The development of inhibitors to Factor VIII (FVIII) is a major complication of hemophilia A (HA) treatment. Most preclinical studies in HA animals have been limited to a xenoprotein response, e.g., evaluating the murine immune response to human FVIII (hFVIII). Severe HA dogs are a naturally occurring outbred model that recapitulates the spontaneous bleeding phenotype of the disease; their severe HA is due to intron 22 inversion (INV22) of the canine F8 gene, analogous to the INV22 found in 45% of severe HA patients. Importantly, similar to HA INV22 patients, about 20% of these animals develop an anti-canine FVIII (cFVIII) inhibitor response. This ability of the HA dogs to develop inhibitors in a species-specific manner provides an opportunity to probe the molecular ontogeny of an alloimmune anti-FVIII response.Here, we use antibody phage display to capture the humoral anti-cFVIII IgG repertoire in a dog with severe HA. This dog developed and maintained high titers of cFVIII inhibitor (89 Bethesda Units, BU) after exposure to cFVIII protein. The dog then received AAV liver-directed gene therapy encoding a cFVIII transgene, and after an initial anamnestic response characterized by rapid increase of cFVIII inhibitor titer (peak titer of 182 BU), the dog exhibited successful inhibitor eradication and immune tolerance induction, similar to our previous report (Finn et al., Blood 2010).Canine peripheral blood mononuclear cells were used to construct a single-chain variable fragment (scFv) phage display library which was sequentially selected four times against immobilized recombinant B-domain deleted (BDD) cFVIII protein. The amount of enrichment per round of selection plateaued during the third round from which 55 scFv/phage clones were isolated for characterization. Individual phage clones were sequenced by Sanger sequencing and screened for binding to both canine and human BDD-FVIII by ELISA. ScFv/phage clones were tested for inhibitory activity in a modified Bethesda assay.We identified 26 distinct scFv clones binding cFVIII based on heavy chain/light chain composition comprising 16 distinct CDRH3 sequences and 23 distinct CDRL3 sequences. The heavy chains of the 26 clones were derived from six canine germline IGHV genes, namely IGHV3-2, IGHV3-5, IGHV3-9, IGHV3-19, IGHV3-38, and IGHV3-41. The 16 distinct CDRH3 sequences had a mean length of 10.1 ± 3.6 amino acids, shorter than the 13.5 ± 3.6 amino acids previously reported for total canine CDRH3 repertoires (Steiniger et al., Mol. Immunol. 2014). Only 10 of the 26 FVIII-specific clones contained a lambda light chain, despite canine IgG repertoires dominated by the use of lambda light chains (Steiniger et al.). Eight of the 26 clones exhibited inhibitory activity. Although this dog was never exposed to hFVIII, 12/26 (46%) clones bound both cFVIII and hFVIII, consistent with the high degree of homology between the two orthologues with 85% identity. For three clones, framework region mutations and different light chain pairings resulted in altered inhibitory activity and hFVIII binding. Interestingly, one clone in particular comprised the majority of randomly screened scFv from the third and fourth rounds of phage library selection.These results suggest that cFVIII-specific B cells arise from multiple germline VH genes and exhibit high CDRH3 diversity. While the IgG-FVIII interaction has previously been suggested to be mostly influenced by the IgG heavy chain, our data suggest that for some antibodies, the IgG light chain may influence both inhibitory activity and epitope reactivity. Future studies will assess B cell clonal evolution via deep sequencing and longitudinal epitope specificity during multi-year immune tolerance induction by continuous exposure to cBDD-FVIII gene therapy. This is the first in-depth assessment of FVIII immunogenicity in a large HA model that avoids the use of a non-species specific antigen. Direct comparison with ongoing studies in HA inhibitor patients will inform the immunodominant epitopes of FVIII across species and thus provide insights on FVIII immunogenicity. DisclosuresDoshi: Janssen: Consultancy; Spark Therapeutics: Speakers Bureau. Samelson-Jones: Spark: Research Funding; Pfizer: Consultancy, Research Funding. Siegel: Verismo Therapeutics, Inc: Other: Co-Founder and Equity Holder; Vetigenics, LLC: Other: Co-Founder and Equity Holder.