A sensitized model of thrombosis validates known multigenic relationships and suggests novel modifiers of hemostasis

Venous thromboembolism is a major cause of morbidity and mortality. Despite identification of risk factors, not all individuals with thrombophilia develop thrombosis. Understanding the multigenic factors modifying this incomplete penetrance would help guide patient care.

The zebrafish has a conserved hemostatic system and is amenable to large genetic studies. Loss of antithrombin III (At3) in zebrafish leads to an early consumptive coagulopathy and lethality in adulthood. Using this genetic background as a sensitized model we performed a dominant unbiased genome-wide N-ethyl-N-nitrosourea (ENU) mutagenesis screen followed by whole genome sequencing (WGS). We used survival studies, laser-mediated endothelial injury, and ex vivo protein assays to validate hits.

ENU-treated at3 ^+/- males were crossed with at3 ^+/- females to produce 4,030 total offspring (1.5x genome coverage). Four permanent lines transmitting a survival benefit beyond 7 months were identified and sequenced. A candidate screen of 63 known coagulation-related loci revealed a missense mutation, C504F, in a highly conserved residue of the prothrombin (F2) heavy chain, which was validated through genetic and biochemical studies. Evaluation of UK Biobank electronic health record (EHR) data was underpowered to detect interactions between F2 and AT3 due to minmal deleterious mutations. Mutations produced through genome editing revealed that heterozygosity for factor X and plasminogen also modified at3 ^-/- , resulting in reduced lethality. The three remaining lines had no coagulation-related variants segregating with survival, suggesting the presence of novel modifier loci.

Unbiased genome-wide screening identified a modifier of thrombosis. This demonstrated that re-balancing of hemostasis to mitigate thrombosis is conserved in zebrafish, including an unexpected role for fibrinolysis. This interaction was not detected even in a large human dataset, establishing the continued benefit of the zebrafish model. Finally, we found evidence for novel loci outside of the canonical coagulation cascade that may be new targets for diagnosis or treatment.