609. Long Term Persistence of Canine Factor IX-Specific Th1 Lymphocytes in Regional Lymph Nodes after AAV1-Mediated Intramuscular Gene Transfer in a Hemophilia B Dog

Abstract

Previous studies on AAV-mediated gene therapy for hemophilia B showed that underlying factor IX (F.IX) mutation, vector dose, serotype, and route of administration all play a role in the risk of anti-F.IX antibody formation. Intramuscular (IM) administration of an AAV2 vector in normal and hemophilia B (HB) mice invariably activates F.IX-specific Th2 lymphocytes with production of mainly IgG1 and IgG2b antibody subclasses. Similarly, HB dogs with a missense mutation usually do not develop an immune response against canineF.IX when injected IM with low doses of an AAV2 vector. At high vector doses, a neutralizing antibody response has been observed, characterized by the production of both IgG1 and IgG2 subclasses, with levels of IgG2>IgG1. Although previous observations ruled out the involvement of CD8+ T cells, a recent study on the ovalbumin mouse model of AAV-mediated muscle-directed gene transfer showed a CTL response against the transgene protein at high vector doses (Wang et al., Blood, in press). An HB dog from the Chapel Hill colony (E57) received by IM injection an AAV1 vector expressing canine F.IX under the control of the CMV promoter at a dose of 2.4 |[times]| 10^11vg/kg, vector was co-administered with immunosuppression (IS) (Arruda et al., Blood 2004). One week after vector administration the whole blood clotting time (WBCT) fell to the normal range (18min) and circulating F.IX levels reached 104 ng/ml. However, after discontinuing IS a long-lasting neutralizing antibody response was observed (6 BU >3 years after injection) with production of both anti-cF.IX IgG1 and IgG2 and levels of IgG1>IgG2. In order to study in detail T cell responses against the cF.IX, a peptide library was produced spanning the F.IX protein sequence. PBMCs were isolated from whole blood of E57 and, from two HB dogs that received an AAV2-cF.IX vector delivered to the muscle. Lymphocytes were tested by ELISpot for secretion of IL-4, IL-10 and IFN-|[gamma]| and, interestingly, dogs injected with the AAV2 vector showed a positive IL-10 response while the AAV1 injected dog did not. Next, E57 and a normal dog were sacrificed and lymphocytes from spleen, draining lymph nodes (LN) and non-draining LN were isolated. Lymphocytes from E57 draining LN showed positive results on ELISpot for IFN-|[gamma]| for 2 cF.IX peptide pools. No response was detected for the other tissues and for the control dog. IF staining for CD8+ cells and H&E staining of muscle sections from E57 were negative. These results suggest a long-lived immune response detectable locally in the draining LN. Cytokine and antibody subclass profiles are typical of a Th1 response vs. the Th2 response seen in AAV2 injected dogs. Whether the shift from a Th2 to a Th1 response is driven by higher levels of antigen expression of by the switch in serotypes is currently under study. We conclude that in AAV-mediated, muscle-directed gene transfer, there may be long-term persistence of F.IX-specific Th1 lymphocytes in the regional LN.