Phenotype-genotype correlations in hemophilia A carriers are consistent with the binary role of the phase between F8 and X-chromosome inactivation

C P Radic,L C Rossetti,M M Abelleyro,T Tetzlaff,M Candela,D Neme,G Sciuccati,M Bonduel,E Medina-Acosta,I B Larripa,M de Tezanos Pinto,C D De Brasi
DOI: https://doi.org/10.1111/jth.12854
Abstract:Background: The recessive X-linked disorder hemophilia A (HA) is rarely expressed in female carriers, most of whom express about half of normal factor VIII activity ( Fviii: C). Objective: To propose an integrative assessment model for the binary role of the phase between the mutated F8 and the active X-chromosome (Xa) in Fviii: C in HA carriers. Methods: We studied 67 females at risk of severe HA, comprising five symptomatic females ( Fviii: C < 1.5 IU dL(-1) ) and 14 controls. A correlation study between Fviii: C (observed vs. expected) and X-chromosome inactivation (XCI) patterns (XIPs; androgen receptor gene [AR] system) in blood leukocyte DNA was performed in carriers, by comparison of a model correlating Fviii: C and XIP with arbitrary models devoid of biological significance, and with Fviii: C levels in non-carriers (mean model) as a proxy from background data dispersion not influenced by XIP. Results: We provide proof-of-concept example from a family presenting with extremely skewed XIPs in which the severe HA phenotype appeared in a heterozygous carrier of a crossover between AR and F8 loci that phased the mutated F8 with the maternally inherited Xa. Furthermore, four cases of severe HA affected women who had a combination of a heterozygous F8 mutation and extremely skewed XIPs in leukocytes or oral mucosa are presented. Correlation analyses between Fviii: C levels and XIPs in carriers (n = 38) but not in non-carriers (n = 20) showed highly significant differences between the proposed correlation model and models without biological significance. The data support a binary influence of XCI, either increasing or decreasing the Fviii: C, subject to the underlying phase set between the F8 mutation and XCI. Conclusions: Our evidence suggests that the phase between XCI and mutated F8 acts as a molecular switch conditioning Fviii: C levels and HA expression in carriers.
What problem does this paper attempt to address?