Abstract
Progress towards an effective gene therapy for hemophilia B (HB) has been facilitated by large animal studies. Previous work has shown intravascular delivery of an adeno-associated viral (AAV) serotype 2 vector expressing canine factor IX (cF.IX) to skeletal muscle in HB dogs resulted in long-term expression of cF.IX at levels of 4–20% of normal, which nearly corrected the disease phenotype. However, occurrence of inhibitors to F.IX in some animals raises concerns of a potential immune response to the transgene product, prompting a more thorough examination of T cell responses in this setting. Early work revealed that transient immunosuppression (IS) with cyclophosphamide was required to avoid inadvertent antibody formation to F.IX. Here we report in detail the nature and the duration of T cell responses against the transgene product. Six HB dogs from the Chapel Hill colony received AAV2 vector at three different doses (1x1012 vg/kg, n=3; 3 x 1012 vg/kg, n=2; 8 x 1012 vg/kg, n=1) in addition to weekly infusion with cyclophosphamide (6 doses total). PBMCs were isolated from whole blood prior to vector infusion, during IS and after removal from IS, and used to measure the T cell response to F.IX by ELISpot assay for IL-10 and IFN-γ secretion using a cF.IX peptide library composed of 15-mers overlapping by 10 amino acids, spanning the entire protein sequence. Peptides were arranged into a matrix of pools, such that each peptide was contained in two orthogonal pools. Interestingly, in the IL-10 assay, one common T cell epitope corresponding to peptide 68 in the cF.IX library was found in all intravascularly-administered dogs from each of the three dose groups. The same epitope was also detectable in naïve HB dogs. Another epitope, corresponding to peptide 84, was found in a dog injected with the highest dose of vector after it developed a non-neutralizing antibody response against the cF.IX transgene product. Peptide 84 spans the region of the protein that contains the missense mutation responsible for HB (Chapel Hill mutation, Glu379 → Gly), which is a key difference in the newly introduced transgene product. Furthermore, the lack of any IFN-γ secretion coupled with the marked IL-10 response gives a cytokine profile that is characteristic of a Th2 response. This is in contrast with the Th1 response seen in previous studies with direct intramuscular injection of an AAV serotype 1 expressing cF.IX in dogs from the same colony. IS successfully reduced T cell responses to undetectable levels, while IL-10 secretion was detectable in PBMCs before vector delivery and one month after IS was discontinued. Overall circulating cF.IX levels did not seem to be affected by late restoration of T cells responses. No T cell responses against AAV capsid were detectable by ELISpot on PBMCs in any of the dogs studied. Interestingly, the relatively mild IS regimen also appeared to reduce the formation of neutralizing antibodies against AAV capsid, regardless of the route of administration. In summary, the nature of T cell responses (Th2 vs. Th1) suggests that route of administration, and/or AAV serotype, may play a role in the determination of the immune response elicited. We conclude that IS may provide a means to decrease T cell responses to the transgene following intravascular delivery of AAV-F.IX to skeletal muscle.
Corrigendum to “MIG (CXCL9) is a more sensitive measure than IFN-γ of vaccine induced T-cell responses in volunteers receiving investigated malaria vaccines” [J. Immunol. Methods 340 (2009) 33–41]
