982. Improved Transduction of Primary Murine Hepatocytes by Recombinant Adeno-Associated Virus 2 Vectors In Vivo

Abstract

Adeno-associated virus 2 (AAV) vectors are in use in clinical trials for gene therapy of hemophilia B. We have reported that following tail-vein injections, AAV vectors selectively target the liver in mice (Gene, 190: 203–210, 1997). Although 100% of hepatocytes can be targeted by AAV vectors, the transgene expression is limited to ~5% of hepatocytes. Because the viral genome is a single-stranded DNA, and single strands of both polarities are encapsidated with equal frequency, it has been suggested that failure to undergo DNA strand-annealing accounts for the lack of efficient transgene expression. We and others, on the other hand, have proposed that failure to undergo viral second-strand DNA synthesis attributes to the observed low efficiency of transgene expression. We have documented that a cellular protein, designated FKBP52, in its phosphorylated forms, inhibits the viral second-strand DNA synthesis, and consequently, transgene expression in non-hepatic cells (J. Virol., 75: 8968–8976, 2001; J. Virol., 77: 2741–2746, 2003). To evaluate whether phosphorylated FKBP52 is also involved in regulating AAV-mediated transgene expression in murine hepatocytes, we generated transgenic mice over-expressing the cellular T cell protein tyrosine phosphatase (TC-PTP) protein, known to catalyze dephosphorylation of FKBP52, as well as mice deficient in FKBP52. Approximately 1×1011 particles of recombinant AAV-EGFP vectors were injected via the tail-vein in cohorts of normal C57BL6, TC-PTP transgenic (TC-PTP-TG), and FKBP52-knockout (FKBP52-KO) mice. Liver tissues were harvested 1-, 2-, and 4-weeks post-injections and sections were analyzed for transgene expression using confocal microscopy. Consistent with previously published reports, whereas ~5% of hepatocytes from normal mice were transduced, the transduction efficiency increased ~15-fold and ~4-fold, 1-week post-injection; ~17-fold and ~7-fold, 2-weeks post-injection; and ~16-fold and ~12-fold, 4-weeks post-injection in TC-PTP-TG and FKBP52-KO mice, respectively. We also examined whether the increase in AAV transduction efficiency in hepatocytes from TC-PTP-TG and FKBP52-KO mice was due to more efficient strand-annealing or viral second-strand DNA synthesis. Low Mr DNA samples isolated from liver tissues 1-, 2-, and 4-weeks post-injection were electrophoresed on denaturing gels and analyzed on Southern blots. Whereas almost all of the input AAV vector genomes were present as single-stranded DNA in hepatocytes from control mice, a significant fraction of the viral genomes was converted to dimer-length DNA forms 1-week after injection, and most of them were converted to dimer length forms 2- and 4-weeks post-injection in TC-PTP-TG mice as well as in FKBP52-KO mice, albeit at a lower level. Thus, our data strongly support the contention that the viral second-strand DNA synthesis, rather than DNA strand-annealing, is the rate-limiting step in the efficient transduction of hepatocytes, which should have implications in the optimal use of recombinant AAV vectors in human gene therapy,1