Circulating soluble P-selectin must dimerize to promote inflammation and coagulation in mice

Abstract

Leukocyte adhesion to P-selectin on activated platelets and endothelial cells induces shedding of the P-selectin ectodomain into the circulation. Plasma soluble P-selectin (sP-selectin) is elevated threefold to fourfold in patients with cardiovascular disease. Circulating sP-selectin is thought to trigger signaling in leukocytes that directly contributes to inflammation and thrombosis. However, sP-selectin likely circulates as a monomer, and in vitro studies suggest that sP-selectin must dimerize to induce signaling in leukocytes. To address this discrepancy, we expressed the entire ectodomain of mouse P-selectin as a monomer (sP-selectin) or as a disulfide-linked dimer fused to the Fc portion of mouse immunoglobulin G (sP-selectin-Fc). Dimeric sP-selectin-Fc, but not monomeric sP-selectin, triggered integrin-dependent adhesion of mouse leukocytes in vitro. Antibody-induced oligomerization of sP-selectin or sP-selectin-Fc was required to trigger formation of neutrophil extracellular traps. Injecting sP-selectin-Fc, but not sP-selectin, into mice augmented integrin-dependent adhesion of neutrophils in venules, generated tissue factor-bearing microparticles, shortened plasma-clotting times, and increased thrombus frequency in the inferior vena cava. Furthermore, transgenic mice that overexpressed monomeric sP-selectin did not exhibit increased inflammation or thrombosis. We conclude that elevated plasma sP-selectin is a consequence rather than a cause of cardiovascular disease.

Figure 1.

Mouse dimeric sP-selectin-Fc binds with higher avidity to PSGL-1 than mouse monomeric sP-selectin. (A) Schematics of mouse sP-selectin and sP-selectin-Fc. (B) SDS-PAGE of sP-selectin and sP-selectin-Fc under reducing conditions, followed by Coomassie blue staining. (C) Western blots (WB) of sP-selectin and sP-selectin-Fc electrophoresed under reducing and nonreducing conditions, probed with polyclonal anti-P-selectin IgG or mAb HPC4. (D) Elution profiles of the indicated proteins applied to a Superdex 200 column, with fractions detected by absorbance at 280 nm. The column was calibrated by the elution profile of the indicated proteins. (E) Overlays of increasing concentrations of sP-selectin or sP-selectin-Fc binding to 2-GSP-6, a surrogate for PSGL-1, on a sensor surface. The horizontal line indicates when sP-selectin or sP-selectin-Fc in running buffer was injected. (F) Nonlinear curve fits of specific binding data for proteins injected in running buffer (Buffer) or in plasma diluted 1/4 in running buffer (Plasma). The data represent the mean ± standard deviation (SD) from 3 experiments. (G-H) Flow cytometry of mouse neutrophils (G) and monocytes (H) incubated with FITC-labeled sP-selectin or sP-selectin-Fc with or without oligomerization by mAb HPC4, in Ca²⁺-containing buffer or in buffer containing EDTA. The data in panels B-D and G-H are representative of 3 experiments. DF, dye front; mAU, milliabsorbance unit; RU, resonance unit.

Figure 2.

sP-selectin must dimerize or oligomerize to induce leukocyte signaling in vitro. (A-B) Static adhesion of mouse leukocytes to immobilized ICAM-1 or fibrinogen, in the presence of the indicated concentration of sP-selectin or sP-selectin-Fc with or without blocking anti-P-selectin mAb. (C) Fluorescence micrographs of purified mouse neutrophils incubated with PMA (positive control, top row); with buffer (Untreated), sP-selectin, sP-selectin-Fc, or control IgG (Control-Fc) (middle 2 rows); or with sP-selectin or sP-selectin-Fc oligomerized with mAb HPC4 in the presence or absence of neutralizing anti-P-selectin mAb (bottom 2 rows). After incubation, the cells were fixed and stained with anticitrullinated histone IgG or with Sytox Green to label DNA. The insets in the top row are higher magnifications. The red arrow marks a neutrophil without staining for citrullinated histones and with staining for DNA only in the nucleus. The white arrow marks a neutrophil stained for citrullinated histones and DNA only in the nucleus. The yellow arrowhead marks a neutrophil with extracellular staining for both citrullinated histones and DNA, indicating NET release. Original magnification ×20; inset magnification 2.7-fold. Scale bar, 50 μm. (D-E) Quantification of cells stained for citrullinated histones or forming NETs (stained for both citrullinated histones and extracellular DNA). The data in panels A, B, D, and E represent the mean ± SEM from 5 experiments. *P < .05, **P < .01, ***P < .001.

Figure 3.

sP-selectin and sP-selectin-Fc have long half-lives in the circulation. (A) Plasma levels of sP-selectin in WT or Selp^−/− mice. (B) Relative plasma levels of sP-selectin or sP-selectin-Fc in Selp^−/− mice at the indicated times after injecting 10 to 20 μg of the indicated protein intravenously (i.v.) or intraperitoneally (i.p.). The circulating half-life was determined using a 2-phase linear regression fit. (C) Representative absolute plasma concentrations of sP-selectin or sP-selectin-Fc in Selp^−/− mice at the indicated times after injecting 10 μg of the indicated protein i.p. or i.v. The data in panels A and B represent the mean ± SEM from 3 to 5 experiments.

Figure 4.

Injecting sP-selectin-Fc, but not sP-selectin, promotes inflammation and coagulation in vivo. (A-B) Number of rolling cells and arrested cells in trauma-stimulated venules of WT mice injected IV with saline, sP-selectin, sP-selectin-Fc, or control-Fc in the presence or absence of blocking anti-P-selectin mAb or anti-β2 integrin mAb. (C) Percentage of tissue factor–positive microparticles in plasma of WT mice 22 hours after intraperitoneal injection of saline, sP-selectin, sP-selectin-Fc, or control-Fc. (D) Plasma-clotting times of WT mice 22 hours after intraperitoneal injection of saline, sP-selectin, sP-selectin-Fc, or control-Fc, in the presence or absence of blocking anti-tissue factor IgG or nonimmune IgG. (E) Plasma levels of sP-selectin or sP-selectin-Fc 22 hours after intraperitoneal injection. The data in panels A and B represent the mean ± SEM from 15 to 20 venules from 4 to 5 mice in each group. The data in panels C-E represent the mean ± SEM from 7 to 12 experiments in each group. *P < .05, **P < .01, ***P < .001.

Figure 5.

Injecting sP-selectin-Fc, but not sP-selectin, increases the frequency of deep vein thrombi. WT mice were injected intraperitoneally every 24 hours for 3 days (time 0, 24 hours, or 48 hours) with saline, sP-selectin, sP-selectin Fc, or control-Fc. Twenty-two hours after the first injection, the inferior vena cava was ligated to induce thrombosis. After 72 hours (48 hours after ligation), the incidence, size, and weight of thrombi were measured. (A) Frequency of thrombus. For each experimental group, the number of mice forming thrombi relative to the total number of mice is shown. (B) Thrombus length. (C) Thrombus weight. (D-E) Plasma levels of sP-selectin or sP-selectin-Fc at the indicated time after injection. The data represent the mean ± SEM from 10 to 16 experiments in each group. *P < .05.

Figure 6.

Generation of transgenic mice (Tg-sP-selectin) expressing chronically elevated plasma levels of sP-selectin. (A) Schematic of the sP-selectin transgene. (B) Plasma levels of total sP-selectin (measured with anti-P-selectin IgG) or transgenic sP-selectin (measured with mAb HPC4) in WT or Tg-sP-selectin mice. The data represent the mean ± SEM from 8 experiments in each group. (C) Western blots (WB) of P-selectin from platelet lysates or from immunoprecipitates of plasma from mice of the indicated genotype, or of recombinant sP-selectin (right lane). The data are representative of 5 experiments.

Figure 7.

Chronic elevation of circulating sP-selectin does not promote inflammation and coagulation. (A-B) Number of rolling cells and arrested cells in trauma-stimulated venules of WT or Tg-sP-selectin mice in the presence or absence of blocking anti-P-selectin mAb or anti-β2 integrin mAb. (C) Plasma-clotting times of WT or Tg-sP-selectin mice without challenge or 4 hours after intraperitoneal injection of lipopolysaccharide (LPS). (D-F) Frequency of thrombus, thrombus length, and thrombus weight in WT or Tg-sP-selectin mice 48 hours after ligation of inferior vena cava. The data in panels A and B represent the mean ± SEM from 11 to 15 venules from 5 to 6 mice in each group. The data in panels D-F represent the mean ± SEM from 15 experiments in each group. *P < .05.