F9 missense variant hot spots associated with qualitative protein defects causing hemophilia B

Abstract

Hemophilia B, a rare bleeding disorder, is caused by genetic variations in F9. Although quantitative factor IX (FIX) deficiencies are successfully treated by protein replacement therapy, qualitative defects may result in dysfunctional proteoforms in the circulation, potentially interfering with prophylactically or therapeutically administered recombinant FIX (rFIX) concentrates. To delineate the F9 missense variants associated with qualitative defects, we integrated genotype and phenotype from the European Association for Haemophilia and Allied Disorders F9 Coagulation Factor Variant Database. Of the 663 patients for whom activity (FIX:Act) and antigen (FIX:Ag) data were available, 40% of patients with severe hemophilia (n = 248), 50% of patients with moderate hemophilia (n = 244), and 47% of patients with mild hemophilia (n = 171) had cross-reactive material defined as FIX:Ag ≥40%. Variants associated with qualitative defects were predominantly localized (1) at proteolytic sites for FIX processing and activation, (2) within exosite II of the serine protease domain, and (3) at calcium ion coordinating residues within the Gla/EGF-1 domain of the FIX light chain. To study the effect of dysfunctional FIX proteoforms on thrombin generation in the presence of rFIX, we investigated 2 individuals with hemophilia B harboring a variant of unknown significance with an unexplained bleeding phenotype despite prophylaxis. Ex vivo therapeutic monitoring using patient plasma supplemented with rFIX concentrates, bypassing agent or emicizumab, enabled head-to-head comparison and revealed limited normalization of the prolonged initiation phase in coagulation. Alignment of information on genotype with functional proteotype may clarify the heterogeneity in bleeding phenotype and treatment response and provide a stepping stone for personalized care.

Graphical abstract

Figure 1.

Overview of FIX variants associated with hemophilia B disease phenotype. (A) Most hemophilia B cases listed in the EAHAD F9 Variant Database harbored a missense mutant (n = 3127 cases; 66% of all 4715 cases; accessed on 6 May 2024). The number of cases per unique variant (n = 628 cases) is displayed over the FIX protein sequence to identify genotype hot spots associated with hemophilia B. (B) Allele frequencies of F9 missense mutations in the reference population reported in the Genome Aggregation Database (n = 241 missense mutations, accessed on 8 December 2024). Variants are displayed over the functional domains of FIX and colored by ClinVar germ line classification simplified in benign/likely benign, pathogenic/likely pathogenic, conflicting pathogenicity, and NA. AP, activation peptide; EGF, epidermal growth factor–like domain; Gla, Gla domain; LR, linking region; NA, not available; PP, propeptide; SP, signal peptide.

Figure 2.

Contribution of CRM in hemophilia B and FIX variant hot spots associated with qualitative defects. (A) Relation between FIX:Act and FIX:Ag reported for patients with hemophilia B harboring a missense mutation (n = 663 cases with FIX:Act and FIX:Ag data available). Each dot represents 1 hemophilia B case, and the colored rectangles indicate the disease classification based on FIX:Act (blue = mild hemophilia, yellow = moderate hemophilia, and red = severe hemophilia). Cases with CRM+ with FIX:Ag ≥40 U/dL (dashed line) were indicated by the dashed line and colored rectangles of disease severity. (B) Distribution of reported FIX:Ag levels in patients with mild (n = 171), moderate (n = 244), or severe (n = 248) hemophilia B based on their FIX activity levels. Median and interquartile range of FIX:Ag and proportion (%) of patients with FIX:Ag ≥40 U/dL are provided per severity subclass. (C) Skyline of variant location hot spots associated with qualitative defects in FIX. For hemophilia B cases with FIX:Act and FIX:Ag levels available (n = 663), the FIX:Ag/FIX:Act ratio was determined to unveil variants associated with qualitative defects. Each dot represents an individual patient case, disease severity is indicated by color, and cases are arranged by variant location within FIX. (D) Hemophilia B CRM+ cases with residual systemic levels of FIX:Ag ≥40% were selected (n = 303), and the median FIX:Ag/FIX:Act was provided per variant location. Data are colored by the domains signal peptide (SP), propeptide (PP), Gla domain (Gla), EGF-like domains (EGF-1 and -2), linker sequence, activation peptide (AP), and serine protease domain. Residues located in the Gla and EGF-1 domains involved in calcium ion positioning are marked ( | ), and proteolytic cleavage sites are visualized by dashed lines. All data are based on the EAHAD F9 Variant Database (accessed on 6 May 2024).

Figure 3.

Qualitative defects observed for patients with hemophilia B with variants at residues coordinating calcium ions in the Gla and EGF-1 domains. (A) Hemophilia B cases harboring an amino acid substitution variant in the Gla, EGF-1, or EGF-2 domain classified by disease severity and evaluated for their residual FIX:Ag level in the plasma. (B) FIX:Ag/FIX:Act determined per case and grouped per unique variant in the Gla domain. Dots represent the individual case values, bars represent median values per unique variant, and residues involved in calcium coordination are marked by asterisks. (C) FIX:Ag/FIX:Act determined per case and grouped per unique variant in the EGF-1 and EGF-2 domains. (D) Zymogen FIX model with incorporation of the EGF-1 crystal structure. Visualization in PyMOL with colors representing functional domains and calcium ion coordination residues in the EGF-1 is in orange. (E) Graphical overview of calcium ion coordinating residues in the EGF-1 domain. (F) Amino acid sequence homology of EGF-1 domain between FIX, FVII, and FX (National Center for Biotechnology Information basic local alignment search tool EGF-1 FIX; Homo sapiens) with residues in the consensus sequence indicated by an asterisk. All hemophilia B case data are based on the EAHAD F9 Variant Database (accessed on 6 May 2024).

Figure 4.

Hemophilia B caused by an F9 mutant, initially classified as variant of unknown significance, which results in a qualitative defect and impaired thrombin generation. (A) Pedigree with inherited F9 C.287A>G mutant resulting in p.(Gln96Arg) variant. (B-C) The p.(Gln96Arg) variant is located in the EGF-1 domain and introduces a positively charged arginine within the calcium ion binding cavity. (D) Other deficiencies or abnormalities in proteins contributing to the severe bleeding phenotype were excluded by quantitative mass spectrometry–based screening of plasma proteins corresponding to the tier 1 diagnostic gene list, updated and endorsed at the International Society of Thrombosis and Hemostasis 2023 conference. Normal concentration of coagulation and fibrinolysis proteins in the plasma was confirmed, except for mild FXI deficiency (39 U/dL) identified in 1 sibling (III.1). (E) Thrombin-generation profile with plasma from the carrier (II.I). (F-G) Thrombin generation with plasma from siblings (III.I) and (III.II) before and 1 hour after 25 U/kg Alprolix as prophylaxis. (H) Inhibitor testing by mixing normal plasma pool and patient plasma (III.II) followed by thrombin generation with (I) lag time and (J) peak height. Data were obtained by calibrated automated thrombography (CAT) assay with 1 pM TF for coagulation initiation. Dashed lines indicate the peak height in the control condition (pooled normal human plasma).

Figure 5.

Evaluation of in vitro efficacy of recombinant FIX concentrates on thrombin generation in patient plasma. Thrombin-generation profiles on in vitro supplementation of Alprolix in (A) normal pooled plasma, (B) FIX-depleted plasma, (C) plasma of a hemophilia B carrier harboring p.(Gln96Arg), and (D-E) plasma of her 2 sons with hemophilia B. Thrombin-generation profiles on in vitro supplementation of Rixubis in (F) normal pooled plasma, (G) FIX-depleted plasma, (H) plasma of a hemophilia B carrier harboring p.(Gln96Arg), and (I-J) plasma of her 2 sons with hemophilia B. Thrombin-generation profiles on in vitro supplementation of Benefix in (K) normal pooled plasma, (L) FIX-depleted plasma, (M) plasma of a hemophilia B carrier harboring p.(Gln96Arg), and (N-O) plasma of her 2 sons with hemophilia B. Data were obtained by CAT assay with 1 pM TF for coagulation initiation. Dashed lines indicate the peak height in the control condition (pooled normal human plasma).

Figure 6.

Evaluation of in vitro efficacy of emicizumab and inhibitor bypassing coagulation complex in patient plasma. Thrombin-generation profiles on in vitro supplementation of emicizumab (Hemlibra) in (A) normal pooled plasma, (B) FVIII-depleted plasma, (C) FIX-depleted plasma, (D) plasma of the carrier II.I, (E) plasma of sibling III.I harboring p.(Gln96Arg), and (F) plasma of sibling III.II harboring p.(Gln96Arg). (G) Thrombin-generation profiles on in vitro supplementation of anti-inhibitor coagulant complex factor eight inhibitor bypassing activity in patient (III.II) plasma. Data were obtained by CAT assay with 1 pM TF for coagulation initiation. Dashed lines indicate the peak height in the control condition (pooled normal human plasma).