The use of recombinant adeno-associated serotype (rAAV2) as a vector for the liver has been hindered by the relatively low efficiency of stable transduction of hepatocytes (approximately 5%). Alpha-1 antitrypsin(AAT) deficiency is a common single gene disorder in which a mutation in the AAT gene (PI*Z) leads to misfolding of the protein, loss of the protective antiprotease effect of AAT for the lungs and a toxic effect on the hepatocyte where it is produced. Therefore, optimal therapy for AAT deficiency will require a high percentage of hepatocyte transduction in order to be effective for both liver and lung disease. Recently, rAAV genomes have been successfully pseudotyped with capsids from novel serotypes (eg. rAAV7 and AAV8), and AAV8 was shown to be quite efficient for hepatic transduction (Gao et al. PNAS 2002; 99:11854-59). We hypothesized that upon portal vein injection to target hepatocytes, serotype 8 would better transduce target cells and therefore express AAT in both a greater percentage of cells and in greater overall amounts. 1 × 1010 particle doses of serotypes 1, 2, 5 and 8 vectors carrying the human alpha-1-antitrypsin (hAAT) gene were injected in C57Bl/6 female mice at six weeks of age. Circulating hAAT in mouse serum was measured by a human-specific sandwich ELISA. RESULTS: In the serotype comparison studies, rAAV8 vectors express significantly sooner after injection than serotypes 1 or 5. This was demonstrated in samples from 1 week post-injection where serotypes 1 and 5 had no detectable hAAT expression, while the rAAV8 group had a mean level of 574 ug/ml and the rAAV2 group had a mean level of 2 ug/ml hAAT. By week 3, rAAV8 injected animals had circulating hAAT levels 30 times higher than AAV2(607 ug/ml versus 22 ug/ml). Subsequently, a dose escalation study was performed with doses of 1 × 109, 1×1010 and 1×1011 particles of the rAAV8-hAAT vector. At three weeks post-portal vein injection, the animals showed a clear dose-response with mean serum levels of 917, 7,065 and 806,449 ng/ml, respectively in the three groups. These results confirm the impression that rAAV8 may be a more promising gene vector for therapy of liver disorders. Further studies will be necessary to quantify the percent transduction of hepatocytes with rAAV8 and the functional effect in a liver disease model, as well as to evaluate the safety and biodistribution of the rAAV8-hAAT vector.