Pathophysiology of Myasthenia Gravis

Benjamin W. Hughes, Ph.D.,1 Maria Luisa Moro De Casillas, M.D.,1 and Henry J. Kaminski, M.D.1,2

ABSTRACT

Myasthenia gravis (MG) is arguably the best understood autoimmune disease, and its study has also led to fundamental appreciation of mechanisms of neuromuscular transmission. MG is caused by antibodies against the acetylcholine receptor (AChR), which produce a compromise in the end-plate potential, reducing the safety factor for effective synaptic transmission. It is clear that AChR antibody destruction of the postsynaptic surface is dependent on complement activation. A muscle-specic kinase has been recently found to be an antigenic target in MG patients without antibodies against the AChR. Autoantibody production in MG is a T-cell-dependent process, but how a breakdown in tolerance occurs is not known. In MG there is an interesting differential involvement of muscle groups, in particular, the extraocular muscles. This article reviews normal neuromuscular transmission, mechanisms of the autoimmune process of MG, and differential susceptibility of eye muscles to MG.
KEYWORDS: Myasthenia gravis, neuromuscular transmission, acetylcholine receptor,

neuromuscular junction

Objectives: On completion of this article, the reader will be able to summarize the anatomy and physiology of the neuromuscular junction, the immunopathogenesis of myasthenia gravis, and the differential susceptibility of eye muscles to myasthenia gravis. Accreditation: The Indiana University School of Medicine is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. Credit: The Indiana University School of Medicine designates this educational activity for a maximum of 1 Category 1 credit toward the AMA Physicians Recognition Award. Each physician should claim only those hours of credit that he/she actually spent in the educational activity. Disclosure: Statements have been obtained regarding the authors relationships with nancial supporters of this activity, use of trade names, investigational products, and unlabeled uses that are discussed in the article. The authors have nothing to disclose.

o understand the pathophysiology of myasthenia gravis (MG), a thorough appreciation of normal neuromuscular transmission and the underlying anatomy of the neuromuscular junction (NMJ) is imperative.15 Although basic mechanisms of how the junction functions to transmit signals from nerve to muscle have been well established for decades, the last 10 years has seen an explosion of information regarding the molecular under-

pinning of transmission. In concert with these developments, appreciation of the intricacies of the autoimmune reaction underlying destruction of the postsynaptic surface of the muscle has expanded. This review will discuss the structure of the NMJ, how neuromuscular transmission occurs, and the pathogenesis of MG, as well as a summary of why certain muscle groups may be differentially involved by the disease.

Myasthenia Gravis; Editor in Chief, Karen L. Roos, M.D.; Guest Editor, Robert M. Pascuzzi, M.D. Seminars in Neurology, Volume 24, Number 1, 2004. Address for correspondence and reprint requests: Henry J. Kaminski, M.D., Department of Neurology, University Hospitals of Cleveland, 11100 Euclid Avenue, Cleveland, OH 44106. Departments of 1Neurology and 2Neurosciences, Case Western Reserve University, Louis Stokes Cleveland DVA Medical Center, University Hospitals of Cleveland, Cleveland, Ohio. Copyright # 2004 by Thieme Medical Publishers, Inc., 333 Seventh Avenue, New York, NY 10001, USA. Tel: +1(212) 584-4662. 0271-8235,p;2004,24,01,021,030,ftx,en;sin00285x.

21

22

SEMINARS IN NEUROLOGY/VOLUME 24, NUMBER 1

2004

ANATOMY AND PHYSIOLOGY OF THE NEUROMUSCULAR JUNCTION Presynaptic Surface As the motor nerve approaches a muscle it branches, innervating many muscle bers and providing a single, unmyelinated nerve terminal to each of the bers. The NMJ is a specialized synapse designed to transmit nerve impulses from the nerve terminal to muscle, via the chemical transmitter, acetylcholine (ACh), which is stored at the terminal in synaptic vesicles (Fig. 1). These vesicles are aligned near release sites called active zones where calcium channels are arranged in parallel double rows. When an action potential reaches the nerve terminal, calcium channels of the P/Q-type (N-type channels also are localized to the presynaptic membrane) are activated, calcium enters the presynaptic terminal, and the local calcium concentration rises signicantly, triggering the vesicles to release their contents. In the vertebrate NMJ, a typical nerve terminal action potential does not fully activate nerve terminal calcium channels, as the duration of the action potential is 1 ms, while nerve terminal calcium channels are activated with a time constant of 1.3 ms.6 Therefore, enhancement of ACh release from the presynaptic nerve terminal can be achieved by increasing the nerve terminal action potential duration via blockade of delayed rectier potassium

channels with 3,4-diaminopyridine or inhibiting calcium uptake with quanidine, agents used in treatment of Lambert-Eaton syndrome. The mechanism of relaying the calcium signal to synaptic vesicle fusion involves conformational changes in multiple proteins on the synaptic vesicle membrane and the plasma membrane of the nerve terminal.4,7,8 The events leading to release of synaptic vesicle contents are being increasingly dened. Vesicles initially undergo a process called docking, in which they come into close proximity with the nerve terminal membrane and then undergo priming that allows them to respond to the calcium signal. During docking, munc18 dissociates from syntaxin and synaptophysin from synaptobrevin allowing the synaptic core complex to form. Three proteins, two on the plasma membrane (syntaxin and synaptic vesicle associated protein 25 or SNAP 25) and one on the synaptic vesicle membrane (synaptobrevin), are thought to form the docking complex. N-ethylmaleimide sensitive factor (NSF) and soluble NSF attachment protein (SNAP) associate to form a fusion complex with the docking proteins. N-ethylmaleimide sensitive factor, which is an ATPase, cross-links multiple core complexes into a network, and ATP hydrolysis leads to hemifusion of the vesicle and presynaptic nerve terminal membranes. Synaptotagmin likely acts as the calcium sensor, but other synaptic proteins may act in concert to perform

Figure 1 Schematic of the mammalian neuromuscular junction, illustrating the major proteins located at the mammalian neuromuscular junction. The presynaptic nerve terminal contains synaptic vesicles (striped ovals), voltage-gated potassium (VGKC, triangles), and calcium channels (VGCC, gray rectangles). Acetylcholine receptor illustrated here in its pentameric structure. Voltage-gated sodium channels (striped rectangles), muscle-specic kinase (MuSK), rapsyn, utrophin, and the sarco-dystroglycan complex are located at the postsynaptic membrane. Details are discussed throughout the article. Objects are not drawn to scale.

PATHOPHYSIOLOGY OF MYASTHENIA GRAVIS/HUGHES ET AL

23

this function. The cytoplasmic portion of synaptotagmin includes two regions with high homology to the calcium and phospholipid-binding domains of a protein kinase. Synaptotagmin likely binds phospholipid of the plasma membranes and syntaxin. Binding of calcium to synaptotagmin may alter interactions of synaptotagmin with membrane lipids and syntaxin, allowing the membranes to fully fuse. Botulinum toxins are proteases that target several of the synaptic vesicle proteins and their investigation has greatly enhanced understanding of transmitter release.9 How the synaptic release mechanism is capable of discharging synaptic vesicle contents in the extremely rapid fashion that it does is not known. After releasing their contents to the synaptic cleft, synaptic vesicle membrane is recycled by a clathrinmediated mechanism. After reuptake of the vesicles, the clathrin-coated vesicles shed their coats and translocate into the interior. The vesicle membranes fuse with endosomes in the nerve terminal and new vesicles bud from the endosome. The new vesicles package ACh and other chemicals by active transport and translocate back to the active zone either by diffusion or by a cytoskeletal transport process.

concentration of ACh in the synaptic cleft, and play a role in modulation of neuromuscular transmission.13,14

Synaptic Cleft Between the nerve and muscle plasma membranes lies a space of  50 nm called the synaptic cleft (Fig. 1). The extracellular matrix of the synaptic cleft is a complex collection of proteins that regulate the synthesis of postsynaptic proteins and the concentration of acetylcholine esterase (AChE). The basement membrane is enriched with collagen and contains several forms of laminin, all of which bind to a-dystroglycan in the postsynaptic membrane. The laminins form a network in the synaptic space that anchors other extracellular matrix proteins.10 Once released from the synaptic vesicles, diffusion of ACh across the cleft is rapid because of the small distance to cross and the high diffusion rate of ACh.2 AChE is concentrated in the basal lamina of the postsynaptic membrane, and its action in hydrolyzing ACh, as well as the diffusion of ACh out of the cleft, leads to a rapid decline in ACh concentration.11 This prevents the AChR from being activated more than once in response to ACh. AChE inhibitors, such as pyridostigmine and edrophonium, prolong the duration of action of ACh on the postsynaptic membrane, slowing the decay of the ACh-induced endplate current.12 In postsynaptic diseases of neuromuscular transmission, AChE inhibition will serve to enhance ACh action and promote achievement of an end-plate potential that will lead to action potential generation. Schwann cells surrounding the nerve terminal have been found to secrete an acetylcholine binding protein, which may reduce the effective

Postsynaptic Surface Once ACh traverses the synaptic cleft it binds the AChR concentrated across from the nerve terminal active zones (Fig. 1). This binding results in opening of the AChR ion channel and the entry of cations, mainly sodium, into the muscle, leading to end-plate potential generation. When a certain threshold depolarization is achieved, voltage-gated sodium channels, at the bottom of the postsynaptic folds, open, allowing the entry of more sodium ions and generating the muscle action potential and contraction.4 The postsynaptic skeletal muscle surface is characterized by invaginations of the plasma membrane, termed secondary synaptic folds.2 The folds increase the surface area of the postsynaptic membrane and their architecture allows end-plate depolarization to be focused on their depths.15 AChR concentrated at the tops of the secondary synaptic folds are anchored to the dystrophin-related protein complex through rapsyn, a protein involved in clustering of AChR at the NMJ during synapse development.2 AChR clusters are connected to the cytoskeletal elements via associations with the dystroglycan and sarcoglycan protein complexes and utrophin.16 Mice decient in rapsyn do not effectively cluster AChR at the synapse. Some congenital myasthenia patients have been found to have rapsyn mutations or alterations in utrophin expression.1719 In addition to the anchoring system serving to concentrate AChR, muscle ber nuclei near the NMJ preferentially express AChR subunit genes.20 In addition to AChR, other proteins are differentially expressed at the NMJ. Na channels are highly concentrated in the membrane at the base of the synaptic folds by interaction with ankyrin, the sarcoglycan complex, the dystroglycan complex, dystrobrevins, and dystrophin/utrophin (Fig. 1). Utrophin and dystrobrevin also connect to a1 syntrophin and a2 syntrophin, which in turn associate with nitric oxide synthetase.21,22 Nitric oxide produced locally on the postsynaptic surface could serve a signaling function and has been shown to inuence synaptogenesis.23

The Acetylcholine Receptor The AChR is composed of four subunits and exists in two isoforms.2426 The adult AChR is composed of two a-subunits and one copy of each of the b-, d-, and e-subunits, while the fetal AChR has a g-subunit in place of the e-subunit. High amino acid sequence homology exists among the subunits of the AChR with each containing four a-helices, designated M1 to M4, that span the plasma membrane. The extracellular portions of

24

SEMINARS IN NEUROLOGY/VOLUME 24, NUMBER 1

2004

the subunits consisting of the N- and C-terminal regions and the region between M2 and M3 form a large extracellular vestibule that surrounds the channel extracellular orice. The regions between M1 and M2 and between M3 and M4 form a smaller vestibule around the intracellular orice of the ion channel.27,28 The skeletal muscle nicotinic AChR has a large extracellular surface with the potential to form many epitopes for autoantibody binding and accommodate simultaneous binding of different antibodies. In developing muscle, the fetal AChR is expressed along the entire ber surface and contributes to the muscle being able to produce spontaneous contractions, which is critical to development.29 With innervation fetal AChR is down-regulated and the adult receptor is expressed at the synapse. Only mammalian extraocular muscle is known to express fetal AChR protein at the synapse in adulthood30,31; however, reexpression of the fetal AChR in certain patients with congenital myasthenia prevents lethality of some mutations of AChR subunit genes.32 Adult and fetal AChR can be distinguished by their electrophysiological characteristics. Adult channels have shorter mean open times and a single channel conductance that is  50% larger than that found with fetal channels. The differences between adult and fetal AChR channels are due to replacement of the g- with the e-subunit. Phosphorylation and other forms of posttranslational modication can alter properties of the AChR, and in particular, subunit phosphorylation appears to regulate agonistinduced desensitization.24

rapid tyrosine phosphorylation of MuSK in myotubes, consistent with the idea that MuSK is a receptor, or a component of a receptor complex, for agrin. On the other hand, forced expression of MuSK in broblasts or myoblasts does not lead to its phosphorylation by agrin.38 Also, recombinant agrin and MuSK in vitro do not bind, indicating that additional components that may be expressed in myotubes and skeletal muscle bers, and not in myoblasts or broblasts, are essential for agrin to activate MuSK. As will be described later, MuSK has been identied as an antigenic target in MG patients seronegative for antibodies against the AChR.

Neuromuscular Junction Formation and Muscle-Specic Kinase The development of the NMJ requires a complex series of interactions between developing motor neurons and muscle bers.2,11 Present understanding indicates that agrin, a protein synthesized by motor neurons and stably deposited into the synaptic basal lamina, stimulates a muscle-specic kinase (MuSK), a receptor tyrosine kinase, that is expressed selectively in skeletal muscle. This signal is thought to cluster signicant postsynaptic proteins, including AChRs, at the NMJ (Fig. 1).33 This formulation is based on data showing that agrin can stimulate multiple aspects of postsynaptic differentiation in cultured myotubes, including the clustering and tyrosine phosphorylation of AChR.3436 Mice lacking agrin or MuSK fail to form neuromuscular synapses, and as a result, die at birth because of a failure to move or breathe.37,38 MuSK mutant mice do not concentrate AChR or other synaptic proteins at the NMJ, but rather these proteins are instead uniformly expressed along the muscle ber. The precise mechanism by which agrin activates MuSK is poorly understood.39,40 Agrin stimulates the

SAFETY FACTOR FOR NEUROMUSCULAR TRANSMISSION Compromise of the safety factor for neuromuscular transmission is common to all neuromuscular transmission disorders.15 The safety factor is dened as the ratio of the end-plate potential amplitude to the difference between the membrane potential and the threshold potential for initiating an action potential. Normally, the nerve terminal releases enough ACh to induce an excitatory end-plate potential, which is greater than the threshold needed to initiate an action potential, and therefore, the safety factor is quite large. Quantal release, AChR conduction properties, AChR density, and AChE activity contribute to the end-plate potential and the safety factor. Sodium channel concentration at the postsynaptic surface affects the safety factor by making the action potential threshold easier to achieve. In addition, the synaptic folds architecture, which allows current to be concentrated on the sodium channels, enhances the end-plate potential. AChE activity terminates the action of ACh and its inhibition will enhance AChR activation. Postsynaptic factors also inuence the safety factor. Repetitive stimulation reduces transmitter release, which under normal conditions will reduce the end-plate potential but not enough to prevent action potential generation. In pathological situations, such as MG, this reduction may be enough to produce transmission failure.

IMMUNOPATHOLOGY MG meets strict criteria that dene antibody-mediated, autoimmune disorders: (1) antibodies against the AChR can be detected in 80 to 90% of patients41; (2) immunoglobulin G (IgG) is deposited at the NMJ, the site of pathology42; (3) administration of IgG from MG patients to experimental animals reproduces the characteristic clinical features of the disease43; (4) therapies that reduce the serum concentration of anti-AChR antibodies improve weakness44; and (5) MG can be induced in animals by immunization with puried AChR,

PATHOPHYSIOLOGY OF MYASTHENIA GRAVIS/HUGHES ET AL

25

further supporting the concept that an immune response directed against the AChR is responsible for the disease in patients.45

Autoantibodies in MG AChR antibodies are polyclonal, vary in light chain composition and subclass structure, and identify a heterogeneous epitope repertoire that varies among patients.25,26,46,47 AChR antibodies produce decient neuromuscular transmission by three different mechanisms: (1) they bind to the AChR and alter function; (2) they promote endocytosis and accelerate the degradation rate of AChR; the effect of these antibodies in accelerating degradation is thought to result from their ability to cross-link the receptors48; and (3) antibodies activate complement leading to destruction of the postsynaptic surface. The most important mechanism of neuromuscular transmission compromise is antibody-mediated complement activation, and this statement is supported by several experimental observations. C3 activation fragments, C9, and the membrane attack complex can be detected at the NMJ in patients and experimental autoimmune MG of animals.49,50 Methods that deplete complement, block activation, or engineer defects in complement components protect animals against disease induction.5155 Skeletal muscle is protected from autologous complement-mediated injury by cell surface regulators located on their plasma membranes.56,57 These regulators are three proteins: the decay accelerating factor, the membrane cofactor protein, and the membrane inhibitor of reactive lysis. Collectively, these control proteins accelerate the decay of autologous C3 convertases and membrane attack complex that inappropriately assembles on self cell surfaces, promote the cleavage of uncomplexed autologous cell-bound complement fragments, and inhibit the formation of autologous attack complex, which brings about lysis of the postsynaptic membrane. Deciency of decay accelerating factor in mice increases the severity of experimentally induced MG, suggesting that these complement regulators could modulate disease severity in humans.58 Most pathogenic antibodies bind epitopes formed by the a-subunit in a region termed the main immunogenic region (MIR).59 The most critical immunogenic area within the MIR is located on the extracellular N-terminal segment of the a-subunit. The MIR may be a particularly pathogenic region because (1) its location on surface of the AChR is easily accessed by antibody, (2) its amino acid structure is highly immunogenic, and (3) its location in densely packed AChR of the NMJ allows cross-linking and antigenic modulation.25,26 The immunodominance of this region might also be facilitated by the molecular mimicry between the MIR and epitopes on certain retroviruses.60

In general, direct effects on AChR function are not considered to contribute signicantly to MG pathology, but some patients have antibodies that bind the ACh binding site. These antibodies can contribute to myasthenic weakness and could cause acute myasthenic crisis. In animals, they produce a unique form of severe experimental autoimmune MG.61 The rst autoantibodies detected in MG patients were the striational antibodies, when MG patient serum was found to bind skeletal muscle in immunohistochemical studies.62 The majority of these striational antibodies have been found to bind the structural protein titin.63,64 Antibodies against other skeletal muscle proteins, including the ryanodine receptor, myosin, tropomyosin, troponin, a-actinin, and actin, have been found among patients.65 The pathogenic signicance of these antibodies is not known, although ryanodine antibodies can compromise in vitro muscle contractility. The antibody prole of MG patients suggests differences in immunopathogenesis. Older-onset patients have higher frequencies of antibodies against the structural protein titin, and the disease is more severe.64 Patients with thymoma also tend to be positive for not only titin antibodies but also ryanodine receptor antibodies, and such patients have been found to have associated myocarditis or myositis.66 Up to 20% of patients with generalized MG are seronegative for AChR antibodies with  30% of these patients having autoantibodies against MuSK.6769 Ocular MG patients who are seronegative for the AChR are also negative for MuSK. MuSKpositive patients exhibit more severe disease than the AChR-positive patients, and some have been found to have muscle atrophy. Some MuSK antibodies compromise AChR channel function, but specic pathogenic mechanisms have not been dened. One would predict that MuSK antibodies compromise neuromuscular transmission by affecting AChR clustering at the NMJ.

Cellular Autoimmunity Autoantibody production in MG is a T-cell-dependent process and a breakdown in tolerance toward self-antigens is the primary abnormality in MG. CD4 T-helper cells that are AChR-specic are crucial for the activation of B cells and for the synthesis of high-afnity IgG autoantibodies.46 CD4 T cells in patients with MG regulate the production of anti-AChR antibodies, and suppressor T cells are present, which can decrease antibody production. Several properties of the anti-AChR CD4 T cells are fundamental for understanding the pathogenesis of MG. CD4 T-helper cell receptors respond to antigens processed by antigen-presenting cells (APC) and associate with self major histocompatibility

26

SEMINARS IN NEUROLOGY/VOLUME 24, NUMBER 1

2004

complex (MHC) class II molecules. T-cell receptor binds AChR only after it is processed by proteolytic cleavage and association with a MHC molecule by APC. The T-cell receptor grants antigen specicity to CD4 T cells. The requirement for MHC association restricts the number of epitopes that a T cell can recognize. Differences in MHC class II molecules have been shown to affect the severity of symptoms in experimental autoimmune MG and could be true in humans.70 The AChR a-subunit contains the majority of T-cell recognition sites, although other subunits are recognized by T cells from patients. The epitopes on the a-subunit recognized by the T cell are distinct from the MIR recognized by antibodies. Similar to antibody production, the T-cell response is polyclonal and varies among MG patients. The T-cell-dependent nature of MG explains its association with human leukocyte antigen (HLA) antigens.7173 An increased association of HLA-A1, -B8, and -DRw3 (and -B12 antigen in Japan) occurs among MG women under 40 without thymoma. HLA-A3, -B7, and -DRw2 (and -A10 in Japan) occur with greater frequency in men after 40 without a thymoma. Restriction fragment length polymorphism analysis has demonstrated a link between MG and HLA-DQ, suggesting that HLA site is close to the coding region for the epitope of the AChR binding site. With increasing disease duration, CD4 sensitization spreads to other parts of the AChR. The epitope repertoire of anti-AChR CD4 cells in MG patients is complex and characteristic of an individual patient.46 A few sequence regions of anti-AChR CD4 cells are recognized in most or all MG patients. These sequences form CD4 epitopes that are both universal and immunodominant. These universal epitopes sensitize pathogenic CD4 cells and drive the synthesis of AChR antibodies.47 CD4 T cells from MG patients respond in a remarkably heterogeneous fashion74 and recognize epitopes from each of the AChR subunits including the adult e-subunit, the d-subunit, and the fetal g-subunit. Patients who have generalized weakness for greater than 5 years usually have CD4 T cells that recognize all subunits, whereas patients with weakness for shorter periods will only recognize few subunits of the AChR. These ndings suggest enlargement of the CD4 epitope repertoire with time in generalized MG. Studies of anti-AChR CD4 cells T-cell receptors from MG patients do not allow distinction between an initial pathogenic mechanism, which entails molecular mimicry, or the intervention of a superantigen. Superantigens are proteins found on viruses or bacteria that stimulate particularly powerful proliferative responses. Superantigens are not processed but directly bind to the class II MHC and to the T-cell receptor.

Cytokine Inuences Cytokines are peptides with intercellular signaling properties that regulate local and systemic immune responses, as well as other biological processes. Cytokines, secreted by different CD4 T-helper cell subsets after antigenstimulated activation, appear to play an important role in the pathogenesis of MG.46 T-helper cells can be distinguished on the basis of the array of cytokines secreted. Th1 cells secrete proinammatory cytokines such as interferon-g, interleukin-2, and tumor necrosis factor (TNF)-a. Th2 cells express anti-inammatory/ regulatory cytokines including interleukin-4 and interleukin-10. Both Th1 and Th2 cell groups induce antibody synthesis by secreting cytokines that induce proliferation and differentiation of B cells. Some studies of cellular cytokine secretion in MG patients have shown that MG is associated with involvement of both Th1 and Th2 cells.75,76 Th1 cells are involved in the antibody response in MG and have an intricate epitope repertoire. In MG, the AChR is the sensitizing antigen and the target of the autoimmune Th1 cells. However, molecular mimicry between one AChR epitope and a microbial antigen might also play a role in the launching of the autoimmune response. Interleukin-4, secreted by Th2 cells, may be protective in experimental MG of animals as interleukin-4 has anti-inammatory effects on both Th1 cells and APCs and acts as a growth-factor for transforming-growth factor (TGF)-b secreting cells. TGF-b, which inhibits both cellular and humoral immunity, is also up-regulated in MG patients. Other cytokines that are being investigated in MG pathogenesis include TNF-a and TNF-b. Their function still needs further clarication; most likely they participate within the cytokine network, modulating the immune response.46

The Thymus in MG Pathogenesis The thymus plays a key role in tolerance induction to self-antigens and in responsiveness of lymphocytes to foreign antigens.46,74 During development bone marrow stem cells appear in the thymic subcapsular epithelium, where a random process of gene rearrangements occurs in the regions that will code for the T-cell antigen receptor. The immature T cells pass through the thymic cortex, and those that recognize MHC antigens pass through to the medulla. The majority of immature T cells that do not recognize self MHC antigens are eliminated. During this stage in the thymic cortex, T cells, which would react toward self-antigens, are also removed; however, the mechanisms by which this occurs are not known. Once in the medulla, the T cells differentiate into helper and suppressor cells and eventually are released to the periphery. Several ndings indicate that the thymus is involved in MG pathogenesis. Pathological changes,

PATHOPHYSIOLOGY OF MYASTHENIA GRAVIS/HUGHES ET AL

27

specically lymphoid follicular hyperplasia and thymomas, are frequent in patients with MG.77 MG thymic tissue contains abnormally elevated amounts of mature T cells. Thymomas and myasthenic thymus are enriched for AChR-reactive T cells.78 When myasthenic thymic tissue fragments are transplanted to severe combined immunodeciency mice, they produce human antibodies, which bind to the AChR.79 Most myasthenic thymuses contain B cells capable of producing AChR antibodies, particularly those hyperplastic thymuses with germinal centers. Normal and myasthenic human thymuses express gene transcripts and epitopes of all AChR subunits.8082 The expression of antigenically similar proteins to the AChR in the thymus would provide a source of antigen for autosensitization of T cells toward the skeletal muscle AChR. Further, thymectomy appears to improve the clinical course of patients with MG.83

THE DIFFERENTIAL INVOLVEMENT OF MUSCLE GROUPS BY MG MG demonstrates a remarkable variation in muscle group involvement among patients. There is no better example of this than the involvement of ocular muscles. Weakness of the extraocular muscle (EOM) and the levator palpebrae are the initial signs of MG in the majority, ultimately occurs in nearly all, and may remain restricted to these muscles in 10 to 15% of patients.84 One of the authors (H.J.K.) has recently provided detailed reviews of the preferential involvement of the EOM by MG,85,86 and the arguments are briey summarized here. The most simple, and least scientically exciting, is that ocular muscle dysfunction leads to dramatic symptoms of double or blurred vision, which brings the patient to medical attention early.87 Because of the precision required to maintain alignment of the visual axes, even a minor reduction of force generation produces visual compromise, in contrast to a small reduction in force generation of a limb muscle, which may not be readily appreciated or be ignored by a patient. Also, in extremity muscle, proprioceptive input may compensate to some extent by recruitment of additional motor units to perform a task. In ocular motor control, proprioception may not compensate in the same manner for reductions of force generation. Several factors specic to the EOM place them at risk for MG involvement. Unique physiological properties of the oculomotor system may be of paramount importance in targeting of EOM by neuromuscular transmission abnormalities. The ocular motor neuron ring frequencies are extremely high, at times 400 to 500 Hz, which would exacerbate a transmission defect. EOM singly innervated bers possess anatomical characteristics that would be predicted to make them susceptible to neuromuscular transmission block. The bers

have less prominent synaptic folds, and therefore one would predict fewer AChRs and sodium channels on the postsynaptic membrane.88 A reduction in AChR and sodium channels would reduce the safety factor predisposing EOM to neuromuscular transmission failure. Na channel density varies between fast-twitch and slow-twitch bers with fast-twitch bers having higher densities in the region of the NMJ.89 These differences could contribute to clinical involvement of muscle groups other than EOM by MG.90 EOM also contains multi-innervated bers that are similar to reptile tonic and intermediate bers that have tonic contractile characteristics with the functional consequence that force generation is directly proportional to the membrane depolarization caused by the end-plate potential. Therefore, a safety factor does not exist for multi-innervated bers, and any reduction of AChR would decrease contractile force of these bers.85,86,88 Also, the neuromuscular junctions have a scarcity (to complete absence) of junctional folds.91 The autoimmune process of patients with ocular myasthenia does differ from patients with generalized disease. Ocular myasthenia patients have lower levels of AChR antibodies.72 The intensity of T-cell responses to AChR epitopes of ocular myasthenia patients are lower than that of generalized patients, and they uctuate over time.92 Complement regulatory proteins are expressed at lower levels in EOM; therefore, a complement-mediated disease, such as MG, could be more likely to induce damage of these muscles.93 The combination of mild autoimmune disease and physiologic susceptibility may be the explanation for the existence of purely ocular myasthenia. The reasons for levator palpebrae involvement are not as clearly understood. The muscle is more similar to other skeletal muscle than to the EOM. The levator muscle bers have twitch characteristics and some are highly fatigue-resistant, based on histological appearances, but there are no multi-innervated bers. Nonimmune disorders of neuromuscular transmission also have a predilection for producing ptosis, suggesting that physiological reasons predispose them to neuromuscular transmission failure. The levator bers are under constant neuronal stimulation during eye opening. This makes neuromuscular fatigue more likely to occur compared with other skeletal muscles. The junctional folds of levator NMJ are also sparse as in EOM and suggest a lower AChR number and a reduction in safety factor.85

SUMMARY This review discussed the basic physiology of neuromuscular transmission and the structure that underlies communication between the nerve and muscle, the NMJ. Diseases of the NMJ and ion channels are of general interest because an understanding of their

28

SEMINARS IN NEUROLOGY/VOLUME 24, NUMBER 1

2004

pathophysiology may provide insight into disorders at other neural synapses, such as epileptic or psychiatric diseases. MG serves as a model for the investigation of all autoimmune disorders.

ACKNOWLEDGMENTS

This work was supported by NIH grants EY013238 and EY014837 to Dr. Kaminski, who is also supported by the Department of Veterans Affairs Medical Research Service.

REFERENCES
1. Engel AG. Anatomy and molecular architecture of the neuromuscular junction. In: Engel AG, ed. Myasthenia Gravis and Myasthenic Disorders. Oxford: Oxford University Press; 1999:339 2. Sanes JR, Lichtman JW. Induction, assembly, maturation and maintenance of a postsynaptic apparatus. Nat Rev Neurosci 2001;2:791805 3. Boonyapisit K, Kaminski HJ, Ruff RL. The molecular basis of neuromuscular transmission disorders. Am J Med 1999;106: 97113 4. Ruff RL. Neurophysiology of the neuromuscular junction: overview. Ann N Y Acad Sci 2003;998:110 5. Kaminski HJ. Myasthenia Gravis and Related Disorders. Totowa, NJ: Humana Press; 2003:1396 6. Stanley EF. Presynaptic calcium channels and the transmitter release mechanism. Ann N Y Acad Sci 1993;681:368 372 7. Poage RE, Meriney SD. Presynaptic calcium inux, neurotransmitter release, and neuromuscular disease. Physiol Behav 2002;77:507512 8. Fon EA, Edwards RH. Molecular mechanisms of neurotransmitter release. Muscle Nerve 2001;24:581601 9. Schiavo G, Matteoli M, Montecucco C. Neurotoxins affecting neuroexocytosis. Physiol Rev 2000;80:717766 10. Abbas L. Synapse formation: lets stick together. Curr Biol 2003;13:R25R27 11. Arber S, Burden SJ, Harris AJ. Patterning of skeletal muscle. Curr Opin Neurobiol 2002;12:100103 12. McMahan UJ, Sanes JR, Marshall LM. Cholinesterase is associated with the basal lamina at the neuromuscular junction. Nature 1978;271:172174 13. Smit AB, Syed NI, Schaap D, et al. A glia-derived acetylcholine-binding protein that modulates synaptic transmission. Nature 2001;411:261268 14. Brejc K, van Dijk WJ, Klaassen RV, et al. Crystal structure of an ACh-binding protein reveals the ligand-binding domain of nicotinic receptors. Nature 2001;411:269276 15. Wood SJ, Slater CR. Safety factor at the neuromuscular junction. Prog Neurobiol 2001;64:393429 16. Vincent A. The neuromuscular junction and neuromuscular transmission. In: Karpati G, Hilton-Jones D, Griggs RC, eds. Disorders of Voluntary Muscle. Cambridge: Cambridge University Press; 2001:142167 17. Sieb JP, Dorer P, Tzartos S, et al. Congenital myasthenic syndromes in two kinships with end-plate acetylcholine receptor and utrophin deciency. Neurology 1998;50:5461

18. Gautam M, Noakes PG, Mudd J, et al. Failure of postsynaptic specialization to develop at neuromuscular junctions of rapsyn-decient mice. Nature 1995;377:232236 19. Burke G, Cossins J, Maxwell S, et al. Rapsyn mutations in hereditary myasthenia: distinct early- and late-onset phenotypes. Neurology 2003;61:826828 20. Merlie JP, Sanes JR. Concentration of acetylcholine receptor mRNA in synaptic regions of adult muscle bers. Nature 1985;317:6668 21. Brenman JE, Chao DS, Gee SH, et al. Interaction of nitric oxide synthase with the postsynaptic density protein PSD-95 and a1-syntrophin mediated by PDZ domains. Cell 1996;84: 757767 22. Kusner LL, Kaminski HJ. Nitric oxide synthase is concentrated at the skeletal muscle endplate. Brain Res 1996;730: 238242 23. Wang T, Xie Z, Lu B. Nitric oxide mediates activity dependent synaptic suppression at developing neuromuscular synapses. Nature 1995;374:262266 24. Kaminski HJ, Ruff RL. Insights into possible skeletal muscle nicotinic acetylcholine receptor (AChR) changes in some congenital myasthenias from physiological studies, point mutations, subunit substitutions of the AChR. Ann N Y Acad Sci 1993;681:435450 25. Lindstrom J. Acetylcholine receptors and myasthenia. Muscle Nerve 2000;23:453477 26. Lindstrom J. Acetylcholine receptor structure. In: Kaminski HJ, ed. Myasthenia Gravis and Related Disorders. Totowa, NJ: Humana Press; 2003:1552 27. Unwin N. Projection structure of the nicotinic acetylcholine receptor: distinct conformations of the a subunits. J Mol Biol 1996;257:586596 28. Unwin N, Toyoshima C, Kubalek E. Arrangement of the acetylcholine receptor subunits in the resting and desensitized states, determined by cryoelectron microscopy of crystallized torpedo postsynaptic membranes. J Cell Biol 1988;107:1123 1138 29. Brehm P, Henderson L. Regulation of acetylcholine receptor channel function during development of skeletal muscle. Dev Biol 1988;129:111 30. Kaminski HJ, Kusner LL, Block CH. Expression of acetylcholine receptor isoforms at extraocular muscle endplates. Invest Ophthalmol Vis Sci 1996;37:345351 31. MacLennan C, Beeson D, Buijs A-M, Vincent A, NewsomDavis J. Acetylcholine receptor expression in human extraocular muscles and their susceptibility to myasthenia gravis. Ann Neurol 1997;41:423431 32. Engel AG, Ohno K, Bouzat C, Sine SM, Griggs RC. Endplate acetylcholine receptor deciency due to nonsense mutations in the epsilon subunit. Ann Neurol 1996;40:810817 33. Ruegg MA, Bixby JL. Agrin orchestrates synaptic differentiation at the vertebrate neuromuscular junction. Trends Neurosci 1998;21:2227 34. Wallace BG. Agrin-induced specializations contain cytoplasmic, membrane, and extracellular matrix-associated components of the postsynaptic apparatus. J Neurosci 1989;9:1294 1302 35. Wallace BG, Qu Z, Huganir RL. Agrin induces phosphorylation of the nicotinic acetylcholine receptor. Neuron 1991;6: 869878 36. Wallace BG. The mechanism of agrin-induced acetylcholine receptor aggregation. Philos Trans R Soc Lond B Biol Sci 1991;331:273280

PATHOPHYSIOLOGY OF MYASTHENIA GRAVIS/HUGHES ET AL

29

37. Gautam M, Noakes PG, Moscoso L, et al. Defective neuromuscular synaptogenesis in agrin-decient mutant mice. Cell 1996;85:525535 38. Glass DJ, Bowen DC, Stitt TN, et al. Agrin acts via a MuSK receptor complex. Cell 1996;85:513523 39. Wallace B. scFvs get down to basics: how MuSK makes synapses. Nat Biotechnol 1997;15:721722 40. Wickelgren I. Synapse-making molecules revealed. Science 1996;272:1100 41. Lindstrom JM, Seybold MD, Lennon VA, Whittingham S, Duane DD. Antibody to acetylcholine receptor in myasthenia gravis: prevalence, clinical correlates, and diagnostic value. Neurology 1976;26:10541059 42. Engel AG, Lambert EH, Howard FM. Immune complexes (IgG and C3) at the motor end-plate in myasthenia gravis. Ultrastructure and light microscopic localization and electrophysiological correlations. Mayo Clin Proc 1977;52:267280 43. Toyka KV, Drachman DB, Grifn DE, Pestronk D. Myasthenia gravis study of humoral immune mechanisms by transfer to mice. N Engl J Med 1977;296:125131 44. Pinching AJ, Peters DK, Newsom-Davis J. Remission of myasthenia gravis following plasma exchange. Lancet 1976;2: 13731376 45. Patrick J, Lindstrom J. Autoimmune response to acetylcholine receptor. Science 1973;180:871872 46. Conti-Fine BM, Kaminski HJ. Autoimmune neuromuscular transmission disorders: myasthenia gravis and Lambert-Eaton myasthenic syndrome. Continuum 2001;7:5693 47. Conti-Fine B, Bellone M, Howard JJ, Protti M. Myasthenia Gravis: The Immunobiology of an Autoimmune Disease. Georgetown, TX: Neuroscience Intelligence Unit, R.G. Landes; 1997 48. Drachman D, Angus CW, Adams RN, Kao I. Effect of myasthenic patients immunoglobulin on acetylcholine receptor turnover: selectivity of degradation process. Proc Natl Acad Sci USA 1978;75:34223426 49. Sahashi K, Engel AG, Lambert EH, Howard FM Jr. Ultrastructural localization of the terminal and lytic ninth complement component (C9) at the motor end-plate in myasthenia gravis. J Neuropathol Exp Neurol 1980;39:160172 50. Nakano S, Engel AG. Myasthenia gravis: quantitative immunocytochemical analysis of inammatory cells and detection of complement membrane attack complex at the endplate in 30 patients. Neurology 1993;43:11671172 51. Lennon VA, Seybold ME, Lindstrom JM, Cochrane C, Ulevitch R. Role of complement in the pathogenesis of experimental autoimmune myasthenia gravis. J Exp Med 1978;147:973983 52. Corey AL, Richman DP, Agius MA, Wollmann RL. Refractoriness to a second episode of experimental myasthenia gravis. Correlation with AChR concentration and morphologic appearance of the postsynaptic membrane. J Immunol 1987;138:32693275 53. Corey AL, Richman DP, Shuman CA, Gomez CM, Arnason BG. Use of monoclonal antiacetylcholine receptor antibodies to investigate the macrophage inammation of acute experimental myasthenia gravis: refractoriness to a second episode of acute disease. Neurology 1985;35:14551460 54. Biesecker G, Gomez CM. Inhibition of acute passive transfer experimental autoimmune myasthenia gravis with fab antibody to complement C5. J Immunol 1989;142:26542659 55. Tuzun E, Scott BG, Goluszko E, Higgs S, Christadoss P. Genetic evidence for involvement of classical complement

56.

57. 58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

pathway in induction of experimental autoimmune myasthenia gravis. J Immunol 2003;171:38473854 Miwa T, Song WC. Membrane complement regulatory proteins: insight from animal studies and relevance to human diseases. Int Immunopharmacol 2001;1:445459 Walport MJ. Complement. N Engl J Med 2001;344:1058 1066 Lin F, Kaminski H, Conti-Fine B, Wang W, Richmonds C, Medof M. Enhanced susceptibility to experimental autoimmune myasthenia gravis in the absence of decay-accelerating factor protection. J Clin Invest 2002;110:12691274 Tzartos S, Seybold M, Lindstrom J. Specicities of antibodies to acetylcholine receptors in sera from myasthenia gravis patients measured by monoclonal antibodies. Proc Natl Acad Sci USA 1982;79:188192 Manfredi AA, Protti MP, Bellone M, Moiola L, ContiTronconi BM. Molecular anatomy of an autoantigen: T and B epitopes on the nicotinic acetylcholine receptor in myasthenia gravis. J Lab Clin Med 1992;120:1321 Gomez CM, Richman DP. Monoclonal anti-acetylcholine receptor antibodies with differing capacities to induce experimental autoimmune myasthenia gravis. J Immunol 1985;135: 234241 Strauss AJL, Seegal JC, Hsu JC, Burkholder PM, Nastuk W, Ossermann KE. Immunouoresence demonstration of a muscle binding, complement xing serum globulin fraction in myasthenia gravis. Proc Soc Exp Biol Med 1960;105:177 184 Aarli JA, Stefanson K, Marton LSG, Wollman RL. Patients with myasthenia gravis and thymoma have in their sera IgG antibodies against titin. Clin Exp Immunol 1990;82:284 288 Skeie GO, Mygland A, Aarli JA, Gilhus NE. Titin antibodies in patients with late onset myasthenia gravis: clinical correlations. Autoimmunity 1995;20:99104 Skeie GO, Lunde PK, Sejersted OM, Mygland A, Aarli JA, Gilhus NE. Myasthenia gravis sera containing antiryanodine receptor antibodies inhibit binding of [3H]-ryanodine to sarcoplasmic reticulum. Muscle Nerve 1998;21:329335 Mygland A, Vincent A, Newsom-Davis J, et al. Autoantibodies in thymoma-associated myasthenia gravis with myositis or neuromyotonia. Arch Neurol 2000;57:527531 Hoch W, McConville J, Helms S, Newsom-Davis J, Melms A, Vincent A. Auto-antibodies to the receptor tyrosine kinase MuSK in patients with myasthenia gravis without acetylcholine receptor antibodies. Nat Med 2001;7:365368 Evoli A, Tonali PA, Padua L, et al. Clinical correlates with anti-MuSK antibodies in generalized seronegative myasthenia gravis. Brain 2003;126:23042311 Vincent A, Bowen J, Newsom-Davis J, McConville J. Seronegative generalised myasthenia gravis: clinical features, antibodies, and their targets. Lancet Neurol 2003;2:99 106 Yang B, McIntosh KR, Drachman DB. How subtle differences in MHC class II affect the severity of experimental myasthenia gravis. Clin Immunol Immunopathol 1998;86: 4558 Vincent A, Newsom-Davis J. Acetylcholine receptor antibody characteristics in myasthenia gravis. I. Patients with generalized myasthenia or disease restricted to ocular muscles. Clin Exp Immunol 1982;49:257265 Limburg PC, The TC, Hummel-Teppel E, Oosterhuis H. Anti-acetylcholine receptor antibodies in myasthenia gravis. I.

30

SEMINARS IN NEUROLOGY/VOLUME 24, NUMBER 1

2004

73.

74. 75.

76.

77. 78.

79.

80.

81.

82.

Relation to clinical parameters in 250 patients. J Neurol Sci 1983;58:357370 Compston DAS, Vincent A, Newsom-Davis J, Batchelor JR. Clinical, pathological, HLA antigen, and immunological evidence for disease heterogeneity in myasthenia gravis. Brain 1980;103:579601 Vincent A, Palace J, Hilton-Jones D. Myasthenia gravis. Lancet 2001;357:21222128 Yi Q, Lefvert AK. Idiotype- and anti-idiotype-reactive T lymphocytes in myasthenia gravis. Evidence for the involvement of different subpopulations of T helper lymphocytes. J Immunol 1994;153:33533359 Link J, Navikas V, Yu M, Fredrikson S, Osterman PO, Link H. Augmented interferon-gamma, interleukin-4 and transforming growth factor-beta mRNA expression in blood mononuclear cells in myasthenia gravis. J Neuroimmunol 1994;51:185192 Hohlfeld R, Wekerle H. The thymus in myasthenia gravis. Neurol Clin 1994;12:331342 Sommer N, Willcox N, Harcourt GC, Newsom-Davis J. Myasthenic thymus and thymoma are selectively enriched in acetylcholine receptor-reactive T cells. Ann Neurol 1990;28: 312319 Schonbeck S, Padberg F, Hohlfeld R, Wekerle H. Transplantation of thymic autoimmune microenvironment to severe combined immunodeciency mice. J Clin Invest 1992;90: 245250 Zheng Y, Wheatley LM, Liu T, Levinson AI. Acetylcholine receptor alpha subunit mRNA expression in human thymus: augmented expression in myasthenia gravis and upregulation by interferon-gamma. Clin Immunol 1999;91:170177 Kaminski HJ, Fenstermaker RA, Abdul-Karim FW, Clayman J, Ruff RL. Acetylcholine receptor subunit gene expression in thymic tissue. Muscle Nerve 1993;16:13321337 Navaneetham D, Penn AS, Howard JFJ, Conti-Fine BM. Human thymuses express incomplete sets of muscle acetylcholine receptor subunit transcripts that seldom include the delta subunit. Muscle Nerve 2001;24:203210

83. Gronseth GS, Barohn RJ. Practice parameter: thymectomy for autoimmune myasthenia gravis (an evidence-based review): report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 2000;55: 715 84. Grob D, Arsura EL, Brunner NG, Namba T. The course of myasthenia gravis and therapies affecting outcome. Ann N Y Acad Sci 1987;505:472499 85. Kaminski HJ, Li Z, Richmonds C, Ruff RL, Kusner L. Susceptibility of ocular tissues to autoimmune diseases. Ann N Y Acad Sci 2003;998:362374 86. Ubogu EE, Kaminski HJ. Preferential involvement of extraocular muscle by myasthenia gravis. Neuroophthalmology 2001;25:219228 87. Leigh RJ, Zee DS. The Neurology of Eye Movements. Contemporary Neurology Series. Philadelphia: F.A. Davis; 1999:561 88. Porter JD. Extraocular muscle: cellular adaptations for a diverse functional repertoire. Ann N Y Acad Sci 2002;956: 716 89. Ruff RL. Na current density at and away from end plates on rat fast- and slow-twitch skeletal muscle bers. Am J Physiol 1992;262:C229C234 90. Ruff RL, Lennon V. End-plate voltage-gated sodium channels are lost in clinical and experimental myasthenia gravis. Ann Neurol 1998;43:370379 91. Khanna S, Richmonds C, Kaminski H, Porter J. Molecular organization of the extraocular muscle neuromuscular junction: partial conservation of and divergence from skeletal muscle prototype. Invest Ophthalmol Vis Sci 2003;44:1918 1926 92. Wang Z, Diethelm-Okita B, Okita D, Kaminski H, Howard J, Conti-Fine B. T-cell recognition of muscle acetylcholine receptor in ocular myasthenia gravis. J Neuroimmunol 2000; 108:2939 93. Porter JD, Khanna S, Kaminski HJ, et al. Extraocular muscle is dened by a fundamentally distinct gene expression prole. Proc Natl Acad Sci USA 2001;98:1206212067