The ability of the endothelium to release NO can be up-regulated or down-regulated in the intact organism by a number of chronic factors.

Up-regulation. Shear stress: Both acute and chronic increases in flow, and the resulting increasing force of shearing (shear stress) of the blood on the endothelial cells, augment the expression and the activity (in a Ca2+-independent way) of eNOS, and thus the release of EDRF/NO (Fig. 2) (Rubanyi et al. 1986, Miller & Vanhoutte 1988, Davies 1995, Davis et al. 2001, Stepp et al. 2001, Busse & Fleming 2003, Bellien et al. 2006, Spier et al. 2007, Yan et al. 2007). This immediate effect of an increase in shear stress on the release of NO explains flow-mediated dilatation, a phenomenon often used to estimate the functional state of the endothelium in humans. In the coronary circulation, the effect of shear stress involves the local production of the autacoid bradykinin that stimulates the release of NO through a Gq-dependent mechanism (Fig. 6) (Flavahan et al. 1989, Mombouli & Vanhoutte 1991, 1995, Shimokawa et al. 1991, Roves et al. 1995). The chronic effect of shear stress is due to an up-regulation of eNOS and a greater activation (phosphorylation) of the enzyme, leading to a larger release of NO for each given stimulation, explaining the beneficial effects of regular exercise on endothelial function (Miller & Vanhoutte 1988, Mombouli et al. 1996, Hambrecht et al. 2003, Suvorava et al. 2004, Watts et al. 2004, Lauer et al. 2005, Gertz et al. 2006, Rakobowchuk et al. 2008).

Model of endothelial dysfunction in the hypercholesterolaemic mouse. Left: in the normal mouse aortic endothelium, l-arginine (l-Arg) is transformed by eNOS to NO, which exerts its well-documented beneficial effects (most are not shown for the sake of clarity), including inhibition of the oxidation of LDLs to oxy-LDL. The by-product of the reaction, l-citrulline (l-Cit), inhibits arginase II (AaII), which is constrained to the microtubules (MT). Right: in the aortic endothelium of the ApoE−/− and the wild-type hypercholesterolaemic mice, the accumulation of oxy-LDL dislocates arginase II from the microtubules and augments its activity. Arginase II competes with endothelial NO synthase for the common substrate l-arginine, leading to uncoupling of NO synthase and the production of superoxide anions (O2−), which further enhance the production of oxy-LDL. The latter also facilitates dissociation of eNOS from the caveolae and reduces the genomic expression of the enzyme, leading to further reduction in the production of NO. This model does not account for the biological effects, if any, of l-ornithine (l-Om) and urea produced by arginase II. It also does not account for endothelium-derived relaxing signals other than NO, or for the generation of endothelium-derived contracting substances. CM indicates cell membrane; +, facilitation; −, inhibition (modified from Vanhoutte 2008).

Oestrogens and gender. Although ovariectomy does not alter or even increase the mRNA expression and the presence of eNOS (Wassmann et al. 2001, Okano et al. 2006), the reintroduction of physiological levels of oestrogens in ovariectomized animals augments endothelium-dependent relaxations to muscarinic agonists (Gisclard et al. 1988, Wassmann et al. 2001, Santos et al. 2004, Scott et al. 2007) and accelerates endothelial healing after injury (Filipe et al. 2008). The potentiating effect of oestrogens on endothelium-dependent relaxations involves both genomic (Fig. 2) and non-genomic effects (see Tostes et al. 2003, Keung et al. 2005, Miller & Duckles 2008). It depends presumably both on a reduction in oxidative stress leading to an increased bioavailability of the endothelium-derived mediator and an increased responsiveness of the vascular smooth muscle cells to vasodilator stimuli (Wassmann et al. 2001, Han et al. 2007, Li et al. 2007a,b, Scott et al. 2007). In the intact organism, a reduced production of the endogenous inhibitor of eNOS, asymmetric dimethyl arginine (ADMA) may contribute (Filser 2005, Monsalve et al. 2007). Phyto-oestrogens and selective oestrogen receptor modulators also potentiate endothelium-dependent relaxations/vasodilatations (Lee & Man 2003, Sbarouni et al. 2003, Wong et al. 2006, Chan et al. 2007, Leung et al. 2007). In coronary arteries, the potentiating effect of chronic treatment with oestrogens is seen only with stimuli that activate Gi-coupled receptors on the endothelial cells and is counteracted by the chronic administration of progesterone (Miller & Vanhoutte 1991). It is likely that this potentiating effect of oestrogens on NO release, presumably resulting from lower oxidative stress, helps to explain why endothelium-dependent relaxations are more pronounced in arteries from female than male animals (Kauser & Rubanyi 1995, Kähönen et al. 1998, Dantas et al. 2004) and thus why women are protected against coronary disease, at least until the age of menopause. The opposing effects of oestrogens and progesterone could explain why hormone replacement therapy has not always had the expected beneficial effect on the occurrence of cardiovascular events.

Insulin. Insulin facilitates NO-dependent vasodilatations in vivo (Steinberg et al. 1994, Taddei et al. 1995b, Lembo et al. 1997). It enhances the expression of eNOS in native endothelial cells in vitro (Fisslthaler et al. 2003).

Adiponectin. Adiponectin improves endothelial function and protects the endothelium by promoting eNOS activity and the bioavailability of NO (Chen et al. 2003, Hattori et al. 2003, Tan et al. 2004, Li et al. 2007b, Wang & Scherer 2008, Zhu et al. 2008).

Other hormones. In postmenopausal women, testosterone appears to potentiate endothelium-dependent vasodilatation (Montalcini et al. 2007). Thyroid hormone up-regulates eNOS and augments the endothelial production of NO in the animal (Spooner et al. 2004). Adrenalectomy augments the expression of eNOS (Li et al. 2007a) and aldosterone acutely augments NO-dependent relaxations by a non-genomic action (Uhrenholt et al. 2003, 2004, Skott et al. 2006, Nietlispach et al. 2007). Glucagon-like peptide-1 enhances the vasodilator response to acetylcholine (Basu et al. 2007).

Diet. The chronic intake of ω3-unsaturated fatty acids potentiates the endothelium-dependent relaxations of coronary arteries to aggregating platelets and other stimuli and have antiatherogenic properties (Shimokawa et al. 1987, 1988a,b, Shimokawa & Vanhoutte 1989a, Shepherd & Vanhoutte 1991, von Schacky & Harris 2007, Sekikawa et al. 2008, Sena et al. 2008). The same holds true for the intake of flavonoids (Machha & Mustafa 2005, Machha et al. 2007, Xu et al. 2007) and other polyphenols, whether present in red wine (in particular resveratrol) (Stockley 1998, Leikert et al. 2002, Wallerath et al. 2002, da Luz & Coimbra 2004, Dell’Agli et al. 2004, Soares de Moura et al. 2004, Coimbra et al. 2005, Boban et al. 2006, Sarr et al. 2006, Das et al. 2007, Lefèvre et al. 2007, Aubin et al. 2008, Chan et al. 2008a,b, Csiszar et al. 2008, Lopez-Sepulveda et al. 2008), in green tea (Kuriyama et al. 2006, Alexopoulos et al. 2008), grape juice (Anselm et al. 2007), in pomegranate juice (Nigris et al. 2006, 2007a,b) or in dark chocolate (Fisher et al. 2003, Engler et al. 2004, Grassi et al. 2005, Schroeter et al. 2006, Flammer et al. 2007, Taubert et al. 2007).

Arginine. Although the acute administration of l-arginine can favour endothelium-dependent responses in humans (e.g. Bode-Böger et al. 1996, Taddei et al. 1997b, Perticone et al. 2005), its chronic supplementation offers no therapeutic benefit in patients with vascular disease (Wilson et al. 2007), reinforcing the early suspicion (Schini & Vanhoutte 1991a,b) that the semi-essential amino acid is rarely a limiting factor for the endothelial production of NO. An exception may be when the endothelial arginases, which compete with eNOS for this substrate, are more active (Fig. 6) (Ming et al. 2004, Brandes 2006, Ryoo et al. 2006, 2008, Holowatz & Kenney 2007, Katusic 2007, Santhanam et al. 2007, Romero et al. 2008, Vanhoutte 2008).

Down-regulation. Oxygen-derived free radicals: Several enzymes in the endothelial cells can produce superoxide anions (Fig. 7). They include NADPH oxidase, xanthine oxidase, cyclooxygenase and eNOS itself, when it is uncoupled by lack of substrate (l-arginine) or shortage of the essential co-factor tetrahydrobiopterin (BH4) (see Kojda & Harrison 1999, Stuehr et al. 2001, Fleming et al. 2005). Superoxide anions can be dismutated by superoxide dismutase (SOD) to hydrogen peroxide (H2O2) which can act as an EDHF and contribute to endothelium-dependent relaxations (Fig. 2) (Matoba et al. 2000, Morikawa et al. 2003, Shimokawa & Matoba 2004; see Félétou & Vanhoutte 2006a,b,c, 2007), or be broken down by catalase. However, superoxide anions also scavenge NO avidly with the resulting formation of peroxynitrite (Gryglewski et al. 1986, Rubanyi & Vanhoutte 1986, Auch-Schwelk et al. 1992, Cosentino et al. 1994, Tschudi et al. 1996a, DeLano et al. 2006, Kagota et al. 2007, Miyagawa et al. 2007, Macarthur et al. 2008). This reduces considerably the bioavailability of NO (see Kojda & Harrison 1999). Hence, increases in oxidative stress have been consistently associated with reduced endothelium-dependent relaxations, and antioxidants shown to acutely improve such responses in vitro and in vivo both in animals (e.g. Aubin et al. 2006, Liu et al. 2007) and humans (e.g. Kanani et al. 1999, Taddei et al. 2001, Holowatz & Kenney 2007). However, the therapeutic relevance of these findings is questionable as chronic treatment with antioxidants usually fails to improve endothelial function in people (e.g. Duffy et al. 2001, Pellegrini et al. 2004), with maybe the exception of the chronic administration of low doses of folic acid (Moat et al. 2006).

Two major contributors of reactive oxygen species in the vascular wall. Left: l-arginine-endothelial NOS (eNOS) pathway. Synthetic pathway of tetrahydrobiopterin (BH4), an essential cofactor, is also shown and some of the most common inhibitors of NOS, analogues of l-arginine, are indicated. FMN, flavin mononucleotide; GTP, guanosine 5′-triphosphate. Right: activation of the NAD(P)H oxidase (NOX). Endothelial cells express NOX1, NOX2 (gp91phox), NOX4 and NOX5 isoforms, whereas vascular smooth muscle cells express the NOX1, NOX4 and NOX5 and in resistance arteries NOX2 isoforms. Apocynin inhibits NOX by preventing translocation of cytosolic subunits and their association with the membrane located subunits, whereas diphenyleneiodonium (DPI), a flavoprotein inhibitor, is a nonspecific inhibitor of NOX (from Félétou and Vanhoutte 2006b. By permission of the American Physiological Society).

Hormones. Long-term exposure to aldosterone has a detrimental effect on NO-dependent relaxations, presumably by reducing the production of the essential cofactor for eNOS, tetrahydrobiopterin and increasing oxidative stress (Mitchell et al. 2004, Hashikabe et al. 2006, Nagata et al. 2006, Skott et al. 2006, Nietlispach et al. 2007, Sartorio et al. 2007). Melatonin also inhibits the endothelial formation of NO (Silva et al. 2007). Castration of male animals augments the vasodilator response to acetylcholine (Ajayi et al. 2004).

Ageing: Both in animals and in humans, increasing age reduces the ability of the endothelium to elicit endothelium-dependent vasodilatations in vitro and in vivo (see Moritoki et al. 1986, Hongo et al. 1988, Koga et al. 1988, Charpie et al. 1994, Kung & Lüscher 1995, Davidge et al. 1996, Chauhan et al. 1996, Taddei et al. 1997b, 2001, Cernadas et al. 1998, Yasuro et al. 1999, Heymes et al. 2000, Csiszar et al. 2002, 2007, Vanhoutte 2002, Subramanian & MacLeod 2003, Spier et al. 2007, Bulckaen et al. 2008). This is due to an increased activity of arginase, competing with eNOS for the common substrate arginine (Katusic 2007, Santhanam et al. 2007), an augmented production of oxygen-derived free radicals reducing the bioavailability of NO (Tschudi et al. 1996a, Taddei et al. 2001, Csiszar et al. 2002, 2007), a reduced expression/presence of eNOS (Challah et al. 1997, Chou et al. 1998, Csiszar et al. 2002), a reduced activity of the enzyme (Cernadas et al. 1998) and ultimately a lesser release of NO (Tschudi et al. 1996a). In addition, the expression of soluble guanylyl cyclase in ageing vascular smooth muscle is reduced (Klößet al. 2000). However, an important part of the endothelial dysfunction with ageing is due to the endothelial release of vasoconstrictor prostaglandins (see section EDCF).

Smoking and environment. Active and passive smoking blunt endothelium-dependent vasodilatations. This appears to be due to an action of nicotine causing a greater formation of ADMA and to an increased production of oxygen-derived free radicals, both resulting in a lesser availability of NO (Sousa et al. 2005, Michaud et al. 2006, Gamboa et al. 2007, Argacha et al. 2008, Celermajer & Ng 2008, Csiszar et al. 2008, Heiss et al. 2008, Lang et al. 2008). Chronic exposure to air pollution decreases endothelium-dependent vasodilatations (Briet et al. 2007).

Homocysteinaemia. Increased levels of homocysteine impair eNOS-dependent relaxations/vasodilatations both in vitro and in vivo, presumably by increasing oxidative stress (e.g. Bellamy et al. 1998, Chambers et al. 1999, Kanani et al. 1999, Lang et al. 2000, Hanratty et al. 2001, Liu et al. 2007, Looft-Wilson et al. 2008).

Hypercholesterolaemia. Both in animals and in humans, hypercholesterolemia reduces endothelium-dependent relaxations/dilatations and the normalization of the cholesterol level with treatment restores the response (Shimokawa & Vanhoutte 1989a,b, Vanhoutte 1991, Trochu et al. 2003, Kaul et al. 2004, Landmesser et al. 2005, August et al. 2006, Fichtlscherer et al. 2006, Inoue & Node 2007, Aubin et al. 2008, Knight et al. 2008, Sena et al. 2008). This is explained best by an increased oxidative stress leading to a reduced bioavailability of NO, an impairment of the turnover rate of eNOS and an increased presence of ADMA (Bode-Böger et al. 1996, Böger & Bode-Böger 2001, Böger et al. 2004, August et al. 2006, Palm et al. 2007).

Obesity. Obese animals and humans exhibit reduced responses to endothelium-dependent vasodilators (Karagiannis et al. 2003, Van Guilder et al. 2006, 2008, Bouvet et al. 2007, Kagota et al. 2007). A major reason for the blunted endothelium-dependent relaxation is the production of EDCF (see section EDCF). Weight loss alone or exercise training improve endothelium-dependent responses (Watts et al. 2004, Focardi et al. 2007, Pierce et al. 2008, Ungvari et al. 2008).

Sleep apnoea. Intermittent hypoxia, as occurring with obstructive sleep apnoea reduces endothelium-dependent responsiveness (Budhiraja et al. 2007).