Search

We evaluated a panel of AAV serotypes for their ability to transduce hematopoietic stem and progenitor cells of human and murine origin. rAAV-2 genomes encoding either dsRed1 or firefly luciferase under the control of the hybrid CBA promoter were packaged into capsids derived from AAV-1, AAV-5, AAV-7, AAV-8 and AAV-9 and purified by successive rounds of cesium chloride gradient centrifugation. Hematopoietic cells were transduced overnight at low cytokine concentrations, following which cells were assayed in vitro for transgene expression or transplanted into sublethally irradiated NOD/SCID mice. For murine marrow cells, AAV-7 and AAV-showed the highest levels of expression as compared with the other serotypes. For analysis of human cells, hematopoietic stem/progenitor cells were purified from cord blood, cytokine-primed peripheral blood and bone marrow samples obtained using IRB approved protocols. CD34 cells were immunomagnetically purified and transduced with rAAV-dsRed. Flow cytometric analyses revealed that AAV-7 consistently showed the highest level of transgene expression in vitro in CD34 cells derived from cord blood and bone marrow with levels ranging from 36-64% 24 hours after transduction. This was followed by AAV-8 and AAV-1, at 10-43% expression at 24 hours post-transduction, while the other serotypes exhibited lower levels of transgene expression. We next compared transduction of the CD34+CD3//Q-population, which is highly enriched for stem cells. CD34+CD3//Q-cells were flow sorted to purity prior to transduction with the different serotypes. Notably, this primitive population yielded higher levels of transduction than the more heterogeneous CD34 population, with approximately 60-70% of CD34+CD3//Q-cells from cord blood and cytokine primed peripheral blood showing dsRed expression 24 hours after transduction with AAV-7. As with CD34 cells, transduction with AAV-8 and AAV-1 were the next most efficient. We have previously demonstrated that although rAAV-2 transduction in vitro results in poor expression at early time points, possibly due to defined limitations in uncoating, nuclear entry and second strand synthesis, long term studies and in vivo assays demonstrate that rAAV-2 efficiently transduces primitive human hematopoietic stem cells capable of serial engraftment in immune deficient mice. Since in vitro assays may not reflect true stem cell activity, we initiated in vivo transplantation studies of stem cells transduced with different AAV serotypes to determine the relative long term engraftment and gene expression potentials of AAV genomes packaged in the different serotype capsids. Initial results following live imaging of luciferase-transduced recipients confirm both the early onset of gene expression with AAV-7 and AAV-8 and the hierarchy of transgene expression seen with the serotypes in vitro. Overall, our results suggest that rAAV genomes pseudotyped in AAV-7 and AAV-8 capsids are very promising for gene transfer into hematopoietic stem cells.

The search for the ideal vector for stem cell gene therapy continues as recognized problems remain with all existing viral vectors. The recent identification of over 100 new serotypes of AAV significantly expands the repertoire of available gene transfer vectors. Additionally, the capacity to pseudotype rAAV2 genomes in novel capsids facilitates evaluation. We and others have analyzed rAAV2-mediated gene transfer into hematopoietic stem cells (HSC). However, despite efficient gene-marking of primitive HSCs, there are identified limitations to transgene expression especially at early time points. The use of novel capsids has been shown to overcome these limitations in other tissues and expand the host range of AAV. We tested transduction of CD34+ cells from cytokine-primed peripheral blood, CD34+CD38|[minus]| cells from cord blood (CB) and the CD34+CD38- cell line TF1a with AAV2-dsRed pseudotyped in capsids of AAV1, 2, 5, 7, 8 and 9. Transduced cells were analyzed by flow cytometry 24 and 48 hours post transduction. Results revealed that the highest level of transgene expression was observed with AAV7 in HSCs from all sources while AAV 2 and 9 showed the lowest. The hierarchy of transgene expression in vitro was: AAV7>AAV8|[ge]|AAV1>AAV5>AAV2=AAV9 regardless of the source of CD34+cells. These results indicate that the use of certain novel capsids allows circumvention of limitations to expression from AAV2. We next evaluated in vivo engraftment of human HSCs transduced with pseudotyped rAAV in a NOD/SCID xenotransplant model. Transduced CD34+ cells purified from either CB or cytokine primed peripheral blood were transplanted into sublethally irradiated mice. The pseudotyped rAAV2 genome encoded the firefly luciferase gene under the control of the CBA promoter. Serial live imaging of recipients revealed transgene expression from each of the 6 serotypes tested. However the levels and trends for long term expression variedby serotype. At early time points, the highest initial expression was observed with AAV9 followed by AAV7 and 8. AAV2 showed the lowest level of initial expression. However differences emerged over time. AAV7, 9 and 8 capsids emerged as the best serotypes for long-term expression in vivo, followed by AAV2. AAV5 and 1 showed a rapid decline in expression after initial moderately high levels of expression. These differences suggest that the processing of pseudotyped rAAV in vivo varies with the capsid. AAV serotypes also appear to have different transduction properties in vivo as compared with in vitro. The mechanisms underlying these differences are being investigated. The rapid and robust expression seen in vivo with certain serotypes is highly promising and underscores the need for further evaluation of pseudotyped rAAV for stem cell gene transfer.

Apo A-1 Milano, an Arg173 to Cys point mutation of apo A-1, has demonstrated marked efficacy in the reduction of atherosclerotic lesions in animal models. A recent human trial testing recombinant apo A-1Milano infusions has similarly shown significant and rapid reduction in coronary atheroma volumes. In this project, we tested the utility of rAAV vectors encoding apo A-1 Milano for gene therapy of atherosclerosis in vivo following either intra-muscular injection or transplantation of rAAV-transduced bone marrow cells. ApoE-/- mice were injected with approximately 1012 vector genomes/kg at 6–8 weeks of age and placed on a high fat diet two weeks later. No major differences in body weight were observed between the treatment groups. 20–24 weeks after injection, the mice were harvested, aortas were cleaned, fixed, mounted, stained with oil red O and atherosclerotic plaque areas quantified. Control untreated mice or mice treated with an irrelevant rAAV vector showed marked atherosclerotic lesion formation. In contrast, the rAAV-2-apo A-1Milano and rAAV-5-apo A-1Milano injected groups showed significant reductions in plaque formation, 58% and 50% respectively, despite a lack of major differences in total plasma cholesterol levels. Since marrow-derived monocytes and macrophages play a pivotal role in atherogenesis, we tested the hypothesis that transplantation of rAAV apo A-1 Milano-transduced marrow cells would also result in a decline in plaque formation, possibly due to the localization of transduced macrophages within atherosclerotic lesions. Lethally irradiated Apo E-/- mice were transplanted with rAAV transduced cells, and high fat diet was initiated two weeks later. Twenty two to 24 weeks after transplantation, the mice were euthanized, and aortas were analyzed for atherosclerotic plaques. Untransplanted mice, and mice transplanted with untransduced and control rAAV transduced marrow served as negative controls. Wild type B6 to Apo E-/- transplants served as positive controls for plaque reduction. Results revealed a 41% and 47% reduction in atheroma formation following transplantation with cells transduced with rAAV-2 and rAAV5 vectors encoding apo A-1 Milano, respectively. These results suggest that either intramuscular injection of rAAV Apo A-1 Milano or transplantation of rAAV transduced marrow cells results in significant long-term inhibition of atherogenesis following a single treatment with an rAAV vector.