Antigen Trafficking and Possibly Intrinsic Immunostimulatory Properties Shape High Immunogenicity of Factor VIII

Inhibitor development is the most serious complication of factor (F)VIII replacement therapy for hemophilia A. The apparent high immunogenicity of FVIII has prompted several hypotheses about the impact of FVIII structure and function on inhibitor formation. This study aimed to pinpoint the differences between FVIII and other strong and weak immunogens.

To determine which antigen presenting cells (APCs) can capture FVIII in vivo, we labeled a B domain-deleted (BDD) FVIII with Alexa Fluor (AF)488 and intravenously (IV) injected 5-35 µg of the labelled protein into B6/129 hemophilia A (HA) mice (n=3-4). Control animals received chicken ovalbumin (OVA)-AF488 (5-200 µg). After 30-180 min, spleens were isolated for flow cytometry. After 1 hour, either protein was detectable in dendritic cells (DCs), macrophages (MFs) and B cells, but they differed in distribution across the analyzed cell types. FVIII-AF488 preferentially localized in conventional DC type 1 (cDC1s), with the percentage of FVIII-AF488⁺ cDC1s being 5-fold higher than FVIII-AF488⁺ conventional DC type 2 (cDC2s) (p<0.0001). Red pulp MFs (RPMFs) had little or no detectable FVIII-AF488. In contrast, in the OVA-AF488-injected animals, RPMFs internalized the most antigen, with 29-fold and 26-fold more OVA-AF488⁺ RPMFs than cDC1s and cDC2s, respectively (p<0.0001). Neither protein was detectable in plasmacytoid DCs (pDCs), regardless of the dose. The distribution of either protein between the analyzed APCs was similar at different doses. FVIII-AF488⁺ APCs reached the highest percentages already 30 min after administration, with declining numbers at later time points. In contrast, OVA-AF488⁺ APCs were not detectable yet at 30 min, only appeared at 60 min and then declined at 180 min except in cDC2s.

We also compared helper T cell responses to FVIII and OVA using an in vivo antigen presentation assay involving a single injection of FVIII-OVA (FOVA), which contained an MHC-II I-A^b peptide epitope of OVA in place of the B domain of FVIII, and an adoptive transfer of 5E6 transgenic OT-II CD4⁺ T cells, which express chicken ovalbumin-specific T cell receptor (TCR), into C57BL/6J mice (n=2-4). Spleens were collected for flow cytometry 4 days after the adoptive transfer. FOVA shows identical specific activity to that of B domain-deleted FVIII. Animals that received 5 µg FOVA had a stronger response (mean T cell proliferation 29%) than mice that received 10-fold more OVA (18%).

Next, we compared antibody responses between FVIII and three other T cell-dependent antigens, including OVA, factor (F)IX ("weak" immunogens), and keyhole limpet hemocyanin (KLH, a potent immunogen), by repeat dosing of each protein (300 ng) to B6/129 HA mice (n=7-9), followed by measurement of specific IgG1 antibody titers using ELISA. We adjusted the antigen dose to a clinically relevant FVIII dose. After 4 weekly injections, only the FVIII and KLH groups developed IgG1 antibodies (8.7 and 33.7 µg/ml, respectively) in response to the antigens.

We also employed C57BL/6J MyD88^-/- mice to evaluate inhibitor responses in the absence of inflammatory signaling pathways (n=15). After 4 weekly IV injections of 2.5 IU FVIII, we measured total and neutralizing anti-FVIII responses using ELISA and Bethesda assay, respectively. Average inhibitor and total anti-FVIII titers were not significantly different between the control and signaling-deficient mice, but the MyD88^-/- group had a markedly larger number of non-responders. The odds of MyD88^-/- mice developing FVIII inhibitors were 50-fold lower [OR=0.13; CI95(0.017;0.71)] compared to the control animals [OR=6.5; CI95(1.8;42)]. Animals that did mount antibody responses had similar inhibitor and total anti-FVIII titers to controls.

We propose that bypassing RPMFs en route to splenic T-cell zones partially accounts for high immunogenicity of FVIII. Since RPMFs filter out and process large amounts of self-antigens, they can efficiently suppress helperT cell responses to prevent autoimmunity. Perhaps failure to piggyback on that tolerogenic mechanism shapes T cell activation patterns in response to FVIII. Finally, we find that wild-type mice are more likely to develop FVIII inhibitors than animals with disrupted inflammatory signaling, suggesting that FVIII has intrinsic immunostimulatory properties, which impact the decision of whether to mount a response.

Kaczmarek:Bayer: Research Funding. George:Pfizer: Consultancy; AskBio Therapeutics: Other: Licensing Fees, Research Funding; STRM.Bio: Membership on an entity's Board of Directors or advisory committees; Spark: Consultancy; Avrobio: Consultancy, Membership on an entity's Board of Directors or advisory committees; Intellia: Consultancy; Biomarin: Consultancy; Bayer: Consultancy. Camire:Bayer: Consultancy, Research Funding; Pfizer: Consultancy, Research Funding. Herzog:BioMarin: Consultancy; Regeneron Pharma/Intellia Therapeutics collaboration: Consultancy; Spark Therapeutics: Research Funding; Pfizer: Consultancy.