Hoch et al. antibody-mediated autoimmune disease in which the nicotinic acetylcholine receptor (AChR) at neuromuscular junctions (NMJs) is the major autoantigen (1). AChR-specific antibodies are detected in 90% of nonimmunosuppressed patients with generalized MG. However, Hoch et al. found antibodies to a novel antigen, muscle-specific kinase (MuSK), in approximately 66% of patients with generalized MG that were lacking detectable AChR autoantibodies (seronegative MG) (2). Subsequent studies have reported MuSK antibody frequencies of 4C47.4% in MG patients seronegative for AChR antibodies (3C9). MG patients with MuSK antibodies tend to develop severe facial weakness and bulbar symptoms, including dysphagia, dysarthria, and respiratory cirsis with some atrophy of facial muscles, that are often hard to treat effectively with immunosuppressive therapies (3, 7). The pathogenic mechanisms of MG caused by AChR antibodies are well delineated, but pathogenicity has not been exhibited for MuSK antibodies (10). Furthermore, no reports have explained the induction of MG by immunization of animals with purified MuSK protein. The present study was undertaken to explore this issue. Here we describe the development of myasthenia and reduction of AChR density in rabbits immunized with the ectodomain of MuSK. The molecular pathogenesis of MG was further SGC 0946 investigated using an in vitro assay of AChR clustering on myotubes that was mediated by MuSK antibodies. MuSK is an AChR-associated transmembrane protein. During development of skeletal muscle mass, MuSK is in the beginning required for organizing a primary synaptic scaffold to establish the postsynaptic membrane (11, 12). Prior to muscle innervation, AChR clusters form at the central regions of muscle mass fibers, creating an endplate zone that is somewhat broader than that in innervated muscle mass (13, 14). MuSK and rapsyn, which is a 43-kDa, membrane-associated cytoplasmic protein, must be expressed before the endplate zone forms (11, 15C17). Subsequent contact of the motor-neuron growth cone SGC 0946 with the muscle mass extinguishes extrasynaptic AChR clusters, resulting in a thin, distinct endplate zone in the midmuscle that is marked by a high density of AChR clustering (13, 14). In this step, agrin released from motoneurons activates MuSK and redistributes AChR clusters to synaptic sites (13, 14, 17C20). Therefore the formation of NMJs either in the absence or presence of agrin requires the expression of MuSK at the endplate membrane. The extracellular segment of MuSK comprises 5 unique domains, i.e., 4 immunoglobulin-like domains and 1 cysteine-rich region (21C25). All 5 domains are conserved in agglutinin (VVA-B4) without activation of MuSK (32C36). Neither the receptor nor the activation mechanisms of AChR clustering induced by agrin-independent inducers has been recognized with certainty. Even so, these mechanisms may also play important functions in the maintenance of NMJs via agrin-independent pathways and in their formation, as shown by genetic studies (13, 14). The data we present herein demonstrate that MuSK autoantibodies inhibit AChR clustering by agrin itself and also by all known agrin-independent pathways. Results Immunization with purified MuSK protein causes flaccid weakness in rabbits. Rabbit antibodies were raised against a purified chimeric protein composed of the MuSK ectodomain and the Fc region of human IgG1 (MuSK-Fc). All of 4 recipient rabbits manifested flaccid weakness after 3 or 4 4 repeated injections with MuSK-Fc. Three of these rabbits developed flaccid weakness within 3 weeks after the last injection of MuSK protein, and the fourth rabbit manifested flaccid weakness 9 weeks after the third injection. Two rabbits that manifested flaccid weakness (M1 and M2 paretic rabbits) are shown in Figure ?Physique1A1A and Supplemental Movies 1 and 2 (supplemental material available online with this short article; doi:10.1172/JCI21545DS1). Two of 4 SGC 0946 paretic rabbits developed severe exhaustion (Physique ?(Physique1A1A and Supplemental Movie 2; M2 paretic rabbit). Histological studies of the muscle tissues in the paretic rabbits revealed that this angular atrophic muscle mass fibers in Rabbit Polyclonal to RPL26L the M2 paretic rabbit were intermingled with normal fibers, whereas the M1 rabbit experienced only subtle changes in the muscle tissue (Physique ?(Figure1B).1B). No muscle mass regeneration was observed in M1 and M2 paretic rabbits (Physique ?(Figure1B).1B). The histological changes.