Introduction

Reactive oxygen species (ROS) are ubiquitous reactive derivatives of O2 metabolism found in the environment and in all biological systems. Reactive oxygen species from mitochondria and other cellular sources were traditionally considered as injurious cellular by-products with the potential to damage lipids, proteins and DNA (Freeman and Crapo 1982). However, there is now convincing evidence that ROS are not only toxic consequences of cellular metabolism but also essential participants in cell signaling and regulation (Griendling et al. 2000a; Sauer et al. 2001; Reth 2002; Chiarugi and Cirri 2003). Reactive oxygen species are implicated in many intracellular signaling pathways leading to changes in gene transcription and protein synthesis and consequently in cell function.

Within the cardiovascular system, ROS play an essential physiological role in maintaining cardiac and vascular integrity and a pathophysiological role in cardiovascular dysfunction associated with conditions such as hypertension, diabetes, atherosclerosis, ischemia–reperfusion injury, ischemic heart disease, and congestive cardiac failure (Landmesser and Harrison 2001; Zalba et al. 2001a). Among the major ROS important in these processes are superoxide anion (·O2−), hydrogen peroxide (H2O2), hydroxyl radical (·OH), and the reactive nitrogen species, nitric oxide (NO) and peroxynitrite (ONOO−). Under physiological conditions, ROS are produced in a controlled manner at low concentrations and function as signaling molecules regulating vascular smooth muscle cell (VSMC) contraction–relaxation and VSMC growth (Rao and Berk 1992; Cosentino et al. 1994; Zafari et al. 1998; Touyz and Schiffrin 1999). Under pathological conditions increased ROS production leads to endothelial dysfunction, increased contractility, VSMC growth and apoptosis, monocyte migration, lipid peroxidation, inflammation, and increased deposition of extracellular matrix proteins, major processes contributing to vascular damage in cardiovascular disease (Rao and Berk 1992; Harrison 1997).

In experimental and clinical hypertension, ROS generation is increased (Kerr et al. 1999; Romero and Reckelhoff 1999; Schnackenberg et al. 1999; Chen et al. 2001). Treatment with antioxidants improves vascular function and structure, prevents target-organ damage, and reduces blood pressure in animal models (Romero and Reckelhoff 1999; Schnackenberg et al. 1999; Chen et al. 2001; Hoagland et al. 2003) and in human hypertension (Sharma et al. 1996; Duffy et al. 1999; Fotheby et al. 2000; Boshtam et al. 2002; Mullan et al. 2002). Mouse models deficient in ROS-forming enzymes have lower blood pressure compared with wild-type counterparts and angiotensin II (Ang II) infusion in these mice fails to induce hypertension (Bendall et al. 2002; Li and Shah 2003). Furthermore, in cultured VSMCs and isolated arteries from hypertensive rats and humans, ROS production is augmented and antioxidant capacity is reduced (Schnackenberg et al. 1999; Chen et al. 2001; Touyz and Schiffrin 2001). Accordingly, evidence at both experimental and clinical levels supports a pathophysiological role for ROS and oxidative stress in the development and progression of hypertension and its associated end-organ damage.

In the present review, we focus on the role of ROS in vascular biology and implications of oxidative stress in hypertensive vascular damage. Although the cardiac, renal, endocrine, and central nervous systems are also major targets for oxidative damage by ROS, these systems will not be discussed here and the reader is referred to excellent recent reviews on these systems (Wilcox 2002; Zanzinger 2002; Cantor et al. 2003). In this chapter we discuss mechanisms whereby ROS are formed in vascular cells, especially relating to non-phagocytic NAD(P)H oxidase, how ROS influence vascular function, and what the implications of oxidative stress are in vascular injury in hypertension.

Reactive oxygen species and oxidative stress in the vasculature

Reactive oxygen species are formed as intermediates in reduction–oxidation (redox) processes, leading from oxygen to water (Fridovich 1997). The univalent reduction of oxygen, in the presence of a free electron (e), yields ·O2−, H2O2, and ·OH. Superoxide has an unpaired electron, which imparts high reactivity and renders it unstable and short-lived. Superoxide is water soluble and acts either as an oxidizing agent, where it is reduced to H2O2, or as a reducing agent, where it donates its extra electron to form ONOO− with NO (Darley-Usmar et al. 1995; Fridovich 1997). In physiological conditions in aqueous solutions at a neutral pH, the favored reaction of ·O2− is the dismutation reaction yielding H2O2. However, when produced in excess, a significant amount of ·O2− reacts with NO to produce ONOO− (Darley-Usmar et al. 1995). Superoxide is membrane-impermeable, but can cross cell membranes via anion channels (Schafer and Beuttner 2001; Han et al. 2003).

Hydrogen peroxide is produced mainly from dismutation of ·O2−. This reaction can be spontaneous or it can be catalyzed by superoxide dismutase (SOD), of which there are three isoforms, CuZnSOD, MnSOD, and extracellular SOD (EC-SOD) (Fridovich 1997). The SOD-catalyzed dismutation is favored when the concentration of ·O2− is low and when the concentration of SOD is high, which occurs under physiological conditions. Unlike ·O2−,H2O2 is not a free radical and is a much more stable molecule. Hydrogen peroxide is lipid soluble, crosses cell membranes, and has a longer half-life than ·O2−. In biological systems, it is scavenged by catalase and by glutathione peroxidase (Schafer and Beuttner 2001). Hydrogen peroxide can also be reduced to generate the highly reactive ·OH (Haber-Weiss or Fenton reaction) in the presence of metal-containing molecules such as Fe2+ (Fridovich 1997). Hydroxyl radical is extremely reactive and unlike ·O2− and H2O2, which travel some distance from their site of generation, ·OH induces local damage where it is formed.

In the vasculature, ·O2−, H2O2, NO, OONO−,and ·OH are all produced to varying degrees (Fig. 1). These pro-oxidants are tightly regulated by anti-oxidants such as SOD, catalase, thioredoxin, glutathione, anti-oxidant vitamins, and other small molecules (Stralin et al. 1995; Halliwell 1999; Channon and Guzik 2002; Yamawaki et al. 2003). Under normal conditions, the rate of ROS production is balanced by the rate of elimination. However, a mismatch between ROS formation and the ability to defend against them by antioxidants results in increased bioavailability of ROS leading to a state of oxidative stress (Griendling et al. 2000a; Landmesser and Harrison 2001; Zalba et al. 2001a). The pathogenic outcome of oxidative stress is oxidative damage (Griendling et al. 2000a; Schafer and Buettner 2001; Zalba et al. 2001a), a major cause of vascular injury in hypertension.

Fig. 1
figure 1

Generation of superoxide (·O2−) and H2O2 from O2 in vascular cells. Many enzyme systems, including NAD(P)H oxidase, xanthine oxidase, and uncoupled nitric oxide synthase (NOS) among others, have the potential to generate reactive oxygen species (ROS). Superoxide acts either as an oxidizing agent, where it is reduced to H2O2 by superoxide dismutase (SOD), or as a reducing agent, where it donates its extra electron (e−) to form ONOO− with NO. Hydrogen peroxide is scavenged by catalase, glutathione and thioredoxin systems, and can also be reduced to generate ·OH in the presence of Fe2+

Full size image

Production of ROS in vessels

Non-phagocytic, vascular NAD(P)H oxidase(s)

Vascular ROS are produced in endothelial, adventitial, and VSMCs (Stralin et al. 1995; Rajagopalan et al. 1996a; Halliwell 1999; Channon and Guzik 2002; Sorescu et al. 2002; Yamawaki et al. 2003) and derived predominantly from NAD(P)H oxidase, which is a multi-subunit enzyme (Jones et al. 1996; Azumimi et al. 1999; Griendling et al. 2000b; Lassegue and Clempus 2003) that catalyzes the production of ·O2− by the one electron reduction of oxygen using NAD(P)H as the electron donor: 2O2 + NAD(P)H → 2O2− + NAD(P)H + H+. The prototypical and best characterized NAD(P)H oxidase is that found in phagocytes (neutrophilic and eosinophilic granulocytes, monocytes, and macrophages; Leusen et al. 1996; Babior et al. 2002; Vignais 2002). Phagocytic NAD(P)H oxidase comprises at least five components: (phox for PHagocyte OXidase), p47phox, p67phox, p40phox, p22phox, and gp91phox (Babior et al. 2002; Vignais 2002). Additional components include the small G proteins Rac2 (Rac1 in some cells) and Rap1A. In unstimulated cells, p40phox, p47phox, and p67phox exist in the cytosol, whereas p22phox and gp91phox are located in the membranes, where they occur as a heterodimeric flavoprotein, cytochrome b558. Upon cell stimulation, p47phox becomes phosphorylated, the cytosolic subunits form a complex, which then migrates to the membrane where it associates with cytochrome b558 to assemble the active oxidase, which now transfers electrons from the substrate to O2 leading to ·O2− generation (Leusen et al. 1996; Babior et al. 2002). Whereas phosphorylation of p47phox and p67phox is critically involved in the activation of NAD(P)H oxidase (Touyz et al. 2003a), phosphorylated p40phox is not essential for activation and has recently been reported to be a negative regulator of the oxidase (Lopes et al. 2004).

It is now clearly established that NAD(P)H oxidase is also functionally important in non-phagocytic cells. In fact NAD(P)H oxidase is the primary source of ·O2− in the vasculature (Berry et al. 2000; Channon and Guzik 2002; Touyz et al. 2002a; Lassegue and Clempus 2003) and is functionally active in all layers of the vessel wall: in the endothelium (Muzaffar et al. 2003), the media (Berry et al. 2000; Touyz et al. 2002a), the adventitia (Rey and Pagano 2002), and in cultured VSMCs, fibroblasts, and endothelial cells (Griendling et al. 1994; Seshiah et al. 2002; Touyz et al. 2002a). Unlike phagocytic NAD(P)H oxidase, which is activated only upon stimulation and which generates ·O2− in a burst-like manner extracellularly (De Leo et al. 1996; Babior et al. 2002), vascular oxidases are constitutively active, preassembled, and produce ·O2− intracellularly in a slow and sustained fashion and act as intracellular signaling molecules (Li and Shah 2002; Lassegue and Clempus 2003; Table 1).

Table 1 Characteristics of phagocytic and vascular NADPH oxidase
Full size table

All of the phagocytic NAD(P)H oxidase subunits are expressed, to varying degrees, in vascular cells. In endothelial and adventitial cells p47phox, p67phox, p22phox, and gp91phox are present (Rey and Pagano 2002; Lassegue and Clempus 2003; Touyz et al. 2003a). The situation is more complex in VSMCs, where the major subunits are not always detected. Only p47phox and p22phox seem to be consistently expressed (Lassegue and Clempus 2003). In rat aortic VSMCs, p22phox and p47phox, but not gp91phox, are present, whereas in human resistance arteries, all of the major subunits, including gp91phox, are expressed (Azumimi et al. 1999; Touyz et al. 2002a; Yamawaki et al. 2003). Recent studies demonstrated that the newly discovered gp91phox (Nox2) homologs, Nox1, Nox4, and Nox5 (Nox for NAD(P)H Oxidase) are found in the vasculature (Suh et al. 1999; Cheng et al. 2001; Ago et al. 2004; Hilenski et al. 2004). Nox1 mRNA is expressed in rat aortic VSMCs and may be a substitute for gp91phox in these cells (Suh et al. 1999; Griendling et al. 2000b; Touyz et al. 2002a). Although initial studies suggested that Nox1 is a subunit-independent low capacity ·O2−-generating enzyme involved in the regulation of mitogenesis (Banfi et al. 2003), recent data indicate that Nox1 requires p47phox and p67phox and that it is regulated by NoxO1 (Nox organizer 1) and NoxA1 (Nox activator 1) (Banfi et al. 2003). The exact role of NoxO1 and NoxA1 in vascular cells is currently unknown. Nox1 may be important in pathological processes as it is significantly upregulated in vascular injury (Lassegue and Clempus 2003). Nox4 appears to be abundantly expressed in all vascular cell types (Wingler et al. 2001; Yamawaki et al. 2003) and may play an important role in constitutive production of ·O2− in non-proliferating cells (Lassegue et al. 2001; Wingler et al. 2001). Ago et al. (2004) recently reported that Nox4 is the major catalytic component of endothelial NAD(P)H oxidase. A unique p67phox homolog has also been identified, but it is not yet known whether this isoform is present in vascular cells (Gauss et al. 2002). The functional significance of NAD(P)H oxidase subunit homologs in the vasculature is presently unclear and awaits further clarification.

Vascular NAD(P)H oxidase is regulated by many humoral factors, including cytokines, growth factors, and vasoactive agents (Lassegue and Clempus 2003; Fig. 1, 2). Physical factors, such as stretch, pulsatile strain, and shear stress also stimulate NAD(P)H oxidase activation (Grote et al. 2003; Lassegue and Clempus 2003). Of particular importance, with respect to hypertension, is Ang II, which stimulates activation of NAD(P)H oxidase, increases expression of NAD(P)H oxidase subunits, and induces ROS production in cultured VSMCs, endothelial cells, adventitial fibroblasts (Lassegue and Clempus 2003), and in intact arteries (Griendling et al. 1994; Seshiah et al. 2002; Touyz et al. 2002a; Yamawaki et al. 2003). Oxidase activation occurs acutely by stimulation of intracellular signaling molecules (Griendling et al. 2000b; Lassegue and Clempus 2003) and chronically by upregulation of NAD(P)H oxidase subunits (Touyz et al. 2002a; Lassegue and Clempus 2003). These effects are mediated via AT1 receptors (Privratsky et al. 2003). Interestingly ROS regulate AT1 receptor gene expression, which in turn modulates ROS formation (Nickenig et al. 2000).

Fig. 2
figure 2

Upstream regulators of NAD(P)H oxidase in vascular cells. Vasoactive agents, such as angiotensin II, which signal through G protein-coupled receptors (GPCR), growth factors and cytokines, which signal through receptor tyrosine kinases and physical factors, such as stretch and pressure, stimulate NAD(P)H oxidase through multiple signaling cascades. AA Arachidonic acid, PLA2 phospholipase A2, PLD phospholipase D, PKC protein kinase C

Full size image

Mechanisms linking Ang II/AT1 to NAD(P)H oxidase and upstream signaling molecules regulating the oxidase in vascular cells have not been fully elucidated, but PLD, PLA, PKC, c-Src, PI3 K, RhoA, and Rac have been demonstrated to be implicated in AT1 signaling to NAD(P)H oxidase (Seshiah et al. 2002; Touyz et al. 2003a; Fig. 2). Platelet-derived growth factor (PDGF), transforming growth factor-β (TGF-β), tumor necrosis factor (TNF)-α and thrombin also activate NAD(P)H oxidase in VSMCs (Marumo et al. 1997; De Keulenaer et al. 1998; Gorlach et al. 2001; Brandes et al. 2002). Endothelin-1 increases NAD(P)H oxidase activity in human endothelial cells (Duerrschmidt et al. 2000) and, in intact vessels, this effect is mediated via ETA receptors (Li et al. 2003a). Activators of peroxisome proliferator-activated receptors (PPARs), statins, and antihypertensive drugs, such as β-blockers, Ca2+ channel blockers, ACE inhibitors, and AT1 receptor blockers, downregulate expression of oxidase subunits and decrease NAD(P)H oxidase activity (Dandona et al. 2000; Dhalla et al. 2000; Mantle et al. 2000; Ohtahara et al. 2001; Taddei et al. 2001; Diep et al. 2002; Wassman et al. 2002). These actions may have therapeutic potential in cardiovascular disease.

Other enzymatic sources of ROS in the vasculature

Nitric oxide synthase (NOS), the enzyme primarily responsible for NO production, can also generate ·O2− in conditions of substrate (arginine) or cofactor (tetrahydrobiopterin; BH4) deficiency (Milstien and Katusic 1999; Cosentino et al. 2001). These findings have led to the concept of “NOS uncoupling”, where the activity of the enzyme for NO production is decreased in association with an increase in NOS-dependent ·O2− formation. eNOS uncoupling has been demonstrated in atherosclerosis (Vasquez-Vivar et al. 2002), diabetes (Bagi and Koller 2003), hyperhomocysteinemia (Virdis et al. 2003), and hypertension (Landmesser et al. 2003; Podjarny et al. 2003). Landmesser et al. (2003) reported that, in hypertension, increased NAD(P)H oxidase-derived ·O2− leads to augmented ROS bioavailability, which causes oxidation of BH4 and consequent uncoupling of eNOS, further contributing to ROS production. Gene transfer of GTP cyclohydrolase (GTPCH) I, the enzyme responsible for regenerating BH4, restored arterial GTPCH I activity and BH4 levels, reduced ROS, and improved endothelium-dependent relaxation and NO release in DOCA-salt hypertensive rats, in which endothelial dysfunction results from NAD(P)H-dependent oxidant excess (Zheng et al. 2003). The potential role of uncoupling of NOS as a source of ROS in hypertension is also supported in human studies where increased endothelial ·O2− production in vessels from diabetic and hypertensive patients is inhibited by sepiapterin, a precursor of BH4 (Guzik et al. 2002; Higashi et al. 2002a). The relative importance of NOS- versus NAD(P)H oxidase-mediated ·O2− generation in hypertension probably relates, in part, to the magnitude of endothelial dysfunction, since most conditions in which ·O2− is derived from NOS are associated with marked endothelial dysfunction (Channon and Guzik 2002.

Other enzymatic sources capable of generating ROS in the vasculature are xanthine oxidase, cytochrome P450, mitochondrial respiratory chain enzymes, and phagocyte-derived myeloperoxidase (Taniyama and Griendling 2003; Fig. 1). However, the contribution of these enzymes to vascular generation of ROS is relatively minor compared with NAD(P)H oxidase.

Signaling pathways and molecular targets of ROS

Although ROS have been shown to be involved in many signal transduction pathways (Fig. 3), the exact molecular targets have not yet been clearly defined. Addition of exogenous ROS activates mitogen-activated protein (MAP) kinases, including ERK1/2, p38MAP kinase, JNK, and ERK5, important in cell growth, inflammation, apoptosis, and cell differentiation, respectively (Torres 2003). In cultured VSMCs, generation of endogenous ROS by Ang II influences activation of p38MAP kinase, JNK, and ERK5, but not of ERK1/2 (Ushio-Fukai et al. 1998; Viedt et al. 2000; Touyz et al. 2003b, 2004). However, serotonin-mediated ERK1/2 activation in smooth muscle cells is redox-sensitive, but in fibroblasts it is not (Lee et al. 1999), suggesting that redox-regulation of MAP kinases may be ligand- and cell-specific. Although MAP kinases are redox-sensitive, they are probably not direct substrates of ·O2− and H2O2. Upstream modulators such as MEKs, tyrosine kinases, and phosphatases are likely direct targets (Lee and Esselman 2002).

Fig. 3
figure 3

Redox-dependent signaling pathways in vascular cells. Intracellular ROS influence the activity of protein tyrosine phosphatases (PTP) by modifying cysteine residues. Oxidation of the cysteine residue to sulfenic acid by H2O2 renders PTPs inactive, whereas reduction renders PTPs active. Activated PTP decreases activity of protein tyrosine kinases (PTK) and mitogen-activated kinases (MAPK), whereas inactivated PTP have opposite actions. ROS also influence gene and protein expression by activating transcription factors, such as NFκB and AP-1. ROS stimulate ion channels, such as plasma membrane Ca2+ and K+ channels, leading to changes in cation concentration and matrix metalloproteinases (MMPs), which influence extracellular matrix proteins (ECM) degradation. Activation of these redox-sensitive pathways results in many cellular responses, which, if uncontrolled, could contribute to altered vascular tone, increased vascular smooth muscle cell (VSMC) growth, inflammation, and increased deposition of extracellular matrix protein, leading to vascular remodeling in hypertension. ↓ Decreased effect, ↑ increased effect

Full size image

Receptor and non-receptor tyrosine kinases are also influenced by ROS (Yang et al. 2000). Exogenous H2O2 induces tyrosine phosphorylation and activation of PDGFR and EGFR, probably due to ROS-mediated inhibition of dephosphorylation of PDGFR and EGFR by inactivation of membrane-associated protein tyrosine phosphatases (Yang et al. 2000; Droge 2001). Oxygen intermediates, which are produced in response to tyrosine kinase receptor activation, are also involved in transactivation of PDGFR and EGFR by Ang II. Under pathological conditions associated with oxidative stress, such as hypertension, ROS may directly activate cell surface receptors, thereby amplifying the process of ·O2− generation. Non-receptor tyrosine kinases such as Src, JAK2, Pyk2, and Akt, all of which have been implicated in cardiovascular remodeling and vascular damage, are also regulated by ROS (Griendling et al. 2000a; Touyz and Schiffrin 2000; Droge 2001; Touyz et al. 2002b).

The best-established direct targets of ROS signaling are protein tyrosine phosphatases (Lee et al. 1998; Meng et al. 2002) and transcription factors (Haddad 2002; Turpaev 2002). All tyrosine phosphatases have a conserved 230-amino acid domain that contains a reactive and redox-regulated cysteine, which catalyses the hydrolysis of protein phosphotyrosine residues by the formation of a cysteinyl-phosphate intermediate (Brigelius-Flohe et al. 2004). Oxidation of this cysteine residue to sulfenic acid by H2O2 renders the tyrosine phosphatase inactive (Brigelius-Flohe et al. 2004). Thus ROS inhibit activity of tyrosine phosphatases, resulting in increased tyrosine phosphorylation, which influences oxidative stress-induced activation of receptor protein tyrosine kinases, such as the EGFR, IGF-1R, and PDGFR (Kamata et al. 2000), and non-receptor tyrosine kinases, such as Src, FAK, PI3 K, and JAK2.

Transcription factors, including nuclear factor κB (NFκB), activator protein 1 (AP-1), c-Myb, Sp-1, p53, early growth response-1 (egr-1), and hypoxia-inducible factor (HIF-1) are directly activated by ROS (Haddad 2002; Turpaev 2002). NFκB and AP-1 induce expression of pro-inflammatory genes, including monocyte chemotactic protein-1 (MCP-1), adhesion molecules, and interleukins (Brigelius-Flohe et al. 2004), that play a role in vascular inflammation associated with hypertension and atherosclerosis. Most redox-sensitive transcription factors possess conserved cysteines, which are susceptible to oxidative modification (Haddad 2002). It has been suggested that the reactive cysteines constitute redox-sulfhydryl switches that directly regulate gene expression (Kamata et al. 2000). Another mode of redox regulation of transcription factor activity is by the redox sensitivity of protein degradation. Increased activation of vascular NFκB and AP-1 and associated inflammatory and mitogenic responses have been demonstrated in hypertensive rats (Taddei et al. 2001). These actions have been attributed, in part, to oxidative excess.

In addition to influencing cellular processes associated with growth and inflammation, ROS modulate intracellular free Ca2+ concentration ([Ca2+]i), a major determinant of vascular contraction/dilation. Superoxide and H2O2 increase [Ca2+]i in VSMCs and endothelial cells (Lounsbury et al. 2000). These effects have been attributed to redox-dependent inositol trisphosphate-induced Ca2+ mobilization, increased Ca2+ influx, and decreased Ca2+-ATPase activation (Lounsbury et al. 2000; Ermak and Davies 2001). Plasma membrane K+ channels in VSMCs controlling hyperpolarization-elicited relaxation are opened by mechanisms associated with thiol oxidation by ROS (Touyz and Schiffrin 2000; Droge 2001; Ermak and Davies 2001; Touyz et al. 2002b). These redox-regulated Ca2+ processes may be more important in stress responses than in receptor-mediated signaling by growth factors or cytokines and may play a role in altered vascular contractility in hypertension. Contractile responses to H2O2 are exaggerated in arteries from spontaneously hypertensive rats (SHR) compared with normotensive counterparts (Gao and Lee 2001), suggesting that in addition to impaired endothelium-dependent vasodilation (due to increased quenching of NO by ·O2−), redox-sensitive Ca2+ changes could contribute to altered vascular contractility in hypertension.

Vascular effects of ROS

Vascular growth and inflammation

Reactive oxygen species stimulate growth factor-like cellular responses, such as intracellular alkalinization, MAP kinase phosphorylation, tyrosine kinase activation, DNA synthesis, and increased expression of proto-oncogenes (Rao and Berk 1992; Droge 2001). During vascular damage in hypertension when oxidative stress is increased, redox-sensitive growth processes may lead to accelerated proliferation and hypertrophy, further contributing to vascular injury and remodeling (Rao and Berk 1992; Griendling et al. 2000a; Touyz 2003a; Fig. 4). In addition to growth-promoting actions, ROS induce apoptosis and differentiation under certain circumstances. This differential response appears to relate to the specific species generated, the concentration of ROS, and the cellular localization of ROS (Deshpande et al. 2002; Li et al. 2003b). At high concentrations (>100 μmol/l) H2O2 and peroxynitrite are pro-apoptotic and induce anoikis (cell detachment and shedding), whereas at lower concentrations they stimulate growth and differentiation (Deshpande et al. 2002; Li et al. 1999, 2003b).

Fig. 4
figure 4

Vascular effects of reactive oxygen species (ROS). Increased bioavailability of ROS influences cellular processes leading to vascular smooth muscle cell (VSMC) growth, inflammation, migration, and extracellular matrix (ECM) protein deposition as well as endothelial damage. MMP Matrix metalloproteinases

Full size image

Reactive oxygen species also modulate vascular structure in hypertension by increasing deposition of extracellular matrix proteins, such as collagen and fibronectin. Superoxide anion and H2O2 influence activity of vascular MMP2 and MMP9, which promote degradation of basement membrane and elastin, respectively (Rajagopalan et al. 1996b). Redox-sensitive inflammatory processes, including expression of proinflammatory molecules, such as MCP-1, osteopontin, and interleukin-6, expression of adhesion molecules, including vascular cell adhesion molecule-1 (VCAM-1) and intracellular adhesion molecule-1 (ICAM-1), lipid peroxidation, and cell migration, further contribute to vascular remodeling in hypertension (Muller et al. 2000; Luft 2001; Suematsu et al. 2002).

Oxygen radicals induce endothelial permeability with extravasation of plasma proteins and other macromolecules, and recruitment of inflammatory proteins and cells, which also impair endothelial function and aggravate vascular damage (Alexander 1995; Kristal et al. 1998). Peripheral polymorphonuclear leukocytes, which generate ·O2−, participate in oxidative stress and inflammation in patients with hypertension (Alexander 1995; Rajagopalan et al. 1996b; Kristal et al. 1998). The co-existence of an inflammatory reaction with oxidative stress induces endothelial dysfunction. Many of the redox-sensitive vascular changes that occur in hypertension also exist in atherosclerotic vessels.

Vascular contraction/dilation

Impaired endothelium-mediated vasodilation in hypertension and hypercholesterolemia has been linked to decreased NO bioavailability. This may be secondary to decreased synthesis of NO and/or to increased degradation of NO because of its interaction with ·O2− to form ONOO− (Tschudi et al. 1996; List et al. 1997; Somers and Harrison 1999). Peroxynitrite is a weak vasodilator compared with NO and has pro-inflammatory properties (Szabo 2003). In experimental models of hypertension, hypercholesterolemia, and diabetes and in hypertensive patients, endothelial function is improved by anti-oxidant vitamins, probucol, SOD, or sepiapterin (a stable precursor of BH4) (Virdis et al. 2003; Mitchell et al. 2004).

Vasomotor tone may also be modulated through direct ROS effects. Reactive oxygen species appear to elicit both contraction and dilation, depending on the vascular bed and type of species generated. Hydrogen peroxide causes vasodilation of pulmonary, coronary, and mesenteric arteries and has been considered to be an endothelium-derived relaxing factor (Somers and Harrison 1999; Yada et al. 2003). In rat aorta, Ang II stimulates vasoconstriction via H2O2-dependent mechanisms (Torrecillas et al. 2001), whereas in human and porcine vessels, acute vasoconstriction by Ang II is not mediated via ROS (Schuijt et al. 2003; Touyz 2003b). In aortic and mesenteric arteries from SHR, redox-mediated contractile effects are enhanced (Gao and Lee 2001). Major factors underlying these differential vascular responses to activated oxygen metabolites could relate to the blood vessel studied, the presence or absence of the endothelium, the concentration and species of free radical studied, and the compartment in which ·O2− or H2O2 predominate (Touyz 2003b). At present it is still unclear exactly what the functions of ·O2− and H2O2 are with respect to vascular contraction/dilation in physiological and pathophysiological conditions.

Reactive oxygen species in hypertension

Oxidative stress in genetic models of hypertension

Reactive oxygen species play an important pathophysiological role in hypertension. This is evidenced by findings that oxidative stress is increased in hypertension and that treatment with antioxidants or agents that inhibit NAD(P)H oxidase-driven generation of ROS reduces, and may even prevent, blood pressure elevation in hypertensive animals. Genetic models of hypertension, such as SHR (Zalba et al. 2000) and stroke-prone SHR (SHRSP) (Chen et al. 2001) exhibit enhanced NAD(P)H oxidase-mediated ·O2− generation in resistance arteries (mesenteric), conduit vessels (aorta), and kidneys (Zalba et al. 2000; Chen et al. 2001; Fig. 5). These processes are associated with increased expression of NAD(P)H oxidase subunits, particularly p22phox and p47phox, and increased activity of the enzyme (Lassegue and Clempus 2003). 8-Hydroxy-2′-deoxyguanosine, a marker for oxidative stress-induced DNA damage, and protein carbonylation, a marker for oxidation status of proteins that are enhanced in aorta, heart, and kidney, are markedly suppressed in SHR and SHRSP compared with normotensive Wistar Kyoto rats, as is the expression of the redox regulating protein thioredoxin (Tanito et al. 2004). Male SHR have a higher vascular ·O2− concentration than female counterparts, a phenomenon that has been linked to upregulation of AT1 receptors in male SHR arteries (Dantas et al. 2004). Several polymorphisms in the promoter region of the p22phox gene have been identified in SHR, which could contribute to enhanced NAD(P)H oxidase activity (Zalba et al. 2001b). These findings may have clinical relevance since an association between a p22phox gene polymorphism and NAD(P)H oxidase-mediated ·O2− production in the vascular wall of patients with atherosclerosis and hypertension has been described (Diez et al. 2003; Moreno et al. 2003). Increased expression of p47phox has been demonstrated in the renal vasculature, macula densa, and distal nephron from young SHR, suggesting that upregulation of renal NAD(P)H precedes development of hypertension (Chabrashvili et al. 2002). The importance of p47phox was demonstrated in p47phox−/− mice, which failed to develop hypertension in response to Ang II infusion (Landmesser et al. 2002). Diminished NO bioavailability as a consequence of enhanced vascular ·O2− generation may also contribute to oxidative stress in SHR and SHRSP. Treatment with antioxidant vitamins, NAD(P)H oxidase inhibitors, SOD mimetics, BH4, and AT1 receptor blockers decrease vascular ·O2− production and attenuate, to varying degrees, the development of hypertension in these genetic models of hypertension (Sharma et al. 1996; Schnackenberg et al. 1999; Chen et al. 2001; Hong et al. 2001). Lifelong treatment with antioxidants can even prevent development of hypertension in SHR (Zhan et al. 2004).

Fig. 5
figure 5

Detection of vascular superoxide by dihydroethidine fluorescence in hypertensive rats. Shown are baseline superoxide levels in aorta from normotensive control Wistar Kyoto rats (WKY) and stroke-prone spontaneously hypertensive rats (SHR-SP) treated or not with the NAD(P)H oxidase inhibitor, apocynin. The increase in superoxide involves all layers within the vessel wall. E Endothelium, M media, A adventitia

Full size image

Oxidative stress and experimentally induced hypertension

Oxidative excess has been demonstrated in various models of experimental hypertension, including Ang II-induced hypertension (Laursen et al. 1997; Virdis et al. 2004), Dahl-salt-sensitive hypertension (Tojo et al. 2002), lead-induced hypertension (Ding et al. 2001), obesity-associated hypertension (Dobrian et al. 2001), mineralocorticoid hypertension (Wu et al. 2001; Virdis et al. 2002), 2-kidney, 1-clip hypertension (Welch et al. 2003), and postmenopausal hypertension (Fortepiani et al. 2003; Table 2). Increased activation of vascular NAD(P)H oxidase (Lassegue and Clempus 2003) and xanthine oxidase (Fortepiani et al. 2003) and uncoupling of eNOS (Milstien and Katusik 1999; Cosentino et al. 2001; Vasquez-Vivar et al. 2002) have been implicated in enhanced ·O2− generation in experimental hypertension. Inhibition of ROS generation with apocynin or allopurinol and scavenging of free radicals with antioxidants or SOD mimetics decreases blood pressure and prevents development of hypertension in most models of experimental hypertension (Sharma et al. 1996; Chen et al. 2001; Frenoux et al. 2002; Park et al. 2002; Tanito et al. 2004). These beneficial effects have been attributed to improved endothelial function, vascular regression, and reduced vascular inflammation (Touyz 2000; Wilcox 2002). Interestingly, norepinephrine-induced hypertension is not associated with enhanced vascular oxidative stress and SOD does not decrease blood pressure in this model (Laursen et al. 1997). These findings suggest that blood pressure itself may not be the primary cause of oxidative excess in hypertension.

Table 2 Clinical and experimental forms of hypertension which exhibit oxidative stress. (↑ Increase, – no change, SHR spontaneously hypertensive rats, SHRSP stroke-prone SHR)
Full size table

Oxidative stress in human hypertension

Clinical studies have demonstrated that essential hypertensive patients produce excessive amounts of ROS (Prabha et al. 1990; Sagar et al. 1992; Lacy et al. 2000; Minuz et al. 2002; Stojiljkovic et al. 2002) and have decreased antioxidant capacity (Russo et al. 1998). In most of these studies, hypertensive patients are salt-sensitive and exhibit some degree of renal dysfunction (Manning et al. 2003). Oxidative stress has also been demonstrated in patients with renovascular hypertension (Higashi et al. 2002b), malignant hypertension (Lip et al. 2002), and in pre-eclampsia (Lee et al. 2003). Most of these findings are based on increased levels of plasma and urine TBARS and 8-epi-isoprostanes, systemic markers of lipid peroxidation and oxidative stress (Sagar et al. 1992; Minuz et al. 2002; Stojiljkovic et al. 2002). In never-treated mild-to-moderate hypertension lipid peroxidation is not increased (Cracowski et al. 2003), suggesting that oxidative stress may not be important in mild hypertension.

Decreased antioxidant activity and reduced levels of ROS scavengers such as vitamin E, glutathione, and SOD (Sagar et al. 1992) and increased activation of vascular NAD(P)H oxidase may contribute to oxidative excess in hypertensive patients (Berry et al. 2000; Bengtsson et al. 2003). Activation of the renin–angiotensin system has been proposed as a major mediator of NAD(P)H oxidase activation and ROS production in human hypertension (Touyz 2003a). In fact some of the therapeutic blood pressure-lowering effects of AT1 receptor blockers and ACE inhibitors have been attributed to inhibition of NAD(P)H oxidase activity and decreased ROS production (Ghiadoni et al. 2003). It has also been suggested that p22phox polymorphisms may play a role in altered NAD(P)H oxidase-generated ·O2− production in human cardiovascular disease (Schachinger et al. 2001; Zalba et al. 2001b; Moreno et al. 2003; Dantas et al. 2004). In particular the -930(A/G) polymorphism in the p22(phox) promoter may be a novel genetic marker associated with hypertension (Zalba et al. 2001b). However, to confirm that these polymorphisms are indeed markers for hypertension, studies in large populations are necessary.

Based on experimental evidence and clinical studies that oxidative stress plays a key role in vascular damage, there has been great interest in developing strategies that target ROS in the treatment of hypertension and other cardiovascular diseases. Therapeutic approaches that have been considered include mechanisms to increase antioxidant bioavailability through diet or supplementation and/or to reduce generation of ROS by decreasing activity of ·O2−-generating enzymes and by increasing levels of BH4 (Brown and Hu 2001; Cai et al. 2003). Findings from clinical trials have been conflicting. Until definitive data become available, antioxidants should not be recommended in the prevention and management of hypertension (Galley et al. 1997; Taddei et al. 1998; Chappell et al. 1999; HOPE Investigators 2000; Brown and Hu 2001; Digiesi et al. 2001; Khaw et al. 2001; Kim et al. 2002; Wu et al. 2002; Vivekananthan et al. 2003).

Conclusions

Reactive oxygen species are produced in the vessel wall in a controlled and tightly regulated manner. Superoxide and H2O2 have important signaling properties, mainly through oxidative modification of proteins and activation of transcription factors that maintain vascular function and structure. In hypertension, dysregulation of enzymes such as NAD(P)H oxidase, NOS, xanthine oxidase, mitochondrial enzymes, or SOD that generate ·O2−, H2O2, and ·OH, altered thioredoxin and glutathione systems, or reduced scavenging by anti-oxidants, results in increased formation of ROS, which has damaging actions on the vasculature. Reactive oxygen species in hypertension contribute to vascular injury by promoting VSMC growth, extracellular matrix protein deposition, activation of matrix metalloproteinases, inflammation, endothelial dysfunction, and increased vascular tone. In experimental hypertension oxidative stress is increased. Clinical data suggest that hypertensive patients, especially those with severe hypertension, salt-sensitive hypertension, and renovascular hypertension, exhibit oxidative excess. Although inconclusive at present, treatment strategies to alter ROS bioavailability by decreasing production and/or by increasing radical scavenging, may regress vascular remodeling, prevent further vascular injury, and reduce blood pressure and associated target organ damage in hypertensive patients. With greater insights and understanding of processes regulating vascular ROS metabolism and identification of molecular pathways that tip the equilibrium to states of oxidative stress which cause vascular damage, it may be possible to target therapies more effectively so that detrimental actions of vascular oxygen free radicals can be reduced and beneficial effects of NO· can be enhanced. Such therapies may be useful in the treatment of hypertension and in the prevention of target-organ damage.