Cell adhesion molecules in hypertension: endothelial markers of vascular injury and predictors of target organ damage?

In health, the vascular endothelium forms a multifunctional interface between the circulating blood and various tissues and organs of the body. It functions as a selectively permeable barrier, and also is a source of a host of biologically active substances, including cytokines, growth factors, coagulation and fibrinolytic factors, eicosanoids and nitric oxide. This single-cell-thick lining of the circulatory system is a vital ‘organ', whose health is essential to normal cardiovascular physiology and whose dysfunction can be a critical factor in the pathogenesis of cardiovascular disease [1]. The vascular endothelium is a dynamically mutable interface, whose structural and functional properties are responsive to a variety of stimuli, both local and systemic, and whose phenotypic modulation to a dysfunctional state can constitute a pathogenic risk factor for vascular disease. Thus, the adherence and subsequent transendothelial migration of circulating leukocytes into the intima is an important early mechanism in the initiation and progression of atherosclerosis [2–4]. For example, it has been observed [1] that treatment of cultured human endothelial cells with proinflammatory cytokines, such as interleukin 1 (IL-1) and tumour necrosis factor-α (TNF-α), or endotoxins (from Gram-negative bacteria), results in a dramatic alteration in many functional properties of endothelial cells, including their non-thrombogenic surface and low adhesiveness for blood leukocytes. The molecular mechanisms of this cytokine activation of endothelial cells involved ‘endothelial-leukocyte adhesion molecules’ (ELAM), including E-selectin (ELAM-1) and ‘ATHERO-ELAM’ (later designated vascular adhesion molecule 1, VCAM-1), which constituted the first examples of inducible endothelial cell-expressed adhesion molecules, and whose interactions with counterligands on various types of leukocytes could mediate their selective recruitment at sites of acute and chronic inflammation. Other manifestations of endothelial cell activation by humoral factors include alterations in various parameters of the ‘pathophysiological balances’ that regulate smooth muscle tone and growth, procoagulant/fibrinolytic activities and chemoattractant and inflammatory cytokine expression [5]. Thus, leukocyte adhesion to the intimal surface is mediated by the expression of leukocyte-selective adhesion molecules such as intercellular adhesion molecule 1 (ICAM-1), VCAM-1 and E-selectin [6]. ICAM-1 is expressed constitutively, but its expression is extremely low under normal conditions, whereas the expression of VCAM-1 and E-selectin is inducible, so that when the endothelium is activated in response to acute and chronic inflammatory diseases [7], cytokines [8] and developing atherosclerotic lesions [8,9], the expression of all these cell surface adhesion molecules (CAMs) is markedly increased on the surface of the vascular endothelium [8–10]. Hypertension is a well-established risk factor for atherosclerotic disease [11,12]. However, the precise mechanisms by which hypertension promotes atherosclerosis are unclear. It has been suggested that a common feature of all major risk factors for atherosclerosis, including hypertension, is the stimulation of leukocyte chemotaxis and adhesion to endothelial cells [13]. For example, Gimbrone et al. [14] have shown that when cultured endothelial cells are exposed to physiological levels of fluid mechanical forces in specially designed flow chambers in vitro, striking changes in their shape, alignment and cytoskeletal architecture are observed, as well as more subtle changes in membrane deformability and cell cycle rate. In addition, exposure of endothelial cells to well-defined fluid shear stresses, similar to those generated by the flow of blood over the endothelial cell lining in vivo, results in the enhanced production of a diverse array of biologically potent mediators, including growth factors, coagulation/fibrinolytic factors, cytokines and prostaglandins. Many of these changes appear to involve gene regulation at the transcriptional level, and thus are analogous to endothelial cell activation by cytokines and other humoral factors. Examination of this ‘biomechanical paradigm’ of endothelial cell activation has led to the definition of a ‘shear-stress response element’ in the promoters of various shear-inducible endothelial cell genes and certain of the corresponding transactivating factors, such as NF-κB, Egr-1 and c-fos, that play a role in regulating their transcription [15]. Thus, shear stress could be involved, at least partly, in the expression of CAMs during essential hypertension with or without associated risk factors. Furthermore, in response to hypertension-related increases in haemodynamic forces and tissue damage, the cells in the vessel wall become activated and release cytokines [16], such as IL-1 and TNF, which are known to upregulate the expression of ICAM-1 and VCAM-1 on endothelial cells, smooth muscle cells, leucocytes and in atherosclerotic lesions [7–9]. Until recently, little information was available regarding the effect of essential or secondary hypertension on the expression of CAMs. However, the presence of circulating forms of these molecules has made their clinical investigation possible. Indeed, soluble isoforms of ICAM-1 (sICAM-1), VCAM-1 (sVCAM-1) and E- selectin (sE-selectin) are present in the circulation and have been reported to be elevated in a number of pathological conditions, including idiopathic fibrosis [6], diabetes [17], dyslipidaemia [18], acute ischaemic stroke [19] and atherosclerosis [20,21]. The appearance of soluble CAMs (sCAMs) in the circulation is thought to result from a proteolytic process which leads to the release of CAMs by activated endothelial cells simultaneously with the expression of their membrane-associated forms [7,22]. Considering that the expression of CAMs is tightly regulated in order to avoid uncontrolled cell emigration [7], the level of sCAMs in the circulation may reflect endothelial inflammation/activation and increased expression on the cell surface [18,22,23]. A few years ago, several authors observed an increase in the endothelial expression of CAMs, in several experimental models of hypertension. For example, as early as 1996, using spontaneously hypertensive rats (SHR) versus two normotensive rat strains as controls, it was demonstrated that hypertension is associated with an elevation of ICAM-1 in the carotid artery from stroke-prone SHR, and with an increased endothelial adhesion and accumulation of monocytes/macrophages (MΦ) via cytokine production. This interaction between MΦ and endothelial activation could underlie the increased risk of stroke in hypertension [24]. Simultaneously, Mai et al. [25] demonstrated, in two-kidney one-clip renovascular hypertensive rats which present hypertensive renal injury, that the development of hypertension resulted in a progressive increase in ICAM-1 expression in the perivascular and interstitial areas of the renal cortex and on proximal tubular brush borders. Subsequently, hypertension and kidney damage were studied with the double transgenic rat (dTRG) harboring both human renin and human angiotensinogen genes, a model in which hypertension is dependent on the human components of the renin–angiotensin system (RAS) [26]. These authors found significant MΦ infiltration in the renal perivascular space and an increased expression of ICAM-1 and VCAM-1 in the interstitium, intima and adventitia of the small renal vessels, and of their corresponding ligands, αLβ1 and α4β2 integrins, on the infiltrating MΦ. The blockade of the RAS by antihypertensive molecules (an angiotensin-converting enzyme inhibitor and an AT1 receptor blocker) prevented the overexpression of CAMs, monocyte recruitment, the inflammatory response and the development of albuminuria in the dTRG. Likewise, Tummala et al. [27] suggest a potential link between the RAS and atherosclerosis. Indeed, in angiotensin II-induced hypertension, a hypertension model contributing to atherosclerosis, the authors showed an increase in aortic VCAM-1 and mRNA expression in rats, whereas norepinephrine, infused at a dose which induced similar hypertensive responses, had no effect. They have demonstrated that angiotensin II per se may contribute to atherogenesis by increasing expression of vascular inflammatory genes and, moreover, that this expression is mediated by transactivation of an NF-κB-driven VCAM-1 promoter. Furthermore, in the dTGR, a model of angiotensin II-induced hypertension known to induce impairment of renal function [28], Muller et al. [29] described an inflammatory process followed by damage to the small renal vessels mediated by an NF-κB-induced endothelial cell activation. Their study reported that the antioxidant pyrrolidine dithiocarbamate, an inhibitor of the NF-κB DNA binding activity, reduces inflammation and protects against angiotensin II-induced renal and cardiac end-organ damage. Concerning hypertensive patients, two questions may be asked: (i) do the sCAMs and soluble selectins [and eventually other soluble endothelial activation-related markers, such as von Willebrand factor (vWF) and endothelin-1] present a link with hypertension? (ii) Are these markers, which are predictors of the associated inflammatory risks for target organs, mainly in the cases of extreme elevation of blood pressure, accompanied by other risk factors? Concerning the first question, several authors have shown that circulating concentrations of sCAMs were increased in primary hypertension [30–36]. Particularly DeSouza et al. [32] demonstrated, in a well-conducted study, that: (i) in the overall population (normotensive plus hypertensive groups), both the sVCAM-1 and sICAM-1 correlated significantly with systolic and diastolic blood pressure; (ii) the systolic blood pressure was more likely to be involved in hypertension-related endothelial damage and dysfunction than diastolic blood pressure; (iii) the older sedentary patients with mild and uncomplicated hypertension (blood pressure less than 180/110 mmHg) had significantly higher circulating levels of the two sCAMs compared to their normotensive counterparts, providing a mechanistic link between chronically elevated blood pressure, endothelial cell adhesiveness and the development of cardiovascular disease; (iv) it is possible that the level of hypertension in this study may not have been severe enough, or that other risk factors such as dyslipidemia may have accompanied hypertension and would thus have contributed to the increase in E-selectin expression. In answer to this last point, in this issue of the journal, Preston et al. [37] studied the link between the risk factors associated with severe hypertension and the endothelial activation-related markers in three groups of patients: (i) subjects with severe uncontrolled hypertension (SHT group) showing a diastolic blood pressure higher than 120 mmHg, and/or a systolic blood pressure higher than 220 mmHg; (ii) subjects with mild hypertension (MHT group) with a diastolic blood pressure of 95–100 mmHg; (iii) normotensive control subjects (NT group) with a blood pressure of 140/90 mmHg. They showed that (i) the circulating level of sCAMs and vWF are correlated with systolic and diastolic blood pressure only when MHT and NT subjects were considered together as one group, confirming the previous observations [32]; (ii) there was no correlation between the concentrations of these soluble markers and blood pressure levels either when the three blood pressure groups were considered together, or when the SHT group was considered separately; (iii) the plasma sCAMs and vWF levels were significantly greater in the SHT and MHT group than in the NT control group, but they did not differ between the SHT and MHT group. This demonstrated that, while endothelial markers are strongly activated with an increase in blood pressure, this activation is already maximal at pressures in the 145/95 mmHg range. These findings are of particular interest in that they show that a relatively low rise of blood pressure caused a maximal endothelial activation above which further elevation of blood pressure has no further influence on the concentration of the endothelial activation markers. Therefore, endothelial activation appears not be responsible for target organ injury in severe hypertension and other mechanisms should be investigated. Hypertension coexists with numerous pathologies associated with structural changes in the vessel wall, such as hyperlipidemia, atherosclerosis, obesity, nephropathy, diabetes mellitus and smoking in which soluble adhesion markers are elevated [18,36,38–41]. However, in these cases, hypertension is the consequence and not the cause. Although hypertension produces endothelial activation and marker elevation, and although the contributing role of vascular endothelium in the development of hypertension-related vascular damage is well accepted, the associated risk factors per se are more important determinants of the endothelial dysfunction than hypertension itself [36,38,39]. In addition, the absolute level of blood pressure becomes quantitatively less important than the factors in the hypertensive microenvironment. Elsewhere, Preston et al. [37] have evaluated clinical and biochemical variables known to affect endothelial activation. They found a positive correlation between sVCAM-1 and proteinuria, diabetes mellitus, age and smoking. Unfortunately, their study did not allow them to show the influence of these risk factors on the relationship between blood pressure levels and soluble CAM concentrations within the two hypertensive groups. For example, it has been reported that obesity per se induced early endothelial activation in hypertensive and normotensive men [36].Very subtly, Verhaar et al. [35] observed that the progression of vascular damage in essential and renovascular hypertensives is associated with a rise in circulating levels of P-selectin and, to a lesser extent, E-selectin, but not of sCAM and vWF, whereas levels of soluble selectins, sICAM-1, sVCAM-1 and vWF are elevated only in diseases associated with acute severe vascular damage, including malignant hypertension. In conclusion, it appears that the plasma concentrations of the endothelial cell surface adhesion molecules may, to a certain extent, be considered as markers of vascular injury during hypertension, but they are not factors of prognostic value concerning the development of target organ damage. Acknowledgement The author expresses his gratitude to Dr Mary Osborne-Pellegrin for her helpful assistance during the preparation of the manuscript.