The periaxin gene (Prx) is specifically expressed in myelinating Schwann cells. Prx -/- mice initially produce compact myelin, but later demyelinate due to disruption of a novel dystroglycan complex (Sherman et al. 2001). Phenotypically, the mice show tremor, inappropriate clasping reflexes, and reduced peripheral nerve conduction velocity (Gillespie et al. 2000). We have now examined the physiology and morphology of Prx -/- neuromuscular junctions, to assess their contribution to the phenotype. Null-mutant mice and their wild-type littermates, aged 7 weeks (asymptomatic) to 8 months (overt phenotype in the mutants), were killed by cervical dislocation and intracellular recordings of EPPs were made from isolated flexor digitorum brevis (FDB), with a constant nerve conduction distance of approximately 1.5 cm, after blocking muscle action potentials with µ-conotoxin (2 µM). Triangularis sterni (TS) muscles were either vitally stained with FM1-43 to visualise recycling synaptic vesicles, or immunostained for axons and terminals (neurofilament/SV2), or Schwann cells (S100; P0; L-Periaxin). Acetylcholine receptors were stained with TRITC-α-bungarotoxin.

Most mutant junctions showed normal synaptic responses, with similar amplitude, time course and quantal contents compared with wild-type at all ages. However, EPP latency increased significantly with age: by 8 months it was 6.31 ± 0.21 ms in Prx -/- muscles (mean ± S.E.M.; n = 9 fibres, 2 muscles) compared with 2.56 ± 0.06 in wild-type (n = 11, 2; P < 0.001, t test). About 40 % of Prx -/- endplates (10/25 fibres in three muscles) responded intermittently to repetitive stimulation at 30 Hz. Vital staining and immunostaining revealed endplates with axonal sprouting, focal swellings, thinning of the preterminal axon and retraction of the myelin sheath. The number of preterminal axon branches was 2.3 ± 0.5 (n = 27, 3) in Prx -/-, compared with 1.3 ± 0.5 in wild-type (n = 13, 2); the number of terminal Schwann cells per endplate was reduced, to 1.2 ± 0.1 in Prx -/- from 2.5 ± 0.1 in wild-type (n = 43, 68 endplates, respectively; P < 0.001, Mann-Whitney test). These features could not be explained by absence of Prx gene expression directly, because periaxin immuno-staining, though present in wild-type myelin, was absent from terminal Schwann cells.

The intermittent nature and increased latency of EPPs, as a consequence of Schwann cell disassociation, provide a simple explanation for the abnormal reflex responses and tremor in Prx -/- mice. The data thus support a critical role for terminal Schwann cells in maintaining normal structure-function relationships at neuromuscular synapses.This work was supported by The Wellcome Trust. F.A.C. holds an Edinburgh University Faculty of Medicine Scholarship.

Gillespie, C.S., Sherman, D.L., Fleetwood-Walker, S.M., Cottrell, D.F., Tait, S., Garry, E.M., Wallace, V.C., Ure, J., Griffiths, I.R., Smith, A. & Brophy, P.J. (2000). Neuron 26, 523-531.

Sherman, D.L., Fabrize, C., Gillespie, C.S. & Brophy, P.J. (2001). Neuron 30, 677-687.