242. The Use of AAV Serotype 1 Vector for Expression of α-Galactosidase A in Fabry Knockout Mice

Abstract

Fabry disease is a lysosomal storage disorder due to a deficiency of |[alpha]|-galactosidase A (|[alpha]|-galA) which leads to the systemic of neutral glycosphingolipids predominantly ceramide trihexoside (Gb3). Fabry disease is an important target for gene therapy. First, |[alpha]|-galA is secreted and recaptured by the surrounding and distant cells. This phenomenon is called cross-correction and is the rational for enzyme replacement therapy. Second, Gb3 accumulates in all organs, but the neurological involvement is minimal. Therefore, it is not necessary to treat the brain guarded by the blood brain barrier. Overexpression of |[alpha]|-galA in any organs such as muscle or liver as reservoirs may be sufficient for treatment of Fabry disease. Using type 2 AAV vector, we have previously shown that expression of |[alpha]|-galA in muscle can achieve long-term correction of biochemical, histological, and functional abnormalities of all organs in |[alpha]|-galA knockout mice. These results strongly indicate that clinical gene therapy of Fabry disease is highly realistic. To optimize the clinical protocol, we examined the utility of type 1 AAV vector carrying the human |[alpha]|-galA cDNA driven by the CAG promoter and the GFP marker expression unit. The vector (1.5 |[times]| 1011 vector genomes) was injected into the quadriceps muscle (IM) or the tail vein (IV) of Fabry knockout mice and |[alpha]|-galA activity in plasma was monitored periodically. The time course of plasma |[alpha]|-galA levels is remarkably different between AAV subtype 2 and 1. In a previous study using AAV2 vector, the enzyme levels increased slowly and reached to a plateau at 6-8 weeks after IM. The plasma enzyme level at the plateau was approx. 25 % of that in normal mice (8|[ndash]|10 nmole/hr/ml: units) and persisted for more than 30 weeks. In contrast, after IM of the AAV1 vector, the enzyme activity in plasma was detectable (3-6 units) even at 1 week after injection, but then quickly decreased to the background level by 4 weeks. However, when the AAV1 was administered by IV, extremely high levels of enzyme activity, 40|[ndash]|1500 units, were observed at 2-4 weeks and these non-physiological high concentrations of |[alpha]|-galA enzyme persisted at least up to 10 weeks without any significant phenotypical changes. These results indicate that expression of |[alpha]|-galA mediated by AAV1 vector is very efficient compared with that by AAV2 vector but the duration of expression is affected by the routs of administration. A number of different subtypes of AAV vectors have been recently developed. Careful characterization of each subtype is required for optimization of clinical protocols of AAV mediated gene therapy.