Transient and intensive pharmacological immunosuppression fails to improve AAV-based liver gene transfer in non-human primates

BACKGROUND: Adeno-associated vectors (rAAV) have been used to attain long-term liver gene expression. In humans, the cellular immune response poses a serious obstacle for transgene persistence while neutralizing humoral immunity curtails re-administration. Porphobilinogen deaminase (PBGD) haploinsufficiency (acute intermittent porphyria) benefits from liver gene transfer in mouse models and clinical trials are about to begin. In this work, we sought to study in non-human primates the feasibility of repeated gene-transfer with intravenous administration of rAAV5 vectors under the effects of an intensive immunosuppressive regimen and to analyze its ability to circumvent T-cell immunity and thereby prolong transgene expression. METHODS: Three female Macaca fascicularis were intravenously injected with 1x10(13) genome copies/kg of rAAV5 encoding the human PBGD. Mycophenolate mofetil (MMF), anti-thymocyte immunoglobulin, methylprednisolone, tacrolimus and rituximab were given in combination during 12 weeks to block T- and B-cell mediated adaptive immune responses in two macaques. Immunodeficient and immunocompetent mice were intravenously injected with 5x10(12) genome copies/kg of rAAV5-encoding luciferase protein. Forty days later MMF, tacrolimus and rituximab were daily administrated to ascertain whether the immunosuppressants or their metabolites could interfere with transgene expression. RESULTS: Macaques given a rAAV5 vector encoding human PBGD developed cellular and humoral immunity against viral capsids but not towards the transgene. Anti-AAV humoral responses were attenuated during 12 weeks but intensely rebounded following cessation of the immunosuppressants. Accordingly, subsequent gene transfer with a rAAV5 vector encoding green fluorescent protein was impossible. One macaque showed enhanced PBGD expression 25 weeks after rAAV5-pbgd administration but overexpression had not been detected while the animal was under immunosuppression. As a potential explanation, MMF decreases transgene expression in mouse livers that had been successfully transduced by a rAAV5 several weeks before MMF onset. Such a silencing effect was independent of AAV complementary strand synthesis and requires an adaptive immune system. CONCLUSIONS: These results indicate that our transient and intensive pharmacological immunosuppression fails to improve AAV5-based liver gene transfer in non-human primates. The reasons include an incomplete restraint of humoral immune responses to viral capsids that interfere with repeated gene transfer in addition to an intriguing MMF-dependent drug-mediated interference with liver transgene expression.