Abstract

Recombinant adeno-associated virus (rAAV) has been extensively studied as a vector for hemophilia gene transfer. AAV serotype 2 targeting the liver and skeletal muscle has been used in clinical trials for hemophilia B patients. Previously we reported that the use of recombinant non-AAV2 serotype vectors generated persistent expression of supra-normal levels of canine factor IX (cFIX) in immunodeficient mice (Mol Ther ‘2000), and resulted in sustained and complete hemophilia B phenotype correction in immune competent hemophilia B mice (Mol Ther ‘2001). In those studies AAV serotypes 1–5 were tested by intramuscular injection. In this study we tested rAAV serotypes 1–8 for hepatic transduction and FIX production in both C57BL/6 and FIX knockout mice. Animals received intrapotal vein injection of 1x1011 virion particles of rAAV 1 thru 8 carrying a human FIX cDNA linked to a chicken beta actin enhancer-CMV promoter and bovine growth hormone polyA. Animals were followed for 8 months and assessed for plasma levels of human FIX by ELISA. A rapid, sustained and maximal rate of hFIX expression was observed within 1–2 weeks with all vectors except AAV2 where the peak expression occurred at 8–10 weeks. hFIX expression was sustained for all serotyped vectors (n=5 animals tested at each AAV serotype). Animals that received serotypes 7 and 8 maintained hFIX at physiological levels (100% hFIX ~ 5000 ng/ml). A differential hFIX expression pattern emerged with rAAV7 (6213 ng/ml) > 8 (5111 ng/ml) > 5 (2367 ng/ml) > 1 (1090 ng/ml) > 4 (377 ng/ml) > 2 (314 ng/ml) > 3 (232 ng/ml). rAAV7 and rAAV8 generated 20 times more hFIX per virion particle than the rAAV2 vector. We did not detect anti-human FIX antibody in any of the experimental mice. To assess why hFIX production differed between each serotype we performed immunohistochemical staining of the mouse liver using a fluorescent-tagged anti-human FIX antibody. Based on this assay, rAAV7 and AAV8 transduced 30% of hepatocytes while rAAV2 transduced less than 2% of the hepatocytes. We subsequently tested FIX production and effect on hemophilia phenotype of rAAV 7 and 8 (1x1011 virion particles) in hemophilia B mice; the data was similar to that observed using the C57Bl/6 mice. Human FIX levels (>5000 ng/ml) were again sustained during the 6–8 month observation period. No anti-human FIX antibody was detected in these hemophilia B mice. Results of clotting function by aPTT testing demonstrated a normalization of clotting time in all the rAA8/hFIX treated hemophilia mice. Survival of all treated mice by tail clip (lethal for all non-treated knockout animals) confirmed that phenotypic correction had been achieved. Our results again demonstrate that AAV serotypes have differential transduction rates and that new serotypes with greater mouse hepatocyte transduction rates are more efficient for hemophilia gene transfer.

Author notes

Corresponding author