A Novel Factor VIII Mutation with Increased Activity, Severe Thrombosis and Resistance to Activated Protein C

Background: An 18-year-old male presented with recurrent, life-threatening thromboses since birth including arterial (perinatal stroke, splenic infarct) and venous (bilateral lower extremity, mesenteric vein, portal vein and pulmonary emboli). He was found to have markedly elevated factor VIII (FVIII) activity ranging from 300 to 900%.

Objective: The objective of this study was to investigate this abnormal FVIII function and structure.

Methods: Assays performed included: prothrombin time (PT)/international normalized ratio (INR), activated partial thromboplastin time (aPTT), fibrinogen, complete blood and platelet count (CBC); one-stage clotting and chromogenic FVIII (Chromo using bovine and human substrates); antithrombin, protein C, FII, FV, FIX, plasminogen and von Willebrand factor (VWF) activities; lupus anticoagulant (dRVVT and Staclot, Stago); FVIII, VWF and free protein S antigens; anticardiolipin and anti-b2GP1 antibodies (ELISA). Thrombin production and reduction were assayed using Calibrated Automated Thrombogram (CAT) and CAT-Thrombomodulin (TM) performed per manufacturer’s instructions using 5 pm tissue factor and TM (Stago). The CAT and CAT-TM generated lag time, peak thrombin, endogenous thrombin potential (ETP) and velocity index (VI). The activated protein C resistance clotting assay for factor V Leiden (APC-R), was performed and curated for FVIII in our lab by substituting FVIII deficient plasma for FV deficient plasma. Whole Genome Sequencing (WGS) was performed and put into Integrative Genomics Viewer (IGV). Docking predictions used ClusPro, while other amino acid models used Pymol.

Results: Coagulation screening tests and blood count were normal. Procoagulant factors, thrombophilia evaluation, plasminogen and antiphospholipid antibodies were normal/negative. One stage clotting FVIII activity on the sample studied was 615%, with FVIII antigen of 105%; Chromo FVIII activity with Bovine and Human substrates were normal at 150 and 88%, respectively. The standard CAT showed an increased VI of 196 nM/min (242% of control) but other indices were normal. In the CAT-TM the VI was increased at 189 nM/min (527% of control) and the peak thrombin was 350 nM (250% of control). The addition of TM in the CAT-TM reduced thrombin ETP by 2.3 and 10.8% on 2 samples (mean control 53%, 2SD range, 32-74%). The clotting APCR assay for factor V was normal while the APCR assay for FVIII showed resistance (decreased ratio of CAT ETP after/before addition of TM). With WGS, IGV was used to identify an unique exonic point mutation, R590A in Exon 12.

ClusPro was used to simulate docking and produced stability statistics on the interactions of wild-type (WT) and R590S FVIII with various proteins. Results suggest that while the mutated FVIIIa has a higher affinity for APC based on scores, it appears to alter the binding site, leading to interactions in non-canonical regions. The APC docking predictions gave FVIII R590S a lower favorability score indicating that it interacts with APC poorly compared with WT, supporting the CAT-TM and APCR for FVIII results. Docking predictions performed for FVIII interactions with thrombin, VWF and factor IXa showed no difference from WT. Docking predictions of FVIIIa with FX or FXa and simulations using the Xase complex with FX or FXa predicted a stronger binding affinity and higher likelihood of binding with R590S, although the mutation appeared to disrupt the normal binding site resulting in non-canonical binding. Amino acid bond imaging using Pymol showed WT FVIII 590Arg produced multiple hydrogen bonds with surrounding amino acids, while FVIII 590Ser produced almost none.

Conclusion: We report a novel FVIII mutation associated with recurrent thrombosis. The mutation confers highly increased FVIII activity. Our studies suggest that R590S FVIII activates thrombin rapidly and is not responsive to APC. Docking studies predict weaker affinity of R590S to APC and stronger, yet abnormal, affinities for FX or FXa alone, or in the Xase complex. Further investigations using a FVIII expression system to study binding kinetics of R590S FVIII are in process to better elucidate this mutation. This high-activity FVIII shares some analogy to FIX Padua and could possibly have therapeutic implications for the treatment of hemophilia A.