Association of FXI activity with thrombo-inflammation, extracellular matrix, lipid metabolism and apoptosis in venous thrombosis

Abstract

Animal experiments and early phase human trials suggest that inhibition of factor XIa (FXIa) safely prevents venous thromboembolism (VTE), and specific murine models of sepsis have shown potential efficacy in alleviating cytokine storm. These latter findings support the role of FXI beyond coagulation. Here, we combine targeted proteomics, machine learning and bioinformatics, to discover associations between FXI activity (FXI:C) and the plasma protein profile of patients with VTE. FXI:C was measured with a modified activated partial prothrombin time (APTT) clotting time assay. Proximity extension assay-based protein profiling was performed on plasma collected from subjects from the Genotyping and Molecular Phenotyping of Venous Thromboembolism (GMP-VTE) Project, collected during an acute VTE event (n = 549) and 12-months after (n = 187). Among 444 proteins investigated, N = 21 and N = 66 were associated with FXI:C during the acute VTE event and at 12 months follow-up, respectively. Seven proteins were identified as FXI:C-associated at both time points. These FXI-related proteins were enriched in immune pathways related to causes of thrombo-inflammation, extracellular matrix interaction, lipid metabolism, and apoptosis. The results of this study offer important new avenues for future research into the multiple properties of FXI, which are of high clinical interest given the current development of FXI inhibitors.

Figure 1

Time-related comparison of factor XI activity (FXI:C) in A, individuals in the acute VTE event and at 12 months of follow-up, and B, stratified by type of anticoagulant. In this figure panel A shows time-differences of FXI:C values available at acute VTE event and 12 months follow-up (n = 187). The p-value was derived from a paired t-test. Panel B shows time differences stratified by type of anticoagulant drug, in red for the acute VTE event and in blue for the 12 month follow up time point.

Figure 2

Shared and specific FXI:C-related proteins of the acute VTE event and 12 months after the acute event. This figure displays the shared and specific proteins selected by LASSO-regularized regression models. Panels A and B depict the effect size and association with FXI:C at the acute VTE event and 12 months follow-up, respectively. Proteins selected in both models were shown indicated with a symbol (*).

Figure 3

Sensitivity analyses excluding DOAC users, and individuals with normal and high FXI:C (%) and prolonged APTT. This figure shows the proteins selected by LASSO-regularized regression models. Panels A and C depict the size effect and association with FXI:C of the proteins when DOAC users are excluded in the acute VTE event and 12 month follow-up models. Panels B and D depict the selected proteins when individuals with normal and high FXI:C (%) and a prolonged APTT are excluded for each model.

Figure 4

Shared and specific enriched pathways of FXI:C related proteins in the acute VTE event and 12 months follow up (A), and protein–protein interaction networks of proteins included in both models (B). This figure shows the shared and specific enriched pathways related with FXI:C associated proteins (A). The vertical black line depicts the cut-off (p-value < 0.05). Orange and red lines depict the p-value of a pathway in the acute event and 12 months follow up, respectively. Panel B shows the protein–protein interaction enrichment analysis where only physical interactions are depicted. Networks with 3 or more proteins (densely connected nodes) were color coded and labeled with a functional term based on a Gene Ontology enrichment analysis (https://metascape.org/).

Figure 5

Summary representation of the main pathways and proteins related with FXI:C in the setting of VTE. This summary figure illustrates the main processes and biomarkers related to FXI:C associated proteins. Proteins highlighted in red were associated in the acute VTE event, in orange with the 12-month post-index event and in blue at both time points. (A) Cytokine receptors are located on the surface of endothelial cells, enhance the expression of proinflammatory proteins and their release, promote leukocyte recruitment, platelet activation, neutrophil degranulation and netosis (DNA/RNA and histones released), and local inflammation, which could act as a procoagulant environment. (B) During apoptosis, apoptotic bodies are released which, together with phosphatidylserine exposure, support a negatively charged surface that can contribute to the activation of FXII and FXI, resulting in increased coagulation activity. (C) Extracellular matrix components such as glycosaminoglycans and collagen support a procoagulant surface against different coagulation factors and cell types. Extracellular matrix turnover by proteins involved in degradation of matrix components is a relevant process that prevents the binding and aggregation of platelets and proteins involved in coagulation. (D) The products of LDL oxidation (OxLDL) interact with various components of the coagulation system, but are also capable of promoting platelet activation and endothelial dysfunction. (E) β-Amyloid protein aggregation may facilitate the aggregation of platelets and coagulation factors, which may lead to thrombus formation in small vessels. (F) FXI has the ability to cleave ADAMTS13, which promotes activation of the intrinsic coagulation cascade and platelet aggregation via vWF. Created with BioRender.com.

Figure 6

Methods overview and flow chart of the study investigating FXI:C related proteins.