Immunogenic implications of translational readthrough modulate the association of F8 nonsense mutations with inhibitors in Hemophilia A.
BACKGROUND: Among F8 mutation types, main determinants of inhibitor development after replacement therapy in Hemophilia A (HA), nonsense mutations display wide variation in the associated risk. Translational readthrough (Rdthr) at premature termination codons (PTCs) produces traces of full-length factor VIII (FVIII) including missense and wild-type molecules (WT Rdthr), and may influence the immune response involved in inhibitor formation. METHODS: For inhibitor association analysis, we investigated F8 genotypes, inhibitor status and Rdthr features in 335 PTCs (1048 patients), recorded in the European Association for Haemophilia and Allied Disorders (EAHAD) database, and exploited expression of F8 PTCs variants. Mean-differences in affinity of HLA-DR alleles for FVIII WT peptides and their missense counterparts were bioinformatically calculated. RESULTS: WT Rdthr was higher for patients affected by PTCs not associated with inhibitor and was not predicted at all in patients with PTCs detected in at least three inhibitor positive cases (n = 136, p = 0.0001). WT Rdthr was lower for PTCs in the inhibitor prone FVIII light chain. Among FVIII PTCs fused with luciferase, quantitative output of WT Rdthr negative and positive groups did not differ, potentially highlighting for the positive group the importance of WT Rdthr to decrease inhibitor association. Readthrough output was the lowest among WT Rdthr negative PTCs, highly associated with inhibitors. The WT Rdthr prediction was extended to all PTCs that could arise by single nucleotide variations for the entire F8 coding sequence. Estimated WT Rdthr was higher in PTCs reported in EAHAD than those predicted (n = 662), foreseeing a higher risk of developing inhibitors. In silico mean differences in affinity of HLA-DR alleles for WT FVIII peptides and their missense counterparts, potentially arising from Rdthr of PTCs without WT formation (n = 297), were higher for missense variants predicted in patients with inhibitors than without (p < 0.0001), and increased for PTCs present in more than one patient with inhibitor, potentially supporting immunogenic features. CONCLUSIONS: The new genetic classification of HA PTCs may improve our knowledge about their relationship with inhibitors. It deserves to be explored for estimating inhibitor PTC association in HLA genotyped patients as well as in other human diseases.