PBMCs from Hemophilia a Subjects with and without an Inhibitor Show Robust Cytokine Secretion in Response to Factor VIII C2 Domain Epitopes

Hemophilia A (HA) is caused by mutations in the F8 gene encoding blood coagulation factor VIII (FVIII), and treatment consists of FVIII infusions. Unfortunately, neutralizing antibodies develop in 20-30% of severe HA patients, causing significant morbidity and a risk of life-threatening bleeds. Almost half of severe HA cases are caused by an inversion mutation in intron 22 of the F8 gene. These patients express mRNA encoded by F8 exons 1-22 but have no circulating FVIII antigen. Humans, including almost all HA patients, express a second transcript termed F8B encoded by a short exon within F8 intron 22 spliced to F8 exons 23-26, which encode the FVIII C2 domain. It has been hypothesized that two partial FVIII proteins are translated from these transcripts, which together span the entire F8 sequence, in patients with an intron-22 mutation and that this expression confers immune tolerance to FVIII in most of them. According to this hypothesis, almost all HA patients would have tolerance to the FVIII C2 domain, and the only region of infused FVIII that would be recognized as "non-self", i.e. would contain potential T-cell epitopes, in patients with an intron-22 inversion would be short sequences spanning the region encoded by F8 exons 22-23.

To test the hypothesis that the putative F8B protein confers immune tolerance to the FVIII C2 domain, we carried out EliSpot assays using peripheral blood mononuclear cells (PBMCs) from HA subjects to detect cytokine secretion in response to stimulation with (a) FVIII; (b) a recombinant FVIII C2 domain. Use of proteins instead of synthetic FVIII peptides tests responses to epitopes that are naturally processed from the protein antigen. In addition, the potential immunogenicity of FVIII sequences encoded by the exon 22-23 junction region was tested by quantitative competition binding assays measuring the affinities of overlapping 20-mer peptides corresponding to FVIII residues 2103-2146 for recombinant MHC (HLA-DRB1) proteins representing 10 common HLA-DRB1 alleles. Effective MHC binding of peptides derived from a protein antigen is a prerequisite for stimulation of T-effector cells.

EliSpot assaysof PBMCs from 19 HA subjects, both with and without a current inhibitor or inhibitor history, were carried out to detect secretion of Th1 cytokines (interferon-gamma), Th2 cytokines (IL-5) and the regulatory cytokine IL-10. More than 60% of these subjects showed robust secretion of Th1 and/or Th2 cytokines following stimulation with FVIII, consistent with a T-effector response. PBMCs from 10 of the 12 subjects tested using FVIII-C2 protein secreted interferon-gamma. Several subjects also secreted IL-10, indicating regulatory as well as effector cytokines were secreted. These results indicate that stimulation by naturally-processed epitopes within the FVIII C2 domain is a common feature of anti-FVIII immune responses. They are also consistent with earlier studies from the Conti-Fine group that employed synthetic FVIII peptides. Furthermore, many patients with no measurable neutralizing antibody activity (inhibitor) clearly maintained a robust cellular response to FVIII. Peptide-MHC binding assays showed moderate-to-strong affinities of exon 22-23 junction peptides for proteins encoded by 4 of the 10 HLA-DRB1 alleles. Cellular responses to these peptides are currently being tested.

We conclude that the postulated intracellular expression of an F8B protein does not modulate FVIII immunogenicity in HA patients and that, on the contrary, cytokine secretion in response to epitopes in the C2 domain are a prevalent feature of the anti-FVIII immune response in many patients with and without a clinically problematic inhibitor.