448. A Genetic Strategy to Inhibit Complement Activation Arising from Adenoviral Vectors

Abstract

Recent reports suggest that recombinant adenoviral (Ad) vectors for gene therapy activate the classical and alternate complement pathways (Gene Ther 11:1482-86, 2004; Mol Ther 10:1140-42,2004). These innate immune responses may contribute to acute toxicity, increase infection of liver cells following systemic delivery, and also enhance the humoral immune responses. To address these problems, a recombinant Ad5 vector was genetically modified to display a peptide sequence (homology to C4) known to inhibit the classic complement pathway. The replication defective vectors, Ad5HVR2rH17d and Ad5HVR5rH17d, were produced with the insert (LGS-HEVKIKHFSPY-HEVKIKHFSPY-GS-HHHHHH-LGS) into either HVR2 or HVR5 of the hexon, respectively. The insert included a homodimer of the 11 C-terminal amino acids of Sh-CRIT-ed1 (previously denoted rH17d, Immunology 110, 73-29, 2003). Control Ad5 vectors included a 6His insert (LGS-HHHHHH-LGS) that was placed either in HVR2 or HVR5, and are denoted as Ad5HVR2His and Ad5HVR5His, respectively. All vectors also encoded firefly luciferase and GFP under control of CMV. The 4 vectors were evaluated in TIB76 mouse liver cells to compared infectivity as a function of incubation conditions (3 experiments). For in vivo studies, immune competent BL/6 mice (4 groups, 5-6/group) were i.v. injected with 4|[times]|109 viral particles and repeatedly imaged over 30 d with an IVIS-100 imaging system (Xenogen, Inc.). Pooled sera from the groups were analyzed at 14 or 28 days for IgG and IgM levels. Sequence analyses established that the inserts were correct and in expected locations. For in vitro studies in the presence of PBS, similar luciferase levels were observed for cells infected with Ad5HVR2rH17d or Ad5HVR2His, and the same trend was found for cells infected with Ad5HVR5rH17d or Ad5HVR5His. In contrast, preincubation of Ad5HVR2His and Ad5HVR5His with fresh sera significantly increased luciferase levels as compared with PBS. This sera-enhanced effect was also observed for Ad5HVR2rH17d, but not for Ad5HVR5rH17d. Mice injected with Ad5HVR2rH17d or Ad5HVR5rH17d always showed significantly less liver luciferase expression as compared with the respectively control vectors. For Ad5HVR2rH17d the reduction in liver expression ranged from 6-fold less (8 d) to 83-fold less (24 d). For Ad5HVR5rH17d the reduction in liver expression ranged from 70-fold less (at 27 d) to 3900-fold less (at 1d). Mice dose with control vectors showed |[sim]|25% higher IgG and |[sim]|3-fold higher IgM levels, as compared with mice dosed with vectors that included the rH17d inserts. This strategy to genetically modify the Ad hexon and display a complement-inhibitory sequence appears to be viable to reduce complement activation, and may be applicable with other gene therapy vectors. Ad vectors displaying the rH17d insert showed greater reduction of liver transfection in vivo after i.v. injection, as compared with the in vitro tests. As rH17d primarily inhibits the classical pathway, the classical pathway may be more important for liver transfection.