Interactions between integrin α9β1 and VCAM-1 promote neutrophil hyperactivation and mediate poststroke DVT

Abstract

Venous thromboembolic events are significant contributors to morbidity and mortality in patients with stroke. Neutrophils are among the first cells in the blood to respond to stroke and are known to promote deep vein thrombosis (DVT). Integrin α9 is a transmembrane glycoprotein highly expressed on neutrophils and stabilizes neutrophil adhesion to activated endothelium via vascular cell adhesion molecule 1 (VCAM-1). Nevertheless, the causative role of neutrophil integrin α9 in poststroke DVT remains unknown. Here, we found higher neutrophil integrin α9 and plasma VCAM-1 levels in humans and mice with stroke. Using mice with embolic stroke, we observed enhanced DVT severity in a novel model of poststroke DVT. Neutrophil-specific integrin α9-deficient mice (α9fl/flMrp8Cre+/-) exhibited a significant reduction in poststroke DVT severity along with decreased neutrophils and citrullinated histone H3 in thrombi. Unbiased transcriptomics indicated that α9/VCAM-1 interactions induced pathways related to neutrophil inflammation, exocytosis, NF-κB signaling, and chemotaxis. Mechanistic studies revealed that integrin α9/VCAM-1 interactions mediate neutrophil adhesion at the venous shear rate, promote neutrophil hyperactivation, increase phosphorylation of extracellular signal-regulated kinase, and induce endothelial cell apoptosis. Using pharmacogenomic profiling, virtual screening, and in vitro assays, we identified macitentan as a potent inhibitor of integrin α9/VCAM-1 interactions and neutrophil adhesion to activated endothelial cells. Macitentan reduced DVT severity in control mice with and without stroke, but not in α9fl/flMrp8Cre+/- mice, suggesting that macitentan improves DVT outcomes by inhibiting neutrophil integrin α9. Collectively, we uncovered a previously unrecognized and critical pathway involving the α9/VCAM-1 axis in neutrophil hyperactivation and DVT.

Graphical abstract

Figure 1.

Stroke leads to increased neutrophil integrin α9, higher plasma VCAM-1 levels, and increased DVT severity. (A) Neutrophil integrin α9 and (B) plasma VCAM-1 levels from patients with stroke and healthy controls. (C) Schematic of experimental design. (D) Representative image of flow-cytometric analysis of integrin α9 for each group (left) and quantification of α9 expression in peripheral neutrophils after stroke or sham surgery in mice (right). (E) Expression of α9 relative to Actb in peripheral neutrophils after stroke or sham surgery. (F) Plasma VCAM-1 levels from mice with stroke and mice with sham surgery. (G) Schematic of experimental design for stroke-DVT studies. (H) Representative IVC thrombus harvested 48 hours after stenosis from each group (left) and thrombus weight (mg; right). Only mice that exhibited thrombosis were included to quantify the thrombus weight. Each dot represents a single mouse. (I) Thrombosis incidence. (J) Expression of α9 relative to Actb in peripheral neutrophils after DVT and stroke-DVT. Data are mean ± standard error of the mean (SEM) and analyzed using the Mann-Whitney test (A-B,D-F); Fisher exact test (I); and Kruskal-Wallis test followed by the corrected method of Benjamini and Yekutieli (J); n = 8 (A-B); n = 6 (D-F); n = 20 (H-I); and n = 4-6 (J). hr, hour; mRNA, messenger RNA.

Figure 2.

Neutrophil integrin α9 promotes poststroke DVT severity. (A) Schematic of experimental design. (B) Representative IVC thrombus harvested 48 hours after stenosis from each group (left) and thrombus weight (mg; right). Only mice that exhibited thrombosis were included to quantify the thrombus weight. Each dot represents a single mouse. (C) Thrombosis incidence. (D) Representative cross-sectional immunofluorescence image (left) of the isolated IVC thrombus (48 hours after stenosis) from each group for Ly6G (neutrophils, green) and DAPI (4′,6-diamidino-2-phenylindole; blue); magnification, 20×; scale bar, 50 μm; and quantification (right). (E) Representative cross-sectional immunofluorescence image (left) of the isolated IVC thrombus (48 hours after stenosis) from each group for the antihistone H3 (citrulline R2 + R8 + R17) (NETs, red) and DAPI (blue); magnification, 20×; scale bar, 50 μm; and quantification (right). Data are mean ± SEM and analyzed by repeated measures analysis of variance (ANOVA) followed by the corrected method of Benjamini and Yekutieli (B,D-E); and Fisher exact test (C); n = 20 (B-C); and n = 6-7 (D-E). Cit H3, citrullinated histone H3; DAPI, 4',6-diamidino-2-phenylindole.

Figure 3.

Integrin α9 and VCAM-1 interactions promote gene expressions related to neutrophil inflammation, exocytosis, NF-κB signaling, and chemotaxis. (A) Schematic of experimental design. (B) Principal component analysis was performed based on RNA-seq of stimulated and unstimulated WT neutrophils and (C) stimulated neutrophils from littermate controls and neutrophil–specific integrin α9^−/− mice. (D) Volcano plots of differentially expressed genes (DEGs) based on RNA-seq analysis of stimulated and unstimulated WT neutrophils and (E) stimulated neutrophils from littermate controls and neutrophil–specific integrin α9^−/− mice. (F) Log fold-change of all the shared DEGs from stimulated and unstimulated WT neutrophils and stimulated neutrophils from littermate controls and neutrophil–specific integrin α9^−/− mice. (G) Log fold-change of selected genes from DEGs of stimulated and unstimulated WT neutrophils. (H) Log fold-change of selected genes from DEGs of neutrophils from littermate controls and neutrophil–specific integrin α9^−/− mice. (I) GSEA was performed based on RNA-seq of stimulated and unstimulated WT neutrophils and (J) stimulated neutrophils from littermate controls and neutrophil–specific integrin α9^−/− mice. The significance of the enriched pathways was determined by right-tailed Fisher exact test followed by Benjamini-Hochberg multiple testing adjustment; n = 5 (B,D,G,I); and n = 4-5 (C,E,F,H,J). GSEA, gene set enrichment analysis.

Figure 4.

Interactions between integrin α9 and VCAM-1 promote neutrophil hyperactivation, mediate neutrophil adhesion at venous shear rate, and enhance endothelial cell apoptosis. (A) Schematic of experimental design. (B) Elastase and (C) MPO levels in cell-culture media 6 hours after stimulation. (D) Experimental design. (E) Representative images of the neutrophil adhesion to the VCAM-1 coated slides at venous shear rate (left); magnification, 20×; scale bar, 50 μm; and qualification (right). (F) Representative cross-sectional immunofluorescence image of terminal deoxynucleotidyl transferase biotin-dUTP nick end labeling-positive mouse endothelial cells incubated with cell supernatant of stimulated neutrophils from α9^fl/fl and α9^fl/flMrp8Cre^+/− mice (left); and quantification (right); magnification, 10×; scale bar, 100 μm. Data are mean ± SEM and analyzed by repeated measures ANOVA followed by the corrected method of Benjamini and Yekutieli (B-C) or the Mann-Whitney test (E-F); n = 5 (B-C); n = 5 (E); and n = 5 (F).

Figure 5.

Pharmacogenomic profiling and virtual screening revealed macitentan as a potent inhibitor of the integrin α9/VCAM-1 interaction. (A) The top 10 concordant perturbagens with concordance scores >0.321 are shown using iLINCS portal based on RNA-seq of neutrophils of littermate controls and neutrophil–specific integrin α9^−/− mice. (B) The template (green, α4 subunit) and target (cyan, α9 subunit) models. (C) The predicted α9β1 structure with an inhibitor. (D) List of top-19 compounds with binding energy and percentage inhibition of integrin α9 to VCAM-1. (E) Chemical structure of macitentan. (F) Docked pose of macitentan with integrin α9β1. (G) Representative images of neutrophil adhesion to VCAM-1 in presence of different concentration of macitentan (left); magnification, 10×; scale bar, 100 μm; and quantification (right). (H) Representative images of the mouse neutrophil adhesion to the activated mouse venous endothelial cells coated slides at venous shear rate (left); magnification, 20×; scale bar, 50 μm; and qualification (right). Data are mean ± SEM and analyzed by 1-way ANOVA followed by Sidak multiple comparisons test (G) or Mann-Whitney test (H); n = 6 (G); and n = 5 (H).

Figure 6.

Macitentan pretreatment reduces poststroke DVT severity. (A) Schematic of experimental design. (B) Representative IVC thrombus harvested 48 hours after stenosis from each group (left); and thrombus weight (mg; right). Only mice that exhibited thrombosis were included to quantify the thrombus weight. Each dot represents a single mouse. (C) Thrombosis incidence. (D) Representative cross-sectional immunofluorescence image (left) of the isolated IVC IVC thrombus (48 hours after stenosis) from each group for Ly6G (neutrophils, green) and DAPI (blue). magnification, 20×; scale bar, 50 μm; and quantification (right). (E) Representative cross-sectional immunofluorescence image (left) of the isolated IVC thrombus (48 hours after stenosis) from each group for the antihistone H3 (citrulline R2 + R8 + R17) (NETs, red) and DAPI (blue); magnification, 20×; Scale bar, 50 μm; and quantification (right). Data are mean ± SEM and analyzed by repeated measures ANOVA followed by the corrected method of Benjamini and Yekutieli (B,D,E) or Fisher exact test (C); n = 20 (B-C); and n = 5-6 (D-E). DAPI, 4',6-diamidino-2-phenylindole.

Figure 7.

The positive effects of macitentan on DVT outcomes are, at least in part, mediated by neutrophil integrin α9. (A) Schematic of experimental design. (B) Representative IVC thrombus harvested 48-hour post-stenosis from each group. (C) Thrombus weight (mg). Only mice that exhibited thrombosis were included to quantify the thrombus weight. Each dot represents a single mouse. (D) thrombosis incidence. Data are mean ± SEM and analyzed by repeated measures ANOVA followed by the corrected method of Benjamini and Yekutieli (C) or Fisher exact test (D); n = 20 (C-D).