4-Hydroxynonenal Contributes to Angiogenesis through a Redox-Dependent Sphingolipid Pathway: Prevention by Hydralazine Derivatives

Abstract

The neovascularization of atherosclerotic lesions is involved in plaque development and may contribute to intraplaque hemorrhage and plaque fragilization and rupture. Among the various proangiogenic agents involved in the neovascularization process, proatherogenic oxidized LDLs (oxLDLs) contribute to the formation of tubes via the generation of sphingosine 1-phosphate (S1P), a major mitogenic and proangiogenic sphingolipid mediator. In this study, we investigated whether 4-hydroxynonenal (4-HNE), an aldehydic lipid oxidation product abundantly present in oxLDLs, contributes to their proangiogenic properties. Immunofluorescence analysis of human atherosclerotic lesions from carotid endarterectomy showed the colocalization of HNE-adducts with CD31, a marker of endothelial cells, suggesting a close relationship between 4-HNE and neovessel formation. In vitro, low 4-HNE concentration (0.5-1 µM) elicited the formation of tubes by human microvascular endothelial cells (HMEC-1), whereas higher concentrations were not angiogenic. The formation of tubes by 4-HNE involved the generation of reactive oxygen species and the activation of the sphingolipid pathway, namely, the neutral type 2 sphingomyelinase and sphingosine kinase-1 (nSMase2/SK-1) pathway, indicating a role for S1P in the angiogenic signaling of 4-HNE. Carbonyl scavengers hydralazine and bisvanillyl-hydralazone inhibited the nSMase2/SK1 pathway activation and the formation of tubes on Matrigel® evoked by 4-HNE. Altogether, these results emphasize the role of 4-HNE in the angiogenic effect of oxLDLs and point out the potential interest of pharmacological carbonyl scavengers to prevent the neovascularization process.

Figure 1

4-HNE is colocalized with CD31 in human atherosclerotic lesions. Paraffin sections of human carotid plaques from endarterectomy were analyzed. In (a), hematoxylin/eosin (H/E) staining and immunostaining for 4-HNE-adduct (HNE) and CD68 expression. In (b), immunofluorescence analysis of 4-HNE-adduct expression (green) and CD31 (red), with nuclei counterstaining by DAPI. Int: intima. These pictures are representative of analysis for 3 separate advanced carotid plaques.

Figure 2

Dual effect of 4-HNE on neocapillary formation by HMEC-1. (a) Dose-response effect of 4-HNE on capillary tubes formed by HMEC-1. Cells were grown on Matrigel in MCDB131 culture medium supplemented with 0.1% FBS and PBS (negative control) or 4-HNE (in PBS) varying from 0.1 to 20 µM. After 18 h incubation, the cells were stained with calcein (1 μmol/1, 30 min) and photographed (Nikon Coolpix 995 camera) under a fluorescence microscope. Tube formation was expressed as linked cells per 100 cells. Results are means ± SEM of 6 to 8 experiments. Right panel, representative pictures of the experiments. ^∗P < 0.05; ns: not significant. (b) Live-dead experiment on HMEC-1 stimulated by increasing 4-HNE concentrations and performed using the fluorescent DNA probes, permeant green Syto13 (0.6 µM) and nonpermeant red propidium iodide (1 µM). The results are expressed as the number (%) of living, apoptotic, or necrotic cells versus total cells. Right panels, representative pictures of fluorescence microscopy of HMEC-1, incubated for 18 h without (control) or with 4-HNE 1 µM or 20 µM. Means ± SEM of 3 experiments. ^∗P < 0.05; ns: not significant.

Figure 3

Implication of ROS in tube formation by 4-HNE. (a) Time-course of intracellular ROS production evoked by 4-HNE (0.5 µM) in HMEC-1 and measured fluorometrically using the H2DCFDA probe (5 µM final concentration). Results are expressed as % of the unstimulated control. (b) Effect of the antioxidant trolox (10 µM) and of NADPH oxidase inhibitors DPI and Vas2870 (10 µM each) and of the anti-Lox-1 antibody (5 µg/ml) on ROS generated by HMEC-1 after 30 min of contact with 4-HNE (0.5 µM). (c) Effect of trolox, DPI, and Vas2870 and anti-Lox-1 antibody, on tube formation elicited by 4-HNE (0.5 µM). Representative pictures of tube formation in the presence of 4-HNE (0.5 µM) and without (none) or with inhibitors Vas2870 (Vas) or anti-Lox-1 Ab (aLox1). (d) Effect of the anti-Lox-1 Ab on tube formation elicited by oxLDL (20 µg/ml). Note that the anti-Lox-1 Ab has no effect on tubes formed by 4-HNE-stimulated HMEC-1 but inhibits tubes formed by oxLDL-stimulated cells. These data are means ± SEM of 5 separate experiments. ^∗P < 0.05; ns: not significant.

Figure 4

Implication of nSMase2 and SK1 in tube formation elicited by 4-HNE. ((a), (b)) Time-course of nSMase2 activation by 4-HNE (0.5 µM) (a) and inhibitory effect of trolox (10 µM) and of the nSMase2 inhibitor GW4869 (5 µM) on nSMase2 activation induced by 4-HNE (b) after 90 min incubation. (c), (d) Time-course of SK1 activation by 4-HNE (0.5 µM) (c) and effect of trolox, GW4869 and of the SK1 inhibitor DMS (1 µM) on SK1 activation by 4-HNE (d), measured after 90 min incubation. (e) Effect of trolox, GW4869, and DMS on capillary tube formation on Matrigel elicited by 4-HNE (0.5 µM). Means ± SEM. ^∗P < 0.05; ns: not significant.

Figure 5

Effect of hydralazine and BVH on the angiogenic signaling of 4-HNE. (a) Chemical structures of hydralazine (Hdz), bisvanillin (BV), and bisvanillyl-hydralazone (BVH). (b) Effect of Hdz and BVH (10 µM each), on ROS generated by HMEC-1 after 30 min of contact with 4-HNE (0.5 µM). (c) Effect of Hdz and BVH (10 µM each), on SK1 activation by 4-HNE, measured after 90 min incubation. (d) Effect of Hdz, BVH, and BV (10 µM each) on tube formation elicited by 4-HNE (0.5 µM). These data are means ± SEM of 4 separate experiments. ^∗P < 0.05; ns: not significant.