Why does acetylcholine exacerbate myasthenia caused by anti‐MuSK antibodies?

Abstract

According to current knowledge, MuSK is a transmembrane protein with four immunoglobulin-like domains and a cysteine-rich domain extracellularly and an intracellular kinase domain. It is expressed on the surface of non-innervated myotubes and on the postsynaptic membrane of the neuromuscular junction (NMJ). Activation of MuSK is a prerequisite for local concentration of acetylcholine receptor (AChR) molecules via rapsyn. In non-innervated myotubes, MuSK is activated by cytosolic Dok-7 to form primitive AChR clusters. At the innervated NMJ, MuSK is activated by neural agrin through interaction with its co-receptor, the transmembrane lipoprotein receptor-related protein LRP4. At the NMJ, MuSK also participates in anchoring the collagenic tail of AChE, collagen Q (ColQ), in the synaptic space. By interacting with additional postsynaptic proteins that include Dok-7, MuSK emerges as a master organizer of postsynaptic development of the NMJ (Wu et al. 2010). Most patients suffering from autoimmune myasthenia gravis (MG) harbour anti-AChR antibodies (AChR-MG) but 5–15% do not. Clinical interest in MuSK arose when enzyme-linked immunosorbent assay (ELISA) and immunoprecipitation experiments revealed that a high proportion of AChR-seronegative myasthenic sera immunoreacted with MuSK (Hoch et al. 2001). Subsequent studies of MuSK antibody-positive myasthenia (MuSK-MG) showed it differs from AChR-MG in several respects. First, the acetylcholinesterase (AChE) inhibitor pyridostigmine, which is effective in AChR-MG, is ineffective in, or worsens, MuSK-MG. Second, MuSK-MG selectively affects bulbar, cervical and respiratory muscles whereas generalized AChR-MG also affects limb muscles. Third, intercostal muscle NMJs in AChR-MG are depleted of AChR, have low-amplitude miniature endplate potentials (MEPPs) and show signs of complement-mediated lysis of the junctional folds whereas in MuSK-MG, NMJs of intercostal or biceps brachii muscles have normal AChR content, generate normal-amplitude MEPPs and have well-preserved junctional folds that bind little or no complement. The reasons for the differences between AChR-MG and MuSK-MG are now reasonably well understood. First, the pathogenic MuSK immunoglobulin is predominantly of class 4 that does not fix complement. Second, combined results of a number of studies show that passive transfer to mice of immunoglobulin G (IgG) isolated from sera of MuSK-MG patients as well as active immunization of mice with rat MuSK causes muscle weakness, reduces MuSK expression by dispersal of synaptic AChR and decreases the synaptic response to spontaneous or evoked release of ACh. Importantly, the deleterious effects on the NMJ are greatest in cranial, paraspinal and masseter muscles that express lower levels of MuSK and MuSK RNA than the less affected muscles (Punga et al. 2011). Finally, it was proposed that pyridostigmine is generally ineffective or harmful in MuSK-MG because MuSK autoantibodies interfere with binding of ColQ to MuSK. This would reduce AChE in the synaptic space and result in cholinergic overactivity at the NMJ. Examination of intercostal muscle NMJs in human MuSK-MG did not show reduced AChE expression, but NMJs of more severely affected muscles were not examined (Kawakami et al. 2011). In this issue of the Journal of Physiology, Morsch and co-workers (2013) report that in the antibody-induced mouse model of MuSK-MG, relevant doses of pyridostigmine worsen the postsynaptic depletion of AChR and reduce the synaptic response to ACh. In contrast, 3,4-diaminopyridine (3,4-DAP), which augments quantal release by nerve impulse, improves neuromuscular transmission without worsening the loss of AChR from the NMJ. What are the possible reasons for this paradox? A partial explanation would be that pyridostigmine increases the lifetime of ACh in the synaptic space by a few milliseconds, during which ACh continues to bind to multiple AChR molecules as it exits the synaptic space by diffusion. In contrast, 3,4-DAP increases quantal release, but the lifetime of ACh in the synaptic space is less than 0.1 ms, during which ACh binds to AChR; and when ACh is released from the receptor over an exponentially distributed period, it is instantly destroyed by AChE. It is also well recognized that the agrin–LRP4–MuSK–Dok-7 signalling pathway promotes AChR clustering, whereas ACh activates an opposite pathway involving subsynaptic AChRs, inositol trisphosphate (IP3) receptors, calcium and cyclin-dependent kinase-5 to disassemble AChR clusters (Misgeld et al. 2005; Zhu et al. 2011). But is the increased exposure to ACh in the model animals sufficient to enhance the synaptic injury? In the experiments of Morsch and co-workers, treatment with conventional doses of pyridostigmine lasted only 7–9 days whereas in other studies chronic treatment of normal animals with high doses of neostigmine was required to cause disintegration of the synaptic folds with loss of AChR. This strongly implies that the enhanced toxicity of ACh in the model animals as well as in human MuSK-MG is caused by impaired cluster stabilization by MuSK. This, in turn, could be due to antibody-dependent reduced expression of MuSK, or is caused by direct interference of anti-MuSK antibodies with MuSK signalling, or both.