Delphinidin, an active compound of red wine, inhibits endothelial cell apoptosis via nitric oxide pathway and regulation of calcium homeostasis
Abstract
-
Epidemiological studies have suggested that moderate consumption of natural dietary polyphenolic compounds might reduce the risk of cardiovascular disease and also protect against cancer. The present study investigates the effects of delphinidin, an anthocyanin present in red wine, on bovine aortic endothelial cells apoptosis.
-
Based on flow cytometry, terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling analysis and detection of mitochondrial cytochrome c release, we show that delphinidin (10−2 g l−1) alone had no effect either on necrosis or on apoptosis, but it significantly reduced apoptosis elicited by actinomycin D (1 μg ml−1, 24 h) and 7β-hydroxycholesterol (10 μg ml−1, 18 h).
-
The protective effect of delphinidin was abolished by inhibitors of nitric oxide-synthase (NOS) (L-NA, 100 μM and SMT, 100 μM), guanylyl cyclase (ODQ, 100 μM) and MAP kinase (PD98059, 30 μM).
-
Western blot analysis and protein detection by confocal microscopy demonstrate that the antiapoptotic effect of delphinidin was associated with an increased endothelial NOS expression mediated by a MAP kinase pathway.
-
Finally, delphinidin alone had no effect on cytosolic-free calcium ([Ca2+]i), but normalized the changes in [Ca2+]i produced by actinomycin D towards the control values, suggesting that the antiapoptotic effect of delphinidin is associated with the maintenance of [Ca2+]i in the physiological range.
-
All of the observed effects of delphinidin may preserve endothelium integrity, the alteration of which lead to pathologies including cardiovascular diseases, such as atherosclerosis, and is often associated with cancers. In conclusion, the protective effect of delphinidin against endothelial cell apoptosis contributes to understand the potential benefits of a consumption rich in polyphenols.
British Journal of Pharmacology (2003) 139, 1095–1102. doi:10.1038/sj.bjp.0705347
Abbreviations:
-
- BAECs
-
- bovine aortic endothelial cells
-
- [Ca2+]i
-
- cytosolic free calcium
-
- eNOS
-
- endothelial NO-synthase
-
- iNOS
-
- inducible NO-synthase
-
- L-NA
-
- Nω-nitro-L-arginine
-
- NO
-
- nitric oxide
-
- NOS
-
- NO-synthase
-
- ODQ
-
- 1H-[1,2,4] oxadiazolo [4,3a] quinoxalin-1]
-
- RWPC
-
- red wine polyphenolic compounds
-
- SMT
-
- S-methyl-isothiourea
-
- TUNEL
-
- terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling
Introduction
Natural dietary polyphenolic compounds present in fruits, vegetables or red wine have been reported to protect against cardiovascular diseases and cancer because of their multitude biological activities (Hertog et al., 1993; Middleton et al., 2000). Indeed, polyphenolic compounds have been reported to suppress cell proliferation, prevent platelet aggregation, stabilize immune cells and promote vascular smooth muscle cell relaxation (Middleton et al., 2000). The mechanisms by which they act have not been totally elucidated. Polyphenols have also been reported to scavenge reactive oxygen species, inhibit oxidation of low-density lipoproteins and interact with the nitric oxide (NO) generating pathway (Middleton et al., 2000). Previous studies from our laboratory have identified the anthocyanin, delphinidin, as a polyphenol presenting an endothelium-dependent relaxant effect. Its action depends on the ability of delphinidin to stimulate NO production, independently of its antioxidant property (Andriambeloson et al., 1997). NO has vasorelaxant (Furchgott et al., 1984) and antiaggregatory properties and is able to limit the flux of atherogenic plasma proteins into the artery wall (Roberts et al., 1997), all of these activities probably contributing to the protective effects of delphinidin and other red wine polyphenolic compounds (RWPC).
Apoptosis, a genetically controlled cell death program, is of fundamental significance for the maintenance of homeostasis in multicellular organisms. Under certain vascular pathological conditions, such as atherosclerosis, hypertension and restenosis (Kockx & Knaapen, 2000), activation of cellular pathways leading to endothelial cell apoptosis has been described, and can be either beneficial or detrimental. The main features of apoptotic cells are DNA fragmentation, chromatin condensation and cell fragmentation without increased permeability of the plasma membrane. NO exhibits a double-edged role in apoptosis, depending on its concentration. High amounts of NO produced by the inducible NO-synthase (iNOS) appear to be linked to its proapoptotic effect (Terenzi et al., 1995). In contrast, the antiapoptotic effects of NO are mainly mediated by low amounts of NO following the stimulation of the constitutive endothelial NO-synthase (eNOS) (Mannick et al., 1994). Studies on the antiapoptotic mechanisms of NO have identified multiple targets that range from direct and indirect interactions with the apoptotic machinery. These include the induction of cytoprotective stress proteins, cGMP-dependent inhibition of apoptotic signal transduction, including modification of Ca2+ homeostasis, suppression of caspase activity and inhibition of cytochrome c release (Kim et al., 1999).
The endothelium is recognized as having a central role in many cardiovascular diseases, and apoptosis is emerging as a determinant process in disease progression. Here, we report the protective effect of delphinidin on endothelial cell apoptosis through NO-guanylyl cyclase-dependent pathway. The maintenance of Ca2+ homeostasis in endothelial cells may account for the protective effects of delphinidin.
Methods
Materials
The anthocyanin delphinidin chloride was purchased from Extrasynthèse (Genay, France). Delphinidin chloride dry powder was 99.93% pure, according to the manufacturer. Delphinidin was diluted in 10% DMSO/90% H2O. After appropriate dilutions, the final concentration of DMSO did not exceed 0.1%. Actinomycin D, 7β-hydroxycholesterol and propidium iodide (PI) were purchased from Sigma (St Louis, MO, U.S.A.). Anti-eNOS and anti-iNOS were from Transduction Laboratories (Lexington, KY, U.S.A.). Anti-cytochrome c antibody was from BD Biosciences Pharmingen (San Diego, CA, U.S.A.). Anti-mouse HRP-conjugated was from Promega (Madison, WI, U.S.A.). Alexa Fluor™ 488 was purchased from Molecular Probes (Leiden, Netherlands), PD98059 from Tocris (Ballwin, MO, U.S.A.) and TUNEL assay kit from Roche Diagnostics (Mannheim, Germany).
Cell culture and treatment
Bovine aortic endothelial cells (BAECs) were grown as previously described (Schini et al., 1988). Briefly, cells were cultured in plastic flasks precoated with type I collagen (0.06 mg ml−1) in a mixture of DMEM and Ham's F12 (1 : 1,v v−1) with 10% fetal bovine serum, 2 mM L-glutamine, 100 mg ml−1 heparin, 100 U ml−1 penicillin and streptomycin and 10 μM vitamin C. Cultures were maintained at 37°C in a humidified incubator gassed with 5% CO2 in air. Cells were used up to five passages. For apoptosis induction, Ca2+ experiments and Western blotting, cells were seeded at a suitable density (5 × 105 cells per well in six-well plate, and 6 × 106 cells per 100-mm Petri dishes, respectively) for 4 h and apoptosis was then induced in subconfluent and exponentially growing BAECs. For confocal microscopy, cells were subcultured as described previously at a density of 5 × 105 cells per 35-mm Petri dishes in which a 2 cm diameter hole had been cut in the base and replaced by a thin (0.07 mm) glass coverslip (Martin et al., 2002).
Apoptosis was induced by the addition of either actinomycin D (1 μg ml−1) for 24 h or 7β-hydroxycholesterol (10 μg ml−1) for 18 h. Actinomycin D has been used experimentally in models of cell death (Bock et al., 2002; Kim et al., 2002). 7β-hydroxycholesterol is a pathogenic endogenous agent, known to be responsible at least in part for cytotoxicity of oxidized low-density lipoproteins and induces apoptosis of several cell types (Aupeix et al., 1996). Delphinidin was added to cells at the same time as actinomycin D or 7β-hydroxycholesterol. Delphinidin was used at 10−2 g l−1, a concentration eliciting maximal endothelium-dependent relaxation of precontracted rat aorta (Andriambeloson et al., 1998), a transient elevation of [Ca2+]i in endothelial cells leading to NO production (Martin et al., 2002) and the inhibition of serum- and vascular endothelium growth factor-induced proliferation in endothelial cells (Martin et al., 2003). Nω-nitro-L-arginine (L-NA, 100 μM), S-methyl-isothiourea (SMT, 100 μM), 1H-[1,2,4] oxadiazolo [4,3a] quinoxalin-1] (ODQ, 100 μM) and PD98059 (30 μM), which are respectively inhibitors of NOS, guanylyl cyclase and MEK1/2, were also present during the induction of apoptosis when indicated.
Determination of DNA strand breaks
DNA single-strand breaks occurring during apoptosis were analyzed by using the TUNEL assay according to the manufacturer's instructions. Briefly, cell concentration was adjusted to 30 × 106 cells ml−1 and fixed in 4% paraformaldehyde, permeabilized with 0.2% Triton X-100 in phosphate-buffered saline (PBS) and incubated with TUNEL reaction mixture. Breaks could be identified by labeling free 3′-OH termini with fluorescein-labeled nucleotides in a reaction catalyzed by terminal deoxynucleotidyl transferase (TdT). Incorporated fluorescein was analyzed with a Bio-Rad 1024 MRC confocal microscope with a 40 × epifluorescence objective (Nikon). Z-series were collected in 1 μm steps and final images were analyzed after stacking of 22 images corresponding to the total cell volume.
Hypodiploid DNA quantification was performed as described previously (Gidon-Jeangirard et al., 1999). Briefly, cell concentration was adjusted to 5 × 105 cells ml−1 and fixed in 70% ethanol for at least 1 h at 4°C. Cells were washed in Hank's balanced salt solution before resuspension in a solution containing type I RNAse A (0.5 mg ml−1) and were incubated for 10 min at 37°C. PI (0.1 mg ml−1) was then added and samples were allowed to stand 15 min in the dark at room temperature before flow cytometry analysis using CELLQuest software (Becton Dickinson, San Jose, CA, U.S.A.).
Immunostaining of eNOS and iNOS by confocal microscopy
After incubation with actinomycin D, adherent cells were fixed for 10 min with PBS containing 3% paraformaldehyde on ice. After three washes with PBS, cells were treated 30 min with 0.2% Triton X-100 in PBS and blocked 30 min with PBS/3% BSA. Cells were washed three times with PBS/0.2% BSA and then incubated with either anti-eNOS or anti-iNOS antibodies, diluted 1/200 (v v−1) for 1 h at room temperature. After three washes, cells were incubated with the Alexa Fluor™ 488-conjugated goat anti-mouse secondary antibody, diluted 1/200 (v v−1) for 1 h. Cells were observed with a Bio-Rad 1024 MRC confocal microscope as previously described (see TUNEL protocol). The amount of labeling was quantified with NIH-image software.
Western blotting
Cells were harvested and spun down at 1000 × g for 10 min, and were lysed for 1 h in 200 μl of ice-cold lysis buffer (50 mM Tris, 250 mM NaCl, 8 mM MgCl2, 10 μg ml−1 of aprotinin, leupeptin and pepstatin, 1 mM PMSF, 5 mM EDTA, 0.5 mM EGTA, 2 mM NaVO3 and 1% Triton X-100). After quantification by the Bradford method, 50 μg protein was resuspended in Laemmli's buffer, separated on 7% SDS–PAGE and Western blotted. The sample loading was verified by staining membranes with Ponceau red and amidoblack. Blots were probed with anti-eNOS, anti-iNOS and anti-cytochrome c antibodies, followed by the HRP-conjugated antibody. Blots were treated with ECL reagents for 5 min, exposed to CL-Xposure films. Blots were scanned and densitometric analysis was performed of the scanning images using Scion Image-Release Beta 4.02 software (http://www.scioncorp.com).
Determination of cytosolic calcium concentration ([Ca2+]i)
Cells were loaded with Fura-2/AM (5 μM) (45 min at 37°C in culture medium), then washed and placed in physiological salt solution (PSS in mM: 119 NaCl, 4.75 KCl, 1.25 CaCl2, 1.17 MgSO4, 1.18 KH2PO4, 25 NaHCO3, 11 glucose, 20 HEPES, pH 7.4, 37°C). [Ca2+]i was determined by fluorimetric readings performed with F-2000 Hitachi spectrofluorimeter system. The maximum and minimum fluorescence were sequentially determined by the addition of 10 μM ionomycin, followed by the addition of 10 mM EGTA. Basal [Ca2+]i of control cells was 216±19 nM.
Statistical analysis
Data are represented as mean±s.e.m. In all cases, n refers to the number of experiments. Statistical analyses were performed by Student's t-test and analysis of variance (ANOVA). Differences were considered significant at a value of P<0.05.
Results
Actinomycin D and 7β-hydroxycholesterol induced apoptosis of BAECs
Staining with PI revealed nuclei with hypodiploid DNA (sub-G1 peak) corresponding to apoptotic cells. Basal apoptosis, that is, in the presence of 0.1% DMSO, never exceed 5%. Exposure of BAECs to actinomycin D, a chemotherapeutic agent, induced DNA fragmentation in a time-dependent manner with a maximal value of 19±1% reached after 24 h of treatment (Figure 1a, n=30). At 18 h, the major part of the cell population was still attached to the culture flask. At 24 h, cells exhibited morphological characteristics of apoptosis including shrinkage, membrane blebbing and detached (Figure 1b). Cells were also treated with the pathogenic agent, known to be responsible, at least in part, for the cytotoxicity of oxidized low-density lipoproteins, 7β-hydroxycholesterol. Treatment of BAECs for 18 h with 10 μg ml−1 7β-hydroxycholesterol induced 14±2% apoptosis (see Figure 2a). Longer period of incubation time enhanced cell mortality by necrosis determined by trypan blue exclusion, making measurement of the degree of apoptosis almost impossible (not shown).

Actinomycin D induces apoptosis of endothelial cells. (a) BAECs were exposed to actinomycin D (1 μg ml−1) for the time indicated, and cell death was assessed by PI staining using flow cytometry. Values showing the apoptosis induced by actinomycin D (apoptosis in the presence of actinomycin D minus basal apoptosis) are mean±s.e.m. (n=30). (b) Morphological changes observed by light microscopy of actinomycin D-treated cells (right) versus control cells (left). Arrows on the right-hand panel point to cell shrinkage (1) and membrane blebbing (2), characteristic features of apoptotic cell death.

Effect of delphinidin on actinomycin D-and 7β-hydroxycholesterol-induced apoptosis in endothelial cells. (a) Histograms showing the percentage of hypodiploid DNA of cells after 24 h treatment with either DMSO, actinomycin D (1 μg ml−1, Act D) alone (n=30), 7β-hydroxycholesterol (10 μg ml−1, 7β-OH Chol) alone (n=11) or in combination with delphinidin (10−2 g l−1, Del). Values represent the percentage of hypodiploid population for each treatment or control cells, (b) Confocal imaging of TUNEL assay from control (1), and cells treated with actinomycin D (2), delphinidin (3) and actinomycin D+delphinidin (4). Apoptotic cells are detected by green fluorescence caused by incorporation of fluorescein-dUTP. (c) Western blot revealing cytochrome c release from mitochondria in BAECs exposed to either control medium (lane 1), actinomycin D (1 μg ml−1, Act D) (lane 2), delphinidin (10−2 g l−1, Del) (lane 3) or the combination of actinomycin D and delphinidin (lane 4).
Delphinidin prevented actinomycin D- and 7β-hydroxycholesterol-induced apoptosis
The effect of delphinidin on the apoptotic process was tested on actinomycin D and 7β-hydroxycholesterol-treated cells. Delphinidin significantly reduced apoptosis induced by actinomycin D in a time-dependent manner with a maximal effect obtained at 24 h. At this time, delphinidin inhibited 30±3% the apoptosis induced by actinomycin D (Figure 2a, n=30). Delphinidin also significantly reduced 7β-hydroxycholesterol-induced apoptosis by 45±6% (Figure 2a, n=11). The protective effect of delphinidin against actinomycin D-induced apoptosis was also confirmed by TUNEL assay. As observed in Figure 2b, delphinidin treatment decreased the labeling in the TUNEL assay. We also investigated the effect of delphinidin on the apoptotic process regulated at the level of the mitochondria by quantifying the cytosolic level of cytochrome c by Western blot analysis. In fact, the release of the cytochrome c from the mitochondria into the cytosol is considered as a marker of apoptosis (Hengartner, 2000). As shown in Figure 2c, actinomycin D treatment increases the cytochrome c release from the mitochondria, and this increase was reduced in the presence of delphinidin. These results indicate that the antiapoptotic effect of delphinidin is associated with inhibition of the intrinsec apoptotic pathway. Delphinidin alone did not induce apoptosis (i.e. in the absence of apoptogenic agents, results not shown).
Antiapoptotic effect of delphinidin is mediated through the NO and guanylyl cyclase pathway
To investigate the implication of NO pathway in delphinidin-induced inhibition of apoptosis, the effect of the nonselective NOS inhibitor, L-NA was tested. In Figure 3, this inhibitor reduced significantly the apoptosis evoked by actinomycin D, probably due to a cross-reaction between both actinomycin D and L-NA. Moreover, in the presence of L-NA, delphinidin was not able to inhibit actinomycin D or 7β-hydroxycholesterol-induced apoptosis (Figure 3a, b, n=10). These results show that delphinidin acts through the NO pathway to protect cells against apoptosis independently of the inducer. Another nonspecific inhibitor of NOS, SMT (Boer et al., 2000), had no effect by itself on actinomycin D-induced apoptosis. In the presence of 100 μM SMT, the antiapoptotic effect of delphinidn was abolished.

Involvement of the NO/cGMP pathway in the antiapoptotic effect of delphinidin in endothelial cells. BAECs were exposed to either actinomycin D (1 μg ml−1, Act D) or 7β-hydroxycholesterol (10 μg ml−1, 7β-OH Chol) in the absence or presence of delphinidin (Del, 10−2 g l−1) alone or in combination with L-NA (100 μM), SMT (100 μM) or ODQ (100 μM). Histograms show the percentage of hypodiploid DNA of cells. Values are the mean±s.e.m. of 4–10 experiments. *P<0.05, **P<0.01.
To decipher the mechanism(s) by which NO mediated the antiapoptotic effect of delphinidin, actinomycin D-treated cells were incubated with the soluble guanylyl cyclase inhibitor ODQ. Under these conditions, ODQ abolished the inhibitory effect of delphinidin (Figure 3d, n=4). The same results were obtained in 7β-hydroxycholesterol-induced apoptosis (not shown). By themselves, L-NA, SMT and ODQ had no effect on cell viability measured by trypan blue exclusion (not shown). Altogether, these results suggest that the antiapoptotic effect of delphinidin is associated with NO production.
Antiapoptotic effect of delphinidin is related to eNOS overexpression
To get insight into the mechanism involved in NO-induced protection, we investigated the level of eNOS and iNOS expression. Immunostaining of eNOS showed that the protein level was significantly increased in the presence of delphinidin alone (Figure 4a). However, immunostaining of eNOS by confocal microscopy cannot be easily visualized after actinomycin D treatment because a high proportion of cells was detached from the matrix after treatment with actinomycin D. In order to avoid this problem, Western blot analysis was performed. eNOS expression measured by Western blot analysis was dramatically reduced after treatment with actinomycin D (Figure 4b, n=4). Delphinidin increased the level of eNOS expression, compensating the reduction induced by actinomycin D, such that eNOS expression was not significantly different from that of control. No iNOS expression was observed under any conditions (not shown).

Quantification of eNOS expression (a) Confocal imaging of eNOS expression in adherent BAECs exposed to either control medium or delphinidin (Del) for 24 h. (b) Western blot revealing eNOS expression in BAECs exposed to either control medium (lane 1), actinomycin D (1 μg ml−1, Act D) (lane 2), delphinidin (10−2 g l−1, Del) (lane 3) or the combination of actinomycin D and delphinidin (lane 4). Histograms show densitometric analysis of eNOS expression. Data represent the mean±s.e.m. of four separate experiments. *P<0.05.
Antiapoptotic effect of delphinidin involves the MEK1/2 pathway
The MEK1/2 pathway has been shown to play an important role as a prosurvival element (Xia et al., 1995). In order to determine whether this pathway is involved in delphinidin-induced inhibition of apoptosis, cells were treated with the specific inhibitor of MEK1/2, PD98059. By itself, PD98059 had no effect on actinomycin D-induced apoptosis, but this inhibitor abolished the antiapoptotic effect of delphinidin (Figure 5a, n=6). The involvement of MEK1/2 pathway in the regulation of eNOS expression was also studied. As shown in Figure 5b, treatment of BAECs by PD98059 significantly altered the overexpression of eNOS induced by delphinidin. It should be noted that PD98059 by itself had no effect on eNOS expression.

Involvement of MEK1/2 pathway in the antiapoptotic effect of delphinidin in endothelial cells (a) BAECs were exposed to actinomycin D (1 μg ml−1, Act D) alone or in combination with delphinidin (10−2 g l−1, Del). Involvement of MEK1/2 pathway was investigated with a MEK1/2 inhibitor, PD98059 (30 μM). Histograms show the percentage of hypodiploid DNA of cells. Values are the mean±s.e.m. of six experiments. *P<0.05. (b) Western blot revealing eNOS expression in BAECs exposed to either control medium, actinomycin D (1 μg ml−1, Act D), delphinidin (10−2 g l−1, Del), PD 98059 (30 μM) or the combination of actinomycin D, delphinidin and PD98059. Histograms show densitometric analysis of eNOS expression. Data represent the mean±s.e.m. of four separate experiments. *P<0.05.
Antiapoptotic effect of delphinidin requires the maintenance of [Ca2+]i in the physiological range
Basal [Ca2+]i values (t=0 h) were not significantly different throughout the experimental condition used (Figure 6, n=5). In control cells (i.e. in the presence of DMSO), the changes in [Ca2+]i were not significantly different at any time tested. The effect of actinomycin D treatment on [Ca2+]i was biphasic. Indeed, [Ca2+]i was significantly decreased during the early phase of apoptosis (12 h), it was significantly greater at 18 h when apoptosis was near to maximum (see Figure 1), and then it returned to control values after 24 h. Delphinidin alone had no effect on [Ca2+]i at any time tested, but normalized the changes in [Ca2+]i produced by actinomycin D towards the control values. Thus, delphinidin significantly attenuated the decrease in [Ca2+]i induced by actinomycin D at 12 h, and it completely prevented the increase in [Ca2+]i observed at 18 h. No difference in Ca2+ homeostasis was observed at 6 and 24 h at any experimental conditions used.

Involvement of calcium variations in endothelial cells in the antiapoptotic effect of delphinidin. BAECs were exposed to actinomycin D (1 μg ml−1, Act D), delphinidin (10−2 g l−1, Del) or in combination of actinomycin D plus delphinidin at indicated times (i.e. 6, 12, 18, and 24 h). [Ca2+]i values are the mean±s.e.m. of five experiments. *P<0.05, **P<0.01.
Discussion
In the present study, we demonstrate that delphinidin, an anthocyanin contained in natural diet, is able to protect endothelial cells against apoptosis. Of particular interest is the finding that the antiapoptotic effect of delphinidin results from increased eNOS expression via MEK1/2 inhibitor-sensitive pathway. The effect of delphinidin also involves the NO and guanylyl cyclase-dependent pathway and is associated with the maintenance of endothelial [Ca2+]i level in a physiological range and the decrease of cytochrome c release from the mitochondria.
The concentration of delphinidin (10−2 g l−1) used in the present study is identical to that able to promote endothelium-dependent relaxation of rat aortic rings (Andriambeloson et al., 1998), to stimulate the increase in [Ca2+]i in endothelial cells (Martin et al., 2002), to inhibit epidermal growth-factor receptor in human vulva carcinoma cell line A431 (Meiers et al., 2001) and to inhibit serum and vascular endothelial growth factor-induced proliferation in BAECs (Martin et al., 2003). Recent studies concerning bioavailability of polyphenols are in agreement with their potential therapeutic effects. Anthocyanins are present in fruits, beans, cereals, vegetables, or red wines, and are therefore ingested in considerable amounts as a constituent of the human diet (180–215 mg daily) (Kuhnau et al., 1976). According to the fact that the absorption rate of anthocyanins, such as delphinidin and cyanidin, was about 1% after oral administration, the concentration of delphinidin used in these study should be reached in vivo (i.e. 10 μg/ml) (Matsumoto et al., 2001). Also recently, we found that daily feedings of rats with RWPC for 1 week with 20 mg kg−1 or for 3 weeks with 40 mg kg−1 (a concentration 10–20 times greater than that producing maximal relaxation of rat aortic rings ex vivo) results in sufficient circulating concentration of polyphenols including delphinidin to improve cardiovascular functions (Diebolt et al., 2001; Bernatova et al., 2002). Altogether, this suggests that the concentration of delphinidin used in this study might be reached in plasma after dietary ingestion of polyphenols and may therefore be physiologically relevant.
A therapeutically relevant mechanism by which natural dietary polyphenolic compounds may exert their beneficial effects against cardiovascular diseases and the thrombotic tendency associated with cancer is the protection of endothelial cell against apoptosis. Endothelial injury, mainly by oxysterols, is commonly considered as playing a pivotal role in atherosclerosis both in the early stages of lesion formation, and later in the disease development process by inducing atherosclerotic plaque unstability. Thus, inhibition of endothelial apoptosis might prevent or reduce thrombosis. In the present study, it was observed that delphinidin protected endothelial cells against apoptosis. Interestingly, delphinidin treatment led to a significant reduction of both oxysterol and actinomycin D toxicity.
Our previous studies showed that NO participates strongly in the effects of delphinidin. In fact, polyphenols, including delphinidin, have been shown to induce endothelial NO production without any increase in the oxidative stress (Andriambeloson et al., 1998; Diebolt et al., 2001; Bernatova et al., 2002; Martin et al., 2002). Here, no iNOS expression could be detected in endothelial cells treated with actinomycin D alone or in combination with delphinidin. Thus, NO formed from iNOS expression cannot account for the protective effect of delphinidin even though overexpression of iNOS could be able to suppress lipopolysaccharide-induced endothelial cell apoptosis (Tzeng et al., 1997). In contrast, the nonselective NOS inhibitors, L-NA and SMT, prevented the effect of delphinidin on actinomycin D- or 7β-hydroxycholesterol-induced apoptosis. The protective effect of delphinidin in endothelial cells against actinomycin D-induced apoptosis was prevented by the guanylyl cyclase inhibitor, ODQ. Thus, it is likely that NO through the cGMP, resulting from guanylyl cyclase hydrolysis of guanosine triphosphate, plays a significant role in this process. In addition, whereas actinomycin D alone dramatically reduced eNOS protein level within endothelial cells, delphinidin restored eNOS expression at control level. Although actinomycin D is an inhibitor of mRNA synthesis, it was used at low concentrations in this study (i.e. 10–25 times lower as that effective to abolish gene transcription) (Inoue et al., 1995; Hisamoto et al., 2001), and this may explain the fact that delphinidin is still able to increase protein expression, even in its presence. This is consistent with the findings demonstrating that actinomycin D induces an upregulation of several protein expression (Roulston et al., 1998; Kleeff et al., 2000; Kim et al., 2002). Taken together, the results obtained in the present study strongly suggest that NO formed by eNOS isoform, whose expression is considerably enhanced by delphinidin, plays a significant role in protecting endothelial cell against apoptosis. These data are consistent with the hypothesis that levels of NO generated by eNOS can inhibit apoptosis (Dimmeler et al., 1997) and allow maintenance of endothelial cell survival, in contrast to that released from iNOS which is generally described to possess a proapoptotic effect. Our results also shed light on the capacity of delphinidin to reglulate eNOS expression in endothelial cells. The fact that delphinidin counteracts the reduction of eNOS expression in actinomycin D-treated cells suggests that delphinidin probably acts, at least in part, by increasing the transcriptional rate of this enzyme. Cieslik et al. (2001) have recently reported that lysophosphatidylcholine is able to upregulate eNOS via the MEK1/2-dependent pathway. In accordance with these observations, both the protective effect and the overexpression of eNOS induced by delphinidin were abolished in the presence of the MEK inhibitor, PD98059. Hence, these data are consistent with the hypothesis that delphinidin protects endothelial cells against apoptosis through the upregulation of eNOS expression at least via MEK inhibitor-sensitive pathway. The mechanisms by which endothelial NO, upon stimulation by delphinidin, mediates its antiapopotic effects were further investigated. One of the molecular mechanisms by which NO protects endothelial cells against apoptosis is the blocking of the intrinsic/mitochochondrial apoptotic pathway. In this sudy, we reported that delphinidin is able to inhibit the release of cytochrome c into the cytosol induced by actinomycin D. It was recently described that in isolated mitochondria, NO effectively inhibits permeability transition and subsequent cytochrome c release (Brooks et al., 2000). NO can also inhibit cytochrome c release indirectly through upregulation of Bcl-2 (Rossig et al., 2000). We can therefore hypothesize that blocking the release of cytochrome c from the mitochondria could be a mechanism by which NO mediates the antiapoptotic effect of delphinidin. Another mechanism by which NO-cGMP pathway protects endothelial cells against apoptosis is the maintenance of [Ca2+]i homeostasis. The increase of [Ca2+]i has been reported to be one of the key signals leading to apoptotic cell death (Thornberry & Lazebnik, 1998). In the present study, actinomycin D induced an initial reduction, followed by a later elevation of [Ca2+]i. Delphinidin alone did not modify [Ca2+]i values, but it prevented the effect of actinomycin D, so that, delphinidin restores [Ca2+]i in physiological range. The early effect might be explained by the capacity of delphinidin to stimulate an increase of [Ca2+]i in endothelial cells (Martin et al., 2002) which therefore counterbalances the reduction of [Ca2+]i produced by actinomycin D. Once, after delphinidin treatment, the [Ca2+]i level is high enough in the cells, it could stimulate the NO-guanylyl cyclase pathway via the activation of eNOS. In its turn, cGMP could decrease [Ca2+]i through either inhibition of Ca2+ entry, enhancement of Ca2+ extrusion, activation of Ca2+ sequestration into intracellular store or inhibition of inositol triphosphate generation (Kim et al., 1997). The reduction of [Ca2+]i by cGMP may account for the latter effect of delphinidin on the actinomycin D-induced increase in [Ca2+]i and apoptosis. The fact that delphinidin abolished the effects of actinomycin D treatment on [Ca2+]i changes, but only partially reduced actinomycin D-induced apoptosis suggests that other mechanisms might be involved in the delphinidin effect such as the reduction of cytochrome c release from the mitochondria. Nevertheless, the present work shows that delphinidin facilitates the maintenance of [Ca2+]i within the physiological range upon stimulation by actinomycin D, explaining, at least in part, its protective effect against endothelial apoptosis.
In conclusion, we have shown that, delphinidin exerts an antiapoptotic effect in endothelial cells through the NO-guanylyl cyclase pathway, probably by upregulating eNOS expression via MEK1/2 pathway. Its preventive effect against cell death is associated with a strict control of Ca2+ homeostasis and the inhibition of the release of cytochrome c from the mitochondria. Endothelial protection by delphinidin may contribute to the beneficial effects of natural dietary polyphenolic compounds, including those contained in red wine, against cardiovascular disease. This antiapoptotic potential of delphinidin may be of importance in atherosclerosis in which the balance between survival and apoptosis is a major determinant of atherosclerotic plaque development.
Acknowledgments
We are grateful to Dr Jean-Marie Freyssinet for helpful suggestions during the preparation of this manuscript. We also thank Dr Gillian Gray for the careful reading of the manuscript. Sophie Martin is in recipient of a Fellowship from ONIVIN. This study was in part supported by grants from the Société Française de Distilleries (00/903/14/026), the EEC (Fair Program CT97-3261-DG12-SSMI) and ONIVIN (00/903/14/013).