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

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.