10, N O . 1 MARCH 1990

SEMINAKS I N NEUROLOGY-VOLUME

Congenital Disorders of Neuromuscular Transmission

Congenital neuromuscular transmission defects are uncommon but challenging diseases for the clinical neurologist. Some of these disorders entail severe or even life-threatening disability; some are treatable, but effective therapy requires precise diagnosis, and the precise diagnosis is often difficult to make. The diagnosis rests on the combination of clinical data, the electromyogram (EMG), and additional studies that may include microelectrode analysis of neuromuscular transniission, ultrastructural and cytochemical studies of the neuromuscular junction (NMJ), and biochemical studies on muscle specimens. Understandably, these studies often depend on the collaboration of several investigators. In each congenital myasthenic disorder, a genetically determined primary abnormality affects neuromuscular transmission directly or causes secondary derangements that eventually affect transmission. In either case, the safety margin of neuromuscular transmission is eventually compromised by one or more specific mechanisms. Therefore a clear grasp of the factors that affect the safety margin is essential for understanding these disorders.

SAFETY MARGIN OF NEUROMUSCULAR TRANSMISSION Only basic concepts and definitions are presented here. For a more complete discussion, the reader is referred to recent reviews on the NMJ.1-3 In striated muscle, the NMJ consists of a nerve terminal separated from the postsynaptic region by --

the synaptic space. Acetylcholine (ACh) is stored in quanta1 packets (6000 to 10,000 molecules per packet) in synaptic vesicles in the nerve terminal. The vesicles release ACh into the synaptic space by exocytosis. The basal lamina of the synaptic space contains acetylcholinesterase (AChE), distributed evenly at a density of about 2500 sites/pm2. The postsynaptic region contains junctional folds. ACh receptor (AChR) molecules are packed on the terminal expansions of the folds at a density of about 10' sites/~m"The binding of two ACh molecules to an AChK molecule opens the AChK ion channel allowing the inward flux of cations along their electrochemical gradient. 'l'he cation flow through the channel is also regulated by dimensions of the channel pore and the distribution of charged amino acid residues in the channel vestibule. After a finite interval, which is usually of the order of a few milliseconds, the channel closes and ACh dissociates from AChR. T h e open time of the ion channel and the current that flows through it can be evaluated by two special techniques: (1) analysis of endplate current fluctuations induced by iontophoretically applied ACh (noise analysis); and (2) direct recording of currents flowing through single ion channels by means of a patch-clamp. In the resting state, single ACh quanta are randomly released into the synaptic space. Thc high local ACh concentration saturates all AChk sites in the nearby basal lamina so that most ACh molecules can reach postsynaptic AChK. The AChR packing density is so high that ACh needs to dif fuse only 0.3 pm along the top and 0.3 pm dowr~ along the junctional folds before it meets all tht. AChR it can saturate. The resultant depolarizatior~ --

-

~

-

Department of Neurology and Neuromuscular Research Laboratory, Mayo Clinic, Kochester, Minnesota. Work in the author's laboratory was supported by National Institutes of Health grant NS6277 and by a research grant from the Muscular Dystrophy Association. Reprint requests: Dr. Engel, Mayo Clinic, Rochester, M N 55905. Copyright O 1990 by Thieme Medical Publishers, Inc., 381 Park Avenue South, New York, NY 10016. All rights reserved.

Downloaded by: National University of Singapore. Copyrighted material.

Andrew G. Engel, M.D.

CONGENITAL DISORDERS O F NEUROMUSCU1,AR TRANSMISSION-ENGEL

low (2 to 3 Hz) frequency; and (2) tetanic stimulation results in transient improvement and then a worsening of the defect.

FAMILIAL INFANTILE MYASTHENIA CLINICAL ASPECTS

The disease is transmitted by autosomal recessive inheritance. It presents in early infancy or childhood. The typical history is one of fluctuating ptosis since birth; poor suck and cry, feeding difficulty and secondary respiratory infections during infancy; and episodic crises precipitated by fever, excitement, or vomiting throughout infancy and childhood. During crises all symptoms worsen. Apnea from respiratory muscle weakness during a crisis can cause sudden death or anoxic brain injury. During childhood, patients may appear normal or have only minimal weakness between crises, but weakness can be induced by exercise. With increasing age, the crises become less frequent. After age 10 years, some patients complain of easy fatigability only on sustained exertion; others have mild to moderate weakness of cranial, limb, and respiratory muscles even at rest, resembling patients with mild to moderately severe autoimmune myasthenia gravis (MG). The deep tendon reflexes remain normally active. There is no loss of muscle bulk, and a permanent myopathy does not oc-

ELECTROPHYSIOLOGZC FEATURES

The EMG may not show any abnormality in muscles that are not weak when examined. In muscles that are weak, testing may reveal nonspecific features of neuromuscular transmission defects: (1) an abnormal fluctuation in the shape or size of the motor unit potentials during voluntary effort; (2) a progressive decrement in the amplitude of the compound muscle action potential evoked by lowfrequency (2 to 3 Hz) stimulation; (3) improvement of the decremental response immediately after a brief period of maximal voluntary exercise, followed within a minute by an increased decremental response; and (4) single fiber EMG abnormalities consisting of an abnormally prolonged interpotential interval between two muscle fibers in an activated motor unit (increased jitter), and failure of a proportion of the impulses to generate an action potential at one of the two fibers (blocking). In those patients whose muscles are not weak, the weakness and the EMG abnormalities can be induced in some, but not all, muscles either by exercise or by repetitive stimulation at 10 Hz for a

Downloaded by: National University of Singapore. Copyrighted material.

of the muscle fiber and the current that flows into the fiber are known as miniature endplate potential (MEPP) and current (MEPC), respectively. When ACh dissociates from AChR, it is hydrolyzed by AChE to choline and acetate. Choline is taken up by the nerve terminal and is reutilized for ACh synthesis. The MEPP amplitude depends on the number of ACh molecules in the quantum, the number of available AChRs, the geometry of the synaptic space, the average depolarization generated by the opening of an AChR ion channel, and is inversely related to the muscle fiber diameter. The MEPP duration depends on the average channel open time, the functional state of AChE, and the cable properties of the muscle fiber surface membrane. The MEPC is independent of the fiber diameter or the cable properties of the fiber membrane; its amplitude and duration are otherwise governed by the same factors that affect the MEPP. Depolarization of the nerve terminal by nerve impulse opens voltage-sensitive calcium channels in the presynaptic membrane. The calcium influx increases the probability of synaptic vesicle exocytosis, and exocytosis of numerous vesicles occurs adjacent to active zones in the presynaptic membrane. The voltage-sensitive calcium channels are associated with regularly arrayed large membrane particles in the active zones. On repetitive nerve stimulation, the presynaptic release sites, and consequently the postsynaptic AChR sites that become saturated with ACh, vary from impulse to impulse. This prevents unduly prolonged exposure of AChR to ACh that would cause desensitization of AChR, that is, closure of its ion channel while it is liganded by ACh. The quanta released by a nerve impulse generate an endplate potential (EPP). The EPP amplitude depends on the MEPP amplitude and the number of quanta released by nerve impulse (m). The value of m depends on the probability of release (p) and the number of quanta readily available for release (n) according to the relationship m = np. The safety margin of neuromuscular transmission is defined as the difference between the actual EPP amplitude and the EPP amplitude required to trigger the muscle fiber action potential. Repetitive stimulation results in a frequencydependent depression of the EPP amplitude, and of the safety margin, to a certain plateau. The decrease is mainly due to a decrease in n. Repetitive stimulation also can facilitate transmitter release by increasing p, n, or both. The temporal profiles of the opposing processes are such that: (1) a defect of neuromuscular transmission is most readily detected by a train of 5 to 10 stimuli delivered at a

SEMINARS IN NEUROLOGY

PATIENTS

I

VOLUME 10, NUMBER 1 MARCH 1990

NORMAL CONTROL

0

NO ADDITION

0

HEMICHOLINIUM

TIME ( M I N I Figure 1. Familial infantile myasthenia (FIM). Effect of 10 Hz stimulation on the amplitude of the evoked compound muscle action potential in external intercostal muscle strips in vitro in two patients (left panel) and in a normal subject (right panel). In the presence of 1 mg/dl of hemicholinium-3,the evoked action potential in normal muscle (open circles in right panel) declines as rapidly as the evoked action potential in FIM muscle in the absence of hemicholinium. (From Mora et a1.12 Reprinted with permission.)

few minutes.1° T h e EMG decrement, when it is present, can be corrected by the cholinesterase inhibitor edrophonium." In vitro studies by Lambert on external intercostal muscles have elucidated the electrophysiologic basis of the disorder.1° l 2 Stimulation of small muscle bundles at 10 Hz resulted in an abnormal decrease of the amplitude of evoked compound muscle action potential (Fig. 1) and of' the EPP. T h e MEPP amplitude was normal in rested muscle but decreased abnormally after 10 Hz stimulation for 5 minutes. These findings are unlike those of autoimmune MG. From these findings one can infer that in familial infantile myasthenia (FIM) the safety margin of neuromuscular transmission is compromised by an abnormal decrease of the EPP due to an abnormal decrease of the MEPP. The decrease of the MEPP amplitude in the course of prolonged stimulation from an initially normal to an abnormally low level suggests a progressive decrease in the ACh content of the synaptic vesicles. An alternative explanation might be an abnormal desensitization of postsynaptic AChK by physiologic amounts of ACh released in the course of stimulation. However, if the latter were true, then cholinesterase inhibitors should worsen the defect, but in fact they improve it. The idea that synaptic vesicles become depleted of ACh during repetitive stimulation is supported by the observation that the response of FIM muscle to elec-

trical stimulation is like that of normal muscle exposed to hemi~holinium,'"'~ an inhibitor of choline uptake by the nerve terminal (Fig. 1). On this basis, FIM could be caused by a defect in: (1) the facilitated uptake of choline by the nerve terminal; (2) ACh resynthesis by choline acetyltransferase; or (3) the transport of ACh molecules into the synaptic vesicles. MORPHOLOGZC OBSERVATZONS

Muscle biopsy specimens show no histochemical abnormality. In particular, the usual checkerboard distribution of histochemical fiber types is preserved. AChE-reacted sections demonstrate no abnormality of the NMJ. There are no immune complexes, (immunoglobulin g or complement) at the NMJ. The nerve terminals show normal immunoreactivity for choline acetyltransferase.' This, however, does not exclude the possibility of a mutation that alters the kinetic properties of the enzyme. On electron microscopy, the NMJ appears normal on simple inspection (Fig. 2A). There is no morphometric abnormality in the size o r mitochondrial content of the nerve terminal, the postsynaptic area of folds and cleft, or the postsynaptic membrane length per unit area. Ultrastructural localization of AChR with peroxidase-labeled a-bungarotoxin shows normal abundance and distribu-

Downloaded by: National University of Singapore. Copyrighted material.

0

Figure 2. Familial infantile myasthenia. Electron micrographs of NMJ regions in external intercostal muscle. The specimen shown in A was prepared for electron microscopy by conventional methods. The nerve terminal and postsynaptic region appear normal. The specimen shown in B was incubated with peroxidase-labeled a-bungarotoxin before fixation. The black reaction product for peroxidase on the terminal expansions on the junctional folds demonmembrane. (A: x 28,200; B: x 16,000.) (From Engel AG. J strates a normal distribution of AChR on the ~ostsyna~tic Child Neurol 1988;3:233-46. Reprinted with permission.)

tion of AChK on the terminal expansions of the junctional folds (Fig. 2B). Quantitative estimates of the length of the postsynaptic membrane reacting for AChR per nerve terminal (AChR index) and radioimrnunochemical estimates of the AChR content of intercostal muscle are also normal.'" Recently, Mora et all2 searched for a morphologic correlate of the failure of neuromuscular transmission in FIM. In previous studies, Jones and Kwanb~nbumpen'"'~have shown that prolonged stimulation of the isolated rat diaphragm

exposed to hemicholinium results in an abnormal decrease of the MEPP amplitude and of synaptic vesicle size. Thus, if in FIM the abnormal decrease of the MEPP amplitude on nerve stimulation were caused by impaired choline uptake by the nerve terminal, or by another defect in ACh synthesis, then one could expect an abnormal decrease in synaptic vesicle size. To test this notion, intercostal muscles from three patients with FIM and three control subjects were studied before and after 10 Hz stimulation for 10 minutes. In the patients. but

Downloaded by: National University of Singapore. Copyrighted material.

CONGENITAL DISORDERS OF NEUROMUSCULAR TRANSMISSION-ENGEL

SEMINARS IN NEUROLOGY VOLUME 10, NUMBER 1 MARCH 1990

TREATMENT

The muscle weakness, when present, responds well to small o r modest doses of anticholinesterase d r u g s Some patients are asymptomatic or have only minimal weakness except during crises and require anticholinesterase drugs on an emergency basis only. Parents of affected children must be in-

EFFECT OF

STIMULATION ON

SYNAPTIC

VESICLE

VOLUME

AFTER STIMULATION

0

PATIENTS

CONTROLS

Figure 3. Familial infantile myasthenia. Mean synaptic vesicle volumes in deep and superficial regions of nerve terminals before stimulation and after 10 Hz stimulation for 10 minutes. Open and closed symbols indicate patients and controls, respectively. (From Mora et a1.I2 Reprinted with permission.)

doctrinated to anticipate sudden worsening of the weakness and possible apnea with febrile illnesses, excitement, or overexertion. T h e parents also must be familiar with the use of a hand-assisted ventilatory device and should be able to administer appropriate doses of prostigmine intramuscularly during crises. Patients with a febrile illness and a previous history of crisis should be hospitalized for close observation and for ventilatory support as needed.

CONGENITAL ENDPLATE ACETYLCHOLINESTERASE DEFICIENCY CLZNZCAL ASPECTS

This disorder was first described by Engel et alZ1in a boy with lifelong symptoms; recently, Walls and coworkers2*observed the same disorder in two sisters. Therefore the disease is transmitted by autosomal recessive inheritance. Weakness, abnormal fatigability, and a decremental EMG are present since birth. Poor suck, cry, and episodes of respiratory distress occur in infancy. Motor milestones are delayed. The symptoms remain relatively static until the end of the first decade when the functional disability appears to worsen. T h e weakness

Downloaded by: National University of Singapore. Copyrighted material.

not in the controls, neuromuscular transmission failed during stimulation (Fig. 1). Synaptic vesicle densities (no./pm2)and diameters were separately analyzed in a superficial 200 nm wide zone adjacent to the presynaptic membrane, where vesicles are positioned for release, and in the remaining deeper part of the nerve terminal from which vesicles may be mobilized for release. In both patients and control subjects, stimulation had a similar effect on the density of the superficial and of the deep synaptic vesicles, decreasing the former by 20% and the latter by 30 to 50% below the initial value. For a given site, synaptic vesicle densities did not differ significantly between patients and control subjects either before or after stimulation. However, and unexpectedly, the size of the synaptic vesicles was significantly smaller in rested FIM than control muscles. Furthermore, after stimulation, when the MEPP amplitude was markedly reduced in FIM, the FIM vesicles increased or did not change in size. By contrast, after stimulation the control synaptic vesicles decreased or did not change in size (Fig. 3). There is no simple explanation for the lack of correlation between synaptic vesicle size and the MEPP amplitude in FIM. From current knowledge derived from studies of Torpedo electric organ synaptic vesicles, ACh is taken up by the synaptic vesicle by a proton-driven ACh translocase that exchanges protons in the vesicles for cytosolic ACh.I7-I' The proton accumulation in the vesicles, in turn, depends on an adenosine triphosphatasedriven proton pump. The synaptic vesicles contain not only ACh and protons, but also a relatively high concentration of adenosine triphosphate, guanosine triphosphate, Ca2+, Mg2+, and a prot e o g l y ~ a n . ' These ~ . ~ ~ observations imply that: (1) the number of ACh molecules in the vesicles is not the only determinant of vesicle volume; and (2) a defect in any of the mechanisms that regulate the concentrations of any of the osmotically active substances in the synaptic vesicles could affect the vesicle volume. Thus, the fact that in FIM the synaptic vesicles are smaller than normal in rested muscle and increase paradoxically in size after stimulation suggests a defect in synaptic vesicle metabolism, but the character of the defect remains undefined.

Downloaded by: National University of Singapore. Copyrighted material.

CONGENITAL DISORDERS O F NEUROMUSCULAR TRANSMISSION-ENGEL

Figure 4. Two sisters with NMJ AChE deficiency. The lordosis and scoliosis in the older sister (center panel) and the lordosis in the younger sister (right panel) became worse during standing for 15 seconds. Lateral views show maximal arm abduction after 15 seconds. At that time, the younger sister was unable to hold her head erect (right panel).

affects the facial, cervical, axial, and limb muscles. Ophthalmoparesis was present in the first patient but not in the two sisters. The axial muscles are, selectively, severely involved, so that on standing the patient may show increasing lordosis and scoliosis after a few seconds (Fig. 4). The deep tendon reflexes were reduced in the first patient but were preserved in the two sisters. The symptoms are refractory to anticholinesterase drugs.

In vitro microelectrode studies on intercostal muscles demonstrate an abnormal prolongation of the decay phase of the MEPP and EPP. The addition of neostigmine to the bath has no additional effect on the potentials. The MEPP are of normal amplitude. The quanta1 content of the EPP is reduced, and the decrease is due to a reduction in the number of immediately releasable quanta. MORPHOLOGZC AND BZOCHEMZCAL OBSERVATZONS

ELECTROPHYSZOLOGZC FEATURES

The EMG shows a decremental response at 2 Hz stimulation in all muscles tested. Furthermore, there is a repetitive muscle action potential response to a single nerve stimulus (Fig. 5 ) . The repetitive response disappears at stimulation frequencies greater than 0.2 Hz or with mild activity. Therefore it can be overlooked unless a well-rested muscle is tested by single shocks.

Conventional histologic studies of muscle are normal. The basic abnormality is total absence of AChE from the NMJ: no enzyme activity can be demonstrated by light microscopic or electron microscopic cytochemistry (Fig. 6), and no immunoreactivity for AChE is detected by polyclonal and several monoclonal AChE antib~dies.~.'~ Quantitative electron microscopy reveals a statistically significant decrease in nerve terminal size 17

V O L U M E 10, N U M B E R 1 MARCH 1990

when the fiber recovers from the refractory period of the first action potential, the abnormally long EPP can evoke one or more additional muscle fiber action potentials. Consequently, a single supramaximal stimulus applied to a motor nerve can evoke two or more compound muscle action potentials. The reduced number of readily releasable quanta is adequately accounted for by the decrease in nerve terminal size and in presynaptic membrane length. However, the smallness of the nerve terminals is not as constant as the AChE deficiency. For example, at some NMJs the nerve terminals are normal in size, whereas AChE is totally absent from all NMJs. This suggests that the AChE deficiency, and not the smallness of the nerve terminal, is the primary defect. AChR is lost from NMJ at which there is focal degeneration of the junctional folds. 'l'he degenerative changes at the NMJ, in turn, can be accounted for by the ACh e x ~ e s s . ~However, ~.*~ the ACh excess is mild because ACh release is limited by the small size of the nerve terminals. The safety margin of neuromuscular transmission is compromised by lack of releasable ACh quanta and, to a lesser extent, by AChR deficiency. Figure 5. NMJ AChE deficiency. Repetitive compound muscle action potential recorded from extensor digitorum brevis muscle (upper panel) and thenar muscles (lower panel) during 2 Hz stimulation of peroneal and median nerve, respectively. The second response (arrow) decrements more rapidly than the first response.

(Figs. 7, 8), and in the length of the presynaptic membrane available for ACh release. At some NMJ, there is degeneration of the junctional folds and also of the underlying muscle fiber regions. The total number of AChR, determined in one patient with '251-labeleda-bungarotoxin, was mildly reduced, falling 23% below the lower range of normal. The total muscle AChE content is reduced. This is due to absence of the enzyme from the NMJ and to a concomitant decrease of extrajunctional AChE. The kinetic properties of the residual extrajunctional AChE are normal. Erythrocyte AChE activity and kinetic properties are also normal. PATHOGENETIC MECHANISMS

18

Because AChE is absent from the NMJ, AChAChR interaction and the duration of the EPP and MEPP are prolonged. Because the amplitude of the prolonged EPP remains above the threshold needed to evoke a muscle fiber action potential

SLOW-CHANNEL SYNDROME CLINICAL ASPECTS

This syndrome was described in 1982 by Engel and coworkers.25An additional report has been published by Oosterhuis et a1 in 1987." The disease is transmitted by an autosomal dominant gene with high penetrance and variable expressivity. Sporadic cases also occur. T h e age of onset, the initial and eventual pattern of muscle involvement, the rate of progression, and the degree of weakness and fatigability vary from case to case. The disease may present in infancy, childhood, or adult life. It progresses gradually or in an intermittent manner, remaining quiescent for years or decades between periods of worsening. Typically, there is selectively severe involvement of cervical, scapular, and finger extensor muscles (Figs. 9, 10); mild to moderate weakness of the eyelid elevators and limitation of ocular movements, with only occasional double vision; and variable involvement of masticatory, facial, and other upper extremity, respiratory, and trunk muscles. The lower limbs tend to be spared, or may be less severely affected than the upper ones. The clinically affected muscles are weak, atrophic, and fatigue abnormally. The weakness and fatigability can fluctuate, but not as rapidly as in autoimmune MG. T h e deep tendon re-

Downloaded by: National University of Singapore. Copyrighted material.

SEMINARS IN NEUROLOGY

Figure 6. Electron cytochemical localization of AChE in NMJ AChE deficiency. A: Control endplate is greatly overreacted after incubation for 30 minutes at room temperature. B: Patient's endplate shows no reaction after incubation for 1 hour at room temperature. (A: x 8900; B: x 20,500.) (From Engel et aL2' Reprinted with permission.)

Figure 7. NMJ AChE deficiency. Small nerve terminal (N) covers only a fraction of the postsynaptic region. Note numerous endocytotic vesicles in the junctional folds (arrow). ( x 11,500.)

Downloaded by: National University of Singapore. Copyrighted material.

CONGENITAL DISORDERS OF NEUROMUSCULAR TRANSMISSION-ENGEL

SEMINARS IN NEUROLOGY

PATIENT

0

I

2

3

4

5

6

7

8

9 1011121314

Nerve Terminal Area (sq um)

Figure 8. Frequency distribution of the nerve terminal area in 13 control subjects (upper panel) and in a patient with NMJ AChE deficiency (lower panel).

flexes are usually normal, but can be reduced in severely affected limbs. T h e AChR antibody test is negative. AChE inhibitors are ineffective. Oosterhuis et a1"j found flunarizine, a calcium channel blocker, ineffective in one patient. ELECTROPHYSZOLOGZC FEATURES

As in congenital endplate AChE deficiency, single nerve stimuli evoke repetitive compound muscle action potentials in all muscles (Fig. 1 1, upper left panels). The consecutive potentials occur at 5 to 10 msec intervals, each smaller than the preceding one, and disappear after a brief voluntary contraction. Two nerve volleys 1 to 3 msec apart increase rather than inhibit the amplitude of the second action potential, indicating that it is not produced by an axon reflex. A decremental EMG response at 2 to 3 Hz stimulation is present, but only in clinically affected muscles (Fig. 11, upper right panel). The motor unit potentials fluctuate in shape and amplitude during voluntary activity. In vitro microelectrode studies indicate that the decay of the intracellularly recorded MEPP and EPP is markedly prolonged in all muscles (Fig. 11, lower panels) and the duration of the potentials is further increased by AChE inhibitors. T h e amplitude of the MEPP is significantly reduced in some but not all muscles, and the decrease is

greater in the more severely affected muscles. The quanta1 content of the EPP is in the normal range. The prolonged duration of the MEPP and EPP is not due to an alteration in the cable properties of the muscle membrane because: (1) the duration of the extracellularly recorded MEPP, which is independent of the cable properties of the membrane and reflects the duration of the miniature endplate c ~ r r e n t , is ~ 'also markedly prolonged;25and (2) in one patient the time constant of the ACh-induced voltage noise at the NMJ was abnormally pro10nged.'~ MORPHOLOGZC OBSERVATIONS

Light microscopic histochemical studies show type 1 fiber predominance, isolated or small groups of atrophic fibers of either histochemical type, tubular aggregates, and vacuoles in fiber regions near motor endplates. Other biopsy specimens show abnormal variation in fiber size, variable fi-. ber splitting, and, in some instances, mild to mod-. erate increase of endomysial or perimysial connec-. tive tissue. AChE activity is present at all endplates. In the more severely affected muscles, the configuration of the endplates is often abnormal, with multiple, small, discrete regions distributed over an extended length of the muscle fiber. Focal calcium deposits were demonstrated at the NMJ by Engel et aP5 in one of four cases in which this test was performed. On electron microscopy, at many NMJs the junctional folds contain myriad pinocytotic vesicles and labyrinthine membranous networks. In the more severely affected muscles, the junctional folds are frequently degenerating, causing widening of the synaptic space and accumulation of electron dense debris (Fig. 12). Some of the highly abnormal postsynaptic regions are denuded of their. nerve terminals (Fig. 13). Unmyelinated nerve sprouts appear near some NMJs. The intramuscular nerves are normal. AChR is reduced at thosr. NMJ that show degeneration of the junctional folds. Degenerative changes also occur in the junctional sarcoplasm and in nearby fiber regions. These involve the nuclei, myofibrils, mitochondria (Fig. 13), and sarcoplasmic reticulum. In some muscle fibers there are larger vacuoles near the NMJ (Fig. 13). These contain amorphous or granular mate.rial or fragmented membranes and are limited by membranes of transverse tubular origin o r by pro.liferating transverse tubular system networks. Morphometric reconstruction of the NMJ shows a 29 to 43% decrease of nerve terminal size and ;t 25 to 37% increase in synaptic vesicle density. The

Downloaded by: National University of Singapore. Copyrighted material.

CONTROLS

VOLUME 10, NUMBER 1 MARCH I990

postsynaptit membrane length and density are significantly reduced due to degcricration of the junctional folds. BIOCHEMICAL STUDIES OF AChE

Although AChE inhibitors further increased the duration of the MEPP in vitro, and although AChE was present at the endplates by cytochernical criteria, a partial deficiency o r kinetic abnormality of AChE could still account for the prolonged MEPP. This possibility was excluded by demonstrating that the activity and K,,, of AChE were norrrlal iri r~iusclein this syndrome.'3ince the catalytic subunit in all firms of muscle AChE is identical,'"he kinetic properties of total muscle AC:hE are a valid measurc of the kinetic properties of endplate AChE.

Figure 10. Slow-channel syndrome. Patient attempting wrists and fingers as shown by examiner (with to sleeve). Note atrophy of patient's forearm muscles. (From Engel et aLZ5Reprinted with permission.)

Downloaded by: National University of Singapore. Copyrighted material.

Figure 9. Slow-channel syndrome in two generations. From left to right: a 54-year-old patient with his 48-year-old brother and 29-year-oldson. Note atrophy of shoulder and forearm muscles and mild asymmetric ptosis in the brothers. The son shows scoliosis, moderate to marked atrophy of cervical, shoulder, arm, forearm, and torso muscles, anda pronounced ptosis. (From Engel AG et aL8 Reprinted with permission.)

Single stimulus ulnar nerve h v ~ ohenar t response

A

Normal

Repetit~vestimulation suprascapular nerve infrasp~natusresponse 2 /s

Pattent

Figure 11. Upper left panels: Stimulus-linked repetitive compound muscle action potential in the slow-channel syndrome. Upper right panel: Decremental response of infraspinatus muscle action potential during 2 Hz stimulation of suprascapular nerve. Lower panels: The duration and decay time of the MEPP are longer in patient than in normal control. (From Engel at Reprinted with permission.) PATHOGENETIC MECHANISMS

The prolonged EPP can be attributed to the prolonged MEPP. Since AChE is intact by physiologic, histochemical, and biochemical criteria, and since the prolonged MEPP cannot be attributed to abnormal cable properties of the muscle fiber plasma membrane, the prolonged MEPP is caused by a prolonged open time of the AChR ion channel. As in congenital endplate AChE deficiency, the repetitive muscle action potential can be explained by the prolonged EPP. The fact that the repetitive muscle action potential response is present in all muscles indicates that the AChR ion channel abnormality is ubiquitous and represents the primary disturbance. Thus, the weakness, wasting, and fatigability that appear in selected muscles at variable intervals are secondary phenomena. The prolonged endplate currents result in an abnormally increased cation flux into the junctional folds and nearby muscle fiber regions. Because a fraction of the current is carried by calcium,29~30 transient or permanent calcium excess is likely to occur in the junctional folds and nearby fiber regions. The deleterious effects of the focal calcium excess, which include activation of intracellular proteases and stimulation of membranebound ph~spholipases,"~~~ can readily explain the focal degeneration of the junctional folds, the loss of AChK from the folds, and the endplate rnyopathy. Furthermore, some of the findings are similar to those noted in mouse muscle exposed to carbachol, a cholinergic agonist, and the carbachol-

induced changes can be prevented by exclusion of calcium from the extracellular fluid." It is also possible that AChR synthesis is inhibited, as in cultured muscle exposed to cholinergic a g ~ n i s t s . ~ ~ The decrease of AChR and the NMJ explains the reduced MEPP amplitude and the impaired safety margin of neuromuscular transmission. The AChR ion channel is also slow in noninnervated muscle fibers and at newly formed endplates." Thus, hypothetically, the slow-channel syndrome could be due to a developmental failure of the slow to fast conversion of the ion channel. It is now known that this conversion depends on the replacement of the gamma by an epsilon subunit in AChR.J7-3Vhereforeexpression of the gamma instead ofthe epsilon subunit in slow-channel syndrome AChR would be evidence for abnormal developmental regulation. However, in a recent study we found immunoreactivity for the epsilon but not for thegamma subunitat theslow-channelNM-J(Engel, Gu, Hall: Unpublished observations). Therefore the most plausible explanation is a mutation in an AChR subunit that hinders the closure of the AChR ion channel.

CONGENITAL NMJ AChR DEFICIENCY This syndrome is less well characterized and more heterogeneous than the preceding syndromes. In three patients described by Vincent et al in 198140 (cases 2, 4, 5), NMJ AChE was preserved. NMJ AChR, estimated from the number of abungarotoxin binding sites, was reduced. In these patients the AChR deficiency could be due to decreased synthesis, reduced membrane insertion, or accelerated degradation of AChR. In a patient investigated by Lecky et al in 198641 (case 2). the NMJ appeared elongated on light microscopy. but NMJ ultrastructure was normal. The number of NMJ a-bungarotoxin binding sites was markedly reduced. No electrophysiologic studies were done in this case. In anorrler patient studled by Vincent et a14" (case 3), the amount of a-bungarotoxin bound to the NM.1 was reduced only after prolonged washing. This may indicate a reduced affinity of AChR for a-bungarotoxin and, by inference, for ACh. In four other patients, NMJ AChR deficient) and small MEPP were associated with poorly developed junctional folds and a paucity of secondar) synaptic cleft^.^'-^^ In two patients the disordel presented at birth with arthrogryposis, respirator) distress, feeding difficulty, and mild ptosis. Subsequently, they had delayed motor development, cy. anotic episodes, reduced exercise tolerance, and exacerbations with fever."," A milder form of thc

Downloaded by: National University of Singapore. Copyrighted material.

SEMINARS I N NEUROLOGY VOLUME 10, NUMBER 1 MARCH 1990

Downloaded by: National University of Singapore. Copyrighted material.

Figure 12. Intercostal NMJ in the slow-channel syndrome. Junctional folds have degenerated in the region imaged at left. Highly electron dense debris marks position of preexisting folds (asterisk). Presynaptic membrane facing degenerated folds is partially covered by Schwann cell. ( x 20,100.) (From Engel at al." Reprinted with permission.)

syndronle was recognized in two adult sibs, a 64year-old man and his 58-year-old sister."' Both had ptosis since early childhood, exercise intolerance, and gradual worsening of synlptorns in adult life. This disorder is associated with a decrernental EM(; response at low frequencies of stimulation and is responsive to AChE inhibitors. Table 1 lists the syndromes discussed u p to now and the f'actor o r factors that reduce the safety margin of' neuromuscular transmission.

PARTIALLY CHARACTERIZED SYNDROMES AChR DEFICIENCY AND INCREASED AFFINITY FOR d-TUBOCURARINE

A patient studied by Morgan-Hughes et a14' showed a slight decrease in NMJ AChK and a tenfold increase in the affinity of AChK for d-tubo-

curai-ine. T h e manner in which the latter abnormality contributed to the deftct of neuromuscular transmission was not defined. DECREASED MEPP AMPLITUDE WITHOUT AChR DEFICIENCY

In a patient studied by Vincent et al"" (case l ) , the MEPP amplitude was low but the AChK content o f t h e NMJ was normal. This could be d u e to decreased conductance of the AChR ion channel, decreased number of ACh molecules per quantum, o r reduced affinity of AChR for ACh. Specific studies to distinguish between these possibilities were not carried out. In a patient investigated by Lecky et alz' (case l ) , a reduced MEPP amplitude was associated with normal a-bungarotoxin binding to the postsynaptic membrane, endplate structure, d-tubocurarine affinity, ion channel properties, and passive membrane properties; the ef-

23

VOI.IJME. 10, N I I R 1 K k . K I

MAK(,'H 1990

Figure 13. NMJ in finger extensor muscle in the slow-channel syndrome. A highly degenerate postsynaptic region denuded of its nerve terminal (asterisk) overlies a fiber region that shows loss of mitochondria, focal myofibrillar degeneration, a disintegrating nucleus (N), and a large, membrane-bound vacuole (V). ( x 7,600.) (From Engel et al.*= Reprinted with permission.)

fectiveness of ACh to open ion cha~inels,however, was rhoughr ro be reduced. POSSIBLE DEFECT IN ACh SYNTHESIS, MOBILIZATION, O R STORAGE

Iri one irifarit with a corigeriital niyastlienic syndrome there was marked EMG decrement at all stirriulatiori frequericies arid rrioderate to rnarked f : acdltation 15 seconds after 50 Hz stimulation. Single nerve stimuli did not evoke repetitive c:ornpound muscle action potentials and AChE and AChR were normally abundant at the NMJ. Ultrastructural studies of the NM.7 suggested a slight decrease in the mean synaptic vesicle diameter. Ariticholiriestcrascs arid guanidinc wcrc irieffective. Although in vitro electrophysiologic studies of neuomuscular transmission were not carried out, the observations were thought to bc consistent with a dcfect in ACh synthesis, ~nobilization,or storage. I" '

FAMILIAL LIMB -GIRDLE MYASTHENIA

This is an autosornal recessive ~ y n d r o m e . ' ~ ~ ~ ' Weakness of limb-girdle muscles and easy fatigability appear during childhood or in the teens. Ocular and other cranial rnuscles arc riot affected. The symptoms respond to anticholinesterase drugs but not to pr-ednisone. E M G studies show a decrerricrltal response. Joint contractures, cardiac repolarization defect, type 1 fihcr atrophy, and abnormal electrical irritability of' the muscle fibcrs wcre also rioted in one family."" Histochemical studies demonstrated tubular aggregates in the muscle fibers, but the positiori of the aggregates relative to the NMJ was not established. Iletailed morphologic studies of the NMJ, in vitro electrophysiologic studies of neuromuscular transrnissiori, and measurements of NMJ AChK are riot available iri this syndrome.

Downloaded by: National University of Singapore. Copyrighted material.

SEMINAKS I N NI..UKOI.OG\i

Table 1. Factors Affecting Neuromuscular Transmission in Congenital Myasthenic Syndromes* Factor(s) Reducing the Safety Margin of Neuromuscular Transmission

Syndrome Familial infantile myasthenia

Reduced ACh resynthesis; abnormal decrease of the MEPP during stimulation

Congenital AChE deficiency

Small nerve terminals; small rn due to small n. Also mild AChR deficiency from focal degeneration of junctional folds; relatively small MEPP

Slow-channel myasthenic syndrome

Long open time of AChR ion channel; focal postsynaptic calcium excess. In clinically affected muscles focal degeneration of junctional folds with loss of AChR and small MEPP

Congenital AChR deficiency

Decreased AChR on simplified junctional folds; small MEPP

*AChE: acetylcholinesterase; AChR: acetylcholine receptor; MEPP: miniature endplate potential; EPP: endplate potential; m: quantal content of the endplate potential; n: number of readily releasable quanta; p: probability of quantal release.

REFERENCES 1. Salpeter MM, ed. T h e vertebrate neuromuscular junction. New York: Alan R. Liss, 1987 2. Magleby KL. Neuromuscular transmission. In: Engel AG, Banker BQ, eds: Myology. New York: McGrawHill, 1986;393-4 18 3. Engel AG. T h e neuromuscular junction. In: Engel AG, Banker B y , eds: Myology. New York: McGraw-Hill, 1986;209-54 4. Greer M, Schotland M. Myasthenia gravis in the newborn. Pediatrics 1960;26:101-8 5. Conomy JP, Levisohn M, Fanaroff A. Familial infantile myasthenia gravis: a cause of sudden death in young children. J Pediatr 1975;87:428-9 6. Robertson WC, Chun RWM, Kornguth SE. Familial infantile myasthenia. Arch Neurol 1980;37: 117-9 7. Gieron MA, Korthals JK. Familial infantile myasthenia gravis. Report of three cases with follow-up until adult life. Arch Neurol 1985;42:143-4 8. Engel AG, Lambert EH, Mulder DM, Gomez MR. Recently recognized congenital myasthenic syndromes: (A) End-plate acetylcholine (ACh) esterase deficiency. ( 8 ) Putative abnormality of the ACh induced ion channel. (C) Putative defect of ACh resynthesis or mobili7.ation-clinical features, ultrastructure and cytochemistry. Ann NY Acad Sci 1981;377:614-39 9. Engel AG. Myasthenic syndromes. In: Engel AG, Banker BQ, eds: Myology. New York: McGraw-Hill, 1986; 1955-90 10. Engel AG, Lambert EH. Congenital myasthenic syndromes. Electroencephalogr Clin Neurophysiol Suppl 1987;39:91-102

11. Hart Z, Sahashi K, Lambert EH, et al. A congenital, familial, myasthenic syndrome caused by a presynaptic defect of transmitter resynthesis or mobilization. (Abst.) Neurology (NY) 1979;29:559 12. Mora M, Lambert EH, Engel AG. Synaptic vesicle abnormality in familial infantile myasthenia. Neurology (Cleve) 1987;37:206-14 13. Elmqvist D, Quastel DMJ. Presynaptic action of hemicholinium at the neuromuscular junction. J Physiol (Lond) 1965; 177:463-82 14. Jones SF, Kwanbunbumpen S. T h e effects of nerve stimulation and hemicholinium on synaptic vesicles at the mammalian neuromuscular junction. J Physiol (Lond) 1970;207:31-50 15. Jones SF, Kwanbunbumpen S. Some effects of nerve stimulation and hemicholinium on quantal transmitter release at the mammalian neuromuscular junction. J Physiol (Lond) 1970;207:51-61 16. Wolters ChMJ, Leeuwin KS, Van Wijngaarden KV. T h e effect of prednisolone on the rat phrenic nerve-diaphragm preparation treated with hemicholinium. Eur J Pharmacol 1974;29: 165-7 17. Anderson DC, King SC, Parsons SM. Pharmacological characterization of the acetylcholine transport system in purefied Torpedo electric organ synaptic vesicles. Mol Pharmacol 1983;24:48-54 18. Harlos P, Lee DA, Stadler H. Characterization of a Mg2+ATPase and a proton pump in cholinergic synaptic vesicles from the electric organ of Torpedo marmorata. Eur J Biochem 1984;144:441-6 19. Whittaker VB. T h e structure and function of cholinergic synaptic vesicles. Biochem Soc Trans 1984; 12:561-78 20. Wagner JA, Carlson SC, Kelly RB. Chemical and physical characterization of cholinergic synaptic vesicles. Biochemistry 1978;17: 1199-206 21. Engel AG, Lambert EH, Gomez MR. A new myasthenic syndrome with end-plate acetylcholinesterase deficiency, small nerve terminals and reduced acetylcholine release. Ann Neurol 1977; 1:3 15-330 22. Walls TJ, Engel AG, Harper CM, et al. Congenital neuromuscular junction acetylcholinesterase deficiency. (Abstr.) Ann Neurol 1989;26:147 23. Laskowski MB, Olson WH, Dettbarn WD. Ultrastructural changes at the motor end-plate produced by an irreversible cholinesterase inhibitor. Exp Neurol 1975; 47:290-306 24. Salpeter MM, Kasprzak H, Feng H, Fertuck H. Endplates after esterase inactivation in vivo: correlation between esterase concentration, functional response and fine structure. J Neurocytol 1979;8:95-115 25. Engel AG, Lambert EH. Mulder DM, et al. A newly recognized congenital myasthenic syndrome attributed to a prolonged open time of the acetylcholine-induced ion channel. Ann Neurol 1982; 11:553-69 26. Oosterhuis HJGH, Newsom-Davis J, Wokke JHJ, et al. T h e slow channel syndrome. Two new cases. Brain 1987;110:1061-79 27. DelCastillo J, Katz B. Localization of active spots within the neuromuscular junction of the frog. J Physiol (Lond) 1956; l32:630-649 28. Vigny M, Bon S, Massoulie J, Gisiger V. T h e subunit structure of mammalian acetylcholinesterase: catalytic subunits, dissociating effects of proteolysis and disulphide reduction on the polymeric forms. J Neurochem 1979;33:559-65 29. Takeuchi N. Effects of calcium o n the conductance change of the end-plate membrane during the action of the transmitter. J Physiol (Lond) 1963;167: 141-55 30. Evans RH. T h e entry of labelled calcium into the innervated region of the mouse diaphragm muscle. J Physiol (Lond) 1974;240:517-33 31. Miledi R, Parker 1, Schalow G. Transmitter induced calcium entry across the postsynaptic membrane at frog

Downloaded by: National University of Singapore. Copyrighted material.

CONGENITAL UISOKDERS OF NEUROMUSCULAK TRANSMISSION-ENGEL

32. 33.

34. 35.

36. 37. 38.

39. 40.

end-plates measured using arsenazo 111. J Physiol (Lond) 1980;300: 197-2 12 Ebashi S, Sugita H. T h e role of calcium in physiological and pathological processes of skeletal muscle. Excerpta Medica ICS 1979;455:73-84 Jackson MJ, Jones DA, Edwards RHT: Experimental skeletal muscle damage: the role of calcium activated degenerative processes. Eur J Clin Invest 1984;14: 369-75 Leonard JP, Salpeter MM. Calcium-mediated myopathy at neuromuscular junctions of normal and dystrophic muscle. Exp Neurol 1982;46: 121-38 Gardner JM, Fambrough DM. Acetylcholine receptor degradation measured by density labeling: effects of cholinergic ligands and evidence against recycling. Cell 1979;16:661-74 Schuetze SM. Developmental regulation of acetylcholine receptors. Annu Rev Neurosci 1987;10:403-57 Mishina M, Takai T, Imoto K, et al. Molecular distinction between fetal and adult forms of muscle acetylcholine receptor. Nature 1986;321:406-11 Witzemann V, Barg B, Nishikawa Y, et al. Differential regulation of muscle acetylcholine receptor gamma and epsilon-subunit mRNAs. FEBS Lett 1987;223: 104-12 Gu Y, Hall ZW. lmmunological evidence for a change in subunits of the acetylcholine receptor in developing and denervated rat muscle. Neuron 1988; 1: 117-25 Vincent A, Cull-Candy SG, Newsom-Davis J, et al. Congenital myasthenia: end-plate acetylcholine receptors and electrophysiology in five cases. Muscle Nerve 1981:4:306-18

MARCH 1990

41. Lecky BRF, Morgan-Hughes JA, Murray NMF, et al. Congenital myasthenia. Further evidence of disease heterogeneity. Muscle Nerve 1986;9:233-42 42. Smit SME, Jennekens FGI, Veldman H, Sarth PG. Paucity of secondary synaptic clefts in a case of congenital myasthenia with multiple contractures: ultrastructural morphology of a developmental disorder. J Neurol Neurosurg Psychiatry 1984;47:1091-7 43. Smit LME, Hageman G, Veldnian H, et al. A myasthenic syndrome with congenital paucity of secondary synaptic clefts: CPSC syndrome. Muscle Nerve 1988;11:337-48 44. Wokke JHJ, Jennekens FGI, Molenaar PC, et al. Congenital paucity of secondary synaptic clefts (CPSC) syndrome in 2 adult sibs. Neurology (Cleve) 1989;39: 648-54 45. Morgan-Hughes JA, Lecky BRF, Landon DN, Murray NMF. Alterations in the number and affinity of junctional acetylcholine receptors in a myopathy with tubular aggregates. A newly recognized receptor defect. Brain 198 1;4:279-95 46. Albers JW, Faulkner JA, Dorovini-Zis K, et al. Abnormal neuromuscular transmission in an infantile myasthenic syndrome. Ann Neurol 1984;16:28-34 47. McQuillen MP. Familial limb-girdle myasthenia. Brain l966;89: 121-32 48. Johns TR, Campa JF, Crowley WJ, Miller JQ. Familial myasthenia with tubular aggregates. (Abstr.) Neurology (Minneap) 1971;2 1:449 49. Dobkin BH, Verity MA. Familial neuromuscular disease with type 1 fiber hypoplasia, tubular aggregates, cardiomyopathy and myasthenic features. Neurology (NY) l978;28: 1135-40

Downloaded by: National University of Singapore. Copyrighted material.

SEMINARS IN NEUROLOGY VOLUME 10, NUMBER I