Fabry disease is caused by a deficiency of the lysosomal enzyme alpha-galactosidase A (alpha-gal A). We previously engineered a retrovirus encoding human alpha-gal A and demonstrated enzymatic correction of patient cells. Further, we demonstrated metabolic cooperativity, in that corrected cells secrete alpha-gal A that can be taken up and utilized by bystander cells in vitro. In the present study, we created a system to examine and quantitate this phenomenon in vivo. To differentiate from endogenous alpha-gal A, we constructed a retroviral vector (pUMFG/alpha-gal A/FLAG) containing a fusion form of alpha-gal A with a specific tag sequence at the carboxy terminus. The catalytic activity of the fusion protein was identical to wild-type alpha-gal A. The fusion protein was overexpressed in and secreted by transduced patient cells. In uptake studies, the fusion protein was detected in the lysosome-enriched fraction of recipient cells. We then examined the effectiveness of the pUMFG/alpha-g A/FLAG retroviral vector in vivo. Murine bone marrow (BM) cells were transduced and transplanted into irradiated hosts. After 9 weeks, proviral DNA was detected by PCR in peripheral blood and BM mononuclear cells. More importantly, specific fusion protein enzymatic activity could be demonstrated in those cells and in plasma. Thus, we have demonstrated that overexpressed alpha-gal A enters the circulation from transduced BM cells and is stable over a significant period of time.