciesThe global multiply innervated ﬁber Differences between the rectus and oblique musclesGlobal MIFs (Figs. 3C and 4F, and Table 2) con- in the same species appear to be largely attributablestitute the remaining 10% of ﬁbers in the global to variations in the total number of ﬁbers in eachlayer. These ﬁbers contain very few, small mito- muscle. Such muscle-to-muscle variability in myo-chondria that are arranged singly between the myo- ﬁber number, however, is primarily the result ofﬁbrils. Myoﬁbrils are very large and sarcoplasmic differences in orbital, but not global, layers (Ohreticulum development is so poor that myoﬁbril et al., 2001). These authors attributed this ﬁnding toseparation is often indistinct (Fig. 5F). The large rectus muscle sharing of a similar mechanical loadmyoﬁbrils mean that the calcium source, the sarco- on the eye-mover global layers, but rectus muscleplasmic reticulum, and the contractile ﬁlaments are dissimilarities in load on pulley-mover orbital lay-spatially far apart, resulting in very slow contrac- ers, because of pulley differences in elasticity.tions. Consistent with slow excitation–contraction EOMs of the same species can differ in relativecoupling in this ﬁber type, the fast calcium ATPase proportions of the six muscle ﬁber types (Ringelfound in all other EOM ﬁber types is absent from et al., 1978; Vita et al., 1980; Carry et al., 1982;global MIFs (Jacoby and Ko, 1993). The ultra- McLoon et al., 1999). Similar to the ﬁber countstructural proﬁle of this ﬁber resembles that of data, human EOMs show the largest same-speciesslow, tonic muscle ﬁbers in amphibians. Global variation in the orbital layer (medial recti havingMIFs express slow-twitch (type I or Myh7) the highest and lateral recti the lowest percentage of(Brueckner et al., 1996; Rubinstein and Hoh, orbital SIFs; Ringel et al., 1978). Finally, while2000), but do not appear to be immunoreactive complexity of myosin heavy chain expression mayfor avian slow-tonic myosin heavy chain (Pierobon- be conﬁned to the orbital layer in rat (i.e., globalBormioli et al., 1980). There are variable reports layer ﬁbers may express single myosin isoforms;that it expresses the a-cardiac myosin heavy chain Rubinstein and Hoh, 2000), myoﬁbers of the globalisoform (Myh6). Global MIFs express the tradi- layers of rabbit and human EOM may be moretional skeletal muscle slow isoform of myosin light heterogeneous in myosin heavy chain expressionchain 1, but not the skeletal/atrial isoform found in patterns (McLoon et al., 1999; Briggs and Schachat,orbital MIFs (Bicer and Reiser, 2004). 2002; Kjellgren et al., 2003b). Taken together,
62individual muscle variations in the proportion of species in the global muscle ﬁber types is theirdifferent muscle ﬁber types, and variability in myo- diameters (although the ratio of global MIF tosin expression patterns of single types, might ex- global SIF diameter appears to be considerablyplain differences in the rate- and tension-related higher in rodents than in higher species). Collec-contractile properties of different EOMs (Meredith tively, differences in global myoﬁber size andand Goldberg, 1986). number could account for observed differences in Differences between the same muscle in different isometric tension in the cat (Barmack et al., 1971)species are more difﬁcult to assess, since compar- and monkey (Fuchs and Luschei, 1971).isons between studies are based on the interpreta-tion of comparable ﬁber types with differentnomenclatures and examined by different meth- An integrated view of EOM biologyods. Based upon the range of data published todate, it is reasonable to conclude, however, that Current knowledge of EOM biology is clearly in-analogous ﬁber types exist across mammalian complete. For example, there is a complex patternspecies. There are, however, suggestions that ﬁb- of myoﬁbril size variation in both orbital and glo-er types in human EOMs may be more complex bal ﬁber types (Davidowitz et al., 1996a, b) thatthan those of other species (Kjellgren et al., 2003a, has not yet been accounted for in modeling EOMb). While the number of muscle ﬁbers, their diam- function. However, new data from approacheseters, and possibly the proportion of different ranging from orbital anatomy to EOM cell andmuscle ﬁber types may vary between species, the molecular biology now allow a more integratedextent of development of individual ﬁber types al- view of EOM. Here, we review the implications ofso may be an important factor that underlies their new data and concepts in EOM and oculomotorphysiological differences. physiology and EOM molecular biology and es- The most dramatic difference between species is tablish the importance of arriving at an integratedin the morphology of the orbital SIF. While the view of EOM myoﬁber and whole muscle biology.contractile elements of this ﬁber are similarbetween species, mitochondrial content varies con-siderably (Fig. 7). Orbital SIFs appear to attain Insights from EOM and oculomotor physiologytheir most extensive mitochondrial development inthe primate. Comparisons between SIFs of the One goal of correlative anatomical, molecular, andrabbit and cat suggest that orbital SIF differences physiological studies of EOM has been to uncoverin mitochondrial content may not be related solely any association of specific muscle ﬁber types withto frontal versus lateral eye placement. A more deﬁned eye movement functions. The segregationparsimonious interpretation is that the morpholo- of function among different EOM motor unitgy of this ﬁber type is directly related the recent types has been a long-debated issue. An early con-linkage of orbital layer to muscle pulley function cept of EOM suggested that the distinct EOM ﬁber(Demer et al., 2000), with increased oxidative func- types might be responsible for the different classestion and fatigue resistance necessary for the greater of eye movement. Oculomotor motoneuron activ-pulley development, wider oculomotor range, and ity in alert animals (Robinson, 1970) and intraop-reliance upon eccentric eye positions in primates erative electromyographic studies (Scott andversus rodents (Khanna and Porter, 2001). Consi- Collins, 1973), however, showed that all motoneu-stent with this view, high blood ﬂow may not be a rons and all EOM ﬁber types participate in all eyegeneral feature of EOMs in all species, but rather movement classes. These ﬁndings supported thevaries between species and is especially high in alternative hypothesis, that the heterogeneity ofthose with greater ocular motility (Wilcox et al., the six EOM ﬁber types is a consequence of their1981). By contrast, differences between species in recruitment at specific eye positions, therebythe morphology of the orbital MIF appear to be requiring a range of contractile and fatigabilitymore subtle. The most apparent difference between properties.