Label-free comparative proteomic analysis reveals distinct profiles of direct oral anticoagulants in human endothelial cells

Introduction Direct Oral Anticoagulants (DOACs) are the first-line treatment for thromboembolism prevention in patients with atrial fibrillation. Clinical and preclinical studies suggest that, beyond their anticoagulant properties, DOACs may have additional vascular effects. However, their effects on endothelial protein-abundance profiles remain incompletely characterized. The objective of this study was to explore the protein-abundance profiles associated with exposure of human umbilical vein endothelial cells (HUVECs) to each of the four DOACs in the absence of external stimuli, using a label-free quantification (LFQ) methodology. Methods Cultured HUVECs were exposed for 6 h to concentrations selected to approximate reported therapeutic plasma concentrations of apixaban, rivaroxaban, edoxaban, and dabigatran. All experiments were carried out in serum-free media, after a starvation period of 18 h. After protein extraction using the TRIzol Reagent, label-free quantitative proteomic analysis was performed using nanoLC-MS/MS, followed by the identification of differentially abundant proteins (DAPs) and bioinformatic enrichment analyses. Results We identified 1,643 protein entries, of which 753 were differentially abundant. Of these, 32 showed higher abundance and 26 showed lower abundance than the vehicle control in all four DOAC comparisons. Proteins showing shared abundance changes were mapped to gene identifiers that were enriched for annotations related to autophagy and cardiac development, and cellular activation. Several proteins showing abundance changes in all three FXa-inhibitor comparisons, but not in the dabigatran comparison, were mapped to annotations related to hypoxia response, vascular permeability, and cholesterol homeostasis. However, this comparison is potentially confounded by the absence of an HCl-matched vehicle control for dabigatran. The majority of DAPs were drug specific, with distinct signatures for each DOAC. Exploratory enrichment subnetworks linked mapped gene identifiers from drug specific protein sets to ontology terms including endothelial-cell chemotaxis, sodium-ion transport, necrotic cell death, mRNA splicing, and regulation of gene expression. These cellular functions were not directly measured. Conclusion Under the experimental conditions used, exposure to the four DOACs was associated with common and drug-specific protein-abundance patterns in HUVECs. These protein-abundance patterns provide hypothesis-generating observations for subsequent functional investigation.