Differences in Endothelial Activation and Dysfunction Induced by Antiphospholipid Antibodies Among Groups of Patients With Thrombotic, Refractory, and Non-refractory Antiphospholipid Syndrome

Abstract

Antiphospholipid syndrome (APS) is an autoimmune disorder characterized by pregnancy morbidity or thrombosis and persistent antiphospholipid antibodies (aPL) that bind to the endothelium and induce endothelial activation, which is evidenced by the expression of adhesion molecules and the production of reactive oxygen species (ROS) and subsequent endothelial dysfunction marked by a decrease in the synthesis and release of nitric oxide (NO). These endothelial alterations are the key components for the development of severe pathological processes in APS. Patients with APS can be grouped according to the presence of other autoimmune diseases (secondary APS), thrombosis alone (thrombotic APS), pregnancy morbidity (obstetric APS), and refractoriness to conventional treatment regimens (refractory APS). Typically, patients with severe and refractory obstetric APS exhibit thrombosis and are classified as those having primary or secondary APS. The elucidation of the mechanisms underlying these alterations according to the different groups of patients with APS could help establish new therapies, particularly necessary for severe and refractory cases. Therefore, this study aimed to evaluate the differences in endothelial activation and dysfunction induced by aPL between patients with refractory obstetric APS and other APS clinical manifestations. Human umbilical vein endothelial cells (HUVECs) were stimulated with polyclonal immunoglobulin-G (IgG) from different groups of patients n = 21), including those with primary (VTI) and secondary thrombotic APS (VTII) and refractory primary (RI+), refractory secondary (RII+), and non-refractory primary (NR+) obstetric APS. All of them with thrombosis. The expression of adhesion molecules; the production of ROS, NO, vascular endothelial growth factor (VEGF), and endothelin-1; and the generation of microparticles were used to evaluate endothelial activation and dysfunction. VTI IgG induced the expression of adhesion molecules and the generation of microparticles and VEGF. RI+ IgG induced the expression of adhesion molecules and decreased NO production. RII+ IgG increased the production of microparticles, ROS, and endothelin-1 and reduced NO release. NR+ IgG increased the production of microparticles and endothelin-1 and decreased the production of VEGF and NO. These findings reveal differences in endothelial activation and dysfunction among groups of patients with APS, which should be considered in future studies to evaluate new therapies, especially in refractory cases.

Keywords: antiphosholipid syndrome; antiphospholipid syndrome; beta 2-glycoprotein I; endothelial activation and dysfunction; endothelial cells; immunoglobulin G.

FIGURE 1

Immunoglobulin-G (IgG) from patients with primary antiphospholipid syndrome (APS) [primary thrombotic APS (VTI) and refractory primary obstetric APS (RI+)] induced an increase in the adhesion molecule expression in human umbilical vein endothelial cells (HUVECs). (A,E) Lipopolysaccharide (LPS) at a concentration of 4 μg/ml was included as a positive control, and the increased E-selectin and vascular cell adhesion molecule 1 (VCAM-1) expression was compared with unstimulated cells. (B–D) The median fluorescence intensity (MFI) of E-selectin increased in HUVECs stimulated with beta 2-glycoprotein-I (β2GPI) and IgG from patients with VTI and RI+ compared with the IgG/normal human serum (NHS) control. (F,G) The MFI of VCAM-1 increased in HUVECs stimulated with β2GPI and IgG from patients with RI+ compared with the IgG/NHS control. A two-way ANOVA and the Holm–Sidak post-test (*p < 0.05 and **p < 0.01) were performed. The results were obtained from three independent experiments. (A–C,E,G) Representative histograms. PM/VT, pregnancy morbidity and vascular thrombosis.

FIGURE 2

Immunoglobulin-G from patients with refractory and primary obstetric APS (RI+) induces increased monocyte adhesion to the endothelium. (A) Baseline adhesion. (B) LPS. The adhesion of monocytes to the endothelium in cells stimulated with β2GPI plus: (C) IgG/NHS; (D) IgG/patients negative for aPL with thrombosis without autoimmune disease (VTI/aPL-); (E) IgG/patients negative for aPL with thrombosis and SLE (VTII/aPL-); and (F) IgG/RI+. (A–F) Representative images. (G) The results obtained using the spectrofluorometer are indicated in relative fluorescence units (RFU) emitted by the endothelial monolayer without monocytes and with monocytes attached without stimulus or baseline adhesion and with LPS. (H) IgG from patients with RI + and β2GPI increased the adhesion of monocytes compared with IgG from patients with VTI/aPL- and VTII/aPL-. A two-way ANOVA and the Holm–Sidak post-test (*p < 0.05) were performed. The results were obtained from three independent experiments. The arrow indicates the aggregates of monocytes. PM/VT, pregnancy morbidity and vascular thrombosis.

FIGURE 3

Immunoglobulin-G from patients with refractory and secondary obstetric APS (RII+) induced O₂^– mitochondrial production. (A,B,D) Hydrogen peroxide and LPS increased the MFI of MitoSOX in HUVECs. (C,E) RII + IgG group induced mitochondrial O₂^– production in HUVECs without β2GPI compared with the IgG from NHS and non-refractory primary obstetric APS (NR+) groups. A two-way ANOVA and the Holm–Sidak post-test (*p < 0.05) were performed. The results were obtained from five independent experiments. (A–C) Representative histograms. PM/VT, pregnancy morbidity and vascular thrombosis.

FIGURE 4

Immunoglobulin-G from patients with secondary refractory APS (RII+) increased the number of 1-μm microparticles (MPs). (A) Electronic noise was excluded from all the analyses. (B) The binding buffer for annexin V staining did not present a significant number of events. (C,D) LPS stimulation increased the number of MPs compared with no stimulation. (E) For all the analyses, the approximate size of the MPs was delimited with 0. 5-, 1-, and 2-μm beads. (F–H) IgG of patients with RII + induced the generation of 1-μm MPs in a β2GPI-dependent manner compared with IgG from NHS control. A two-way ANOVA and the Holm–Sidak post-test (*p < 0.05) were performed. The results were obtained from three independent experiments. PM/VT, pregnancy morbidity and vascular thrombosis.

FIGURE 5

Immunoglobulin-G from patients with non-refractory primary obstetric APS (NR+) induced the production of 0.5- and 1-μm endothelial MPs. (A,G) Control indicates MPs without antibodies, which were used to define the location of the negative and positive MPs for Annexin V and CD31. (A–F,K) IgG/NR+ increased the number of 0.5-μm CD31+/annexin V+ MPs in the presence of β2GPI, compared with NHS IgG, secondary thrombotic APS (VTII)/antiphospholipid antibodies (aPL) IgG, VTII IgG, and IgG of patients with NR+ without β2GPI. (G–J,L) IgG from patients with NR+ increased the number of MPs of 1-μm CD31+/annexin V- in a β2GPI-dependent manner compared with the IgG/NHS control and patients with NR+ without β2GPI. (M) IgG from patients with VTI and NR+ increased the number of total MPs of 0.5-(μm annexin V (in a (β2GPI-dependent manner more than IgG/NHS; Annexin V (MPs are the indicators of procoagulant potential. A two-way ANOVA and the Holm–Sidak post-test (*p < 0.05) were performed. The results were obtained from three independent experiments. PM/VT, pregnancy morbidity and vascular thrombosis. ****p < 0.0001.

FIGURE 6

All IgG from patients with pregnancy morbidity and thrombosis (RI+, RII+, and NR+) induced endothelial dysfunction through a decrease in nitric oxide (NO) production in HUVECs. (A) β2GPI alone did not alter NO production in contrast with the basal control. (A,B) NO release is reduced by the addition of NG-nitro-L-arginine methyl ester (L-NAME). (B) Vascular endothelial growth factor (VEGF) induced NO production in a dose-dependent manner in contrast with cells treated with L-NAME, which exerted its antagonistic effect. (C) IgG/RI+ decreased NO bioavailability in contrast with the IgG/NHS control without β2GPI. All IgG from patients with pregnancy morbidity and thrombosis (RI+, RII+, and NR+) plus β2GPI reduced NO production compared with the IgG/NHS control (**p < 0.01). A two-way ANOVA and the Holm–Sidak post-test (*p < 0.05) were performed. (D,E) Aspirin (ASA) had no modulatory effect on the reduction of NO induced by IgG. (F,G) Enoxaparin (ENX) restored NO bioavailability reduced by IgG/RI+ with and without β2GPI. (H,I) Hydroxychloroquine (HCQ) restored NO bioavailability reduced by IgG/RI+ with β2GPI. A t-test (*p < 0.05) was performed in ASA and EXN analyses. The HCQ data were analyzed using the Mann–Whitney test (*p < 0.05). The results were obtained from four independent experiments. PM/VT, pregnancy morbidity and vascular thrombosis.

FIGURE 7

All IgG from patients with APS-induced endothelial dysfunction through an alteration in endothelin-1 or VEGF production in HUVEC supernatants. (A) RII+ and NRI+ IgG groups of patients increased endothelin-1 production compared with IgG/NHS without β2GPI. RI+ IgG with β2GPI increased the production of endothelin-1 compared with this same group without β2GPI. (B) IgG from patients with VTI and VTII increased the VEGF in contrast with NHS IgG without β2GPI. IgG from patients with NR+ decreased the VEGF amount in contrast with NHS IgG without β2GPI. NR+ IgG and β2GPI increased the VEGF production compared with this same IgG without β2GPI. A two-way ANOVA and the Holm–Sidak post-test (*p < 0.05; **p < 0.01) were performed. The results were obtained from three independent experiments. PM/VT, pregnancy morbidity and vascular thrombosis.

FIGURE 8

Endothelial activation and dysfunction in APS: clinical manifestation association and role of β2GPI. The pathogenetic effects associated with endothelial activation induced by aPL with or without β2GPI contribute to thrombosis or pregnancy morbidity. (A) VTI+ IgG: (1) increased the expression of E-selectin; (2) generated positive annexin V procoagulant MPs of 0.5 μm; and (3) increased the VEGF generation. (B) RI+ IgG: (1) increased the expression of E-selectin and VCAM; (2) induced monocyte adhesion; (3) decreased NO bioavailability, which was modulated by HCQ and ENX; and (4) decreased NO bioavailability, which was modulated by ENX. (C) NR+ IgG: (1) decreased NO bioavailability; (2) increased the number of 0.5-μm procoagulant MPs (CD31+/annexinV+ and CD31-/annexinV+); (3) increased CD31+/annexinV, MP of 1 μm; (4) increased endothelin-1 production; and (5) reduced VEGF. (D) RII+ IgG: (1) decreased NO bioavailability; (2) increased MPs of 1 μm; (3) generated endothelin-1 production; and (4) induced oxidative stress (mitochondrial O₂^– production).