? Despite significant efforts to develop gene therapy strategies for erythroid diseases, several roadblocks have limited success in this field including inefficient gene transfer to repopulating hematopoietic stem cells (HSCs) in large animals and humans, and low levels of transgene expression in red blood cells. More recently, insertional mutagenesis leading to leukemia, as observed in the French X-linked SCID trial, has also become a critical concern for gene therapy. Here we propose to overcome these problems by developing a HSC gene therapy for pyruvate kinase (PK) deficiency using foamy retrovirus vectors, and by developing means to target integration of foamy vectors using polydactyl zinc finger proteins. We have recently demonstrated efficient gene transfer to canine long term repopulating HSCs using vectors based on the non-pathogenic foamy virus. In these studies we observed considerable GFP expression in red blood cells. These studies suggest that foamy vectors may be effective for HSC gene therapy for erythroid diseases, and that the Basenji PK-deficient canine model should be an excellent preclinical model to evaluate their potential. We have incorporated erythroid-specific promoters into foamy vectors and will compare transgene expression levels in vitro towards developing effective erythroid-specific foamy vectors. We will also explore means to target the integration of foamy vectors using polydactyl zinc finger DNA binding proteins. Finally, we will attempt to cure the Basenji dog PK-deficiency by transplantation with autologous foamy-transduced CD34+ cells. In the Basenji dog, corrected cells will not have a significant selective advantage in vivo, and high levels of gene marking will likely be required to observe a therapeutic effect. The Basenji canine PK model is thus ideal to evaluate in vivo selection strategies. We will use foamy vectors that contain a P140K mutant O6-methylguanine-DNA- methyltransferase (MGMT) selection cassette in addition to the PK transgene. This will allow us to increase the level of marking post-transplantation to increase our chances of curing PK-deficiency. Unlike mouse models the canine model has been predictive of HSC transplantation outcomes in the clinical setting and of gene transfer levels in humans. These data should thus be directly translatable to future clinical studies to treat PK deficiency and for other erythroid diseases. ? ? ?