Advanced oxidation protein products activate vascular endothelial cells via a RAGE-mediated signaling pathway

Abstract

The accumulation of advanced oxidation protein products (AOPPs) has been linked to vascular lesions in diabetes, chronic renal insufficiency, and atherosclerosis. However, the signaling pathway involved in AOPPs-induced endothelial cells (ECs) perturbation is unknown and was investigated. AOPPs modified human serum albumin (AOPPs-HSA) bound to the receptor for advanced glycation end products (RAGE) in a dose-dependent and saturable manner. AOPPs-HSA competitively inhibited the binding of soluble RAGE (sRAGE) with its preferential ligands advanced glycation end products (AGEs). Incubation of AOPPs, either prepared in vitro or isolated from uremic serum, with human umbilical vein ECs induced superoxide generation, activation of NAD(P)H oxidase, ERK 1/2 and p38, and nuclear translocation of NF-kappaB. Activation of signaling pathway by AOPPs-ECs interaction resulted in overexpression of VCAM-1 and ICAM-1 at both gene and protein levels. This AOPPs-triggered biochemical cascade in ECs was prevented by blocking RAGE with either anti-RAGE IgG or excess sRAGE, but was not affected by the neutralizing anti-AGEs IgG. These data suggested that AOPPs might be new ligands of endothelial RAGE. AOPPs-HSA activates vascular ECs via RAGE-mediated signals.

FIG. 1.

AOPPs-induced ROS production. (A) ROS production detected by DCF fluorescence in HUVECs stimulated by indicated concentrations of AOPPs-HSA, AOPPs-F or native HSA. (B) Time course of AOPPs-HSA (200 μg/ml)-induced ROS production. (C) AOPPs-induced ROS generation in HUVECs pretreated with apocynin, DPI, L-NAME, rotenone, oxypurinol, c-SOD, or Gö6983. (D) NAD(P)H-dependent O₂⁻ generation determined by lucigenin-enhanced chemiluminescence in HUVECs homogenates. (E) Time course of AOPPs-HSA (200 μg/ml)-induced O₂⁻ production. (F) AOPPs-HSA-induced enhancement of O₂⁻ production was markedly attenuated by pretreating the cells with apocynin, DPI and Gö6983. Data from three independent experiments are shown as mean ± SEM. ANOVA, p < 0.001 in A, B, C, D, E, and F; *p < 0.05 vs. HSA group; #p < 0.05 vs. group without respective inhibitors.

FIG. 2.

Interaction of AOPPs-HSA with RAGE: role on AOPPs-HSA-induced ROS generation. (A) Microplate wells were coated with 10 μg/ml of AOPPs-HSA, CML-HSA, GA-HSA, or native HSA. After washing and blocking, the wells were incubated with indicated concentrations of sRAGE. The bound sRAGE was detected with a HRP-conjugated monoclonal α-RAGE IgG. (B) CML-HSA (10 μg/ml) was immobilized onto the wells. sRAGE (4 μg/ml) was pre-incubated with indicated concentrations of AOPPs-HSA, CML-HSA, GA-HSA, or native HSA, and then added to the wells. The binding of sRAGE with immobilized AGEs-HSA was detected as described in A. (C) HUVECs were pre-incubated with a polyclonal α-RAGE IgG, a nonimmune IgG, or excess sRAGE, and then stimulated with AOPPs-HSA. ROS generation was determined by measuring the DCF fluorescence. (D) AOPPs-HSA (200 μg/ml) was pre-incubated with indicated concentrations of an α-AOPPs or a nonimmune IgG, and then interacted with HUVECs. ROS generation was determined as described above (lower panel). The α-AOPPs recognized AOPPs-HSA in Western blot (upper panel, lane 1), but did not react with CML- (lane 2), GA- (lane 3), GC- (lane 4), and RB-derived AGEs (lane 5) or native HSA (lane 6). (E) AOPPs-HSA was pre-incubated with indicated concentrations of 6D12 antibody and then interacted with HUVECs. ROS generation was measured as described above. (F) 6D12 antibody reacted with CML, GA- and GC-derived AGEs, and, to a lesser extent, with RB-derived protein. Data from three independent experiments are shown as mean ± SEM. ANOVA, p < 0.001 in A, B, and C; p < 0.05 in D. *p < 0.05 vs. wells coated with CML-HSA; #p < 0.05 vs. wells coated with AOPPs-HSA; §p < 0.05 vs. group pre-incubated with CML-HSA; II p < 0.05 vs. group pre-incubated with AOPPs-HSA; Δ p < 0.05 vs. group without respective inhibitors.

FIG. 3.

AOPPs-induced activation of ECs NAD(P)H oxidase. (A) HUVECs were incubated with 200 μg/ml of AOPPs-HSA, AOPPs-F, or native HSA for 15 ∼ 30 min. In some experiments, cells were pre-incubated with an α-RAGE IgG, a PKC inhibitor (Gö6983), or a nonimmune IgG and then stimulated with AOPPs-HSA. Phosphorylation of p47^phox was assayed by immunoprecipitation (IP) using an α-p47^phox and detected by immunoblotting (IB) using an α-pan-p47^phox and an α-phosphoserine as the primary antibodies. (B) AOPPs-induced binding of p47^phox to p22^phox. HUVECs were treated as described above. IP was performed by using an α-p22^phox and IB was conducted by using an α-p47^phox. (C) Interaction of p47^phox with Nox 2. HUVECs were treated as described above. IP was performed by using an α-Nox 2 and IB was conducted by using an α-p47^phox. (D) Interaction of p47^phox with Nox 4. HUVECs were treated as described above. IP was performed by using an α-Nox 4 and IB was conducted by using an α-p47^phox. (E) Representative photos of AOPPs-induced membrane translocation of p47^phox. HUVECs were incubated with an α-p47^phox and then with a FITC-conjugated α-mouse immunoglobulin and staining with PI. Data from three independent experiments are shown as mean ± SEM. ANOVA, p < 0.001 in A and B; *p < 0.05 as compared with groups treated with medium alone or HSA; #p < 0.05 vs. group without respective inhibitors. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article at www.liebertonline.com/ars).

FIG. 4.

AOPPs-HSA-induced MAPK activation in HUVECs. HUVECs were incubated with 200 μg/ml of AOPPs-HSA for indicated time. Phospho-ERK 1/2 (p-ERK 1/2) and ERK 1/2 (A), phospho-p38 (p-p38) and p38 (B), and phospho-JNK 1/2 (p-JNK 1/2) and JNK 1/2 (C) were detected by immunoblotting. In other experiments, HUVECs were pre-incubated with the α-RAGE IgG or nonimmune IgG, apocynin, c-SOD, or U0126/SB 203580. AOPPs-induced activation of ERK 1/2 (D) or p38 (E) was determined by immunoblotting. Data from three independent experiments are shown as mean ± SEM. ANOVA, p < 0.001 in A, B, D, and E, and p = 1.000 in C; *p < 0.05 as compared with groups treated with medium alone or HSA; #p < 0.05 vs. group without respective inhibitors.

FIG. 5.

AOPPs-induced NF-κB activation in HUVECs. (A) HUVECs were incubated with 200 μg/ml of AOPPs-HSA for indicated time. NF-κB p65 in the cytoplasmic extracts (c-p65) or in the nuclear extracts (n-p65) were detected by immunoblotting. (B) NF-κB p65 translocation was detected by immunofluorescence and PI staining. (C) NF-κB activation in nuclear extracts of AOPPs-HSA-stimulated HUVECs was detected by EMSA. (D) HUVECs were pretreated with the α-RAGE IgG or nonimmune IgG, apocynin, c-SOD, U0126/SB203580, and SN50 before AOPPs-HSA stimulation. The ratio of n-p65 and c-p65 was determined by immunoblotting. Data from three independent experiments are shown as mean ± SEM. ANOVA, p < 0.001 in A and D; *p < 0.05 as compared with group treated with medium alone; #p < 0.05 vs. group without respective inhibitors. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article at www.liebertonline.com/ars).

FIG. 6.

AOPPs-induced ICAM-1 and VCAM-1 expression in HUVECs. HUVECs were incubated with indicated concentrations of AOPPs-HSA for 6 h (A) or with 200 μg/ml of AOPPs-HSA for indicated time (B). ICAM-1 and VCAM-1 mRNA expression was measured by real time RT-PTCR. Expression of ICAM-1 and VCAM-1 protein was detected by immunoblotting (C and D). (E) Representative immunofluorescent staining for ICAM-1 and VCAM-1 in HUVECs. (F and G) HUVECs were pretreated with the α-RAGE IgG, apocynin, c-SOD, U0126/SB203580, and SN50. AOPPs-induced expression of ICAM-1 (F) and VCAM-1 (G) were determined by immunoblotting. Data from three independent experiments are shown as mean ± SEM. ANOVA, p < 0.001 in A, B, C, D, F, and G; *p < 0.05 as compared with group treated with medium alone; #p < 0.05 vs. group without respective inhibitors. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article at www.liebertonline.com/ars).