Gene transfer in murine MPS IIIA using canine adenoviral vectors.

Abstract

Mucopolysaccharidosis type IIIA (MPS IIIA) is an autosomal-recessively inherited disorder caused by the deficiency of lysosomal sulphamidase (NS) enzyme activity, resulting in the accumulation of the glycosaminoglycan (GAG) heparan sulphate (HS). MPS IIIA patients experience progressive and severe neurological deterioration with death usually occurring in the mid-late teenage years. A naturally-occurring mouse model of MPS IIIA has been characterised and the biochemical, histological and behavioural changes closely parallel the human condition. In order to treat the neurological effects of MPS IIIA, it is anticipated that a continual supply of replacement enzyme to affected cells will be required. Consequently, this study aimed to evaluate the efficacy, longevity and safety of gene therapy as a potential treatment for MPS IIIA. Canine adenoviral vectors (CAV-2) were selected on the basis of several important properties. They are non-integrating, are predicted to be less immunogenic in humans than human-derived viral vectors and mediate transgene expression for at least 1 year in vivo. An E1-deleted (∆E1) CAV-2 vector, CAV-NS, co-expressing recombinant human NS (rhNS) and Green Fluorescent Protein (GFP) was constructed and purified. In vitro testing revealed rhNS produced by CAV-NS significantly decreased sulphated GAG storage in human MPS IIIA fibroblasts in a mannose-6-phosphate-dependent manner. Preliminary studies in young adult guinea pigs with CAV-GFP demonstrated widespread GFP expression in the absence of a humoral response. In contrast, minimal GFP expression was found in CAV-injected adult mice due to formation of neutralising antibodies against the CAV-2 capsid. Consequently, intraventricular delivery of CAV-NS was evaluated in newborn mice at various doses. Widespread and dose-dependent GFP expression was observed and the optimal dose for large-scale studies was determined to be 109 CAV-NS particles/hemisphere. Antibodies against CAV-2, rhNS or GFP were not detected. Concurrently, the cognitive function and anxiety-related behaviours of unaffected and MPS IIIA mice were evaluated. MPS IIIA mice had significantly impaired memory and spatial learning in the Morris Water Maze (16-wks) and reduced anxiety in the Elevated Plus Maze (18-wks) when compared to unaffected animals. In a large therapeutic assessment trial, newborn MPS IIIA or unaffected mice received 109 particles of CAV-NS, saline or remained uninjected. GFP expression was visualised for at least 20-wks post-injection. Reductions in the vacuolation of ependymal and choroidal cells of the lateral ventricle and the cerebral cortex of treated MPS IIIA animals were observed in some GFP-positive (and presumably rhNS-expressing) regions. Furthermore, improvements in reactive astrogliosis, but not in the number of activated microglia, were measured in CAVNS- treated MPS IIIA mice. However, insufficient CAV-NS-mediated rhNS expression was generated to improve functional changes as assessed by a behavioural test battery (motor function, open field activity, Elevated Plus Maze, Morris Water Maze), potentially due to chronic inflammatory responses against the CAV-2 vector. Collectively, these data suggest that early intervention with ∆E1 CAV-NS gene therapy was able to improve several components of neuropathology in MPS IIIA animals but was unable to significantly alter the clinical progression of murine MPS IIIA.