Dominant heterozygous mutations as well as a recently identified homozygous nonsense mutation in the SPTBN2 gene (1), encoding β-III spectrin, are implicated in human cerebellar ataxia. A mouse model lacking β-III spectrin (KO), published by our lab in 2010, mimics the progressive human phenotype displaying motor deficiencies as well as Purkinje cell (PC) degeneration and death (2). Recent immunohistological analysis of cerebellar brain slices from 12 month old mice using anti-calbindin antibody has shown significant PC loss in the posterior cerebellum of KOs when compared to wild-type (WT) animals (55.5±7.8% PCs of WT, N=3, p<0.01, t-test). These findings were supported by the observation that KO PC dendritic trees in posterior lobules filled with Alexa 568 show significant reduction in the surface area occupied (WT 13889±972.3μm2 and KO 6512±1786μm2, N=3, n=8, p<0.01, t-test). We have used a recombinant adeno-associated viral (rAAV) approach to elucidate whether reintroduction of β-III spectrin to the posterior cerebellum can halt, alleviate or reverse the disease phenotype. Unfortunately the full length (FL) β-III spectrin cDNA is too large to be packaged but using HEK 293 cells we have identified that that C-terminus (C-trm) of β-III spectrin localises at the plasma membrane and does not alter the distribution of FL protein. Furthermore we have previously reported increased sodium currents in primary hippocampal cultures transfected with FL β-III spectrin (3) and C-trm β-III spectrin was found to enhance sodium currents to the same degree as FL (untransfected -886.2±117.2pA, FL -1726±277.8pA and C-trm -1712±250pA at -10mV step, N=6, n=20, p<0.01 when compared to untransfected cells, one-way ANOVA). AAV1/2 particles expressing C-trm of β-III spectrin under the control of the enhanced synapsin promoter and GFP from an internal ribosome entry site were introduced via stereotaxic injection into mouse cerebellum (anesthesia induced by 2% isoflurane inhalation at flow rate of 1 l/min). We observed that 5 months after surgery KO animals had similar levels of transduced cells surviving to WTs (WT 204±64.2 cells per animal, KO 109±57.8, N=4) but with higher percentage of transduced cells residing in the posterior lobules (WT 28.3±16.7% and KO 62.3±16.7%). The motor phenotype of β-III spectrin KO mice using the elevated beam task was found not to be reversed (0.25±0.08 slips for WTs and 1.8± 0.32 for KOs, p<0.01, two-way ANOVA), however the motor deficits did not deteriorate beyond 2 months of viral expression. Nevertheless, the weak behavioural effect is likely due to a low viral titre and small number of PCs transduced in this study. Ongoing work is examining whether injection at an earlier age and with more viral particles shows more therapeutic promise. Data reported as mean±SEM. N indicates the number of animals/cultures, n indicates the number cells.