Recombinant adeno-associated virus (AAV) vectors are promising vehicles for achieving stable liver transduction in vivo. We have previously demonstrated that hepatocyte transduction with recombinant AAV serotype 2 (rAAV2) vectors in mice becomes saturated at a maximum level of ~10 % of total hepatocytes, after administration of high doses (1012 vector genomes (vg) / mouse or more). The mechanism(s) that restricts liver transduction at high doses have not been elucidated yet, but it is not related to vector entry into the cells and likely involves some cellular metabolic state that determines permissiveness or non-permissiveness to stable rAAV2 transduction. In order to investigate whether pseudo-serotyped rAAV vectors other than rAAV2 also exhibit similar restricted transduction in mouse liver, we performed a dose response study using rAAV1 and rAAV8 vectors. rAAV1 vector was selected because our previous study demonstrated that a smaller increase of the rAAV1 vector dose resulted in a greater increase of liver transduction (i.e., non-linear response) (Grimm et al, Blood 102: 3412, 2003). rAAV8 vector was chosen because rAAV8 vector can transduce mouse hepatocytes better than rAAV2 and we have recently suggested that the differences in transduction efficiency may be due to the rate of vector uncoating in vivo (Thomas et al., J. Virol., in press). We injected 8-week-old male C57BL/6 rag1 mice via the portal vein with various doses (5.0 × 1010, 3.0 × 1011, 1.8 × 1012 and 7.2 × 1012 vg / mouse, n = 3 to 6 per group) of AAV1-EF1α-nlslacZ (AAV serotype 1 capsid-pseudotyped human elongation factor 1α promoter-driven nlslacZ-expressing rAAV vector) and AAV8-EF1α-nlslacZ. As controls, we injected mice with AAV2-EF1α-nlslacZ. All mice were sacrificed 6 weeks after vector injection, and liver sections were analyzed with Xgal staining to determine transduction efficiency. The results demonstrated that 24% transduction efficiency was achievable with rAAV1 at the highest dose but the liver transduction was not saturated at the dose. In the rAAV8-mediated liver gene transfer, virtually all hepatocytes can be transduced at the high dose. We subsequently compared liver transduction efficiency with AAV8-EF1α-nlslacZ between tail vein injection (TV) and portal vein injection (PV). The results demonstrated that TV was as efficient as PV at any given doses, and ~100% hepatocyte transduction was achievable at a dose of 7.2 × 1012 vg / mouse. A preliminary tissue distribution study with Xgal staining revealed that, rAAV8 vector, when administered via TV or PV, transduced heart muscle and smooth muscle of intestinal walls in addition to the liver, and interestingly, rAAV8 passed the blood brain barrier and transduced brain as well. Detailed histological and molecular analyses are underway, and the results will be presented. In summary, in contrast to rAAV2, the restricted liver transduction was not the case in rAAV8. Approximately 100% liver transduction was achievable with a high dose via intra-vascular infusions of rAAV8 vector, suggesting that it is a robust vector for gene therapy applications. However, we need to take into consideration that the vector is promiscuous and can transduce a variety of tissues when given by this route of administration at a high dose.