Electrophysiological analysis of neuromuscular synaptic function in myasthenia gravis patients and animal models

YEXNR-11917; No. of pages: 14; 4C: Experimental Neurology xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Experimental Neurology journal homepage: www.elsevier.com/locate/yexnr

Review

Electrophysiological analysis of neuromuscular synaptic function in myasthenia gravis patients and animal models Jaap J. Plomp a,⁎, Marco Morsch b, William D. Phillips c, Jan J.G.M. Verschuuren a a b c

Department of Neurology, Leiden University Medical Centre, Leiden, The Netherlands Motor Neuron Disease Research Group, Macquarie University, Sydney, Australia Physiology and Bosch Institute, University of Sydney, Sydney, Australia

a r t i c l e

i n f o

Article history: Received 7 November 2014 Revised 7 January 2015 Accepted 16 January 2015 Available online xxxx Keywords: Acetylcholine receptor Electromyography Compound muscle action potential Electrophysiology Endplate Muscle-specific kinase Mouse models Myasthenia gravis Neuromuscular junction Synapse

a b s t r a c t Study of the electrophysiological function of the neuromuscular junction (NMJ) is instrumental in the understanding of the symptoms and pathophysiology of myasthenia gravis (MG), an autoimmune disorder characterized by fluctuating and fatigable muscle weakness. Most patients have autoantibodies to the acetylcholine receptor at the NMJ. However, in recent years autoantibodies to other crucial postsynaptic membrane proteins have been found in previously ‘seronegative’ MG patients. Electromyographical recording of compound and single-fibre muscle action potentials provides a crucial in vivo method to determine neuromuscular transmission failure while ex vivo (miniature) endplate potential recordings can reveal the precise synaptic impairment. Here we will review these electrophysiological methods used to assess NMJ function and discuss their application and typical results found in the diagnostic and experimental study of patients and animal models of the several forms of MG. © 2015 Elsevier Inc. All rights reserved.

Contents Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Structure and physiological function of the neuromuscular junction . . . . . . . . . . . . . Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electrophysiological function . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electrophysiological methods to detect functional abnormalities at the neuromuscular junction . Electromyography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electromyography in patients with suspected neuromuscular junction dysfunction . Electromyography in animal models of myasthenia gravis . . . . . . . . . . . . Ex vivo electrophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Human myasthenia gravis muscle biopsies . . . . . . . . . . . . . . . . . . . Dissected muscles from myasthenia gravis animal models . . . . . . . . . . . . Microelectrode recordings . . . . . . . . . . . . . . . . . . . . . . . . . . Patch-clamp recordings. . . . . . . . . . . . . . . . . . . . . . . . . . . . Electrophysiological abnormalities in myasthenia gravis . . . . . . . . . . . . . . . . . . . Abnormal electrophysiology in patients . . . . . . . . . . . . . . . . . . . . . . . . AChR myasthenia gravis . . . . . . . . . . . . . . . . . . . . . . . . . . . MuSK myasthenia gravis . . . . . . . . . . . . . . . . . . . . . . . . . . . Abnormal electrophysiology in animal models . . . . . . . . . . . . . . . . . . . . . AChR myasthenia gravis models . . . . . . . . . . . . . . . . . . . . . . . . MuSK myasthenia gravis models . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

⁎ Corresponding author at: Leiden University Medical Centre, Depts. Neurology and MCB Neurophysiology, Research Building, S5-P, P.O. Box 9600, 2300 RC Leiden, The Netherlands. E-mail address: [email protected] (J.J. Plomp).

http://dx.doi.org/10.1016/j.expneurol.2015.01.007 0014-4886/© 2015 Elsevier Inc. All rights reserved.

Please cite this article as: Plomp, J.J., et al., Electrophysiological analysis of neuromuscular synaptic function in myasthenia gravis patients and animal models, Exp. Neurol. (2015), http://dx.doi.org/10.1016/j.expneurol.2015.01.007

2

J.J. Plomp et al. / Experimental Neurology xxx (2015) xxx–xxx

Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Introduction The neuromuscular junction (NMJ) is perhaps the most-studied and best-understood synapse of the nervous system. For many decades it has served as a model synapse, enabling the discovery of universal synaptic principles. Electrophysiological study of synaptic signalling has provided critical insights into the mechanisms of neurotransmission. Here, we will review the electrophysiological methods widely used to measure NMJ function in vivo and ex vivo and discuss their application in the study of the NMJ disorder myasthenia gravis (MG) and its animal models. MG is hallmarked by fluctuating and fatigable muscle weakness. Most patients (~85%) have serum autoantibodies against the acetylcholine (ACh) receptors (AChRs) on the postsynaptic membrane at the NMJ. In recent years it has become apparent that a large proportion (40–70%) of the AChR-negative MG patients are seropositive for antibodies against muscle-specific kinase (MuSK), a postsynaptic membrane tyrosine kinase that forms the core of a multi-protein signalling complex. MuSK is vitally involved in the embryonic development, and subsequent maintenance of AChR clusters at NMJs (Hoch et al., 2001; Koneczny et al., 2014). Some of the remaining MG patients (seronegative for both AChR and MuSK antibodies) instead have antibodies

0 0 0

against low-density lipoprotein receptor-related protein 4 (LRP4) (Higuchi et al., 2011; Pevzner et al., 2012; Zhang et al., 2012). LRP4 is a membrane receptor that gets activated by neurally released agrin and stimulates MuSK activation to drive AChR clustering (Ghazanfari et al., 2011). Very recently anti-agrin antibodies have also been detected in some MG sera, mostly in those patients who are also seropositive for either AChR, MuSK or LRP4 antibodies (Gasperi et al., 2014; Zhang et al., 2014). In addition, some previously AChR ‘seronegative’ MG cases appear to have low-affinity antibodies that can be detected in a sensitive cell-based assay in which AChRs are clustered on the cell surface (Leite et al., 2008). AChR MG can be classified into an early-onset group (age b40 years) with female predominance (often with thymic hyperplasia), and a lateonset group in which males predominate and with mostly a normal (i.e. atrophied) thymus (Sieb, 2014; Verschuuren et al., 2013). Thymoma is present in ~10% of AChR MG patients with generalized weakness. It is seen across all ages, although somewhat more often in the elderly. About 20% of AChR MG patients display weakness of only their extraocular muscles. MuSK MG has distinct features. It occurs at higher frequency in women, with the peak incidence in the 4th age decade as compared to the 3rd decade for AChR MG. Often there is prominent oropharyngeal, facial, neck and respiratory muscle weakness and facial and

Fig. 1. Structure of the neuromuscular junction. (A) Inter-species morphological variation of the neuromuscular junction (NMJ). Axon terminal processes are shown in black on the muscle fibre surface. Taken from (Slater, 2008), with permission. (B) En face image of three pretzel-shaped mouse diaphragm NMJs (confocal microscopy z-projection). Acetylcholine receptors (red) are stained with Alexa Fluor 555‐α‐bungarotoxin and nerve terminals (green) with anti-synaptophysin antibody (scale bar = 20 μm). (C) Electron microscopical picture of a nerve terminal profile. pSC = perisynaptic Schwann cell; M = mitochondria; SV = synaptic vesicles; AZ = active zone; PF = postsynaptic folds (scale bar = 0.5 μm). (D) Schematic drawing of key proteins involved in neuromuscular transmission (ACh = acetylcholine; AChE = acetylcholinesterase; ColQ = collagen-Q; LRP4 = low-density lipoprotein receptor-related protein 4).

Please cite this article as: Plomp, J.J., et al., Electrophysiological analysis of neuromuscular synaptic function in myasthenia gravis patients and animal models, Exp. Neurol. (2015), http://dx.doi.org/10.1016/j.expneurol.2015.01.007

J.J. Plomp et al. / Experimental Neurology xxx (2015) xxx–xxx

3

tongue muscles may be atrophied. Respiratory crisis occurs more often than in AChR MG. Isolated ocular weakness is less often present. Generally there is no thymic involvement in MuSK MG. The main therapies for MG are cholinesterase inhibiting drugs (to enhance neurotransmission at the NMJ) and treatments that influence the autoimmunity such as plasma exchange, human immunoglobulins, immunosuppressants and thymectomy (Sieb, 2014). Most patients receive combinations of these treatments with frequent adjustments to optimize improvement. MuSK MG is more difficult to treat than AChR MG and is complicating by the often poor or worsening response to cholinesterase inhibitors (Guptill et al., 2011). Lambert–Eaton myasthenic syndrome is a closely related disorder featuring NMJ transmission defects in ACh release that are generally caused by autoantibodies against presynaptic Ca2+ channels. For recent overview, see (Titulaer et al., 2011). The electrophysiological studies on this syndrome and its animal models will not be included here. Structure and physiological function of the neuromuscular junction Structure The electrophysiological function of the NMJ is to evoke an action potential on the muscle fibre membrane in response to each motor neuronal impulse. The specific subcellular organization of several ligandand voltage-gated ion channels in pre- and postsynaptic membranes enables the NMJ to perform this important task with high reliability. We will first give a brief structural description. Mammalian skeletal muscles are innervated by axons originating from motor neuron cell bodies lying in the ventral horn of the spinal cord. Within the muscle, distal branching occurs and each axon terminal innervates a single muscle fibre, generally in its middle portion. At the NMJ, the axon terminal is no longer covered by myelin but instead by a few (usually 3–5) perisynaptic Schwann cells. The terminal forms a microbranched presynaptic structure that closely approaches the AChR clusters in the postsynaptic membrane. This nerve-AChR composition forms the ‘pretzel shaped’ structure that is characteristic for mouse NMJs (Fig. 1A,B). Human axon terminals on the other hand form a few spot-like boutons which are interconnected by tiny branches (Fig. 1A) (Slater, 2008). In addition, human NMJs are about three times smaller in area than mouse or rat NMJs, even on muscle fibres of comparable diameter. An excellent overview on the interspecies differences in NMJ structure and function has been published elsewhere (Slater, 2008). The key function of the presynaptic terminal is to release the neurotransmitter ACh in response to each action potential that propagates down from the motor neuron's cell body via the axon. The synaptic contact side of the axon terminal (the presynaptic membrane) contains specialized spots called ‘active zones’ where the neuroexocytotic machinery is concentrated. In both human and mouse NMJs, the membrane density of active zones is ~2.5/μm2 (Clarke et al., 2012; Slater, 2008). Synaptic vesicles, each containing a ‘quantum’ of ACh (~ 10,000 molecules), are present in the cytoplasm of the presynapse (Fig. 1C,D). Some vesicles are held docked to the presynaptic membrane by dedicated molecules so that they are readily available for exocytotic release of their content. Transmembrane voltage-gated Ca2+ channels are present in the active zone (in mammals the Cav2.1 type). The opening of these Ca2+ channels in response to the presynaptic action potential allows for influx of Ca 2 + ions into the presynaptic nerve terminal, which triggers rapid (~ 50 μs) ACh exocytosis. Knowledge on the molecular constituents of the neuroexocytotic machinery and their structural organization at active zones has markedly expanded in recent years (Meriney and Dittrich, 2013; Nagwaney et al., 2009; Nishimune, 2012; Nishimune et al., 2012; Sudhof, 2013; Szule et al., 2012). The released ACh diffuses rapidly across the ~ 50 nm wide synaptic cleft to activate postsynaptic AChRs, translating the chemical signal into an electrical one. Acetylcholinesterase within the synaptic cleft degrades the ACh by hydrolytic cleavage.

Fig. 2. Ex vivo electrophysiological recordings at the neuromuscular junction. (A) A single spontaneous miniature endplate potential (MEPP) and a single nerve stimulation-evoked endplate potential (EPP) displayed at expanded time scale to show their similar kinetics: rapid rise-time and a somewhat slower, exponential decay time. Recordings were made from muscle fibres of a mouse diaphragm muscle. (B) Longer duration example traces showing multiple spontaneous MEPPs, and 40 consecutive EPPs evoked by 1 Hz nerve stimulation. (C) Schematic drawing of the intracellular glass capillary microelectrode recording configuration.

The postsynaptic membrane is also highly specialized, with extensive folding beneath the motor nerve terminal (Fig. 1C,D). AChRs of the nicotinic type are found at high density of ~10,000/μm2 on the crests of these folds, immediately opposing the active zones. Each AChR consists of a ring of five subunits (two α, one β, one δ and one ε subunit at the adult NMJ) to form a ligand-gated heteropentameric cation channel. An ACh molecule binds to each of the α subunits, at the interface with a non-α subunit, causing brief opening of the central pore (Sine, 2012). Human and mouse muscle AChRs have a comparable single

Please cite this article as: Plomp, J.J., et al., Electrophysiological analysis of neuromuscular synaptic function in myasthenia gravis patients and animal models, Exp. Neurol. (2015), http://dx.doi.org/10.1016/j.expneurol.2015.01.007

4

J.J. Plomp et al. / Experimental Neurology xxx (2015) xxx–xxx

channel conductance of ~ 60 pS (Milone et al., 1994; Newland et al., 1995). The ion pore rather non-specifically conducts cations: mainly Na+ and K+, and a little Ca2 + (under physiological conditions). Of note, human AChRs have higher Ca2 + permeability: ~ 7% of the total current, compared to ~ 4% for murine AChRs (Fucile et al., 2006). A second crucial postsynaptic ion channel is the Nav1.4 type voltagegated Na+ channel. It is present in the troughs of the postsynaptic membrane folds at ~ 10 fold higher density than elsewhere in the muscle fibre, thus locally increasing membrane excitability (Fig. 1D) (Ruff and Lennon, 1998; Slater, 2008). The geometry of the membrane folds themselves also promotes excitability. The small cytoplasmatic space forms a relatively high electrical resistance pathway for the ACh-induced ion current, causing a larger local depolarization which stimulates more Na+ channels so that an action potential is triggered more easily (Vautrin and Mambrini, 1989). Electrophysiological function The main electrophysiological events underlying synaptic transmission at the NMJ progress as follows: 1) an axonal action potential invades the motor nerve terminal and causes opening of Cav2.1 channels. The ensuing Ca2 + influx triggers exocytosis of a number of ACh quanta. This number, termed quantal content, varies between NMJs of different muscles and species but is roughly correlated with NMJ size. Human NMJs are small (~100 μm2) and have only a low quantal content (~20), as compared to ~40–100 for adult mouse and rat and ~ 200 for the much larger frog NMJs (N200 μm2) (Plomp et al, 1995; Slater, 2008). 2) The released ACh stimulates opening of AChRs. The resulting inward ionic endplate current (EPC) causes a local depolarization of the membrane called the endplate potential (EPP). The muscle fibre's resting membrane potential lies around − 80 mV. The EPP has an amplitude of ~15–30 mV (depending on muscle type and species), a rapid rise-time (b1 ms) and a near-exponential decay phase of a few ms (Fig. 2A). 3) The EPP causes the opening of a proportion of the densely packed Nav1.4 type Na+ channels in the troughs of the folds. If the EPP amplitude is sufficient to reach the endplate depolarization threshold (the ‘firing threshold’), a Na+ influx-fuelled positive feedback loop of further depolarization and Na+ channel opening ensues. This forms the upstroke of the muscle fibre action potential, which spreads out over the muscle fibre membrane in both directions away from the NMJ. The action potential invades the T-tubuli of the muscle fibre where an excitation-contraction coupling molecular machinery comes into play which, a few ms later, results in contraction of the muscle fibre (Bannister, 2007). The firing threshold at the NMJ is dictated by the density and channel activation characteristics of the postsynaptic Nav1.4 channels. In rat NMJs it has been shown to lie at around − 63 mV (Wood and Slater, 1995). At −75 mV resting membrane potential this means that EPPs need to be at least 12 mV in amplitude to trigger a muscle fibre action potential. However, EPPs in rat and mouse NMJs are generally much larger, i.e. in the range of 20–35 mV. Thus, a substantial safety factor exists in healthy muscles. Among mammals, interspecies comparison reveals a large variability in safety factors, ranging from 1.7 to 12 (1.8 to 6 in mouse/rat). Human NMJs have a low safety factor of only ~ 2 (Wood and Slater, 2001), making neuromuscular transmission in humans relatively more sensitive to agents that block AChRs or reduce ACh release. The safety factor in transmission enables the NMJ to cope with decreasing quantal content during high frequency nerve impulses. To achieve sustained (‘tetanic’) muscle contraction, a motor neuron fires trains of nerve impulses at frequencies in the range of 20–100 Hz, depending on muscle fibre type (Eken, 1998; Hennig and Lomo, 1985). During such tetanic activity, ACh release decays and consequently EPP amplitudes at mouse NMJs run down by 20–30% to a plateau value within 10 impulses (Zitman et al., 2008). This is presumably caused by a combination of inactivation of the presynaptic Cav2.1 channels and a limiting size and replenishment rate of the pool of

releasable ACh vesicles. Without a safety factor, EPPs would become sub-threshold during such trains and muscle fatigue would occur. In human NMJs the rate of EPP rundown is even more pronounced (~40%, Niks et al., 2010). On the other hand, human NMJs have relatively deep postsynaptic membrane folds which, due to their high cytosolic electrical resistance, help to increase the depolarizing effect of ACh quanta, adding to the safety factor and thus partly compensating for the large EPP rundown (Slater, 2008). In addition to the synchronized multi-quantal ACh release evoked by an action potential, quiescent nerve terminals spontaneously release single quanta of ACh with a random timing, each causing a postsynaptic miniature EPP (MEPP). In the mouse, these MEPP events take place at a ~ 1–4/s (Fig. 2B), depending on the muscle type and NMJ size. At the small NMJs in the human intercostal muscle the MEPP frequency is much lower (b 0.1/s) (Niks et al., 2010; Plomp et al., 1995). The amplitude of MEPPs range from about 0.3–1.5 mV depending upon the muscle type. The kinetics of MEPPs are similar to those of nerve action potential-evoked EPPs (Fig. 2A). Individual MEPPs are too small to trigger a muscle fibre action potential and their physiological functions remain enigmatic. There are indications from experiments in cultured hippocampal synapses that spontaneous uni-quantal synaptic events have a role in postsynaptic local protein synthesis (Sutton et al., 2006). In fruit fly larvae NMJs, MEPPs seem necessary for the normal structural maturation of the synapse (Choi et al., 2014). At the mammalian NMJ it remains to be seen if such functions of MEPPs are also present or whether the MEPP phenomenon is just a spill-over of the enormous number of synaptic vesicles (~150,000–300,000) present in a presynaptic nerve terminal (Slater, 2003). Electrophysiological methods to detect functional abnormalities at the neuromuscular junction Many general principles of synaptic function were first defined at the NMJ thanks to its experimental accessibility. Since the 1920–30s the cholinergic synaptic function of the NMJ has been intensively studied and largely elucidated, first through biochemical methods and later (with the advent of electronics and glass micropipette making techniques) in single synapse detail with intracellular electrophysiological recordings. Early recordings from muscle biopsies of MG patients in the 1960s revealed small MEPP and EPP amplitudes (Elmqvist et al., 1964), which we now know is associated with reduced postsynaptic AChR number. In the 1970s MG became known to be of autoimmune origin and the clinical and experimental analysis of NMJ function in this disorder benefited greatly from the established methodology and theory. In this section we will provide an overview on the electrophysiological methods currently used to assess NMJ function. The Section 'Electrophysiological abnormalities in myasthenia gravis' below will then discuss the specific abnormalities that have been found in MG and its animal models using these electrophysiological techniques. Electromyography It is currently not feasible to measure synaptic signals intracellularly at single NMJs in patient skeletal muscles in vivo. However, electromyography (EMG) can provide important (albeit indirect) information on NMJ function. EMG recordings typically employ extracellular needle electrodes inserted in the muscle, or extracorporeal skin surface electrodes. They detect the electrical signal resulting from action potentials propagating along the muscle fibres, either during electrode stimulation of the innervating nerve or voluntarily contraction by the patient. NMJ dysfunction can then be inferred from the characteristics of the recorded signals. EMG is crucial for the clinical diagnosis of MG and related disorders in patients. It can also be used for in vivo assessment of NMJ function in experimental animals.

Please cite this article as: Plomp, J.J., et al., Electrophysiological analysis of neuromuscular synaptic function in myasthenia gravis patients and animal models, Exp. Neurol. (2015), http://dx.doi.org/10.1016/j.expneurol.2015.01.007

J.J. Plomp et al. / Experimental Neurology xxx (2015) xxx–xxx

Electromyography in patients with suspected neuromuscular junction dysfunction The most commonly used electrodiagnostic test for NMJ dysfunction is repetitive nerve stimulation EMG (RNS-EMG) (Howard, 2013; Meriggioli and Sanders, 2005). In this test, repetitive electrical stimuli are applied to the peripheral nerve and the resulting compound muscle action potential (CMAP) is recorded. This is the summed extracellular signal of the action potentials of the firing muscle fibres in the vicinity of the recording electrode. Repetitive activity of the NMJ will cause rundown of EPP amplitude (see above). If the transmission safety factor is compromised (e.g. due to autoantibody-mediated reduction of AChR number such as in MG) EPPs in an increasing proportion of NMJs will become sub-threshold during the repetitive stimulation. This will be reflected in a decrement of the CMAP amplitude as the action potential fails in more and more muscle fibres. If the CMAP decrements by N 10% during a train 5–10 nerve stimuli at 2–5 Hz, this is generally considered as evidence of an NMJ defect, which can be localized either post- or presynaptically. A presynaptic defect can then be tested for by using prolonged high-rate stimulation (5–10 s at 20–50 Hz), or by voluntary muscle activation for 10 s, which is more tolerable than high-rate stimulation for most patients. If this yields a N 2-fold increase of CMAP amplitude, from initially low value, a presynaptic defect is indicated (Howard, 2013; Meriggioli and Sanders, 2005). RNS-EMG should preferably be performed in multiple, clinically weak muscles that can be well immobilized to prevent movement artefacts. Muscles should be

5

warm, because lower temperatures reduce the CMAP decrement thus limiting detection sensitivity. In muscles of a putative MG patient where RNS-EMG appears normal, single-fibre EMG (SF-EMG) can provide a more sensitive test for NMJ dysfunction. This technique measures variation (‘jitter’) in the delay between a nerve stimulus and the resulting action potential in a single muscle fibre during consecutive nerve stimuli. Alternatively, jitter can be determined from pairs of action potentials recorded from two fibres belonging to the same motor unit during voluntary contraction. If an EPP is of just-above-threshold amplitude, the onset of the muscle fibre action potential triggered by it will be somewhat delayed, as compared to when an EPP is well-above-threshold. Consecutive EPPs display some inherent amplitude variation, mainly due to some quantal content fluctuation. Consequently, a fibre with EPP amplitudes that are just-above-threshold will show increased jitter in the muscle fibre action potential timing as compared to a fibre with a healthy NMJ with well-above-threshold EPPs. To measure jitter, a special needle electrode is inserted into the muscle which has an insulating coating with only a very small opening near the fine tip so that signal is sensed from only a very limited extracellular volume compartment. The nerve is activated by either voluntary activation or with repetitive stimulation by an electrode. Due to inherent variation, multiple (N20) measurements must be made in the same muscle (Howard, 2013). Classically, SF-EMG is done with a re-usable specific needle electrode. More recently disposable (multi-purpose) concentric needle electrodes have been

Table 1 Comparison of mouse models of myasthenia gravis. Type of MG model

Advantages

Disadvantages

Laboriousness

Relative costs

Suitability for Suitability for drug studies electrophysiological NMJ analysis

Passive transfer of patient IgG

– High human relevance: tests pathogenicity of MG patient IgG – Muscle weakness reasonably well titratable (in case of MuSK MG)

High (multiple IgG purifications, daily injections over several weeks)

High

High

– Very high, due to possibility of creating stable disease – Suitable for drugs targeting the NMJ, the pathogenic antibodies or complement – Not suitable for drugs aimed at autoimmunity mechanisms

Passive transfer of monoclonal antibodies against AChR

–

Low to moderate

Moderate High

Suitable for drugs targeting the NMJ or the pathogenic antibodies or complement

Active immunization

–

Low to moderate (requires regular monitoring over several months)

Moderate High

– High, particularly suitable for drugs aimed at autoimmunity mechanisms as well as at NMJ and complement – Intrinsic variability may require larger number of mice

Moderate (requires regular injections over several weeks)

Low

–

–

α-Bungarotoxin-induced

–

– Genetic (natural and transgenic mutations)

–

– Need of large amounts (g) of patient IgG, purified from plasmapheresis material – Sudden, stepwise elevation of autoantibody titre, unlike gradual rise in patients – Not all strains equally suitable (e.g. variability in complement activity) Robust weakness – Need to produce large after single injection amounts of monoclonal Reproducibility antibodies – Monoclonals are non-human Involvement of the – Muscle weakness not animal's own titratable, uncontrollaimmune system ble disease can lead to Gradual increase of death autoantibody titre, – Sometimes nonlike in patients responders – Mouse IgG subclass characteristics differ from human ‘Clean’ NMJ effects, Less clinical relevance for i.e. no involvement of autoimmune-mediated MG immune system (e.g. no complement Muscle weakness activation component) easily titratable Breeding of strain needs to Human congenital MG with known be maintained mutation can be exactly mimicked (knock-in)

– Initially high (in case of transgenesis) – Moderate once established (ongoing genotyping and breeding)

High

Moderate High to high

Moderate, no immunological drug targets, only for drugs targeting the NMJ

High, for drugs targeting the NMJ (e.g. the mutated protein) or drugs acting on DNA/RNA

Please cite this article as: Plomp, J.J., et al., Electrophysiological analysis of neuromuscular synaptic function in myasthenia gravis patients and animal models, Exp. Neurol. (2015), http://dx.doi.org/10.1016/j.expneurol.2015.01.007

6

J.J. Plomp et al. / Experimental Neurology xxx (2015) xxx–xxx

and Sanders, 2009). A recent development is the advent of disposable versions of the classical specific SF-EMG electrodes. These are more expensive but seem to perform well (Papathanasiou and ZambaPapanicolaou, 2012). Although SF-EMG is unable to detect permanently blocked NMJs, it provides a sensitive test for detecting critically transmitting NMJs that are operating with almost no safety factor and thus have borderline functionality, such as in MG. In fibres with somewhat more compromised NMJs, intermittent block of transmission can be observed. In the hands of an experienced electromyographer, SF-EMG forms a highly sensitive test for detecting NMJ transmission defects in patients with symptoms of muscle weakness. The (disposable) electrodes used for stimulation and recording, the amplifiers, stimulators, signal display and recording devices used for RNS- and SF-EMG are all commercially available from multiple specialized suppliers.

Fig. 3. Repetitive nerve stimulation electromyography in the anaesthetized mouse. (A) Cartoon showing the electrode configuration used to record compound muscle action potentials (CMAPs). (B) Example of CMAP recording from the gastrocnemius muscle of a myasthenic mouse, demonstrating a decrement in amplitude during 3 Hz repetitive nerve stimulation. Superimposed CMAPs are spaced somewhat for easier visualization of the decrement. (C) Schematic drawing of a single CMAP on an expanded timescale.

used which are cheaper and perhaps safer (i.e. no risk of pathogen transmission between patients), but have a somewhat lower sensitivity in detecting abnormal jitter. There is still debate on which electrode type is to be preferred (Farrugia et al., 2009; Howard, 2013; Stalberg

Electromyography in animal models of myasthenia gravis Animal models of MG are useful to: 1) prove pathogenicity of patient antibodies, 2) study pathophysiology and 3) test potential therapeutic effects of new or existing drugs. Myasthenic animals can be generated either by active immunization with the antigen of study or passive transfer of patient antibodies or plasma, or monoclonal antibodies (see other articles in this issue). For AChR MG, a non-immunological model can additionally be generated by injecting animals with low doses of α-bungarotoxin, a near-irreversible AChR antagonist (Molenaar et al., 1991; Sons et al., 2006). Most often mice or rats are used in model studies, and less frequently rabbits. Furthermore, transgenic or spontaneously mutant animals, most often mice, can model rare congenital forms of MG. Each type of (mouse) model has its specific advantages and limitations (Table 1), but all are equally suitable for in vivo and ex vivo electrophysiological analysis of NMJ function. It is important to assess and quantify NMJ function in these MG models in vivo during the course of the study as well as at termination of the experiment. To this end, EMG can be performed. In principle, procedures in small laboratory animals (Fig. 3) are similar to those in humans. However, one important difference is that animals must be anaesthetized. Many different anaesthetic agents have been used, including ketamine/xylazine, ketamine/medetomidine, pentobarbital and isoflurane/oxygen. Careful dosing is important, especially in weak animals, since extra respiratory depression induced by the anaesthetic agent may be fatal. Perhaps the best management of the depth of anaesthesia is obtained with isoflurane/oxygen inhalation (by carefully adjusting the percentage isoflurane and monitoring the breathing rate), although injectables might be more practical in some situations. Another reason for keeping the dosing of anaesthetic compounds as low as possible is a potential interference with neuromuscular transmission. Some early studies have suggested that volatile anaesthetics such as isoflurane at clinically relevant concentrations may potentiate the effect of non-depolarizing neuromuscular blocking drugs. Presumably this involves an (indirect) effect on AChR function (Waud, 1979). The same may be true for ketamine, albeit to a lesser extent (Cronnelly et al., 1973). Thus, CMAP decrements in MG models may be somewhat potentiated by anaesthetic compounds. In view of the dependence of CMAP decrement on temperature it seems wise to place the mouse on heating mat or under a heating blanket to maintain body temperature, which can drop during anaesthesia. Generally the anaesthetized mouse is placed on its back and limbs are slightly stretched out and fixed with adhesive tape over the paws to the underground to prevent movement artefacts. Very often RNS-EMG is performed on the gastrocnemicus muscle while stimulating the sciatic nerve, but other nerve-muscle combinations are also possible (e.g. trigeminal nerve and masseter muscle (Chroni and Punga, 2012)). Generally, two monopolar needle electrodes are used to stimulate the sciatic nerve. They are inserted subcutaneously into the sciatic notch and advanced while stimulating at low frequency until muscle

Please cite this article as: Plomp, J.J., et al., Electrophysiological analysis of neuromuscular synaptic function in myasthenia gravis patients and animal models, Exp. Neurol. (2015), http://dx.doi.org/10.1016/j.expneurol.2015.01.007

J.J. Plomp et al. / Experimental Neurology xxx (2015) xxx–xxx

contraction of the leg is noted. Then the recording electrodes are positioned, being two additional monopolar needle electrodes. One is placed subcutaneously onto or through the gastrocnemicus muscle belly. The second is placed subcutaneously at or through the muscle tendon near the ankle. The mouse must be electrically grounded through either a further subcutaneous needle electrode or a skin surface gel electrode. This is placed either at the contralateral leg or the abdominal wall. Before starting the recordings, an intensity (voltage) of nerve stimulation should be ascertained that produces supramaximal and stable nerve excitation, resulting in stable CMAPs at low frequency (e.g. 0.2 Hz) stimulation. To this end, several rounds of slight repositioning of the stimulation electrodes and stepwise increase of the stimulus amplitude and/or duration are needed to acquire a stable biphasic CMAP that does not increase further when the stimulus amplitude and/or duration is increased. Stimulation durations in the range of 20–500 μs are generally used. Once a stable supramaximal nerve stimulation condition is established, a protocol of (10–20) nerve stimuli at several frequencies (0.2– 40 Hz) is applied and the resulting CMAPs are recorded. A pause of at least 30 s is included after each stimulus train to allow for complete recovery. In MG model mice that are weak, CMAP decrements can generally be observed upon 3 Hz nerve stimulation. The decrement becomes more pronounced at higher stimulation frequencies. CMAP recordings in animals can be made using clinical grade EMG machines. Some laboratories prefer a set-up of non-medical stimulation and recording apparatus, which is less expensive and provides more flexibility in stimulation protocols, signal display and data analysis (Klooster et al., 2012). For further details on mouse RNS-EMG protocols and examples of applications in recent studies on MG mouse models, see (Chroni and Punga, 2012; Cole et al., 2008; Klooster et al., 2012; Mori et al., 2012b; Wu et al., 2013). It is also possible to perform SF-EMG in mice (to detect muscle fibre action potential jitter and possible intermittent transmission blocking) using a specific SF-EMG needle (Lin and Cheng, 1998). However, SFEMG has not often been employed in MG mouse models because it is technically much more difficult than RNS-EMG and may be too timeconsuming to complete within the time available under anaesthesia (Chroni and Punga, 2012). Fortunately, CMAP decrement is often easily detected with RNS-EMG in MG model mice displaying muscle weakness, so there is usually no need to perform additional SF-EMG. Ex vivo electrophysiology Detailed information about NMJ function at single synapse level can be obtained with electrophysiological study of (unfixed) muscle biopsies freshly dissected from patients or from experimental animals. Muscle specimens dissected quickly and carefully (with or without codissected innervating peripheral nerve trunk) are placed in a suitable physiological solution such as a Ringer's medium. The medium contains glucose, pH buffer(s) and the essential electrolytes at physiological levels and is bubbled with 95% O2/5% CO2 for oxygenation. In this way specimens can be kept viable for many hours at room temperature. The muscles need to be pinned out in a mildly stretched position (roughly 1.5× the unstretched length), using small insect pins in preparation dishes containing a transparent silicone-rubber base. This prevents deterioration of the excised muscle. Electrophysiological recordings are usually made with the preparation visualized with an upright microscope with a fixed stage so that electrode-to-specimen position is maintained during focusing. Objectives with long working distances are needed to allow for moving around the measuring and stimulation electrodes. The microscope must be placed on a vibrationfree table and shielded by a grounded Faraday cage to minimize electromagnetic interference from power lines (50 or 60 Hz ‘hum’, i.e. background alternating current noise). The preparation dish is usually placed in or onto a Peltier element-based heating device, which can be

7

used to regulate incubation bath temperature. Fluid content of the bath can be changed either by pipetting or using a peristaltic pumping system. Human myasthenia gravis muscle biopsies Several muscles are suitable to biopsy for electrophysiological NMJ study. These include vastus lateralis, anconeus and intercostal muscles (Elmqvist et al., 1960; Maselli et al., 1991a; Slater et al., 1992). Care should be taken that the excised tissue is transferred directly into preoxygenized Ringer medium and does not dry out. Parasternal intercostal muscle biopsy is most commonly studied in MG because it can be relatively easily biopsied during the thymectomy that many MG patients undergo. An important advantage of these muscles is the relatively short length of the fibres (~15 mm). Even a small specimen, provided it has been excised entirely from one rib to another, will most certainly contain NMJs. Another advantage is that most muscle fibres in the specimen will likely remain intact, maintaining healthy membrane potentials required for the electrophysiological measurements at the NMJ. Small bundles of intact fibres need to be micro-dissected from the biopsy specimen using a stereomicroscope and ultra-fine dissecting instruments. As with animal muscles it is important to pin out each bundle of human fibres in a mildly stretched position. Ideally the specimen is dissected in such a way as to retain a certain length of intramuscular nerve branch that can be sucked up into the tip of a suction stimulation electrode. Dissected muscles from myasthenia gravis animal models Many different muscles from experimental animals can be dissected for electrophysiological analysis of NMJs. Most suitable are flat muscles that are only a few fibre layers thick and which can be completely excised with both tendons (avoiding damage to the muscle fibres) and some length of the innervating nerve. One of the most commonly used muscles is rat or mouse hemi-diaphragm with phrenic nerve, but many other muscles are suitable including epitrochleoanconeus, levator auris longus and triangularis sterni (Angaut-Petit et al., 1987; Bradley et al., 1989; McArdle et al., 1981). Thicker muscles such as soleus, flexor digitorum brevis or extensor digitorum longus can also be used (Kaja et al., 2007; Wood and Slater, 1995), but it is more difficult to visualize the intramuscular nerves and NMJ-rich areas. Hemi-diaphragm muscle is particularly suitable because the innervating phrenic nerve can be dissected for ~ 2 cm allowing it to be easily positioned on a bipolar stimulation electrode. For other muscle preparations, with much shorter or thinner nerve trunks, a stimulation electrode with a suction tip is required. Such an electrode is more tedious to position, sometimes less stable and prone to causing larger electrical stimulation artefacts. Microelectrode recordings The most straightforward way to monitor synaptic signals at single NMJs is to impale the muscle fibre near the NMJ with the tip of a sharp glass capillary microelectrode and measure the voltage signals intracellularly (Fig. 2C). An Ag/AgCl2 reference electrode is placed in the bath fluid. The ideal width of the capillary microelectrode tip requires a compromise. Small tips impale muscle fibres more easily but produce high background electrical noise. Wide tips have lower noise levels but impalement is difficult. Generally, a tip width of b1 μm is used. When filled with 3 M KCl solution as the conducting electrolyte, the electrical resistance range is 5–15 MΩ. Several electrode pullers are commercially available which can make microelectrodes from commercially available glass or quartz capillaries by applying melting heat (either using an electrical filament or by a CO2 laser beam) in combination with bidirectional pulling steps. The pulled electrode is filled with 3 M KCl and placed in a special holder that inserts a silver wire into the internal solution and connects with the input of a preamplifier headstage. The preamplifier is in turn connected to a signal amplifier/filtering unit of which the output is digitized, visualized and stored on a computer. The preamplifier-electrode combination is

Please cite this article as: Plomp, J.J., et al., Electrophysiological analysis of neuromuscular synaptic function in myasthenia gravis patients and animal models, Exp. Neurol. (2015), http://dx.doi.org/10.1016/j.expneurol.2015.01.007

8

J.J. Plomp et al. / Experimental Neurology xxx (2015) xxx–xxx

manoeuvred with a manual or motorized micromanipulator, enabling very subtle movement of the tip of the electrode in three dimensions. The tip of the microelectrode is lowered to a position near the NMJ, just touching the muscle fibre membrane. The contact of the electrode to the surface is apparent by small changes of the potential. Gentle manual tapping on the back of the electrode holding micromanipulator then causes abrupt penetration of the electrode tip through the membrane, witnessed by a jump of the measured voltage from 0 to the resting membrane potential of the muscle fibre (usually around − 75 mV). Membrane potentials that are much less negative than this or that rise rapidly after the impalement indicate unwanted damage of the fibre and leakage of internal electrolytes. Measurement at such fibres is usually discontinued, the electrode is retracted and another fibre in the preparation is impaled. Successful impalement is represented by a stable resting membrane potential (drift b 5 mV per minute). At fibres with a healthy NMJ, MEPPs can be seen rising from the baseline. Since MEPPs are quite variable in amplitude (variance coefficient of ~0.3) at least 20–30 MEPPs must be sampled to calculate a meaningful average. In addition there is considerable inter-NMJ variation in synaptic potentials so at least 10 NMJs should be sampled at random locations within the muscle preparation allowing a grand mean value to be calculated from the individual NMJs means. To assess evoked ACh release, EPPs are recorded. In order to do this the muscle fibre action potentials must first be blocked since they obscure the EPP and cause contraction which displaces the microelectrode from the fibre. The best method for rat and mouse muscle is to incubate muscles in μ-conotoxin-GIIIB (1–3 μM). This selectively blocks the Nav1.4 Na+ channels on muscle fibres without affecting Nav channels on the nerve (Plomp et al., 1992). Unfortunately μ-conotoxinGIIIB is not selective for muscle Na+ channels in human tissue (Plomp et al., 1995). Mouse or rat muscle becomes paralysed within 20 min. Next, the nerve can be stimulated at desired frequencies (usually a low-rate, 0.3–1 Hz) and multiple EPPs can be recorded reliably. At every NMJ, both MEPPs and EPPs should be recorded to allow for estimation of the quantal content. The actual resting membrane potential, which forms the driving force for the ion current through AChR channels, varies somewhat from fibre to fibre during measurements. This contributes to the inter-NMJ variation of MEPP and EPP amplitudes. However, this source of variability can be minimized by normalizing MEPP and EPP amplitudes to a standard resting membrane potential of − 75 mV (Kaja et al., 2005). The quantal content is estimated by dividing the EPP by the MEPP amplitude. EPP amplitudes must first be corrected for non-linear summation (McLachlan and Martin, 1981; Wood and Slater, 2001). This is the phenomenon that the depolarizing effect of each additive ACh quantum during the rising phase of the EPP gradually lessens because the membrane potential (which is the driving force) declines. To adjust for this, EPPs are corrected according to a formula described by (McLachlan and Martin, 1981) (with factor f = 0.8 for mammalian NMJs). The quantal content is then calculated by dividing the normalized, corrected mean EPP amplitude for a given NMJ by the normalized mean MEPP amplitude of the same NMJ. The rundown of EPP amplitude at physiologically relevant stimulation frequency (e.g. 40 Hz) is another parameter of interest. It gives insight into the presynaptic endurance capacity, which depends on various factors such as Ca v 2.1 channel inactivation, synaptic vesicle pool size and synaptic vesicle recycling capacity. During a train of stimuli the EPP amplitude will decline to a plateau value, which is usually reached after about 10 stimuli. The mean plateau EPP amplitude is expressed as percentage of the amplitude of the first EPP in the train. When stimulating the whole nerve, all NMJs in the muscle preparation will become activated and may suffer some degree of fatigue. Therefore, a pause of at least 1 min should be included between sampling responses from successive NMJs. In some disease models transmission at NMJs can become completely blocked. It is possible to assess the prevalence of such ‘silent’ NMJs by

making many (40–60) fibre impalements within the muscle preparation before the addition of μ-conotoxin GIIIB and shortly inspect at each NMJ whether MEPPs are present and if an action potential (or a sub-threshold EPP) appears after nerve stimulation (Klooster et al., 2012). When performed by an experienced NMJ electrophysiologist this test yields a nearly 100% score of ‘active’ NMJs in healthy control (mouse) muscles. Microelectrodes can also be used to record the endplate current (EPC), the summed electrical current through all the ACh-activated postsynaptic AChRs at a NMJ. To this end, a given fibre must be impaled close to the NMJ with two glass capillary microelectrodes. The first microelectrode is used to record the membrane potential while the second is used to inject current. This so-called ‘two-electrode voltage-clamp’ configuration is used to hold the membrane voltage at an experimenter-determined fixed value (e.g. − 75 mV). Any deviations sensed by the voltage-recording electrode will instantaneously be compensated for by current injection via the second microelectrode. When the nerve is stimulated the injected current required to neutralize the EPP is recorded and displayed. This reflects the EPC. A special amplifier, equipped with a voltageclamp option is required to deliver the large clamp current (N300 nA) of full-size EPCs (Wood and Slater, 1997). Spontaneous miniature EPCs (MEPCs), which underlie the MEPPs can also be recorded. From mean EPC and MEPC amplitudes, the quantal content can be simply calculated without need for normalization or nonlinear correction because the membrane voltage is held constant. This represents a key advantage of two-electrode voltage clamp recordings, compared to MEPP/EPP measurements. EPC/MEPC recordings can also provide more detailed information about AChR channel opening and closing kinetics than EPP/MEPP measurements because the kinetics of the latter signals depend, in part, on the passive electrical capacitance of the muscle fibre membrane. A major disadvantage of the two-electrode voltage clamp is that it is technically much more challenging with a much lower success rate. In addition, with the large dimensions of muscle cells there is considerable risk of inadequate clamping of the voltage, resulting in underestimation of the synaptic currents. Another argument in favour for the simpler EPP/MEPP recordings is that the physiological signals used by the NMJ are potentials, not currents, and therefore EPP/ MEPP data may be more relevant to understanding the physiology. Extracellular field potential recording using a microelectrode can give some extra information about the proportion of active muscle fibres in the ex vivo preparation (Rogozhin et al., 2008). For these recordings the tip of a microelectrode is placed just above a flat muscle. e.g. epitrochleoanconeus, and inched towards the midline NMJ containing zone and the position is then optimized to obtain the largest CMAP response possible. If this technique is carefully and consistently applied signal amplitudes between muscles can be compared. Great care should be taken in maintaining equal distance to the muscle surface because extracellular signal amplitude greatly depends on electrode distance from the current source. Although not very often applied, microelectrodes can also be used to record presynaptic activity at motor nerve terminals. In perineural measurements the tip of a microelectrode is inserted under the perineurium of a nerve fascicle very close to the NMJ. All postsynaptic AChRs need to be blocked by a high concentration D-tubocurarine. Upon nerve stimulation, a complex waveform is recorded from the nerve terminal which can provide information on Na+ current just proximal to the nerve terminal and intraterminal K+ and Ca2 + currents (Brigant and Mallart, 1982). Disadvantages of the technique are that microelectrode placement considerably affects the waveform, and that the complexity of the waveform complicates interpretation. Patch-clamp recordings Patch-clamp is a variation of the voltage-clamp technique where a carefully fire-polished mouth of a relatively wide glass capillary tip (1–

Please cite this article as: Plomp, J.J., et al., Electrophysiological analysis of neuromuscular synaptic function in myasthenia gravis patients and animal models, Exp. Neurol. (2015), http://dx.doi.org/10.1016/j.expneurol.2015.01.007

J.J. Plomp et al. / Experimental Neurology xxx (2015) xxx–xxx

3 μm, 1–5 MΩ resistance) is pressed against a patch of membrane, forming a seal. Suction is then applied to increase the seal resistance (to giga-ohms) and to subsequently detach the membrane patch. It is also possible to stay in the cell-attached mode, leaving the cell membrane intact. Ideally, the membrane patch underneath the electrode tip contains one or only a few ion channels of interest. The electrophysiological characteristics of individual channels can then be studied by current measurement during stimulation by voltage steps or agonist applied via the electrode. Direct study of AChR with patch-clamp at the intact NMJ is difficult because of the overlying nerve terminal and perisynaptic Schwann cells (which first have to be removed by collagenase treatment). Nevertheless, single-channel characteristics of human and frog AChRs have been studied at the NMJ in this way, stimulating the channels with ACh (or analogues) in the pipette solution (Colquhoun and Sakmann, 1981; Milone et al., 1994). The patch-clamp technique is most often used in cellular expression systems in which (mutant) ion channels, including congenital myasthenic syndrome related mutant AChRs, can be expressed artificially. Patch-clamp analysis of mutant AChRs in expression systems, genetic mouse models or human biopsy NMJs from cases with congenital forms of MG has been discussed by others (Engel et al., 1998; Otero-Cruz et al., 2010). The best studied condition in this way is the autosomal dominantly inherited slow-channel syndrome. Using a patch-clamp measurement configuration in the cellattached mode it was demonstrated that AChR channel openings in response to stimulation by ACh were markedly prolonged. This leads to an abnormally slow decay of the EPP, triggering repetitive muscle fibre action potentials and, eventually, causing a persistent depolarization that completely prevents further action potentials due to inactivation of the voltage-gated Na+ channels. In addition, endplate myopathy occurs due to excess Ca2+ ingress through the open AChRs. Once specific mutations were identified in several subunit genes, subsequent singlechannel patch clamp study in expression systems confirmed abnormal AChR opening durations (Engel, 2012; Otero-Cruz et al., 2010; Rodriguez Cruz et al., 2014). Electrophysiological abnormalities in myasthenia gravis In vivo and ex vivo electrophysiological study of MG patients and animal models has contributed much to the understanding of the pathophysiological mechanisms at their NMJs. In addition, clinical EMG helps to diagnose MG patients and may give indications about a

9

possible pre- or postsynaptic localization of the NMJ defects. In recent years it has become clear that MuSK MG and its animal models are electrophysiologically somewhat distinct from AChR MG. Abnormal electrophysiology in patients AChR myasthenia gravis Current electromyographical testing for MG in patients usually includes RNS-EMG and sometimes SF-EMG. In RNS-EMG, the amplitude of the first CMAP in a train is often of normal amplitude, comparable to healthy subjects. However, in severely weak muscles the initial CMAP may already be reduced, suggesting that some NMJs have EPPs that are continuously sub-threshold. Generally, RNS for 5–10 stimuli at 1– 5 Hz reveals a decremental CMAP response (N 10% is considered abnormal) that may recover somewhat after the 4–5th stimulus. Small initial CMAP amplitudes and exaggerated decrements are common after maximum voluntary contraction for N30 s. CMAP decrement is found more often abnormal and more severe in nasal and facial muscles than in distal hand or foot muscles (Niks et al., 2003). Ideally, testing should be performed on multiple muscles, including clinically weak ones. However, RNS-EMG may be normal in a considerable proportion of MG patients, especially in mild generalized cases or when there is limited regional distribution of weakness (Howard, 2013; Meriggioli and Sanders, 2005; Niks et al., 2003). In general the sensitivity to detect a decremental CMAP with RNS-EMG is higher in more severe MG cases. Importantly, abnormal RNS-EMG can sometimes also be found in patients with diseases not primarily concerned with the NMJ (e.g. multiple sclerosis, motor neuron disease or motor neuropathies). EMG on its own cannot diagnose a specific disease (Howard, 2013; Meriggioli and Sanders, 2005). SF-EMG in MG patients is much more sensitive, showing abnormal jitter values in most MG patients, even in clinically unaffected muscles (Benatar, 2006; Padua et al., 2014). In the case of suspected MG with only ocular muscle weakness the orbicularis oculi muscle should be included because this has the highest diagnostic sensitivity (Milone et al., 1993; Oey et al., 1993; Padua et al., 2000; Valls-Canals et al., 2003). It is now well established that in AChR MG weakness is primarily caused by loss of postsynaptic AChRs and that the electrophysiological hallmark in biopsy NMJs is a (severe) reduction of MEPP/MEPC amplitude. However, the earliest ex vivo microelectrode studies of the NMJ in intercostal muscle biopsies of MG patients reported a presynaptic

Fig. 4. Compensatory increase in quantal content missing in MuSK myasthenia gravis. (A) Schematic indication of mean quantal content levels at neuromuscular junctions (NMJs) in AChR and MuSK myasthenia gravis (MG). Quantal content is roughly doubled in AChR MG but in MuSK MG it is equal or even somewhat lower than in healthy controls (for references see Abnormal electrophysiology in animal models). (B) Schematic representation of the inverse relationship of the quantal content vs. MEPP amplitude for individual NMJs. In AChR MG a compensatory mechanism causes more ACh quanta to be released at NMJs with small MEPP amplitudes. This regulatory mechanism seems to be absent in MuSK MG NMJs. Lines indicate the pattern of data points previously found in animal models of AChR MG, MuSK MG and in healthy control animals (Morsch et al., 2013; Plomp et al., 1992, 1995).

Please cite this article as: Plomp, J.J., et al., Electrophysiological analysis of neuromuscular synaptic function in myasthenia gravis patients and animal models, Exp. Neurol. (2015), http://dx.doi.org/10.1016/j.expneurol.2015.01.007

10

J.J. Plomp et al. / Experimental Neurology xxx (2015) xxx–xxx

defect and reduction of MEPP amplitude was initially controversial (Dahlback et al., 1961; Elmqvist et al., 1964). These inconsistent results and partially incorrect conclusions may, in retrospect, have been due to the still relatively primitive experimental techniques and perhaps the small cohorts. Also, cohorts may have included Lambert–Eaton myasthenic syndrome cases, which was not clearly identified as a presynaptic NMJ disorder at the time (Titulaer et al., 2011). Later electrophysiological MG biopsy studies unequivocally showed (severe) reductions of MEPP or MEPC amplitudes and decreased postsynaptic AChR density was proven by radioligand, histological and ultrastructural data (Albuquerque et al., 1976; Cull-Candy et al., 1978, 1980; Engel et al., 1977; Fambrough et al., 1973; Maselli et al., 1991b; Plomp et al., 1995; Ruff and Lennon, 1998; Tsujihata et al., 1980). AChR MG autoantibodies are most often IgG1 and IgG3 isotypes which, when bound to the antigen, activate the complement cascade (Rodgaard et al., 1987; Tuzun and Christadoss, 2013; Verschuuren et al., 2013; Vincent and Newsom-Davis, 1982). This culminates in membrane attack complex-mediated destruction of the postsynaptic folds of the NMJ (Engel et al., 1977; Maselli et al., 1991b), which results in reduced AChR density. Consequently, reduced EPP amplitudes and thus the loss of safety factor in neuromuscular transmission forms the primary electrophysiological defect in MG. In addition, two very important secondary electrophysiological effects occur (which perhaps have been somewhat neglected in the literature): 1) the action potential firing threshold becomes elevated due to loss of Nav1.4 channels that are concentrated in the troughs of the postsynaptic folds and 2) simplified folds have lower cytosolic electrical resistance and therefore the (already small) ACh-induced ion current will cause even less membrane depolarization and will be directed less efficiently at the remaining Nav1.4 channels. These secondary effects of membrane attack complex-mediated damage further lower the safety factor in neuromuscular transmission (Ruff, 2011). As a compensatory presynaptic response to the reduced postsynaptic sensitivity for ACh at AChR MG NMJs, motor nerve terminals increase the amount of ACh released per nerve impulse (Fig. 4). This was first demonstrated in a biochemical study of intercostal muscle biopsies (Molenaar et al., 1979), and confirmed and extended in later microelectrode studies (Cull-Candy et al., 1980; Plomp et al., 1995). The phenomenon has also been shown in a few congenital MG cases (Ohno et al., 1997). The quantal content increase is regulated at individual NMJs and presumably involves retrogradely acting trans-synaptic signalling factors (Plomp et al., 1995). This is most likely an evolutionary conserved, slowly acting homeostatic mechanism that allows synapses to adjust presynaptic neurotransmitter release level in response to changes in postsynaptic sensitivity, thus adjusting synaptic strength under impairing conditions. While studied in more detail in animal AChR MG models (see below), the identity of the retrograde messenger(s) and the presynaptic targets at (myasthenic) human NMJs has so far remained elusive. However, more extensive studies in Drosophila larval NMJs have brought forward interesting candidate molecules, such as the bone morphogenetic protein receptor ligand glass bottom boat or endostatin, a peptide fragment which is enzymatically cleaved from multiplexin, a Drosophila homologue of collagen XV and XVII (Davis, 2013; Frank, 2014; Wang et al., 2014). The elevated ACh release level may nevertheless be of only limited beneficial effect on neurotransmission at the myasthenic NMJ because it is not sustained at physiological repetitive activity of the synapse. EPPs evoked by 30 Hz nerve stimulation at NMJs in AChR MG run down to a level of ~40% of the initial EPP, as compared to 60% in normal controls (Niks et al., 2010). One possibility is that the pool of releasable ACh vesicles is depleted faster due to the elevated release level. MuSK myasthenia gravis MuSK MG is a rare disorder that has been identified only relatively recently in a proportion of AChR-seronegative MG patients (Hoch et al., 2001). The clinical, pharmacological and genetic profile of MuSK

MG differs from that of AChR MG (Evoli et al., 2003; Reddel et al., 2014; Sanders et al., 2003). Most importantly, acetylcholinesterase inhibitors are not effective and can even be contraproductive in a considerable proportion of MuSK MG patients (Evoli et al., 2003; Guptill et al., 2011; Punga et al., 2006). As in AChR MG, RNS-EMG shows a decrement and SF-EMG is abnormal in most MuSK MG patients, especially when done in facial muscles (Guptill et al., 2011; Oh et al., 2006). However, many MuSK MG patients show additional, myopathic EMG features such as fibrillation potentials and repetitive discharges, which are much less often seen in AChR MG (Oh et al., 2006; Sanders et al., 2003). An unusually high sensitivity to acetylcholinesterase inhibitors causing cholinergic neuromuscular hyperactivity may underlie these phenomena (Punga et al., 2006; Shin et al., 2014). Only a few human MuSK MG biopsy NMJ studies have been reported. Two early studies suggested that anti-MuSK autoantibodies might only be bystanding disease markers, given the relatively normal AChR staining and absence of IgG deposits observed at NMJs in intercostal and limb muscle biopsies of MuSK MG patients (Selcen et al., 2004; Shiraishi et al., 2005). However, later passive transfer studies in mice demonstrated clear pathogenic effects of MuSK MG IgG at mouse NMJs (further outlined below). Little or no complement deposits were found at the human NMJs. This is not surprising in view of anti-MuSK antibodies being predominantly of the IgG4 subclass, which does not activate the complement cascade (Bruggemann et al., 1987; McConville et al., 2004; Ohta et al., 2007; Tsiamalos et al., 2009). Electrophysiological investigation in two intercostal muscle biopsies revealed reduced MEPP amplitude (50–70%) and reduced EPP amplitudes (Niks et al., 2010; Selcen et al., 2004). Interestingly, no compensatory increase in quantal content was present. This may suggest that MuSK in some way is involved in the mechanism responsible for compensatory upregulation of quantal content at the myasthenic NMJ. In addition, Niks et al. reported decreased MEPP frequency and, at 30 Hz nerve stimulation, a somewhat exaggerated rundown of EPP amplitude, both indicative of presynaptic impairment. Much less is known about the NMJ electrophysiology of MG patients seronegative for AChR and MuSK antibodies but positive for antibodies against MuSK's binding partner LRP4. One study using RNS-EMG reported a CMAP decrement in facial muscles of all investigated patients (Pevzner et al., 2012). As far as we know, no muscle biopsy NMJ study has yet been published. Abnormal electrophysiology in animal models AChR myasthenia gravis models The first AChR MG animal model was generated in 1973 through active immunization of rabbits with purified AChR from the electric organ of the eel Electrophorus electricus (Patrick and Lindstrom, 1973). Animals developed flaccid paralysis after about four weeks, which could be relieved by an injection of the acetylcholinesterase inhibitor neostigmine. Antibodies against eel AChR were demonstrated in the serum of the rabbits. RNS-EMG showed a decremental CMAP, which was restored to normal by the injected neostigmine. Further studies extended the model to rats and guinea pigs, using eel AChR but also purified AChR from the electric organ of the ray Torpedo californica, and coined it experimental autoimmune MG (EAMG) (Lennon et al., 1975). Similarly, Rhesus monkeys immunized with Torpedo AChR developed acetylcholinesterase inhibitor reversible muscle weakness and showed CMAP decrement (Tarrab-Hazdai et al., 1975). Injection of anaesthetized rats with α-toxin from the Formosan cobra, known to block AChRs, was also shown to cause an MG-like decrement of CMAPs recorded at 3 Hz nerve stimulation. This too was reversed by acetylcholinesterase inhibition (Satyamurti et al., 1975). Most importantly, passive transfer of purified IgG from MG patients to mice subsequently demonstrated that MG IgG was pathogenic at the NMJ (Toyka et al., 1975). Ex vivo electrophysiological analyses showed small amplitude MEPPs at NMJs and a CMAP decrement upon repetitive nerve

Please cite this article as: Plomp, J.J., et al., Electrophysiological analysis of neuromuscular synaptic function in myasthenia gravis patients and animal models, Exp. Neurol. (2015), http://dx.doi.org/10.1016/j.expneurol.2015.01.007

J.J. Plomp et al. / Experimental Neurology xxx (2015) xxx–xxx

stimulation that was relieved by acetylcholinesterase inhibition. A first microelectrode study of NMJs in rat EAMG also clearly revealed small MEPPs and RNS-EMG showed a CMAP decrement which improved by acetylcholinesterase inhibition as well (Engel et al., 1976). Together with the human MG biopsy studies described above, these early animal studies proved that MG was caused by an autoantibody-mediated decrease in AChR density at the NMJ, causing the electrophysiological hallmark of small MEPPs. Later passive transfer studies in mice, rats and guinea pigs with rat monoclonal antibodies against the AChR yielded similar electrophysiological results, confirming and further strengthening this evidence (Lennon and Lambert, 1980; Richman et al., 1980). In the following decades, EAMG models have been extensively used in the study of immunological mechanisms of AChR MG as well as in testing experimental drugs, such as complement inhibitors (Baggi et al., 2012; Fuchs et al., 2014; Fuchs et al., 2008; Janssen et al., 2008; Piddlesden et al., 1996; Soltys et al., 2009; Tuzun and Christadoss, 2013; Zhou et al., 2007). Surprisingly few drug studies in mouse models of MG have included microelectrode and/or EMG electrophysiological analyses in the assessment of experimental drug effects in EAMG models. Perhaps the relative technical difficulty and need of specialized expertise underlie this lack of electrophysiological analyses. Currently some of the experimental drugs previously found beneficial in EAMG models are being tested in clinical trials (e.g. eculizumab, a complement C5 inhibitor in refractory generalized MG, ClinicalTrials.gov Identifier NCT01997229, and rituximab, which depletes B-cells, ClinicalTrials.gov Identifier: NCT02110706). For a recent review on emerging drug therapies in MG we direct the reader to Sieb (Sieb, 2014). The phenomenon of compensatory upregulation of presynaptic ACh release in AChR MG has been studied in some detail in the animal models. In a rat model of AChR MG with AChR loss induced by chronic low-dose injections of α-bungarotoxin, the upregulation of ACh release was shown to occur at the level of individual NMJs. The degree to which quantal content increased depended on the extent by which AChR density was reduced at any given synapse. This is reflected in an inverse correlation between MEPP amplitude and quantal content (Plomp et al., 1992) (Fig. 4B). Further analyses in rat and mouse α-bungarotoxininduced MG models showed that this form of presynaptic adaptation takes 2–4 weeks to fully develop, and that it is dependent on Ca2 + and Ca2+/calmodulin-dependent kinase II. It possibly involves presynaptic factors such as neurotrophic factor receptor tyrosine kinases, neurexins and Munc18-1 (Plomp et al., 1994; Sons et al., 2006; Sons et al., 2003). As in human AChR MG, the adaptation results in an initially enhanced ACh release rate, but also with extra EPP rundown during high frequency use of the synapse. This might partially neutralize the beneficial effect during physiological use of the muscle. The homeostatic upregulation of quantal content at the single NMJ level was also found in rats that were actively immunized with Torpedo AChR (Plomp et al., 1995). The adaptive mechanism does not seem to require direct AChR attack by antibody or α-bungarotoxin per se. NMJs of heterozygous neuregulin null-mutant mice, which have reduced postsynaptic AChR density due to the lowered neuregulin level, also display compensatory increased presynaptic ACh release (Sandrock et al., 1997). MuSK myasthenia gravis models On the basis of MuSK MG biopsy studies mentioned above, doubts were raised as to whether MuSK antibodies were the principal pathogenic agents (Lindstrom, 2004; Selcen et al., 2004; Shiraishi et al., 2005). However, electrophysiological and morphological NMJ analyses of the animal models generated since then have provided strong evidence in favour of such a role. Importantly, the post-developmental importance of MuSK was proven by conditional inactivation of the MuSK gene in early postnatal mouse muscle (Hesser et al., 2006). These young mice became severely myasthenic due to loss of AChRs, and electrophysiologically this was paralleled by severe MEPP and EPP amplitude reduction and some extra EPP rundown during high rate use of the synapse (JJ Plomp, unpublished results). Several laboratories

11

have generated myasthenic animal models through active immunization of mice, rats or rabbits with fragments of the extracellular domain of MuSK (Jha et al., 2006; Punga et al., 2011; Richman et al., 2012; Shigemoto et al., 2006; Ulusoy et al., 2014; Viegas et al., 2012; Xu et al., 2006). In those studies that performed RNS-EMG, clear decrement of CMAPs was observed. Microelectrode electrophysiological studies at NMJs of active immunization MuSK MG mice revealed reduced MEPP amplitudes (Jha et al., 2006; Mori et al., 2012b; Viegas et al., 2012). This was consistent with a reduced number of AChRs and fragmentation of AChR staining area in fluorescence microscopy. Presynaptic ACh release parameters were changed in addition. MEPP frequency was severely reduced (Mori et al., 2012b; Viegas et al., 2012), and the quantal content was either unchanged (Viegas et al., 2012) or reduced, (Mori et al., 2012b). This can be regarded as a defect in view of the homeostatic upregulation of ACh release that would be expected upon reduction of postsynaptic transmitter sensitivity (see AChR MG models above). The failure of presynaptic compensation in MuSK MG models suggests that MuSK, or downstream pathway factors, are needed for this form of synaptic strength homeostasis. Ultimate proof that human MuSK MG IgG is pathogenic came from several passive transfer studies using patient plasma or purified IgG. Looking back, the first study was actually performed in 1994, in mice that were injected with plasma and IgG from seven ‘seronegative’ MG patients (Burges et al., 1994), which later mostly appeared to be MuSK MG cases (Koneczny et al., 2014; Vincent et al., 2008). Here too, MEPP amplitude at NMJs was reduced without increase in quantal content, although no reduction in AChR number was shown with radioactive α-bungarotoxin labelling. It was not reported whether mice became weak. However, in the first passive transfer study in more recent years, using IgG from identified MuSK MG patients, reduction of postsynaptic AChR was shown using quantitative fluorescence microscopy, and CMAP decrement in RNS-EMG indicated NMJ dysfunction. This explained the observed frank muscle weakness and body weight loss (Cole et al., 2008). Later morphological analysis showed that AChRs ‘drift’ away from their original sites and that the normal AChR turnover process is insufficient in replacing them by new ones (Ghazanfari et al., 2014a). A first detailed microelectrode analysis was performed in another passive transfer study, using immunodeficient Nod/scid mice (preventing an immune response against the injected human IgG) and purified IgG4 from MuSK MG patients (Klooster et al., 2012). All tested MuSK MG IgG4 fractions (and not the IgG1–3 fractions from the same patients) induced severe muscle weakness with decrementing CMAPs in RNS-EMG, indicating a reduced safety factor of neuromuscular transmission at NMJs. The microelectrode NMJ study revealed severe reduction of MEPP and EPP amplitudes, reduced MEPP frequency and exaggerated depression of EPPs during physiological, high-rate nerve stimulation. Intriguingly, compensatory ACh release upregulation was found absent in this model as well. Similar synaptic electrophysiological defects were subsequently observed in mice passively transferred with total IgG from MuSK MG patients (Morsch et al., 2012; Viegas et al., 2012). Jointly, the passive and active MuSK MG mouse models show that MuSK autoantibodies (of the IgG4 subclass) are severely pathogenic and cause reduced MEPP amplitude without compensatory increased ACh release, paralleled by low MEPP frequency and extra EPP rundown. These combined post- and presynaptic defects explain the (fatigable) muscle weakness, and have also been observed in the few human biopsy studies (Niks et al., 2010; Selcen et al., 2004), giving high clinical relevance to the MuSK MG mouse models. The models will be instrumental in further pathophysiological analysis as well as in drug studies. For instance, the hypersensitivity to acetylcholinesterase inhibitors in MuSK MG patients has been reproduced and characterized in active immunization and passive transfer mouse models (Chroni and Punga, 2012; Mori et al., 2012b; Morsch et al., 2013). Further mouse studies suggest that 3,4-diaminopyridine and albuterol might be useful in the treatment of MuSK MG (Ghazanfari et al., 2014b; Mori et al., 2012a; Morsch et al., 2013).

Please cite this article as: Plomp, J.J., et al., Electrophysiological analysis of neuromuscular synaptic function in myasthenia gravis patients and animal models, Exp. Neurol. (2015), http://dx.doi.org/10.1016/j.expneurol.2015.01.007

12

J.J. Plomp et al. / Experimental Neurology xxx (2015) xxx–xxx

Only very recently, antibodies against MuSK's binding partner LRP4 have been demonstrated in some MG patients and the first animal models are now emerging. An active immunization mouse model and a passive transfer mouse model (using rabbit anti-LRP4 antibodies) both showed muscle weakness and body weight loss (Shen et al., 2013). AChR density at NMJs was low and areas were more fragmented. Electrophysiological analysis revealed NMJ features that are very similar to those shown in MuSK MG models, i.e. CMAP decrement in RNS-EMG and combined post- and presynaptic NMJ defects (low MEPP amplitude, MEPP frequency, EPP amplitude and quantal content). Conditional knock-out of the lrp4 gene in mouse muscle resulted in similar symptoms of muscle weakness and electrophysiological NMJ defects. This shows that loss of postsynaptic LRP4 from the adult NMJ is sufficient to cause the myasthenia (Barik et al., 2014). Conclusions The electrophysiology of NMJs of patients and animals models with ‘classical’ AChR MG has been studied quite extensively since the 1960– 70s. In conjunction with morphological studies this has yielded in a detailed picture of the primary and secondary functional and morphological defects at the NMJ in AChR MG. However, more electrophysiological characterizations of NMJs in muscle biopsies from patients with the rarer and more recently discovered MuSK and LRP4 MG subforms are clearly needed. On the other hand, a number of electrophysiological studies has been performed in active and passive immunization models of MuSK MG. These revealed that autoantibodies to MuSK cause secondary loss of postsynaptic electrophysiological sensitivity for the transmitter ACh. Interestingly, this was not accompanied by the compensatory upregulation of presynaptic ACh release, which can be readily found in NMJs of AChR MG patients and animal models. Electrophysiological study will remain instrumental in characterizing the detailed effects of autoantibodies causing myasthenia and in the assessment of the efficacy of experimental drugs. Acknowledgments The work of JP and JV is supported by the ‘Prinses Beatrix Spierfonds’, ‘Stichting Spieren voor Spieren’ and ‘L'Association Française contre les myopathies’. The work of MM and WP was supported by previous grants from the Muscular Dystrophy Association (USA) and National Health and Medical Research Council (Australia). References Albuquerque, E.X., Rash, J.E., Mayer, R.F., Satterfield, J.R., 1976. An electrophysiological and morphological study of the neuromuscular junction in patients with myasthenia gravis. Exp. Neurol. 51, 536–563. Angaut-Petit, D., Molgo, J., Connold, A.L., Faille, L., 1987. The levator auris longus muscle of the mouse: a convenient preparation for studies of short- and long-term presynaptic effects of drugs or toxins. Neurosci. Lett. 82, 83–88. Baggi, F., Antozzi, C., Toscani, C., Cordiglieri, C., 2012. Acetylcholine receptor-induced experimental myasthenia gravis: what have we learned from animal models after three decades? Arch. Immunol. Ther. Exp. (Warsz) 60, 19–30. Bannister, R.A., 2007. Bridging the myoplasmic gap: recent developments in skeletal muscle excitation–contraction coupling. J. Muscle Res. Cell Motil. 28, 275–283. Barik, A., Lu, Y., Sathyamurthy, A., Bowman, A., Shen, C., Li, L., Xiong, W.C., Mei, L., 2014. LRP4 is critical for neuromuscular junction maintenance. J. Neurosci. 34, 13892–13905. Benatar, M., 2006. A systematic review of diagnostic studies in myasthenia gravis. Neuromuscul. Disord. 16, 459–467. Bradley, S.A., Lyons, P.R., Slater, C.R., 1989. The epitrochleoanconeus muscle (ETA) of the mouse: a useful muscle for the study of motor innervation in vitro. J. Physiol. 415, 3P. Brigant, J.L., Mallart, A., 1982. Presynaptic currents in mouse motor endings. J. Physiol. 333, 619–636. Bruggemann, M., Williams, G.T., Bindon, C.I., Clark, M.R., Walker, M.R., Jefferis, R., Waldmann, H., Neuberger, M.S., 1987. Comparison of the effector functions of human immunoglobulins using a matched set of chimeric antibodies. J. Exp. Med. 166, 1351–1361. Burges, J., Vincent, A., Molenaar, P.C., Newsom-Davis, J., Peers, C., Wray, D., 1994. Passive transfer of seronegative myasthenia gravis to mice. Muscle Nerve 17, 1393–1400.

Choi, B.J., Imlach, W.L., Jiao, W., Wolfram, V., Wu, Y., Grbic, M., Cela, C., Baines, R.A., Nitabach, M.N., McCabe, B.D., 2014. Miniature neurotransmission regulates Drosophila synaptic structural maturation. Neuron 82, 618–634. Chroni, E., Punga, A.R., 2012. Neurophysiological characteristics of MuSK antibody positive myasthenia gravis mice: focal denervation and hypersensitivity to acetylcholinesterase inhibitors. J. Neurol. Sci. 316, 150–157. Clarke, G.L., Chen, J., Nishimune, H., 2012. Presynaptic active zone density during development and synaptic plasticity. Front. Mol. Neurosci. 5, 12. Cole, R.N., Reddel, S.W., Gervasio, O.L., Phillips, W.D., 2008. Anti-MuSK patient antibodies disrupt the mouse neuromuscular junction. Ann. Neurol. 63, 782–789. Colquhoun, D., Sakmann, B., 1981. Fluctuations in the microsecond time range of the current through single acetylcholine receptor ion channels. Nature 294, 464–466. Cronnelly, R., Dretchen, K.L., Sokoll, M.D., Long, J.P., 1973. Ketamine: myoneural activity and interaction with neuromuscular blocking agents. Eur. J. Pharmacol. 22, 17–22. Cull-Candy, S.G., Miledi, R., Trautmann, A., 1978. Acetylcholine-induced channels and transmitter release at human endplates. Nature 271, 74–75. Cull-Candy, S.G., Miledi, R., Trautmann, A., Uchitel, O.D., 1980. On the release of transmitter at normal, myasthenia gravis and myasthenic syndrome affected human endplates. J. Physiol. 299, 621–638. Dahlback, O., Elmqvist, D., Johns, T.R., Radner, S., Thesleff, S., 1961. An electrophysiologic study of the neuromuscular junction in myasthenia gravis. J. Physiol. 156, 336–343. Davis, G.W., 2013. Homeostatic signaling and the stabilization of neural function. Neuron 80, 718–728. Eken, T., 1998. Spontaneous electromyographic activity in adult rat soleus muscle. J. Neurophysiol. 80, 365–376. Elmqvist, D., Johns, T.R., Thesleff, S., 1960. A study of some electrophysiological properties of human intercostal muscle. J. Physiol. 154, 602–607. Elmqvist, D., Hofmann, W.W., Kugelberg, J., Quastel, D.M., 1964. An electrophysiological investigation of neuromuscular transmission in myasthenia gravis. J. Physiol. 174, 417–434. Engel, A.G., 2012. Current status of the congenital myasthenic syndromes. Neuromuscul. Disord. 22, 99–111. Engel, A.G., Tsujihata, M., Lambert, E.H., Lindstrom, J.M., Lennon, V.A., 1976. Experimental autoimmune myasthenia gravis: a sequential and quantitative study of the neuromuscular junction ultrastructure and electrophysiologic correlations. J. Neuropathol. Exp. Neurol. 35, 569–587. Engel, A.G., Lindstrom, J.M., Lambert, E.H., Lennon, V.A., 1977. Ultrastructural localization of the acetylcholine receptor in myasthenia gravis and in its experimental autoimmune model. Neurology 27, 307–315. Engel, A.G., Ohno, K., Milone, M., Sine, S.M., 1998. Congenital myasthenic syndromes. New insights from molecular genetic and patch-clamp studies. Ann. N. Y. Acad. Sci. 841, 140–156. Evoli, A., Tonali, P.A., Padua, L., Monaco, M.L., Scuderi, F., Batocchi, A.P., Marino, M., Bartoccioni, E., 2003. Clinical correlates with anti-MuSK antibodies in generalized seronegative myasthenia gravis. Brain 126, 2304–2311. Fambrough, D.M., Drachman, D.B., Satyamurti, S., 1973. Neuromuscular junction in myasthenia gravis: decreased acetylcholine receptors. Science 182, 293–295. Farrugia, M.E., Weir, A.I., Cleary, M., Cooper, S., Metcalfe, R., Mallik, A., 2009. Concentric and single fiber needle electrodes yield comparable jitter results in myasthenia gravis. Muscle Nerve 39, 579–585. Frank, C.A., 2014. Homeostatic plasticity at the Drosophila neuromuscular junction. Neuropharmacology 78, 63–74. Fuchs, S., Feferman, T., Meidler, R., Margalit, R., Sicsic, C., Brenner, T., Laub, O., Souroujon, M.C., 2008. Immunosuppression of EAMG by IVIG is mediated by a disease-specific anti-immunoglobulin fraction. Ann. N. Y. Acad. Sci. 1132, 244–248. Fuchs, S., Aricha, R., Reuveni, D., Souroujon, M.C., 2014. Experimental autoimmune myasthenia gravis (EAMG): from immunochemical characterization to therapeutic approaches. J. Autoimmun. 54, 51–59. Fucile, S., Sucapane, A., Grassi, F., Eusebi, F., Engel, A.G., 2006. The human adult subtype ACh receptor channel has high Ca2+ permeability and predisposes to endplate Ca2+ overloading. J. Physiol. 573, 35–43. Gasperi, C., Melms, A., Schoser, B., Zhang, Y., Meltoranta, J., Risson, V., Schaeffer, L., Schalke, B., Kroger, S., 2014. Anti-agrin autoantibodies in myasthenia gravis. Neurology 82, 1976–1983. Ghazanfari, N., Fernandez, K.J., Murata, Y., Morsch, M., Ngo, S.T., Reddel, S.W., Noakes, P.G., Phillips, W.D., 2011. Muscle specific kinase: organiser of synaptic membrane domains. Int. J. Biochem. Cell Biol. 43, 295–298. Ghazanfari, N., Morsch, M., Reddel, S.W., Liang, S.X., Phillips, W.D., 2014a. Muscle-specific kinase (MuSK) autoantibodies suppress the MuSK pathway and ACh receptor retention at the mouse neuromuscular junction. J. Physiol. 592, 2881–2897. Ghazanfari, N., Morsch, M., Tse, N., Reddel, S.W., Phillips, W.D., 2014b. Effects of the beta2adrenoceptor agonist, albuterol, in a mouse model of anti-MuSK myasthenia gravis. PLoS One 9, e87840. Guptill, J.T., Sanders, D.B., Evoli, A., 2011. Anti-musk antibody myasthenia gravis: clinical findings and response to treatment in two large cohorts. Muscle Nerve 44, 36–40. Hennig, R., Lomo, T., 1985. Firing patterns of motor units in normal rats. Nature 314, 164–166. Hesser, B.A., Henschel, O., Witzemann, V., 2006. Synapse disassembly and formation of new synapses in postnatal muscle upon conditional inactivation of MuSK. Mol. Cell. Neurosci. 31, 470–480. Higuchi, O., Hamuro, J., Motomura, M., Yamanashi, Y., 2011. Autoantibodies to lowdensity lipoprotein receptor-related protein 4 in myasthenia gravis. Ann. Neurol. 69, 418–422. Hoch, W., McConville, J., Helms, S., Newsom-Davis, J., Melms, A., Vincent, A., 2001. Autoantibodies to the receptor tyrosine kinase MuSK in patients with myasthenia gravis without acetylcholine receptor antibodies. Nat. Med. 7, 365–368.

Please cite this article as: Plomp, J.J., et al., Electrophysiological analysis of neuromuscular synaptic function in myasthenia gravis patients and animal models, Exp. Neurol. (2015), http://dx.doi.org/10.1016/j.expneurol.2015.01.007

J.J. Plomp et al. / Experimental Neurology xxx (2015) xxx–xxx Howard Jr., J.F., 2013. Electrodiagnosis of disorders of neuromuscular transmission. Phys. Med. Rehabil. Clin. N. Am. 24, 169–192. Janssen, S.P., Phernambucq, M., Martinez-Martinez, P., De Baets, M.H., Losen, M., 2008. Immunosuppression of experimental autoimmune myasthenia gravis by mycophenolate mofetil. J. Neuroimmunol. 201–202, 111–120. Jha, S., Xu, K., Maruta, T., Oshima, M., Mosier, D.R., Atassi, M.Z., Hoch, W., 2006. Myasthenia gravis induced in mice by immunization with the recombinant extracellular domain of rat muscle-specific kinase (MuSK). J. Neuroimmunol. 175, 107–117. Kaja, S., van de Ven, R.C., Broos, L.A., Veldman, H., van Dijk, J.G., Verschuuren, J.J., Frants, R.R., Ferrari, M.D., van den Maagdenberg, A.M., Plomp, J.J., 2005. Gene dosagedependent transmitter release changes at neuromuscular synapses of CACNA1A R192Q knockin mice are non-progressive and do not lead to morphological changes or muscle weakness. Neuroscience 135, 81–95. Kaja, S., van de Ven, R.C., van Dijk, J.G., Verschuuren, J.J., Arahata, K., Frants, R.R., Ferrari, M.D., van den Maagdenberg, A.M., Plomp, J.J., 2007. Severely impaired neuromuscular synaptic transmission causes muscle weakness in the Cacna1a-mutant mouse rolling Nagoya. Eur. J. Neurosci. 25, 2009–2020. Klooster, R., Plomp, J.J., Huijbers, M.G., Niks, E.H., Straasheijm, K.R., Detmers, F.J., Hermans, P.W., Sleijpen, K., Verrips, A., Losen, M., Martinez-Martinez, P., de Baets, M.H., van der Maarel, S.M., Verschuuren, J.J., 2012. Muscle-specific kinase myasthenia gravis IgG4 autoantibodies cause severe neuromuscular junction dysfunction in mice. Brain 135, 1081–1101. Koneczny, I., Cossins, J., Vincent, A., 2014. The role of muscle-specific tyrosine kinase (MuSK) and mystery of MuSK myasthenia gravis. J. Anat. 224, 29–35. Leite, M.I., Jacob, S., Viegas, S., Cossins, J., Clover, L., Morgan, B.P., Beeson, D., Willcox, N., Vincent, A., 2008. IgG1 antibodies to acetylcholine receptors in ‘seronegative’ myasthenia gravis. Brain 131, 1940–1952. Lennon, V.A., Lambert, E.H., 1980. Myasthenia gravis induced by monoclonal antibodies to acetylcholine receptors. Nature 285, 238–240. Lennon, V.A., Lindstrom, J.M., Seybold, M.E., 1975. Experimental autoimmune myasthenia: a model of myasthenia gravis in rats and guinea pigs. J. Exp. Med. 141, 1365–1375. Lin, T.S., Cheng, K.S., 1998. Characterization of the relationship between motor end-plate jitter and the safety factor. Muscle Nerve 21, 628–636. Lindstrom, J., 2004. Is “seronegative” MG explained by autoantibodies to MuSK? Neurology 62, 1920–1921. Maselli, R.A., Mass, D.P., Distad, B.J., Richman, D.P., 1991a. Anconeus muscle: a human muscle preparation suitable for in-vitro microelectrode studies. Muscle Nerve 14, 1189–1192. Maselli, R.A., Richman, D.P., Wollmann, R.L., 1991b. Inflammation at the neuromuscular junction in myasthenia gravis. Neurology 41, 1497–1504. McArdle, J.J., Angaut-Petit, D., Mallart, A., Bournaud, R., Faille, L., Brigant, J.L., 1981. Advantages of the triangularis sterni muscle of the mouse for investigations of synaptic phenomena. J. Neurosci. Methods 4, 109–115. McConville, J., Farrugia, M.E., Beeson, D., Kishore, U., Metcalfe, R., Newsom-Davis, J., Vincent, A., 2004. Detection and characterization of MuSK antibodies in seronegative myasthenia gravis. Ann. Neurol. 55, 580–584. McLachlan, E.M., Martin, A.R., 1981. Non-linear summation of end-plate potentials in the frog and mouse. J. Physiol. 311, 307–324. Meriggioli, M.N., Sanders, D.B., 2005. Advances in the diagnosis of neuromuscular junction disorders. Am. J. Phys. Med. Rehabil. 84, 627–638. Meriney, S.D., Dittrich, M., 2013. Organization and function of transmitter release sites at the neuromuscular junction. J. Physiol. 591, 3159–3165. Milone, M., Monaco, M.L., Evoli, A., Servidei, S., Tonali, P., 1993. Ocular myasthenia: diagnostic value of single fibre EMG in the orbicularis oculi muscle. J. Neurol. Neurosurg. Psychiatry 56, 720–721. Milone, M., Hutchinson, D.O., Engel, A.G., 1994. Patch-clamp analysis of the properties of acetylcholine receptor channels at the normal human endplate. Muscle Nerve 17, 1364–1369. Molenaar, P.C., Polak, R.L., Miledi, R., Alema, S., Vincent, A., Newsom-Davis, J., 1979. Acetylcholine in intercostal muscle from myasthenia gravis patients and in rat diaphragm after blockade of acetylcholine receptors. Prog. Brain Res. 49, 449–458. Molenaar, P.C., Oen, B.S., Plomp, J.J., Van Kempen, G.T., Jennekens, F.G., Hesselmans, L.F., 1991. A non-immunogenic myasthenia gravis model and its application in a study of transsynaptic regulation at the neuromuscular junction. Eur. J. Pharmacol. 196, 93–101. Mori, S., Kishi, M., Kubo, S., Akiyoshi, T., Yamada, S., Miyazaki, T., Konishi, T., Maruyama, N., Shigemoto, K., 2012a. 3,4-Diaminopyridine improves neuromuscular transmission in a MuSK antibody-induced mouse model of myasthenia gravis. J. Neuroimmunol. 245, 75–78. Mori, S., Kubo, S., Akiyoshi, T., Yamada, S., Miyazaki, T., Hotta, H., Desaki, J., Kishi, M., Konishi, T., Nishino, Y., Miyazawa, A., Maruyama, N., Shigemoto, K., 2012b. Antibodies against muscle-specific kinase impair both presynaptic and postsynaptic functions in a murine model of myasthenia gravis. Am. J. Pathol. 180, 798–810. Morsch, M., Reddel, S.W., Ghazanfari, N., Toyka, K.V., Phillips, W.D., 2012. Muscle specific kinase autoantibodies cause synaptic failure through progressive wastage of postsynaptic acetylcholine receptors. Exp. Neurol. 237, 286–295. Morsch, M., Reddel, S.W., Ghazanfari, N., Toyka, K.V., Phillips, W.D., 2013. Pyridostigmine but not 3,4-diaminopyridine exacerbates ACh receptor loss and myasthenia induced in mice by muscle-specific kinase autoantibody. J. Physiol. 591, 2747–2762. Nagwaney, S., Harlow, M.L., Jung, J.H., Szule, J.A., Ress, D., Xu, J., Marshall, R.M., McMahan, U.J., 2009. Macromolecular connections of active zone material to docked synaptic vesicles and presynaptic membrane at neuromuscular junctions of mouse. J. Comp. Neurol. 513, 457–468.

13

Newland, C.F., Beeson, D., Vincent, A., Newsom-Davis, J., 1995. Functional and nonfunctional isoforms of the human muscle acetylcholine receptor. J. Physiol. 489 (Pt 3), 767–778. Niks, E.H., Badrising, U.A., Verschuuren, J.J., van Dijk, J.G., 2003. Decremental response of the nasalis and hypothenar muscles in myasthenia gravis. Muscle Nerve 28, 236–238. Niks, E.H., Kuks, J.B., Wokke, J.H., Veldman, H., Bakker, E., Verschuuren, J.J., Plomp, J.J., 2010. Pre- and postsynaptic neuromuscular junction abnormalities in musk myasthenia. Muscle Nerve 42, 283–288. Nishimune, H., 2012. Active zones of mammalian neuromuscular junctions: formation, density, and aging. Ann. N. Y. Acad. Sci. 1274, 24–32. Nishimune, H., Numata, T., Chen, J., Aoki, Y., Wang, Y., Starr, M.P., Mori, Y., Stanford, J.A., 2012. Active zone protein Bassoon co-localizes with presynaptic calcium channel, modifies channel function, and recovers from aging related loss by exercise. PLoS One 7, e38029. Oey, P.L., Wieneke, G.H., Hoogenraad, T.U., van Huffelen, A.C., 1993. Ocular myasthenia gravis: the diagnostic yield of repetitive nerve stimulation and stimulated single fiber EMG of orbicularis oculi muscle and infrared reflection oculography. Muscle Nerve 16, 142–149. Oh, S.J., Hatanaka, Y., Hemmi, S., Young, A.M., Scheufele, M.L., Nations, S.P., Lu, L., Claussen, G.C., Wolfe, G.I., 2006. Repetitive nerve stimulation of facial muscles in MuSK antibody-positive myasthenia gravis. Muscle Nerve 33, 500–504. Ohno, K., Quiram, P.A., Milone, M., Wang, H.L., Harper, M.C., Pruitt, J.N., Brengman, J.M., Pao, L., Fischbeck, K.H., Crawford, T.O., Sine, S.M., Engel, A.G., 1997. Congenital myasthenic syndromes due to heteroallelic nonsense/missense mutations in the acetylcholine receptor epsilon subunit gene: identification and functional characterization of six new mutations. Hum. Mol. Genet. 6, 753–766. Ohta, K., Shigemoto, K., Fujinami, A., Maruyama, N., Konishi, T., Ohta, M., 2007. Clinical and experimental features of MuSK antibody positive MG in Japan. Eur. J. Neurol. 14, 1029–1034. Otero-Cruz, J.D., Baez-Pagan, C.A., Dorna-Perez, L., Grajales-Reyes, G.E., Ramirez-Ordonez, R.T., Luciano, C.A., Gomez, C.M., Lasalde-Dominicci, J.A., 2010. Decoding pathogenesis of slow-channel congenital myasthenic syndromes using recombinant expression and mice models. P. R. Health Sci. J. 29, 4–17. Padua, L., Stalberg, E., LoMonaco, M., Evoli, A., Batocchi, A., Tonali, P., 2000. SFEMG in ocular myasthenia gravis diagnosis. Clin. Neurophysiol. 111, 1203–1207. Padua, L., Caliandro, P., Di, I.G., Pazzaglia, C., Ciaraffa, F., Evoli, A., 2014. Reliability of SFEMG in diagnosing myasthenia gravis: sensitivity and specificity calculated on 100 prospective cases. Clin. Neurophysiol. 125, 1270–1273. Papathanasiou, E.S., Zamba-Papanicolaou, E., 2012. A comparison between disposable and reusable single fiber needle electrodes in relation to stimulated single fiber studies. Clin. Neurophysiol. 123, 1437–1439. Patrick, J., Lindstrom, J., 1973. Autoimmune response to acetylcholine receptor. Science 180, 871–872. Pevzner, A., Schoser, B., Peters, K., Cosma, N.C., Karakatsani, A., Schalke, B., Melms, A., Kroger, S., 2012. Anti-LRP4 autoantibodies in AChR- and MuSK-antibody-negative myasthenia gravis. J. Neurol. 259, 427–435. Piddlesden, S.J., Jiang, S., Levin, J.L., Vincent, A., Morgan, B.P., 1996. Soluble complement receptor 1 (sCR1) protects against experimental autoimmune myasthenia gravis. J. Neuroimmunol. 71, 173–177. Plomp, J.J., Van Kempen, G.T., Molenaar, P.C., 1992. Adaptation of quantal content to decreased postsynaptic sensitivity at single endplates in alpha-bungarotoxin-treated rats. J. Physiol. 458, 487–499. Plomp, J.J., Van Kempen, G.T., Molenaar, P.C., 1994. The upregulation of acetylcholine release at endplates of alpha-bungarotoxin-treated rats: its dependency on calcium. J. Physiol. 478 (Pt 1), 125–136. Plomp, J.J., Van Kempen, G.T., De Baets, M.B., Graus, Y.M., Kuks, J.B., Molenaar, P.C., 1995. Acetylcholine release in myasthenia gravis: regulation at single end-plate level. Ann. Neurol. 37, 627–636. Punga, A.R., Flink, R., Askmark, H., Stalberg, E.V., 2006. Cholinergic neuromuscular hyperactivity in patients with myasthenia gravis seropositive for MuSK antibody. Muscle Nerve 34, 111–115. Punga, A.R., Lin, S., Oliveri, F., Meinen, S., Ruegg, M.A., 2011. Muscle-selective synaptic disassembly and reorganization in MuSK antibody positive MG mice. Exp. Neurol. 230, 207–217. Reddel, S.W., Morsch, M., Phillips, W.D., 2014. Clinical and scientific aspects of muscle-specific tyrosine kinase-related myasthenia gravis. Curr. Opin. Neurol. 27, 558–565. Richman, D.P., Gomez, C.M., Berman, P.W., Burres, S.A., Fitch, F.W., Arnason, B.G., 1980. Monoclonal anti-acetylcholine receptor antibodies can cause experimental myasthenia. Nature 286, 738–739. Richman, D.P., Nishi, K., Morell, S.W., Chang, J.M., Ferns, M.J., Wollmann, R.L., Maselli, R.A., Schnier, J., Agius, M.A., 2012. Acute severe animal model of anti-muscle-specific kinase myasthenia: combined postsynaptic and presynaptic changes. Arch. Neurol. 69, 453–460. Rodgaard, A., Nielsen, F.C., Djurup, R., Somnier, F., Gammeltoft, S., 1987. Acetylcholine receptor antibody in myasthenia gravis: predominance of IgG subclasses 1 and 3. Clin. Exp. Immunol. 67, 82–88. Rodriguez Cruz, P.M., Palace, J., Beeson, D., 2014. Inherited disorders of the neuromuscular junction: an update. J. Neurol. 261, 2234–2243. Rogozhin, A.A., Pang, K.K., Bukharaeva, E., Young, C., Slater, C.R., 2008. Recovery of mouse neuromuscular junctions from single and repeated injections of botulinum neurotoxin A. J. Physiol. 586, 3163–3182. Ruff, R.L., 2011. Endplate contributions to the safety factor for neuromuscular transmission. Muscle Nerve 44, 854–861.

Please cite this article as: Plomp, J.J., et al., Electrophysiological analysis of neuromuscular synaptic function in myasthenia gravis patients and animal models, Exp. Neurol. (2015), http://dx.doi.org/10.1016/j.expneurol.2015.01.007

14

J.J. Plomp et al. / Experimental Neurology xxx (2015) xxx–xxx

Ruff, R.L., Lennon, V.A., 1998. End-plate voltage-gated sodium channels are lost in clinical and experimental myasthenia gravis. Ann. Neurol. 43, 370–379. Sanders, D.B., El-Salem, K., Massey, J.M., McConville, J., Vincent, A., 2003. Clinical aspects of MuSK antibody positive seronegative MG. Neurology 60, 1978–1980. Sandrock Jr., A.W., Dryer, S.E., Rosen, K.M., Gozani, S.N., Kramer, R., Theill, L.E., Fischbach, G.D., 1997. Maintenance of acetylcholine receptor number by neuregulins at the neuromuscular junction in vivo. Science 276, 599–603. Satyamurti, S., Drachman, D.B., Slone, F., 1975. Blockade of acetylcholine receptors: a model of myasthenia gravis. Science 187, 955–957. Selcen, D., Fukuda, T., Shen, X.M., Engel, A.G., 2004. Are MuSK antibodies the primary cause of myasthenic symptoms? Neurology 62, 1945–1950. Shen, C., Lu, Y., Zhang, B., Figueiredo, D., Bean, J., Jung, J., Wu, H., Barik, A., Yin, D.M., Xiong, W.C., Mei, L., 2013. Antibodies against low-density lipoprotein receptor-related protein 4 induce myasthenia gravis. J. Clin. Invest. 123, 5190–5202. Shigemoto, K., Kubo, S., Maruyama, N., Hato, N., Yamada, H., Jie, C., Kobayashi, N., Mominoki, K., Abe, Y., Ueda, N., Matsuda, S., 2006. Induction of myasthenia by immunization against muscle-specific kinase. J. Clin. Invest. 116, 1016–1024. Shin, H.Y., Park, H.J., Lee, H.E., Choi, Y.C., Kim, S.M., 2014. Clinical and electrophysiologic responses to acetylcholinesterase inhibitors in MuSK-antibody-positive myasthenia gravis: evidence for cholinergic neuromuscular hyperactivity. J. Clin. Neurol. 10, 119–124. Shiraishi, H., Motomura, M., Yoshimura, T., Fukudome, T., Fukuda, T., Nakao, Y., Tsujihata, M., Vincent, A., Eguchi, K., 2005. Acetylcholine receptors loss and postsynaptic damage in MuSK antibody-positive myasthenia gravis. Ann. Neurol. 57, 289–293. Sieb, J.P., 2014. Myasthenia gravis: an update for the clinician. Clin. Exp. Immunol. 175, 408–418. Sine, S.M., 2012. End-plate acetylcholine receptor: structure, mechanism, pharmacology, and disease. Physiol. Rev. 92, 1189–1234. Slater, C.R., 2003. Structural determinants of the reliability of synaptic transmission at the vertebrate neuromuscular junction. J. Neurocytol. 32, 505–522. Slater, C.R., 2008. Structural factors influencing the efficacy of neuromuscular transmission. Ann. N. Y. Acad. Sci. 1132, 1–12. Slater, C.R., Lyons, P.R., Walls, T.J., Fawcett, P.R., Young, C., 1992. Structure and function of neuromuscular junctions in the vastus lateralis of man. A motor point biopsy study of two groups of patients. Brain 115 (Pt 2), 451–478. Soltys, J., Kusner, L.L., Young, A., Richmonds, C., Hatala, D., Gong, B., Shanmugavel, V., Kaminski, H.J., 2009. Novel complement inhibitor limits severity of experimentally myasthenia gravis. Ann. Neurol. 65, 67–75. Sons, M.S., Verhage, M., Plomp, J.J., 2003. Role of Munc18-1 in synaptic plasticity at the myasthenic neuromuscular junction. Ann. N. Y. Acad. Sci. 998, 404–406. Sons, M.S., Busche, N., Strenzke, N., Moser, T., Ernsberger, U., Mooren, F.C., Zhang, W., Ahmad, M., Steffens, H., Schomburg, E.D., Plomp, J.J., Missler, M., 2006. Alphaneurexins are required for efficient transmitter release and synaptic homeostasis at the mouse neuromuscular junction. Neuroscience 138, 433–446. Stalberg, E.V., Sanders, D.B., 2009. Jitter recordings with concentric needle electrodes. Muscle Nerve 40, 331–339. Sudhof, T.C., 2013. Neurotransmitter release: the last millisecond in the life of a synaptic vesicle. Neuron 80, 675–690. Sutton, M.A., Ito, H.T., Cressy, P., Kempf, C., Woo, J.C., Schuman, E.M., 2006. Miniature neurotransmission stabilizes synaptic function via tonic suppression of local dendritic protein synthesis. Cell 125, 785–799. Szule, J.A., Harlow, M.L., Jung, J.H., De-Miguel, F.F., Marshall, R.M., McMahan, U.J., 2012. Regulation of synaptic vesicle docking by different classes of macromolecules in active zone material. PLoS One 7, e33333. Tarrab-Hazdai, R., Aharonov, A., Silman, I., Fuchs, S., Abramsky, O., 1975. Experimental autoimmune myasthenia induced in monkeys by purified acetylcholine receptor. Nature 256, 128–130. Titulaer, M.J., Lang, B., Verschuuren, J.J., 2011. Lambert–Eaton myasthenic syndrome: from clinical characteristics to therapeutic strategies. Lancet Neurol. 10, 1098–1107. Toyka, K.V., Brachman, D.B., Pestronk, A., Kao, I., 1975. Myasthenia gravis: passive transfer from man to mouse. Science 190, 397–399.

Tsiamalos, P., Kordas, G., Kokla, A., Poulas, K., Tzartos, S.J., 2009. Epidemiological and immunological profile of muscle-specific kinase myasthenia gravis in Greece. Eur. J. Neurol. 16, 925–930. Tsujihata, M., Hazama, R., Ishii, N., Ide, Y., Takamori, M., 1980. Ultrastructural localization of acetylcholine receptor at the motor endplate: myasthenia gravis and other neuromuscular diseases. Neurology 30, 1203–1211. Tuzun, E., Christadoss, P., 2013. Complement associated pathogenic mechanisms in myasthenia gravis. Autoimmun. Rev. 12, 904–911. Ulusoy, C., Kim, E., Tuzun, E., Huda, R., Yilmaz, V., Poulas, K., Trakas, N., Skriapa, L., Niarchos, A., Strait, R.T., Finkelman, F.D., Turan, S., Zisimopoulou, P., Tzartos, S., Saruhan-Direskeneli, G., Christadoss, P., 2014. Preferential production of IgG1, IL-4 and IL-10 in MuSK-immunized mice. Clin. Immunol. 151, 155–163. Valls-Canals, J., Povedano, M., Montero, J., Pradas, J., 2003. Stimulated single-fiber EMG of the frontalis and orbicularis oculi muscles in ocular myasthenia gravis. Muscle Nerve 28, 501–503. Vautrin, J., Mambrini, J., 1989. Synaptic current between neuromuscular junction folds. J. Theor. Biol. 140, 479–498. Verschuuren, J.J., Huijbers, M.G., Plomp, J.J., Niks, E.H., Molenaar, P.C., Martinez-Martinez, P., Gomez, A.M., de Baets, M.H., Losen, M., 2013. Pathophysiology of myasthenia gravis with antibodies to the acetylcholine receptor, muscle-specific kinase and low-density lipoprotein receptor-related protein 4. Autoimmun. Rev. 12, 918–923. Viegas, S., Jacobson, L., Waters, P., Cossins, J., Jacob, S., Leite, M.I., Webster, R., Vincent, A., 2012. Passive and active immunization models of MuSK-Ab positive myasthenia: electrophysiological evidence for pre and postsynaptic defects. Exp. Neurol. 234, 506–512. Vincent, A., Newsom-Davis, J., 1982. Acetylcholine receptor antibody characteristics in myasthenia gravis. I. Patients with generalized myasthenia or disease restricted to ocular muscles. Clin. Exp. Immunol. 49, 257–265. Vincent, A., Leite, M.I., Farrugia, M.E., Jacob, S., Viegas, S., Shiraishi, H., Benveniste, O., Morgan, B.P., Hilton-Jones, D., Newsom-Davis, J., Beeson, D., Willcox, N., 2008. Myasthenia gravis seronegative for acetylcholine receptor antibodies. Ann. N. Y. Acad. Sci. 1132, 84–92. Wang, T., Hauswirth, A.G., Tong, A., Dickman, D.K., Davis, G.W., 2014. Endostatin is a transsynaptic signal for homeostatic synaptic plasticity. Neuron 83, 616–629. Waud, B.E., 1979. Decrease in dose requirement of D-tubocurarine by volatile anesthetics. Anesthesiology 51, 298–302. Wood, S.J., Slater, C.R., 1995. Action potential generation in rat slow- and fast-twitch muscles. J. Physiol. 486 (Pt 2), 401–410. Wood, S.J., Slater, C.R., 1997. The contribution of postsynaptic folds to the safety factor for neuromuscular transmission in rat fast- and slow-twitch muscles. J. Physiol. 500 (Pt 1), 165–176. Wood, S.J., Slater, C.R., 2001. Safety factor at the neuromuscular junction. Prog. Neurobiol. 64, 393–429. Wu, B., Goluszko, E., Huda, R., Tuzun, E., Christadoss, P., 2013. Experimental autoimmune myasthenia gravis in the mouse. Curr. Protoc. Immunol. Unit 15, 15.8.1–15.8.26. Xu, K., Jha, S., Hoch, W., Dryer, S.E., 2006. Delayed synapsing muscles are more severely affected in an experimental model of MuSK-induced myasthenia gravis. Neuroscience 143, 655–659. Zhang, B., Tzartos, J.S., Belimezi, M., Ragheb, S., Bealmear, B., Lewis, R.A., Xiong, W.C., Lisak, R.P., Tzartos, S.J., Mei, L., 2012. Autoantibodies to lipoproteinrelated protein 4 in patients with double-seronegative myasthenia gravis. Arch. Neurol. 69, 445–451. Zhang, B., Shen, C., Bealmear, B., Ragheb, S., Xiong, W.C., Lewis, R.A., Lisak, R.P., Mei, L., 2014. Autoantibodies to agrin in myasthenia gravis patients. PLoS One 9, e91816. Zhou, Y., Gong, B., Lin, F., Rother, R.P., Medof, M.E., Kaminski, H.J., 2007. Anti-C5 antibody treatment ameliorates weakness in experimentally acquired myasthenia gravis. J. Immunol. 179, 8562–8567. Zitman, F.M., Todorov, B., Jacobs, B.C., Verschuuren, J.J., Furukawa, K., Furukawa, K., Willison, H.J., Plomp, J.J., 2008. Neuromuscular synaptic function in mice lacking major subsets of gangliosides. Neuroscience 156, 885–897.

Please cite this article as: Plomp, J.J., et al., Electrophysiological analysis of neuromuscular synaptic function in myasthenia gravis patients and animal models, Exp. Neurol. (2015), http://dx.doi.org/10.1016/j.expneurol.2015.01.007