Overexpression of Factor VIII Is Associated with Immune Responses to Factor VIII and Cellular Stress in Hemophilia A Mice

Abstract 22

For many years it has been evident that it is challenging to express recombinant factor VIII (FVIII) in vitro and in vivo. In vitro studies suggest that high levels of FVIII expression can lead to cellular stress. We have demonstrated that adeno-associated viral (AAV) vector delivery results in long-term expression of therapeutic levels of FVIII in hemophilia A (HA) dogs up 10% of normal. In this model transgene expression is restricted to a portion of liver cells resulting in sustained FVIII levels. Notably, in dogs FVIII expression is sustained for periods longer than 7 years. Conceivably, the increased load per cell for cellular processing of FVIII may impact the cell's ability to synthesize and secrete functional FVIII, activation of cellular stress and the potential for immune responses. We sought to determine whether cellular stress and immune responses to the transgene are increased by overexpression of FVIII in mice by AAV vectors. Using AAV to deliver FVIII as a B-domain deleted single chain (single chain approach) or as two separate chains (light and heavy chains) (two chain approach), we analyzed dose-dependent FVIII synthesis and its cellular impact. We hypothesized that high levels of FVIII expression after AAV-mediated delivery of FVIII may induce cellular stress. HA mice were administered AAV serotype 8 (AAV8) vector in 3 groups: (1) two-chain approach (2) single chain approach or (3) AAV empty capsid (control). Three total AAV doses were administered: 1×10¹⁰ vector genomes/mouse (low dose), 5×10¹⁰ vg/mouse (mid-dose) and 2.5×10¹¹ vg/mouse (high dose). FVIII antigen levels (heavy and light chain) and anti-cFVIII IgG antibodies were detected by ELISA at a series of time points post vector administration until the terminal time points (1, 2, 4, 8, 12 and 24 wks). As we previously observed, after single chain delivery we detect equivalent amounts of heavy chain and light chain in the circulation by ELISA that correlates with activity. However, in the two-chain approach we observe that the light chain is >2-fold higher than the heavy chain in the circulation and that the activity correlates with the amount of heavy chain. We observe a dose-dependent increase in FVIII expression and dose-dependent anti-cFVIII IgG antibody titers. The kinetics of FVIII expression and antibody formation are different for the two-chain delivery compared to the single chain delivery. For the two-chain delivery, we observe AAV dose-dependent peak FVIII expression within 1 week (30, 100, or 400 ng/ml, respective of dose) followed by the onset of anti-FVIII IgG2 antibodies by 4 weeks (16, 27, 52 μg/ml, respective of dose). For single chain delivery, we observe FVIII expression within 1 week and low level antibody titers by 4 weeks that increase dramatically by 12 weeks (14, 21, 38 μg/ml, respective of dose). Thus, we observe a more gradual increase in antibody titers in the single chain compared to the two-chain delivery but at late time points the immunogenicity of both approaches is indistinguishable. RNA was isolated from the livers of the mice at the terminal time points for quantitative PCR analysis of two key components of the unfolded protein response signaling network, BiP and CHOP. We compared the two-chain and single chain AAV delivery of FVIII alongside animals injected with AAV empty capsid that did not express FVIII. A tunicamycin treated positive control had an increase in BiP (54-fold) and CHOP (74-fold) expression compared to untreated HA mice. There were mild elevations (2–3-fold) of BiP and/or CHOP in some animals observed at week 1 in all three treatment groups. Only at 12 weeks post vector administration were significant increases in BiP (>70 fold) and CHOP (>4 fold) observed in the single chain treated group at the high dose expressing >300% but not the two-chain approach or empty capsid. These studies suggest that supraphysiological expression (>100%) of BDD-cFVIII may increase the likelihood of an immune response to FVIII and cellular stress. Further studies may determine if there is a relationship between high FVIII expression, cellular stress and immune responses to FVIII. Thus, there is a threshold of FVIII expression levels that may prove unsafe. Moreover, these data are in agreement with the lack of evidence of cellular toxicity and immunogenicity in HA dogs expressing therapeutic levels of FVIII by AAV vectors.