Kidney International, Vol. 63, Supplement 84 (2003), pp. S45–S49

Endothelial dysfunction and inflammation: What is the link?

JAN GALLE, THOMAS QUASCHNING, STEFAN SEIBOLD, and CHRISTOPH WANNER

Department of Medicine, Division of Nephrology, Julius-Maximilians-University, Wurzburg, Germany

Endothelial dysfunction and inflammation: What is the link? cause of death in patients with ESRD [3]. Arteriosclero-Cardiovascular disease, resulting from arteriosclerotic remod- sis is considered to be an inflammatory disease associatedeling of the vasculature, is the main cause of death in end-

with enhanced oxygen radical formation [4, 5]; enhancedstage renal disease (ESRD) patients. Early during the courseoxidative stress is likely to be a major cause for endothe-of arteriosclerosis, endothelial dysfunction can be detected in

various vascular beds, including peripheral forearm arteries, lial dysfunction. It is the aim of this article to highlight theas well as the coronary circulation. Furthermore, endothelial interplay of inflammation, endothelial dysfunction, anddysfunction seems to predict the prognosis of cardiovascular

oxidative stress. Particular emphasis is placed on the roledisease. Therefore, the question deserves attention whetherof angiotensin II (AngII) and atherogenic lipoproteins.endothelial dysfunction is simply a marker of cardiovascular

disease, or an active player in the progress of the disease. Apossible link between arteriosclerosis, endothelial dysfunction,and cardiovascular disease is increased oxidative stress. In- ENDOTHELIAL DYSFUNCTIONflammatory processes involved in the pathogenesis of arterio-sclerosis enhance vascular O2 formation, leading to endothelial Endothelial dysfunction is a sensitive indicator for car-dysfunction. An activated renin angiotensin system, together diovascular disease [6], predicts its prognosis [7], and iswith oxidized low-density lipoprotein, may play a prominent closely associated with the development of arteriosclero-role for enhanced vascular oxidative stress. In this context, the

sis [8]. The term endothelial dysfunction could, of course,endothelium is not only a target of oxygen radicals, but mayalso contribute to O2 formation. It is the aim of this article to implicate a “loss of function” of any of the numeroushighlight the interplay of inflammation, endothelial dysfunc- activities of the endothelium. In the context of vasculartion, and oxidative stress.

diseases, however, endothelial dysfunction usually de-scribes reduced dilatory capacities, particularly reducedNO activity. Probably the most important mechanism

Doubtlessly, one of the major achievements in medi-leading to reduced NO activity is enhanced oxygen radicalcine during the last two decades has been the discoveryformation. Superoxide anion (O�

2 ) scavenges NO, yieldingthat the vascular endothelium is not simply a semiperme-peroxynitrite (ONOO�), which is rather stable but canable membrane, but instead forms a most active organrearrange to form nitrate and the highly reactive OH•with endocrine and paracrine functions [1]. Concerning[9]. Many studies clearly indicate that endothelial dys-cardiovascular medicine, most relevant functions of thefunction is closely correlated with arteriosclerosis, andendothelium are the release of relaxing factors, contrac-that O�

2 formation is enhanced in hypercholesterolemiatile factors, and of antiaggregatory substances. Promi-and arteriosclerosis [10]. Indeed, endothelial dysfunctionnent representatives of relaxing factors are nitric oxidecan be detected (e.g., by means of reduced forearm blood(NO) and prostacycline (PGI2), both of which also have

antiaggregatory properties on platelets. Endothelium- flow, early in the development of arteriosclerosis, andderived hyperpolarizing factor (EDHF) is another vaso- in patients with hypercholesterolemia) [11]. Importantly,dilator released from the endothelium. Endothelin-1 endothelial dysfunction occurs in arteries without visible(ET-1) and thromboxane (TXA2) are the best-character- characteristics of arteriosclerotic lesions such as fattyized endothelial contractile factors so far. Endothelial streaks or plaque formation.function, with its relevance to cardiovascular medicine, In view of the numerous endocrine and paracrine func-has frequently been reviewed [2]. tions of the endothelium, the question deserves attention

Arteriosclerotic cardiovascular disease is the main whether endothelial dysfunction is simply a sensitive in-dicator for cardiovascular disease, or whether it is anactive player in its pathophysiology. To shed some lightKey words: atherosclerosis, inflammation, oxidative stress, antioxi-

dants, oxidized LDL. on these questions, the development of arteriosclerosisshall briefly be summarized. 2003 by the International Society of Nephrology

S-45

Galle, Quaschning, Seibold, and Wanner: Endothelial dysfunction and inflammationS-46

PATHOPHYSIOLOGY OF ARTERIOSCLEROSIS

Arteriosclerosis, a generic term for several diseasesin which the arterial wall becomes thickened and loseselasticity, is not a static condition; rather, it reflects acontinuous development over decades from macroscopi-cally intact arteries to grossly damaged and rupturedsclerotic plaques, with different stages being present si-multaneously within one individual. Important stepsduring atherogenesis include enhanced endothelial per-meability, expression of adhesion molecules, monocyteadhesion and immigration, foam cell formation, fattystreaks, smooth muscle cell migration, plaque formation,and finally plaque rupture and thrombus formation. Ear-lier concepts on the pathophysiology of arteriosclerosiswere based on the assumption of an initial injury leading

Fig. 1. Endogenous and exogenous factors contributing to vascularto intimal lesions, and subsequent repair mechanisms inflammation, oxidative stress, endothelial dysfunction, and vascular(“response to injury” hypothesis) [12]. Today we know remodeling, with particular relevance to patients with end stage renal

disease. Abbreviations are: AngII, angiotensin II; OxLDL, oxidizedthat even in macroscopically intact arteries, processesLDL; LPS, lipopolysaccharide; AGEs, advanced glycation end products;

take place that can be best defined as chronically in- CVC, central venous catheter.flammatory [4]. Arguments that support the hypothesisof a chronic inflammation in arteriosclerosis include thefollowing: Cells found in early atherosclerotic lesions are

nary endothelial dysfunction is a strong predictor fortypically inflammatory cells (monocytes/macrophagesfuture cardiovascular events in patients with coronaryand T-lymphocytes); early in arteriosclerosis, enhancedheart disease [7].vascular oxygen radical formation can be detected, as

Thus, there is strong evidence that arteriosclerosis is anwell as enhanced activity of lipoxygenases; arteriosclero-inflammatory disease, and that endothelial dysfunction issis is associated with enhanced serum levels of inflam-an early event in this process. How could inflammationmation parameters; the arteriosclerotic artery producescontribute to endothelial dysfunction, and how could thedifferent hydrolytic enzymes, adhesion molecules, cyto-endothelium contribute to inflammation?kines, and growth factors, as seen in chronic inflamma-

tion. In addition, large clinical studies clearly show astrong correlation between markers of inflammation and PATHOPHYSIOLOGY OF INFLAMMATIONclinical outcome in patients with coronary heart disease. AND CONSEQUENCES OF ENHANCEDIn particular, high sensitive C-reactive protein (hs-CRP) OXIDATIVE STRESS FOR VASCULARhas been identified as a predictor for cardiovascular AND RENAL FUNCTIONevents in patients with and without known coronary Certainly, many different factors may contribute toheart disease [13, 14]. the state of inflammation in the vasculature, as summa-

As mentioned above, at very early stages of arterio- rized in Figure 1. For patients with ESRD, it makessclerosis (characterized by enhanced endothelial perme- sense to distinguish between endogenous and exogenousability, expression of adhesion molecules, monocyte ad- factors, because the dialysis treatment, in addition tohesion, and immigration), endothelial dysfunction can uremia, may be a source for continuous exposure to pro-be detected. For example, in one study, the intima-media inflammatory stimuli. What these different factors havethickness and plaque formation of the common carotid in common is that they all—albeit by different mecha-artery of 34 men with arteriosclerosis was correlated with nisms—enhance oxidative stress. Endothelial dysfunc-flow mediated dilation (FMD) of the brachial artery. A tion through scavenging of NO by O�

2 is one importantsignificant negative correlation of FMD with the intima- consequence of enhanced oxidative stress, as outlinedmedia thickness of the common carotid artery was found, above. In various vascular or renal diseases, enhancedsupporting the concept that endothelial dysfunction is formation of reactive oxygen species is considered to besignificantly related to atherogenesis [8]. In another pathogenic (e.g., in atherosclerosis, glomerular diseases,study, a close relation between coronary artery endothe- renal failure, pyelonephritis, or aminoglycoside ne-lium-dependent vasomotor responses to acetylcholine phropathy) [15, 16]. Furthermore, oxidative stress impor-and FMD in the brachial artery was found [6]. The rele- tantly influences vascular cell cycle decisions; inductionvance of endothelial dysfunction for the prognosis of of apoptosis [17, 18], as well as induction of cell prolifera-

tion [19], has been described.patients was shown in a study of the Mayo Clinic: coro-

Galle, Quaschning, Seibold, and Wanner: Endothelial dysfunction and inflammation S-47

The scope of this article would be disrupted if all the AngII-induced stimulation of O�2 formation could be

inhibited by diphenylene iodinium, suggesting that O�2influences listed in Figure 1 which contribute to inflam-

was produced by membrane-bound NAD(P)H oxidasesmation and oxidative stress should be discussed in detail.[26]. In the mean time, it has been shown that AngIIInstead, we will focus on two factors, AngII and oxidizedalso induces O�

2 formation in endothelial cells and non-low density lipoprotein (OxLDL), for two reasons: First,vascular tissue [28]. In numerous studies, it has beenAngII and OxLDL are well characterized as potentiallydemonstrated that AngII stimulates vascular NAD(P)Hendothelium-damaging agents, and, second, we possessoxidase-dependent O�

2 formation via the AT1-receptor.pharmacologic tools to interfere with their activity.The NAD(P)H oxidase consists of four major subunits:a plasma membrane spanning cytochrome b558 (com-

ANGII AND OXIDIZED LDL posed of the large subunit gp91phox and the small sub-unit p22phox), and 2 cytosolic components, p47phox andAccumulation of OxLDL in atherosclerotic plaques isp67phox [29]. In smooth muscle cells, the NADPH sub-a well-known event in the development of atherosclero-units p22phox, p47phox, and eventually p67phox, aresis [20]. Only recently did it become apparent that ath-involved in O�

2 formation [29]. In endothelial cells,erosclerotic arteries (human atherectomy preparationsmRNAs for gp91phox, p22phox, p47phox, and p67phoxand arteries of hypercholesterolemic monkeys) show en-have been detected, and the gp91phox [30] and p22phoxrichment with AngII, co-localizing with resident macro-[19] subunits seem to be of particular importance forphages [21]. Thus, AngII accumulates in the same vascu-O�

2 formation in endothelial cells.lar region as OxLDL. Not only do AngII and OxLDLco-localize in atherosclerotic plaques, there is strong ex-perimental evidence that these agents interact with each SPECIFIC EFFECTS OF ANGII-INDUCED

OXYGEN RADICAL FORMATIONother, with relevance for vascular biology and athero-sclerosis. The particular importance of AngII-induced oxidative

stress for vascular biology has been investigated in thecontext of vasomotor tone and of cell cycle regulationEVIDENCE FOR A POTENTIAL INTERACTION(e.g., enhanced O�

2 formation resulted in impairment ofBETWEEN ANGII AND OXIDIZED LDLendothelium-dependent dilations) [31], which could be

Clinical studies suggest that ACE-inhibitors are of par- prevented by liposome-encapsulated superoxide dismu-ticular benefit for endothelial function in hypercholester- tase. Indirect evidence that AngII-induced O�

2 formationolemic patients with high LDL levels [22, 23]. Further- takes place in vivo in humans was provided by a studymore, several studies have shown that the expression of using the forearm plethysmography method, which al-the OxLDL receptor LOX-1, and of the AT1 receptor, lows direct measurement of AngII-induced vasomotoris stimulated by the respective other receptor agonist actions. Constrictor actions of AngII in the human fore-[24, 25]. In cultured smooth muscle cells, LDL induced arm were enhanced during NO inhibition and were at-

tenuated during vitamin C infusion, suggesting AngII-expression of the AT1 receptor [24]. Thus, LDL mayassociated stimulation of endothelial NO and of oxygensensitize the vascular tissue to AngII. On the other hand,radicals, respectively [32].expression of the OxLDL receptor LOX-1 and uptake

The impact of AngII on cell cycle decisions via activa-of OxLDL in human umbilical vein endothelial cellstion of NAD(P)H-dependent oxidases has been demon-(HUVEC) is increased by AngII [25]. Together, thesestrated in vascular and nonvascular tissue, and may bestudies imply that AngII and OxLDL amplify the effectof particular importance for cellular turnover in arterio-of the respective other agonist. But how could AngIIsclerosis. For example, AngII causes vascular smooth mus-and OxLDL contribute to arteriosclerosis, inflammation,cle cell hypertrophy, and inhibition of p22phox mRNAand endothelial dysfunction?expression in vascular smooth muscle cells results in asignificant inhibition of AngII-stimulated NADH(P)H-dependent superoxide production, subsequent hydrogenEXPERIMENTAL EVIDENCE FOR PRO-peroxide production, and H3-leucine incorporation [33].INFLAMMATORY AND PRO-OXIDATIVE

Thus, AngII-induced O�2 formation has important con-EFFECTS OF ANGII

sequences for the NO metabolism and for cell cycle deci-Recently, it has become apparent that AngII is a po-sions, potentially affecting the course of arteriosclerosis.tent stimulator of vascular oxygen radical production,

thereby contributing to endothelial dysfunction and in-OXIDATIVE STRESS INDUCED BY OXLDLflammation [26, 27]. Cell culture studies with rat smoothAND FUNCTIONAL CONSEQUENCESmuscle cell preparations provided first experimental evi-

dence for stimulation of O�2 formation by AngII [26]. In Animal studies with cholesterol-fed rabbits provided

first the indirect evidence for a role of LDL in the induc-experiments with rat smooth muscle cell membranes,

Galle, Quaschning, Seibold, and Wanner: Endothelial dysfunction and inflammationS-48

tion of oxidative stress. Aortas from hypercholesterol- ACKNOWLEDGMENTSemic rabbits produced significantly more superoxide than This work was supported by a grant from The Deutsche Forschungs-

gemeinschaft (DFG, SFB 355 B11).control aortas [10]. Later, our group was able to showdirectly that incubation of cultured human umbilical

Reprint requests to Dr. Jan Galle, Department of Medicine, Divisionvein endothelial cells (HUVEC), and of isolated arteries of Nephrology, University Hospital Wurzburg, Josef-Schneider-Str. 2,

D-97080 Wurzburg, Germany.with oxidized LDL or Lp(a), stimulated O�2 formation

E-mail: J-C.Galle@mail.uni-wuerzburg.de[34]. Functional consequences of OxLDL-induced oxida-tive stress extend to atherosclerosis, vasomotor regula-

REFERENCEStion, and endothelial dysfunction. OxLDL affects endo-

1. Inagami T, Naruse M, Hoover R: Endothelium as an endocrinethelial function and impairs endothelium-dependentorgan. Ann Rev Physiol 57:171–189, 1995

dilations [34]. The impact of OxLDL on apoptotic cell 2. Ignarro LJ: Physiology and pathophysiology of nitric oxide. Kid-ney Int 49(Suppl. 55):S2–S5, 1996death may be a clue to its role in the development of

3. Foley RN, Parfrey PS, Sarnak MJ: Epidemiology of cardio-atherosclerosis. OxLDL induces apoptosis in HUVEC vascular disease in chronic renal disease. J Am Soc Nephrol 9:S16–[17, 18, 35] and in smooth muscle cells of isolated aortas S23, 1998

4. Ross R: Mechanisms of disease. Atherosclerosis: An inflammatory[17]. In all the cited studies, the use of antioxidants (SOD,disease. N Engl J Med 340:115–126, 1999

vitamin C/E, or butylated hydroxytoluene) prevented 5. Lusis AJ: Atherosclerosis. Nat 407:233–241, 20006. Anderson TJ, Uehata A, Gerhard MD, et al: Close relation ofthe induction of apoptosis. Another common effect of

endothelial function in the human coronary and peripheral circula-OxLDL and AngII is the stimulation of endothelial celltions. J Am Coll Cardiol 26:1235–1241, 1995

proliferation [19]. 7. Suwaidi JA, Hamasaki S, Higano ST, et al: Long-term follow-up of patients with mild coronary artery disease and endothelialdysfunction. Circulation 101:948–954, 2000

8. Hashimoto M, Eto M, Akishita M, et al: Correlation betweenCLINICAL RELEVANCE flow-mediated vasodilatation of the brachial artery and intima-

media thickness in the carotid artery in men. Arterioscler ThrombThe latest argument for the relevance of AngII-inducedVasc Biol 19:2795–2800, 1999

oxidative stress for endothelial dysfunction is provided by 9. Halliwell B, Zhao K, Whiteman M: Nitric oxide and peroxyni-trite. The ugly, the uglier and the not so good: A personal viewa study in patients with renovascular hypertension [36].of recent controversies. Free Radic Res 31:651–669, 1999In such patients with activated renin-angiotensin system 10. Ohara Y, Peterson TE, Harrison DG: Hypercholesterolemia

and endothelial dysfunction, the response of forearm increases endothelial superoxide anion production. J Clin Invest91:2546–2551, 1993blood flow to acetylcholine, an endothelium-dependent

11. Zeiher AM, Drexler H, Wollschlager H, Just H: Endothelialvasodilator, was evaluated before and after renal-artery dysfunction of the coronary microvasculature is associated with

impaired coronary blood flow regulation in patients with earlyangioplasty. The forearm blood flow in response to ace-atherosclerosis. Circulation 84:1984–1992, 1991tylcholine was less in subjects with renovascular hyper-

12. Ross R, Glomset JA: Atherosclerosis and the arterial smooth mus-tension than in control subjects. After angioplasty, the cle cell: Proliferation of smooth muscle is a key event in the genesis

of the lesions of atherosclerosis. Science 180:1332–1339, 1973forearm blood flow in response to acetylcholine was in-13. Ridker PM, Cushman M, Stampfer MJ, et al: Inflammation, aspi-creased in the patients with renovascular hypertension. rin, and the risk of cardiovascular disease in apparently healthy

Interestingly, angioplasty decreased serum malondialde- men. N Engl J Med 336:973–979, 199714. Harb TS, Zareba W, Moss AJ, et al: Association of C-reactivehyde-modified LDL, an index of oxidative stress that

protein and serum amyloid A with recurrent coronary events inwas positively correlated with endothelial dysfunction. stable patients after healing of acute myocardial infarction. Am J

Cardiol 89:216–221, 2002The authors conclude that excessive oxidative stress is15. Klahr S: Oxygen radicals and renal diseases. Miner Electrolyteinvolved in endothelial dysfunction in patients with reno- Metab 23:140–143, 1997

vascular hypertension. 16. Gwinner W, Landmesser U, Brandes RP, et al: Reactive oxygenspecies and antioxidant defense in puromycin aminonucleosideglomerulopathy. J Am Soc Nephrol 8:1722–1731, 1997

17. Galle J, Schneider R, Heinloth A, et al: Lp(a) and LDL induceSUMMARY apoptosis in human endothelial cells and in rabbit aorta: Role of

oxidative stress. Kidney Int 55:1450–1461, 1999In early stages of arteriosclerosis, enhanced oxygen 18. Dimmeler S, Haendeler J, Galle J, Zeiher AM: Oxidized lowradical formation already leads to endothelial dysfunc- density lipoprotein induces apoptosis of human endothelial cells

by activation of CPP32-like proteases: A mechanistic clue to thetion. The endothelium is a target of oxidative stress by“response to injury” hypothesis. Circulation 95:1760–1763, 1997

means of attenuated endothelium-dependent dilation, 19. Heinloth A, Heermeier K, Raff U, et al: Stimulation of NADPHoxidase by oxidized LDL induces proliferation of human vascularenhanced cellular turnover, and cell death. However, theendothelial cells. J Am Soc Nephrol 11:1819–1825, 2000endothelium also contributes to the vascular state of

20. Yla-Herttuala S, Palinski W, Butler SW, et al: Rabbit andinflammation because it is one source of O�

2 formation. human atherosclerotic lesions contain IgG that recognizes epitopesof oxidized LDL. Arterioscler Thromb 14:32–40, 1994Beside other factors, AngII and OxLDL are important

21. Potter DD, Sobey CG, Tompkins PK, et al: Evidence that macro-stimuli for vascular oxygen radical formation, endothe- phages in atherosclerotic lesions contain angiotensin II. Circulation98:800–807, 1998lial dysfunction, and cell proliferation.

Galle, Quaschning, Seibold, and Wanner: Endothelial dysfunction and inflammation S-49

22. Mancini GBJ, Henry GC, Macaya C, et al: Angiotensin-con- 30. Gorlach A, Brandes RP, Nguyen K, et al: A gp91phox containingverting enzyme inhibition with quinapril improves endothelial NADPH oxidase selectively expressed in endothelial cells is avasomotor dysfunction in patients with coronary artery disease: major source of oxygen radical generation in the arterial wall. CircThe TREND (Trial on Reversing ENdothelial Dysfunction) study. Res 87:26–32, 2000Circulation 94:258–265, 1996 31. Laursen JB, Rajagopalan S, Galis Z, et al: Role of superoxide

23. Cashin Hemphill L, Holmvang G, Chan RC, et al: Angiotensin- in angiotensin II-induced but not catecholamine-induced hyperten-converting enzyme inhibition as antiatherosclerotic therapy: No sion. Circulation 95:588–593, 1997answer yet. QUIET Investigators. QUinapril Ischemic Event Trial. 32. Dijkhorst-Oei LT, Stroes ESG, Koomans HA, Rabelink TJ:Am J Cardiol 83:43–47, 1999 Acute simultaneous stimulation of nitric oxide and oxygen radicals

24. Nickenig G, Sachinidis A, Michaelsen F, et al: Upregulation of by angiotensin II in humans in vivo. J Cardiovasc Pharmacol 33:vascular angiotensin II receptor gene expression by low-density lipo- 420–424, 1999protein in vascular smooth muscle cells. Circulation 95:473–478, 1997 33. Ushio Fukai M, Zafari AM, Fukui T, et al: p22phox is a critical

25. Morawietz H, Rueckschloss U, Niemann B, et al: Angiotensin component of the superoxide-generating NADH/NADPH oxidaseII induces LOX-1, the human endothelial receptor for oxidized system and regulates angiotensin II-induced hypertrophy in vascu-low density lipoprotein. Circulation 100:899–902, 1999 lar smooth muscle cells. J Biol Chem 271:23317–23321, 199626. Griendling KK, Minieri CA, Ollerenshaw JD, Alexander RW:

34. Galle J, Bengen J, Schollmeyer P, Wanner C: Impairment ofAngiotensin II stimulates NADH and NADPH oxidase activity inendothelium-dependent dilation in rabbit renal arteries by oxidizedcultured vascular smooth muscle cells. Circ Res 74:1141–1148, 1994lipoprotein(a): Role of oxygen-derived radicals. Circulation 92:27. Sohn H-Y, Raff U, Hoffmann A, et al: Differential role of Angio-1582–1589, 1995tensin II receptor subtypes on endothelial O�

2 formation. Br J Phar-35. Harada-Shiba M, Kinoshita M, Kamido H, Shimokado K: Oxi-macol 131:667–672, 2000

dized low density lipoprotein induces apoptosis in cultured human28. Hannken T, Schroeder R, Zahner G, et al: Reactive oxygenumbilical vein endothelial cells by common and unique mecha-species stimulate p44/42 mitogen-activated protein kinase and in-nisms. J Biol Chem 273:9681–9687, 1998duce p27Kip1: Role in angiotensin II-mediated hypertrophy of proxi-

36. Higashi Y, Sasaki S, Nakagawa K, et al: Endothelial functionmal tubular cells. J Am Soc Nephrol 11:1387–1397, 2000and oxidative stress in renovascular hypertension. N Engl J Med29. Griendling KK, Sorescu D, Ushio-Fukai M: NAD(P)H oxidase:

Role in cardiovascular biology and disease. Circ Res 86:494–501, 2000 346:1954–1962, 2002