Number 22, 2004
Endothelial Dysfunction

Endothelial function and dysfunction

Back Back to the Summary

Paul M. Vanhoutte
Department of Pharmacology, University of Hong Kong
Correspondence: P.M. Vanhoutte, Department of Pharmacology, Faculty of Medicine, 21 Sassoon Road, Hong Kong. e-mail: vanhoutte.hku@hku.hk

Abstract

The endothelium mediates a number of responses (relaxations or contractions) of isolated arteries and veins from animals and humans. The endothelium-dependent relaxation results from the release by the endothelial cells of potent nonprostanoid vasodilator substances. Among these, the best characterized is endothelium-derived relaxing factor (EDRF), which most probably is nitric oxide. Nitric oxide is formed by the metabolism of L-arginine by the constitutive nitric oxide synthase of endothelial cells. In arterial smooth muscle, the relaxation evoked by nitric oxide is best explained by the stimulation by nitric oxide of soluble guanylate cyclase, which leads to the accumulation of cyclic 325-guanosine monophosphate. The endothelial cells also release prostacyclin and a substance that causes hyperpolarization of the cell membrane (endothelium-derived hyperpolarizing factor, EDHF). The release of relaxing factors can be initiated by circulating hormones (catecholamines, vasopressin, oxytocin, and estrogens). The release of EDRF from the endothelium can be mediated by both pertussis toxin-sensitive (2-adrenergic activation, serotonin, aggregating platelets, leukotrienes) and pertussis toxin-insensitive (adenosine diphosphate, bradykinin) G proteins. In blood vessels from animals with regenerated and reperfused endothelium, or atherosclerosis, or both, there is a selective loss of the pertussis toxin sensitive mechanisms of EDRF release that favors the occurrence of vasospasm, thrombosis, and cellular growth. Heart Metab. 2004;22:510.

Keywords: Nitric oxide, atherosclerosis, platelet aggregation, G proteins, endothelium-dependent relaxation

Introduction
In 1980, Furchgott and Zawadzki [1] demonstrated that endothelial cells have an obligatory role in the relaxation of isolated arteries in response to acetylcholine. This pivotal observation has revolutionized thinking about the local control of vasomotor function. The endothelium-dependent responses are caused by the release of several diffusible substances (endothelium-derived relaxing [EDRF] and contracting factors) from the endothelial cells. This review briefly summarizes the observations, obtained mainly in the author's laboratory, that have examined how the production of relaxing factors by endothelial cells underlies moment-to-moment changes in the tone of the surrounding vascular smooth muscle cells, and how a lack of this function by endothelial cells eventually initiates atherosclerosis and thus vascular disease. It updates similar, more exhaustive overviews [213].

Endothelium-derived relaxing factors

Endothelium-derived nitric oxide
The short-lived diffusible factor that underlies endothelium-dependent relaxation in response to acetylcholine [1] has been identified as nitric oxide. Endothelial nitric oxide is formed from the guanidine-nitrogen terminal of L-arginine by the action of endothelial constitutive nitric oxide synthase (nitric oxide synthase III, eNOS). The activation of eNOS depends on the intracellular concentration of calcium ions in the endothelial cells, and is Ca2+-calmodulin-dependent (Figure 1). The activity of the enzyme requires cofactors: in particular, reduced nicotinamide adenine dinucleotide phosphate, and 5,6,7,8-tetrahydrobiopterin. eNOS can be inhibited competitively by synthetic L-arginine analogs such as NG-monomethyl-L-arginine or NG-nitro-L-arginine, or by the endogenous inhibitor, asymmetric dimethyl arginine. Nitric oxide diffuses to the underlying smooth muscle cells and, in them, stimulates cytosolic soluble guanylate cylase, which accelerates the formation of cyclic 325-guanosine monophosphate (cyclic GMP). The cyclic nucleotide in turn inhibits the contractile process. Nitric oxide is the major contributor to endothelium-dependent relaxation in large arteries [115]. In the intact organism, both animal and human, the inhibitors of nitric oxide synthase cause vasoconstriction in most vascular beds and an increase in systemic arterial pressure, not only because they prevent the direct inhibitory action of nitric oxide on the vascular smooth muscle, but also because nitric oxide inhibits the production of renin and of endothelin 1 [16].


Figure 1. Role of the increase in cytosolic calcium concentration in the release of endothelium-derived relaxing factor(s). Endothelial receptor activation induces an influx of calcium into the cytoplasm of the endothelial cell; after interaction with calmodulin, this activates nitric oxide synthase (NOS) and cyclooxygenase, and leads to the release of endothelium-derived hyperpolarizing factor (EDHF). Nitric oxide (NO) causes relaxation by activating the formation of cyclic 325-guanosine monophosphate (cGMP) from guanosine triphosphate (GTP). EDHF causes hyperpolarization and relaxation by opening potassium (K+) channels. Prostacyclin (PGI2) causes relaxation by activating adenylate cyclase, which leads to the formation of cyclic adenosine monophosphate (cAMP). Any increase in cytosolic calcium (including that induced by the calcium ionophore, A23187) causes the release of relaxing factors. When agonists activate the endothelial cells, an increase in inositol phosphate may contribute to the increase in cytoplasmic Ca2+ by releasing it from the sarcoplasmic reticulum (SR). AA, arachidonic acid; L-Arg, L-arginine; P-450, cytochrome P-450; R, membrane receptor. (From Vanhoutte et al [42], with permission.)

Nitric oxide is also released in the lumen of the blood vessel. Because it is scavenged by the oxyhemoglobin of the blood, it does not fulfil a hormonal role. However, at the interface between the blood and the blood vessel wall, it inhibits the adhesion of platelets and white cells to the endothelium. It acts (in strong synergy with prostacyclin) to inhibit platelet aggregation [3,4,9,15]. It also inhibits the growth of the vascular smooth muscle cells and prevents the production of adhesion molecules [17] (Figure 2).


Figure 2. Postulated signal transduction processes in an endothelial cell. Activation of the cell causes the release of endothelium-derived relaxing factor nitric oxide (EDRF-NO), which has important protective effects in the vascular wall. , -adrenergic; B, bradykinin receptor; cAMP, cyclic AMP; ET, endothelin receptors; G, coupling proteins; 5-HT, serotonin (5-hydroxytryptamine) receptor; P, purinoceptor. (From Vanhoutte [11], with permission.)

The activity of eNOS can be upregulated acutely. For example, the shear forces exerted by the flowing blood on the endothelial cells are one of the main regulators of the local release of nitric oxide, a mechanism that explains flow dependent vasodilation. Several substances, whether circulating in the blood or produced by the blood vessel wall, can increase the release of nitric oxide through activation of specific receptors on the endothelial cell membrane (Figure 3). They include hormones (eg, estrogen, catecholamines, vasopressin), neurotransmitters (eg, substance P), autacoids (bradykinin, histamine), and products formed during platelet aggregation (serotonin, adenosine diphosphate [ADP] or blood coagulation (thrombin). The cell membrane receptors for these substances are coupled to the activation of eNOS by two different families of G proteins (Figure 2). Thus, in coronary arteries, 2-adrenergic receptors, serotonin receptors, and thrombin receptors are coupled to pertussis toxin-sensitive Gi proteins, whereas, in contrast, the receptors for ADP or bradykinin are not coupled to the production of nitric oxide by pertussis-toxin sensitive G proteins [18]. The activation of eNOS by bradykinin involves low molecular weight G proteins of the Rho family [19]. In coronary and cerebral arteries, aggregating platelets induce endothelium-dependent relaxation, and the presence of a healthy endothelium inhibits the constriction induced by the platelet products (thromboxane A2 and serotonin). Serotonin, acting on 5-HT1D serotonin receptors, plays the major part in this response, whereas ADP, activating P2y-purinoceptors, contributes little (Figure 2). The release of nitric oxide, both toward the underlying smooth muscle and at the interface with the blood, in response to thrombin and platelet-derived serotonin is pivotal for the protective role played by the healthy endothelium against the platelet attack (Figure 4) [3,8,13].


Figure 3. Some of the neurohumoral mediators that cause the release of endothelium-derived relaxing factors (EDRFs) through activation of specific endothelial receptors (encircled). , -adrenergic receptor; A, adrenaline (epinephrine); AA, arachidonic acid; Ach, acetylcholine; ADP, adenosine diphosphate; AVP, arginine vasopressin; B, kinin receptor; E, estrogen; ET, endothelin, endothelin-receptor; H, histaminergic receptor; 5-HT, serotonin (5-hydroxytryptamine), serotoninergic receptor; M, muscarinic receptor; NA, noradrenaline (norepinephrine); P, purinergic receptor; T, thrombin receptor; VP, vasopressinergic receptor. (From Vanhoutte [11], with permission.)


Figure 4. Interaction between platelet products, thrombin, and endothelium. If the endothelium is intact, several of the substances released from the platelets [in particular, the adenine nucleotides (ADP and ATP) and serotonin (5-hydroxytryptamine, 5-HT)] cause the release of endothelium-derived relaxing factor (EDRF) and prostacyclin (PGI2). The same is true for any thrombin formed. The released EDRF will relax the underlying vascular smooth muscle, opening up the blood vessel, and thus flushing the microaggregate away; it will also be released towards the lumen of the blood vessel to brake platelet adhesion to the endothelium and, synergistically with prostacyclin, inhibit platelet aggregation. In addition, monoamine oxidase (MAO) and other enzymes will break down the vasoconstrictor serotonin, limiting the amount of the monoamine that can diffuse toward the smooth muscle. Finally, the endothelium acts as a physical barrier that prevents the access to the smooth muscle of the vasoconstrictor platelet products serotonin and thromboxane A2 (TXA2). These different functions of the endothelium have a key role in preventing unwanted coagulation and vasospastic episodes in blood vessels with a normal intima. If the endothelial cells are removed (eg, by trauma), the protective role of the endothelium is lost locally, platelets can adhere and aggregate, and vasoconstriction follows; this contributes to the vascular phase of hemostasis. +, activation; , inhibition; NO, nitric oxide. (From Vanhoutte [11], with permission.)

Prostacyclin
Prostacyclin, formed primarily in endothelial cells, relaxes vascular smooth muscle by stimulation of adenylate cyclase, with a resulting increased production of cyclic 325-adenosine monophosphate (cyclic AMP). It acts synergistically with nitric oxide to inhibit platelet aggregation (Figure 4) [3,6,14].

Endothelium-dependent hyperpolarizing factor
In large and small arteries from different species (including the human), acetylcholine, and other endothelium-dependent vasodilators, cause endothelium-dependent hyperpolarization which can contribute to endothelium-dependent relaxation. The hyperpolarization has been attributed to a diffusible endothelium-derived hyperpolarizing factor (EDHF) different from nitric oxide and prostacyclin, although these last two can, in certain but not all blood vessels, cause hyperpolarization of vascular smooth muscle. The exact nature of EDHF remains a matter of intense debate. Among the more recent candidates to explain endothelium-dependent hyperpolarization, gap junctions, epoxyeicosatrienoic acids, potassium ions, and hydrogen peroxide appear to have major roles [2023] (Figure 1).
The contribution of hyperpolarization to endothelium-dependent relaxation varies as a function of the size of the blood vessel and thus is more pronounced in smaller than in larger arteries [23,24]. In the latter, although both mediators can contribute to endothelium-dependent relaxation, nitric oxide predominates under normal circumstances. However, in these large arteries, such as the coronaries, EDHF can maintain near normal endothelium-dependent relaxation when the synthesis of nitric oxide is dysfunctional [25]. In certain cases, nitric oxide exerts an inhibitory effect on endothelium-dependent hyperpolarization [26].

Chronic modulation
Chronic modulatory influences that can upregulate the release of relaxing factors by endothelial cells include chronic increases in blood flow, exercise training, estrogen administration, and intake of 3-unsaturated fatty acids, red wine polyphenols, green tea, and other antioxidants [2730].

Endothelial dysfunction
In the course of aging, and in several types of vascular disease and hypertension, the endothelial cells become dysfunctional [3,4,8,1013]. This dysfunction is evident as an impairment of endothelium-dependent relaxation, mainly as the result of a reduced release of EDRFs, in particular nitric oxide, although production of endothelium-derived vasoconstrictor substances may contribute [2,3,5,31,32].

Regenerated endothelium
The normal aging process induces a turnover (apoptotic death, desquamation followed by regeneration) of endothelial cells. Unfortunately, regenerated endothelial cells have lost part of the ability to release nitric oxide in response to platelet aggregation [33,34], because they respond minimally to serotonin and other substances using the Gi protein-dependent pathway controlling the release of nitric oxide (Figure 2); the Gi proteins are present, but exhibit a reduced activity [3539]. The loss of the pertussis toxin sensitive response is selective, and it does not apply, at least initially, to endothelium-dependent responses mediated by Gq-coupling proteins, in particular that to bradykinin [37,38]. It is caused by the greater accumulation of oxidized low density lipoproteins by the regenerated endothelial cells [40,41]. The reduced release of nitric oxide can be compensated in part by the larger contribution of EDHF to the endothelium-dependent relaxation [25].

Hypercholesterolemia and atherosclerosis
Hypercholesterolemia impairs endothelium-dependent relaxation [33,34]. In contrast, endothelium-independent relaxation in response to exogenous nitric oxide remains largely normal. In the initial phase of the atherosclerotic process, endothelial dysfunction is limited to the pertussis toxin sensitive, Gi protein-dependent pathway (Figure 2). Thus the ability of regenerated endothelial cells, chronically exposed to high cholesterol concentrations, to ADP-ribosylate pertussis toxin is reduced [39]. Hence, in coronary arteries from hypercholesterolemic pigs, endothelium-dependent relaxation in response to serotonin, 2-adrenergic agonists, aggregating platelets, or thrombin is depressed, whereas those induced by ADP and bradykinin are maintained [3339]. Oxidized low-density lipoprotein induces, in vitro, a similar selective endothelial dysfunction, whereas at higher concentrations it also inhibits endothelium-dependent relaxation in response to stimuli that are not Gi protein-dependent [40] (Figure 2).

Summary
The most important aspect of endothelial dysfunction is the reduced release or bioavailibility of nitric oxide, which probably is the fundamental, initial step of the atherosclerotic process. This hypothesis implies that aging and prolonged exposure to shear stress, coupled with risk factors such as obesity, diabetes, high blood pressure, and smoking, accelerate endothelial turnover and endothelial regeneration. Thus larger and larger sections of the endothelial lining (particularly in areas of turbulence) can no longer prevent platelet adhesion and aggregation, and become insensitive to thrombin. The negative feedback that nitric oxide, together with prostacyclin, exerts on platelet aggregation decreases steadily, whereas vasoconstrictor and growth-promoting substances (serotonin and thromboxane A2) are released in increasing amounts, together with growth factors such as platelet-derived growth factor. This sequence of events permits the local inflammatory response and initiates the characteristic morphological changes in atherosclerosis, in particular because the local shortage of nitric oxide unleashes the growth process [3,714].

Back Back to the Summary

REFERENCES

1. Furchgott RF, Zawadzki JV.
The obligatory role of the endothelial cells in relaxation of arterial smooth muscle by acetylcholine.
Nature. 1980;288:373376.
PMID: 6253831 [PubMed - indexed for MEDLINE]


2. Furchgott RF, Vanhoutte PM.
Endothelium-derived relaxing and contracting factors.
FASEB J. 1989;3:20072018.
PMID: 2545495 [PubMed - indexed for MEDLINE]


3. Lscher TF, Vanhoutte PM.
The Endothelium: Modulator of Cardiovascular Function. Boca Raton: CRC Press, Inc; 1990: 1228.

4. Vanhoutte PM.
The endothelium  modulator of vascular smooth-muscle tone.
N Engl J Med. 1988;319:512513.

5. Vanhoutte PM.
The other endothelium-derived vasoactive factors.
Circulation. 1993;87(suppl V):V9V17.

6. Vanhoutte PM, Boulanger CM.
Endothelium-dependent responses in hypertension.
Hypertens Res Clin Exp. 1995;18:8798.

7. Vanhoutte PM, Shimokawa H.
Endothelium-derived relaxing factor(s) and coronary vasospasm.
Circulation. 1989;80:19.
PMID: 2661051 [PubMed - indexed for MEDLINE]


8. Vanhoutte PM.
Hypercholesterolaemia, atherosclerosis and release of endothelium-derived relaxing factor by aggregating platelets.
Eur Heart J. 1991;12(suppl E):2532.
PMID: 1790781 [PubMed - indexed for MEDLINE]


9. Vanhoutte PM.
State of the art lecture. Endothelium and control of vascular function.
Hypertension. 1989;13:658667.
PMID: 2661425 [PubMed - indexed for MEDLINE]


10. Vanhoutte PM, Boulanger CM, Mombouli JV.
Endothelium-derived relaxing factors and converting enzyme inhibition.
Am J Cardiol. 1995;76:3E12E.
PMID: 7484885 [PubMed - indexed for MEDLINE]


11. Vanhoutte PM.
Endothelial dysfunction and vascular disease. In: Panza JA, Cannon III RO, eds.
Endothelium, nitric acid oxide and atherosclerosis. Armonk, NY: Futura Publishing Co, Inc; 1999: 7995.

12. Vanhoutte PM.
How to assess endothelial function in human blood vessels.
J Hypertens. 1999;17:10471058.
PMID: 10466457 [PubMed - indexed for MEDLINE]


13. Vanhoutte PM.
Ageing and endothelial dysfunction.
Eur Heart J. 2002;4:A8A17.

14. Vanhoutte PM.
Endothelial control of vasomotor function  from health to coronary disease.
Circulation. 2003;67:572575.

15. Moncada S, Palmer RMJ, Higgs EA.
Nitric oxide: physiology, pathophysiology, and pharmacology.
Pharmacol Rev. 1991;43:109142.
PMID: 1852778 [PubMed - indexed for MEDLINE]


16. Vanhoutte PM.
Say NO to ET.
J Auton Nerv Syst. 2000;81:271277.
PMID: 10869731 [PubMed - indexed for MEDLINE]


17. Scott-Burden T, Vanhoutte PM.
The endothelium as a regulator of vascular smooth muscle proliferation.
Circulation. 1993;87(suppl V):V51V55.

18. Flavahan NA, Vanhoutte PM.
Endothelial cell signaling and endothelial dysfunction.
Am J Hypertens. 1995;8:28S41S.
PMID: 7646842 [PubMed - indexed for MEDLINE]


19. Shibano T, Vanhoutte PM.
Low molecular weight G-proteins of rho-family mediate relaxations to bradykinin in porcine coronary arteries.
Acta Pharmacol Sin. 2003;24:10701076.
PMID: 14627487 [PubMed - indexed for MEDLINE]


20. Fltou M, Vanhoutte PM.
Endothelium-dependent hyperpolarization of vascular smooth muscle cells.
Acta Pharmacol Sin. 2000;21:118.
PMID: 11263241 [PubMed - indexed for MEDLINE]


21. Vanhoutte PM.
Endothelium-derived free radicals: for worse and for better.
J Clin Invest. 2001;107:2325.
PMID: 11134174 [PubMed - indexed for MEDLINE]


22. Busse R, Edwards G, Fltou M, Fleming I, Vanhoutte PM.
EDHF: bringing the concepts together.
Trends Pharmacol Sci. 2002;23:374380.
PMID: 12377579 [PubMed - indexed for MEDLINE]


23. Vanhoutte PM,
ed. EDHF 2002. London and New York: Taylor & Francis; 2003:1427.

24. Nagao T, Illiano S, Vanhoutte PM.
Heterogeneous distribution of endothelium-dependent relaxations resistant to nitro-L-arginine in the arterial tree of the rat.
Am J Physiol. 1992;263:090094.

25. Thollon C, Fournet-Bourguignon MP, Saboureau D, et al.
Consequences of reduced production of NO on vascular reactivity of porcine coronary arteries after angioplasty: importance of EDHF.
Br J Pharmacol. 2002;136:11531161.
PMID: 12163348 [PubMed - indexed for MEDLINE]


26. Olmos L, Mombouli JV, Illiano S, Vanhoutte PM.
cGMP mediates the desensitization to bradykinin in isolated canine coronary arteries.
Am J Physiol. 1995;268:H865H870.
PMID: 7864213 [PubMed - indexed for MEDLINE]


27. Gisclard V, Miller V, Vanhoutte PM.
Effect of 17-estradiol on endothelium-dependent responses in the rabbit.
J Pharmacol Exp Ther. 1988;244:1922.
PMID: 3121846 [PubMed - indexed for MEDLINE]


28. Miller VM, Aarhus LL, Vanhoutte PM.
Modulation of endothelium-dependent responses by chronic alterations of blood flow.
Am J Physiol. 1986;251:H520H527.
PMID: 3752266 [PubMed - indexed for MEDLINE]


29. Mombouli JV, Nakashima M, Hamra M, Vanhoutte PM.
Endothelium-dependent relaxation and hyperpolarization evoked by bradykinin in canine coronary arteries: enhancement by exercise-training.
Br J Pharmacol. 1996;117:413418.
PMID: 8821528 [PubMed - indexed for MEDLINE]


30. Shimokawa H, Lam JY,Chesebro T, Bowie JH, Walter EJ, Vanhoutte PM.
Effects of dietary supplementation with cod-liver oil on endothelium-dependent responses in porcine coronary arteries.
Circulation. 1987;76:898905.
PMID: 3115623 [PubMed - indexed for MEDLINE]


31. Lscher TF, Vanhoutte PM.
Endothelium-dependent contractions to acetylcholine in the aorta of the spontaneously hypertensive rat.
Hypertension. 1986;8:344348.
PMID: 2870025 [PubMed - indexed for MEDLINE]


32. Vanhoutte PM.
Endothelium-dependent contractions: from superoxide anions to TP-receptor agonists.
Dialogues Cardiovasc Med. 2002;7:211225.

33. Shimokawa H, Aarhus LL, Vanhoutte PM.
Porcine coronary arteries with regenerated endothelium have a reduced endothelium-dependent responsiveness to aggregating platelets and serotonin.
Circ Res. 1987;61:256270.
PMID: 3113760 [PubMed - indexed for MEDLINE]


34. Shimokawa H, Flavahan NA, Vanhoutte PM.
Natural course of the impairement of endothelium-dependent relaxations after balloon endothelial removal in porcine coronary arteries. Possible dysfunction of a pertussis toxin-sensitive G protein.
Circ Res. 1989;65:740753.
PMID: 2504508 [PubMed - indexed for MEDLINE]


35. Borg-Capra C, Fournet-Bourguignon MP, Janiak P, et al.
Morphological heterogeneity with normal expression but altered function of Gi proteins in cultured regenerated porcine coronary endothelial cells.
Br J Pharmacol. 1997;122:9991008.

36. Castedo-Delrieu M, Fournet-Bourguignon MP, Bidouard JP, et al.
Phenotypic and functional characterization of regenerated endothelial cells after balloon injury in the pig.
J Vasc Res. 1997;34(suppl 1):10.

37. Shimokawa H, Vanhoutte PM.
Impaired endothelium-dependent relaxation to aggregating platelets and related vasoactive substances in porcine coronary arteries in hypercholesterolemia and atherosclerosis.
Circ Res. 1989;64:900914.
PMID: 2495869 [PubMed - indexed for MEDLINE]


38. Shimokawa H, Flavahan NA, Vanhoutte PM.
Loss of endothelial pertussis toxin-sensitive G protein function in atherosclerotic porcine coronary arteries.
Circulation. 1991;83:652660.
PMID: 1991383 [PubMed - indexed for MEDLINE]


39. Shibano T, Codina J, Birnbaumer L, Vanhoutte PM.
Pertussis toxin-sensitive G-proteins in regenerated endothelial cells after balloon denudation of porcine coronary artery.
Am J Physiol. 1994;267:H979H981.
PMID: 8092301 [PubMed - indexed for MEDLINE]


40. Cox DA, Cohen ML.
Effects of oxidized low-density lipoprotein on vascular contraction and relaxation: clinical and pharmacological implications in atherosclerosis.
Pharmacol Rev. 1996;48:319.
PMID: 8685247 [PubMed - indexed for MEDLINE]


41. Kennedy S, Fournet-Bourguignon M-P, Breugnot C, et al.
Cells derived from regenerated endothelium of the porcine coronary artery contain more oxidized forms of apolipoprotein-B-100 without a modification in the uptake of oxidized LDL.
J Vasc Res. 2003;40:389398.
PMID: 12904688 [PubMed - indexed for MEDLINE]


42. Vanhoutte PM, Boulanger CM, Vidal M, Mombouli JV.
Endothelium-derived mediators and the reninangiotensin system. In: Robertson JIS, Nicholls MG, eds. The ReninAngiotensin System. London: Gower Medical Publishing; 1993: 29.129.15.

Back Back to the Summary

Although great care has been taken in compiling the information given in this website,
the publisher or the sponsor is not responsible for the continued currency of the information,
for any errors or omissions, or for any consequence arising therefrom.
© 2010 Les Laboratoires Servier