Endothelial dysfunction represents one of the earliest events in vascular atherogenesis. Proinflammatory stimuli activate endothelial cells, resulting in an increased expression of adhesion molecules and chemoattractants that mediate leukocyte and monocyte adhesion, migration, and homing. High density lipoproteins (HDL) inhibit endothelial cell expression of adhesion molecules in response to proinflammatory stimuli. In the present work, we demonstrate that the modification of HDL3 (the major and the most antiatherogenic HDL subfraction) by 15-lipoxygenase (15-LO), an enzyme overexpressed in the atherosclerotic lesions, impairs the anti-inflammatory activity of this lipoprotein. The 15-LO-modified HDL3 failed to inhibit TNF-α-mediated mRNA and protein induction of adhesion molecules and MCP-1 in several models of human endothelial cells, and promoted inflammatory response by up-regulating the expression of such mediators of inflammation and by increasing monocyte adhesion to endothelial cells. Moreover, 15-LO-modified HDL3 were unable to contrast the formation of reactive oxygen species in cells incubated with TNF-α, and increased the reactive oxygen species content in unstimulated cells. Activation of NF-κB and AP-1 was mainly involved in the expression of adhesion molecules and MCP-1 induced by 15-LO-HDL3. Altogether, these results demonstrate that enzymatic modification induced by 15-LO impaired the protective role of HDL3, generating a dysfunctional lipoprotein endowed with proinflammatory characteristics.
Inflammation plays a relevant role in the initiation and progression of atherosclerosis (1, 2, 3). Vascular endothelial dysfunction represents one of the earliest events in atherogenesis (4, 5). A number of proinflammatory stimuli can activate endothelial cells, resulting in an increased expression of adhesion molecules (VCAM-1, ICAM-1, and E-selectin) (6) and chemoattractants, such as MCP-1 (2, 7, 8). These molecules allow leukocyte and monocyte adhesion, migration, and homing, thus perpetuating vascular wall inflammation processes (9).
Plasma levels of high density lipoprotein (HDL)3 cholesterol inversely correlate with the incidence of coronary heart disease (10, 11, 12); however, clinical cardiovascular events occur in subjects with normal or even higher HDL cholesterol levels (13, 14), suggesting that quality and function of HDL are a relevant issue. The protective role of HDL is believed to be largely related to its function in reverse cholesterol transport (15). HDL possess, however, a number of additional antiatherogenic properties (16), such as antioxidative, anti-inflammatory, antithrombotic, and vasodilatory activities (17, 18, 19, 20, 21). The atheroprotective properties of HDL are associated with the chemico-physical integrity of the lipoprotein. HDL can become dysfunctional and proinflammatory under specific conditions, including coronary heart disease, metabolic syndrome, and infections (22). Multiple mechanisms can confer proinflammatory features to HDL, mainly impairing its role in reverse cholesterol transport (23, 24, 25).
In a previous work, we established that HDL3 can be enzymatically modified in vitro by 15-lipoxygenase (15-LO), generating a dysfunctional lipoprotein that exhibits a decreased ability to promote cholesterol efflux from lipid-laden macrophages (26). This enzyme has been associated with inflammation and atherosclerosis: specific proinflammatory cytokines up-regulate 15-LO expression in macrophages (27, 28); the active form of 15-LO is overexpressed in atherosclerotic lesions, localizes in macrophage-rich areas, and colocalizes with oxidized low density lipoproteins (LDL) (29, 30, 31, 32), suggesting a role in LDL oxidation. Because 15-LO-mediated modification significantly impairs the ability of HDL3 to function as an efficient cholesterol acceptor (26), we investigated whether this modification also affects the anti-inflammatory activity of HDL3. Experiments were therefore performed to evaluate the ability of 15-LO-modified HDL3 to inhibit TNF-α-mediated expression of adhesion molecules and MCP-1 in endothelial cells and to verify whether 15-LO-modified HDL3 exhibited a proinflammatory profile. The adhesiveness properties of endothelial cells exposed to 15-LO-modified HDL3 were tested using monocyte-binding assay. Finally, we investigated the possible pathways responsible for the effects induced by 15-LO-modified HDL3.
Materials and Methods
Medium M199 was from Invitrogen. MCDB 131, FBS, penicillin-streptomycin, glutamine, hydrocortisone, TNF-α, 2,7-dichlorofluorescein diacetate (DCFH-DA), 2,7-bis(2-carboxyethyl)-5(6)-carboxyfluorescein acetoxymethylester (BCECF-AM), and SP600125 were from Sigma-Aldrich; BAY11-7085 was from Alexis; and PD10 columns and ECL were from Amersham Biosciences. Endothelial cell growth factor and epidermal growth factor were from Boehringer Mannheim. Abs were purchased as follows: anti-phospho-IκBα and anti-c-Jun (detecting endogenous levels of total c-Jun, regardless of phosphorylation state) from Cell Signaling Technology; anti-CD54 (ICAM-1), anti-CD106 (VCAM-1), and anti-CD62 (E-selectin) from Bio Optica; goat anti-mouse IgG FITC from BD Biosciences; anti-β-actin and anti-mouse IgG peroxidase conjugate from Sigma-Aldrich; and anti-rabbit IgG peroxidase conjugate from Bio-Rad.
Expression and purification of 15-LO
Rabbit reticulocyte-type 15-LO was prepared, as described (33). The final preparation was electrophoretically pure (>95%) and exhibited a linoleic acid oxygenase turnover rate of 20 s−1.
The human 15-LO (h15-LO) was expressed in Escherichia coli as recombinant N-terminal his-tag fusion protein. After growth in Luria-Bertani medium containing ampicillin for selection, transformed bacteria were spun down, resuspended in PBS, lysed, and centrifuged to remove cell debris. To purify h15-LO, the supernatant was fractionated sequentially on a Ni-agarose column and on a semipreparative Mono-Q anion exchange column. The final enzyme preparation was electrophoretically homogeneous (SDS-PAGE) and exhibited a molecular turnover rate of linoleic acid oxygenation of ∼4 s−1.
HUVEC were isolated according to established procedures (34) and cultured in medium M199 supplemented with 20% FBS, endothelial cell growth factor (20 μg/ml), heparin (15 U/ml), penicillin-streptomycin (1%), and glutamine (1%). Cells were used between the third and fifth in vitro passage. Primary passage cryopreserved human aortic endothelial cells (HAEC) were obtained from Cambrex and cultured in the same medium used for HUVEC. Human microvascular endothelial cells (HMEC-1) (35) were provided by F. Candal (Centers for Disease Control, Atlanta, GA) and by D. Taramelli (Department of Public Health, Microbiology, and Virology, University of Milan, Milan, Italy). Cells were cultured in MCDB 131 medium supplemented with 10% FBS, 10 ng/ml epidermal growth factor, 1 μg/ml hydrocortisone, 1% penicillin-streptomycin, and 1% glutamine.
The monocytoid cell line U937 was cultured in RPMI 1640 supplemented with 10% FBS, 1% penicillin-streptomycin, and 1% glutamine.
Isolation and modification of HDL3
HDL3 (d = 1.125–1.21 g/ml) were isolated by sequential ultracentrifugation at 4°C from the plasma of healthy volunteers (36). HDL3 were dialyzed using a Sephadex G25 column (PD10) with 0.01% PBS-EDTA (pH 7.4), and then sterilized using a 0.22-μm filter (Corning Glass) and stored at 4°C. The protein content was evaluated by the Lowry method (37), using BSA as a standard.
HDL3 (1 mg protein/ml) were modified with 32 μg of rabbit reticulocyte 15-LO/mg HDL protein at 37°C for 72 h (26). HDL3 oxidation extent was evaluated as the thiobarbituric acid-reactive substances content and as hydroxy polyunsatured fatty acids:polyunsatured fatty acids ratio (%) (26). Modification of HDL3 by h15-LO was conducted, as described above, taking into account the different sp. act.
Apoprotein cross-linking after modification was evaluated by Western blotting, as described (26).
Sparse, proliferating HUVEC were preincubated for 18 h in the presence of native or 15-LO-modified HDL3 (100 μg/ml) before the addition of TNF-α (10 ng/ml). At the end of the incubation, cells were harvested by trypsinization, washed in PBS-BSA (1%), and incubated for 20 min at 4°C with anti-CD54 (ICAM-1) or anti-CD62 (E-selectin) mAbs (dilution 1/20). Cells were then washed with PBS-BSA (1%) and incubated for 20 min at 4°C with goat anti-mouse IgG FITC (1/20). After washing, Ag expression was measured by flow cytometry (FACScan; BD Biosciences). Isotype controls were obtained by incubating cells with goat anti-mouse IgG FITC to determine nonspecific fluorescence. Cells were gated, and data were obtained from fluorescence channels in a logarithmic mode. A total of 10,000 events was analyzed; data were processed using the CellQuest program (BD Biosciences).
Whole-cell lysates were prepared in lysis buffer (2% SDS, 62.5 mM Tris, 50 mM DTT, 1 mM PMSF, and 5 μg/ml aprotinin) and sonicated. Cellular proteins were separated on a 10% SDS-PAGE, and then transferred onto a nitrocellulose membrane. Membranes were saturated with 5% nonfat milk in PBS-T (PBS-0.1% Tween 20) for 1 h at room temperature, followed by overnight incubation with the selected primary Ab at 4°C and 1-h incubation with a 1/1000 dilution of a goat anti-rabbit or anti-mouse IgG-HRP conjugate. Immunocomplexes were detected by ECL, followed by autoradiography. The bands were quantified by a computer-assisted system for image analysis (ISF Image 1.52), and the expression of each Ag was corrected for β-actin cellular content.
HUVEC were preincubated for 1 h with 50 μg/ml DCFH-DA, and then incubated 18 h with native or 15-LO-modified HDL3 and with TNF-α for an additional 1 h. Alternatively, after preincubation with DCFH-DA, cells were incubated with native or 15-LO-modified HDL3 for 1 h. Cells were harvested by trypsinization, and resuspended in 1% PBS/BSA for immediate determination of oxidative stress by flow cytometry. For each sample, 10,000 events were counted and intracellular oxidant stress was monitored by changes in fluorescence intensity deriving from intracellular probe oxidation.
Quantitative real-time PCR
Total RNA was extracted and reverse transcribed. A total of 3 μl of cDNA was amplified by real-time quantitative PCR with 1× SYBR Green universal PCR mastermix (Bio-Rad) (38). The sequence of the primers used for RLP-13A (housekeeping gene), adhesion molecules, and MCP-1 amplification is reported in Table I. Each sample was analyzed in duplicate using the IQ-Cycler (Bio-Rad). For quantification, the target genes were normalized to the RLP-13A content.
|Primer .||Sequence .|
|Primer .||Sequence .|
Monocyte adhesion assay
U937 cells were labeled with the fluorescent dye BCECF-AM at 10 μM final concentration in RPMI 1640 containing 10% FBS at 37°C for 1 h. HUVEC were incubated 18 h with TNF-α (10 ng/ml), native, or 15-LO-modified HDL3 (100 μg/ml), and then cocultured with BCECF-AM-labeled U937 cells for 1 h at 37°C. Alternatively, cells were preincubated overnight with native or 15-LO-modified HDL3 before addition of TNF-α for an additional 4 h. At the end of the experiment, unbound monocytes were removed by washing three times with medium. Cells were then harvested by trypsinization, fixed, and immediately analyzed by flow cytometry. Endothelial cells and monocytes were gated, and 20,000 events in the gate of endothelial cells were counted. Endothelial cell-associated fluorescence was measured as an index of monocyte adhesion, and data were expressed as percentage of control.
Values are expressed as means ± SD. The statistical significance of the differences between groups was determined by Student’s t test, and values of p < 0.05 were considered to be significant.
Low doses of HDL3 reduce TNF-α-induced ICAM-1 expression
Physiological concentrations of HDL (up to 2 mg/ml) efficiently inhibit TNF-α-induced expression of VCAM-1 in endothelial cells (20, 39). We established that, under our experimental conditions, endothelial expression of ICAM-1 in response to TNF-α was efficiently inhibited by 100 μg/ml HDL3 (Fig. 1). This concentration of HDL3 was used in all the following experiments.
Effect of 15-LO-modified HDL3 on TNF-α-induced ICAM-1 expression
To test the effect of 15-LO-mediated HDL3 modification on ICAM-1 expression, HUVEC were incubated overnight with native or 15-LO-modified HDL3 (100 μg/ml) before addition of TNF-α (10 ng/ml). Flow cytometry analysis revealed that modification with 15-LO significantly impaired the ability of HDL3 to inhibit TNF-α-induced ICAM-1, VCAM-1, and E-selectin surface expression (Fig. 2).
Because HUVEC are derived from vessels that are not susceptible to atherosclerosis, we tested the effects of HDL3 modification in two additional models of endothelial cells; 15-LO-modified HDL3 did not inhibit TNF-α-induced ICAM-1 expression in HAEC (Fig. 3,A) or in HMEC-1 (Fig. 3 B), suggesting that the observed effects were not associated with a specific endothelial cell type.
Effect of 15-LO-modified HDL3 on ICAM-1 expression
When HUVEC were incubated with native HDL3, ICAM-1 surface expression was unchanged, as compared with control cells. Conversely, when cells were incubated with 15-LO-modified HDL3, ICAM-1 surface expression was significantly increased (Fig. 4,A). ICAM-1 total protein (surface plus intracellular) was also increased (Fig. 4 B).
Effect of modified HDL3 on adhesion molecules and MCP-1 gene transcription
Incubation of HUVEC with native HDL3 did not affect mRNA levels of adhesion molecules and MCP-1 (Fig. 5). In agreement with protein surface expression data, mRNA levels of adhesion molecules and MCP-1 were significantly increased in HUVEC incubated with 15-LO-modified HDL3 (Fig. 5).
Preincubation of HUVEC with native HDL3 significantly decreased TNF-α-induced mRNA expression of ICAM-1, VCAM-1, and E-selectin (Fig. 6, A–C). When cells were preincubated with 15-LO-modified HDL3, the inhibitory effect was lost (Fig. 6, A–C). In cells pretreated with native HDL3, a decreased MCP-1 mRNA expression in response to TNF-α was observed (Fig. 6 D); once more, 15-LO-modified HDL3 failed to inhibit MCP-1 expression in TNF-α-stimulated cells.
Adhesion molecules and MCP-1 mRNA levels were also decreased in HMEC-1 preincubated with native HDL3 before exposure to TNF-α (Fig. 7); preincubation with 15-LO-modified HDL3 failed to inhibit TNF-α-induced mRNA levels of adhesion molecules and MCP-1 (Fig. 7), confirming that the effects mediated by 15-LO-modified HDL3 are not related to the type of endothelial cell being studied.
Effect of modified HDL3 on U937 monocyte adhesion molecules to endothelial cells
The adhesiveness properties of endothelial cells incubated with 15-LO-modified HDL3 were investigated by monocyte adhesion assay. As expected, incubation of HUVEC with native HDL3 did not affect monocyte adhesion (Fig. 8,A); conversely, when cells were incubated with 15-LO-modified HDL3, a significant increase of monocyte adhesion was observed (Fig. 8 A).
Furthermore, the preincubation with native HDL3 significantly reduced monocyte adhesion to HUVEC in response to TNF-α; preincubation with 15-LO-modified HDL3 failed to inhibit TNF-α-induced monocyte adhesion (Fig. 8 B).
Effect of HDL modification on oxidative stress status in HUVEC
TNF-α stimulates the generation of reactive oxygen species (ROS), which represent a crucial signaling mechanism mediating adhesion molecules expression (40, 41). We addressed the question of whether native or 15-LO-modified HDL3 could affect ROS generation in TNF-α-stimulated HUVEC. Native HDL3 significantly inhibited TNF-α-induced oxidative stress, whereas 15-LO-modified HDL3 did not efficiently contrast ROS formation induced by TNF-α (Fig. 9,A). In addition, 15-LO-modified HDL3 reduced the oxidative stress to a lesser extent than native HDL3 (Fig. 9 B).
The 15-LO-modified HDL3 fails to inhibit phosphorylation of IκB-α and c-Jun induced by TNF-α
ROS can regulate gene and protein expression in endothelial cells by modulating redox-sensitive transcription factors, such as NF-κB and AP-1. Because activation of NF-κB and AP-1 is essential for the up-regulation of adhesion molecules in cytokine-activated endothelial cells (41, 42), we investigated the effect of native or 15-LO-modified HDL3 on IκB-α and c-Jun phosphorylation induced by TNF-α. HDL3 inhibited both IκB-α (Fig. 10,A) and c-Jun phosphorylation (Fig. 10 B), thereby preventing NF-κB and AP-1 activation and, consequently, NF-κB- and AP-1-mediated inflammatory gene transcription; 15-LO-modified HDL3 had no inhibitory effects.
Role of NF-κB and AP-1 in the expression of adhesion molecules induced by 15-LO-modified HDL3
To study the involvement of NF-κB and AP-1 in the modulation of adhesion molecules and MCP-1 mRNA levels by 15-LO-modified HDL3, cells were preincubated with BAY11-7085 (5 μM) or SP600125 (25 μM) before addition of modified lipoprotein. Both inhibitors decreased the mRNA levels of adhesion molecules and MCP-1 in cells incubated with 15-LO-modified HDL3 (Fig. 11). No effects were observed in cells incubated with the inhibitors alone.
Effect of HDL3 modified with h15-LO on U937 monocyte adhesion molecules to endothelial cells
Finally, we evaluated the effect of HDL3 modified with the human form of 15-LO on HUVEC. Modification of HDL3 with rabbit reticulocyte 15-LO or with h15-LO induced a similar degree of apoprotein cross-linking (Fig. 12 A).
As shown for r15-LO-modified HDL3, h15-LO-modified HDL3 were unable to inhibit surface expression of ICAM-1, VCAM-1, and E-selectin in HUVEC stimulated with TNF-α (Fig. 12,B). Accordingly with adhesion molecule expression data, the preincubation with h15-LO-modified HDL3 failed to inhibit TNF-α-induced monocyte adhesion to HUVEC (Fig. 12,C); moreover, h15-LO-modified HDL3 significantly increased monocyte adhesion to endothelial cells (Fig. 12 D).
Endothelial dysfunction contributes to atherogenesis (4, 5); the activation of vascular endothelial cells by a variety of proinflammatory stimuli triggers, in fact, the expression of adhesion molecules and chemoattractant proteins (2, 7, 8), promoting adhesion and transmigration of blood leukocytes. Much evidence supports the concept of a positive effect of HDL on endothelial function through the modulation of different endothelial activities (43). One of the earliest events in the pathogenesis of atherosclerosis is the massive recruitment of inflammatory cells through endothelial-dependent mechanisms: HDL can efficiently contrast this process by decreasing cytokine-induced expression of adhesion molecules (44, 45).
A growing body of experimental evidence suggests that some HDL is dysfunctional or proinflammatory, facilitating leukocyte recruitment and cellular activation (22, 46, 47). The protective properties of HDL can be, in fact, compromised under specific conditions such as dyslipidemia or during inflammation (48, 49, 50, 51, 52), generating HDL with a reduced ability to promote reverse cholesterol transport and impaired antioxidative functions (53, 54, 55, 56). Several mechanisms can modulate HDL properties from anti-inflammatory to proinflammatory. Glycation, a process that appears relevant in diabetes, can impair HDL function (57, 58); myeloperoxidase, an enzyme present in human atherosclerotic lesions, can modify the apoprotein moiety of HDL, reducing its efficiency in reverse cholesterol transport as well endothelial nitric oxide synthase expression and activity (23, 59). We previously reported that 15-LO, an enzyme overexpressed in atherosclerotic lesions and involved in the modification of LDL in the intima, reduces HDL3 efficiency to act as cholesterol acceptor (26). HDL3 modification with 15-LO, in fact, decreases the lipoprotein interaction with scavenger receptor class B type I (SR-BI), thus reducing cholesterol efflux process (26). In the present work, we demonstrate that modification with 15-LO significantly impairs the ability of HDL3 to contrast the effect of a proinflammatory stimulus on endothelial cells. In fact, 15-LO-modified HDL3 fail to inhibit the expression of TNF-α-induced adhesion molecules and MCP-1 in endothelial cells at both transcriptional and protein levels. Furthermore, 15-LO-modified HDL3 induced adhesion molecules and MCP-1 expression in unstimulated cells and significantly increased the adhesion of monocytes to endothelial cells, suggesting that the up-regulation of adhesion molecule levels really affects monocyte binding to 15-LO-HDL3-treated HUVEC. These findings suggest that modification with 15-LO generates a dysfunctional HDL3 that, besides its failure in preventing the effect of TNF-α, activate endothelial cells also in absence of other proinflammatory stimuli.
Enhanced oxidative stress is important in vascular dysfunction, because an increased production of ROS can promote leukocyte adhesion to the endothelium by modulating the activity of redox-sensitive transcription factors, such as NF-κB and AP-1 (60). The activation of NF-κB following exposure to extracellular proinflammatory stimuli results in the expression of several genes such as adhesion molecules in many cell types. HDL3 reduce adhesion molecule expression in TNF-α-treated cells in part by reducing NF-κB activation, as demonstrated by other studies (20, 61, 62) and confirmed in the present work. Conversely, 15-LO-mediated modification significantly reduced the ability of HDL3 to inhibit ROS formation and NF-κB activation in TNF-α-stimulated cells, suggesting a loss of anti-inflammatory activity in 15-LO-modified HDL3.
ICAM-1 and VCAM-1 genes contain other transcription factor-binding elements in their promoters; among them, AP-1 appears to play a relevant role (20, 63). As expected (20), HDL3 decreased TNF-α-induced phosphorylation of c-Jun, a major component of the AP-1 transcription factor complex, whereas after enzymatic modification, this inhibitory activity was significantly decreased. These results indicate that although native HDL3 can reduce the proinflammatory effects of TNF-α by inhibiting the activation of transcription factors involved in the regulation of adhesion molecule expression, 15-LO-modified HDL3 resulted unable to achieve such effect. Moreover, the inhibition of cell adhesion molecules and MCP-1 by NF-κB and AP-1 inhibitors in cells incubated with 15-LO-modified HDL3 suggests a proinflammatory activity of enzymatically modified lipoproteins.
Sphingosine-1-phosphate (S1P) is a lysosphingolipid present in the human plasma, mainly associated to HDL particles (64); a number of recent studies suggest that the S1P content of HDL can account for several antiatherogenic properties of this lipoprotein. For example, HDL-associated S1P decreases NF-κB activation and inhibits endothelial cell expression of VCAM-1 and ICAM-1 in response to TNF-α (61, 65). Also, LDL contain a minor fraction of S1P that is significantly decreased during LDL oxidation (66), whereas cytotoxic compounds such as lysophosphatidylcholine (LPC) and lysophosphatidic acid (LPA) accumulate. Both LPC and LPA stimulate adhesion molecule expression and monocyte binding to endothelial cell surface (67, 68). We can speculate that this might occur also during 15-LO-mediated oxidation of HDL3: S1P content might decrease, and proatherogenic substances such as LPC and LPA increase, so that the altered balance between protective and proatherogenic lipids might shift HDL3 to a dysfunctional lipoprotein with proinflammatory properties. The component(s) of 15-LO-modified HDL3 responsible for the proinflammatory effects observed in the present study remains to be identified. Understanding the role of these bioactive lipids in 15-LO-modified HDL3 also could help the comprehension of the cellular receptors involved in the observed effects.
SR-BI plays a central role in the inhibitory effect of HDL on cytokine-induced adhesion molecule expression (65). We have previously shown that 15-LO-mediated HDL3 modification significantly reduces the interaction of lipoprotein with SR-BI (26); similar results were observed in HUVEC (data not shown), suggesting a diminished association between endothelial cells and 15-LO-modified HDL3 through this receptor. This finding could explain in part the decreased inhibitory effect of 15-LO-modified HDL3 on TNF-α-stimulated cells. However, this cannot explain the proinflammatory effect of 15-LO-modified HDL3, which, although less able to interact with cells via SR-BI, induce a higher oxidative stress and the expression of adhesion molecules both at mRNA and at protein levels. These findings suggest that other cellular receptors might be involved in mediating the effects of 15-LO-modified HDL3, such as LPA receptors or scavenger receptors present on endothelial cells capable of interacting with the modified HDL3 particle.
In conclusion, we demonstrate that in vitro modification of HDL3 with an enzyme overexpressed in human atherosclerotic lesions generates a defective lipoprotein that, besides having a reduced ability to promote cholesterol efflux from macrophages (26), exhibits together a decreased anti-inflammatory activity and the acquisition of proinflammatory features, supporting the assumption that qualitative characteristics of HDL can be as relevant as HDL cholesterol levels. The exact mechanism by which 15-LO-modified HDL3 exerts its proinflammatory effect remains to be elucidated.
We thank Dr. G. D. Norata for primer design.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by grants from Università degli Studi di Milano (Fondo per gli Investimenti e la Ricerca Scientifica e Tecnologica) and Istituto Nazionale per le Ricerche Cardiovascolari.
Abbreviations used in this paper: HDL, high density lipoprotein; 15-LO, 15-lipoxygenase; BCECF-AM, 2,7-bis(2-carboxyethyl)-5(6)-carboxyfluorescein acetoxymethylester; DCFH-DA, 2,7-dichlorofluorescein diacetate; h15-LO, human 15-LO; HAEC, human aortic endothelial cell; HMEC, human microvascular endothelial cell; LDL, low density lipoprotein; LPA, lysophosphatidic acid; LPC, lysophosphatidylcholine; ROS, reactive oxygen species; S1P, sphingosine-1-phosphate; SR-BI, scavenger receptor class B type I.