About the role of shear stress in atherogenesis

See article by Gosnach et al. [22] (pages 486–492) in this issue. As blood flows through a vessel, it exerts a physical force on the vessel wall. This force can be resolved into two principal vectors [1]. One, shear stress, is parallel to the vessel wall. It represents the frictional force that the flow of blood exerts at the endothelial surface of the vessel wall. The other, tensile stress, is perpendicular to the vessel wall. It represents the dilating force of blood pressure on the vessel wall. The whole vessel wall, including the endothelium, smooth muscle cells and the extracellular matrix, is exposed to it. In contrast, only the inner surface of the vessel wall, the endothelial cells, are exposed to the frictional force of shear stress. There is some slow transmural flow around smooth muscle cells in arterial walls, but this force is considered insignificant when compared with the tensile wall stress to which these cells are exposed [2]. Many lines of evidence are consistent with the view that a chronic exposure of endothelial cells to high levels of shear stress with little temporal fluctuations causes them to exhibit an atheroprotective phenotype [3]. In vivo studies in humans have demonstrated that areas of the arterial wall that are exposed to low mean shear stress and oscillatory flow with flow reversal are associated with localised endothelial cell dysfunction and atherosclerotic plaque formation [3–7]. Animal experimental studies have shown that endothelial cells in these low shear stress regions are characterised by a rounded shape, increased proliferation rate and increased permeability [8–15]. In these parts of arterial vessels there is an increased uptake of lipoproteins, appearance of leukocyte adhesion molecules on the surface of endothelial cells and leukocyte migration [3,15]. Secretion of chemotactic factors and growth factors causes influx of monocytes, and proliferation of resident macrophages and smooth muscle cells. Together these factors represent the atherosclerotic inflammatory process. In contrast, steady laminar shear stress promotes release of factors from endothelial cells that inhibit coagulation, migration of leukocytes, and smooth muscle cell proliferation [3]. Apart from this growth inhibitory effect of steady laminar shear stress on the vessel wall, it causes smooth muscle cell relaxation and, consequently, vasodilatation. In animal experimental models stimulation of endothelial nitric oxide production has been identified to play an important role in the vasodilatation that occurs within seconds after the application of increased shear stress [16–18]. The measurement of the magnitude of endothelium-dependent vasodilatation in response to reactive hyperaemia, which is widely used in humans to assess endothelial function, is derived from this physiological phenomenon [19,20]. It has been documented in many animal experimental studies that vasodilatation also occurs in response to a chronic increase in shear stress. However, in this setting more factors are involved than nitric oxide alone. It is accompanied by a shift in the balance of release of vasoactive substances and growth-promoting and inhibiting factors by endothelial cells towards vasodilatation and antiproliferation [3,21]. A stimulation of angiotensin converting enzyme (ACE) activity in endothelial cells by shear stress, as reported by Gosnach et al. in this issue of the journal [22], seems to be inconsistent with the above-described view. A higher ACE activity stimulates the production of angiotensin II, a potent vasoconstrictive and growth-promoting agent. The authors argue that the in vitro finding of a higher ACE activity corresponds with observations of reduced lung-tissue ACE activity in an in vivo study of experimental congestive heart failure [23]. They also argue that similar results have been reported in pulmonary hypertension [24]. However, this latter study made a differentiation between ACE activity in alveolar capillary endothelium and ACE activity in small muscular arteries. Indeed, total lung ACE activity decreased, both in the experimental study of pulmonary hypertension [24], and in the experimental study of congestive heart failure [23]. However, in the pulmonary hypertension model it was demonstrated that the reduction in total lung ACE activity was due to a decrease in ACE activity in alveolar capillary endothelial cells. In sharp contrast, an increase in ACE activity in the endothelial cells lining the small muscular arteries was observed [24]. Apparently, changes in shear stress differently affect ACE activity expression in endothelial cells underlined by vascular smooth muscle cells and endothelial cells lacking such an underlining. One explanation may be that feedback inhibition of angiotensin II formed in pulmonary arteries caused a downregulation of ACE mRNA transcription further downstream in the alveolar capillary endothelial cells [25]. Whatever the cause of the discrepancy between shear stress modulation of ACE activity in endothelial cells in pulmonary arteries and alveoli may be, the observation supports the notion that a decrease in shear stress gives rise to an increase in ACE activity in muscular arteries, and vice versa. The findings of Gosnach et al. [22] oppose those of a previous in vitro study by Rieder et al. [26]. A possible confounding factor that may be involved in this controversy is the timing of the experiments with respect to the moment in time that the cultured endothelial cells reached confluence. The result of the study by Rieder et al. is consistent with the current view on how shear stress modifies endothelial cell function [26]. It was demonstrated that 20 dyne/cm2 of shear stress applied for 18 h did not affect ACE activity when experiments were started 2 days after the endothelial cells reached confluence. However, when the same experiment was performed in cultured endothelial cells that reached confluence 4 days before, ACE activity was suppressed 27% by exposure to shear stress. Importantly, Rieder et al. have also demonstrated ACE activity in cultured endothelial cells to be determined to a much greater extent by the time interval lapsing after confluence of the cells had been reached than by shear stress. When endothelial cells were kept in culture for 4 days after they reached confluence, their ACE activity was 85% higher than when they were kept into culture for only 2 days after they reached confluence [26]. Gosnach et al. report to have used cells in their experiments 2–4 days after they have reached confluence [22]. Gosnach et al. suggest that an increase in ACE activity in vascular smooth muscle cells in response to a shear stress induced basic fibroblast growth factor (bFGF) release could play a role in vascular wall remodelling after endothelial injury [22]. In this context it should be realised that the development of an atherosclerotic plaque occurs below an intact, although probably dysfunctional, endothelial surface. Until the endothelial surface becomes disrupted in the later stages of atherosclerosis, vascular smooth muscle cells in vivo are not exposed to shear stress, but to tensile wall stress [2]. Interestingly, increases in tensile wall stress have been shown to strongly promote bFGF production in vascular smooth muscle cells [27,28]. It has indeed been demonstrated in an animal model that a chronic increase in blood flow leads to sustained increases in bFGF mRNA and bFGF protein in vascular smooth muscle cells in vivo [29]. In such a model only endothelial cells, and not vascular smooth muscle cells are exposed to shear stress, while endothelial cells and vascular smooth muscle cells may be expected to be exposed to increases in tensile wall stress [1,2]. Thus, tensile walls stress is more likely to regulate bFGF and ACE activity in vascular smooth muscle cells with respect to physiological vascular wall remodelling and early stages of atherosclerosis. Vascular smooth muscle cells only become exposed to shear stress in the later stages of atherosclerosis, or, for instance, after denudation of the vascular wall from endothelial cells by a balloon catheter, or vascular surgery. The suggestion of Gosnach et al. that shear stress increases ACE activity through inducing bFGF release may however have important implications for our understanding of the role of shear stress in the early course of the atherosclerotic process if it also applies to endothelial cells. There are good reasons to assume that this is indeed the case. Cultured under static conditions, endothelial cells will have a rounded shape, and are not aligned [3,15]. Exposure to chronic shear stress will cause the endothelial cells to reach a flattened shape, and cause them to become elongated and aligned [8,9,30]. After this rearrangement, a certain flow over the surfaces of individual cells will be accompanied by much lower levels of shear stress on individual cells than without previous exposure to flow. Therefore, a sudden exposure to shear stress may induce “wounding” of individual endothelial cells, inducing temporary disruptions of the plasma membrane [31]. Interestingly, such transient plasma membrane disruptions of aortic endothelial surface have been reported in vivo in those areas that have been subjected to a low mean shear stress and oscillatory flow with flow reversal [32]. The same areas are predilection places for the development of atherosclerotic plaques [3–7]. It has furthermore been shown that such plasma membrane disruptions cause release of bFGF from their cytosolic storage sites in endothelial cells [33]. bFGF is considered a “wound hormone” for rapidly initiating cell growth and matrix deposition required for routine maintenance of tissue integrity and/or repair after injury [32]. Malek et al. demonstrated an increase in bFGF mRNA transcription in response to shear stress [34]. However, their results suggest this response to be transient and only present if shear stress is applied above a certain threshold level [34]. The level of this threshold may differ between endothelial cell lines, and it may be important whether shear stress applied in in vitro models rises from zero to the levels aimed for instantaneously or slowly. Thus, the results of the present study are interesting and thought provoking. Moreover, it emphasises the need for a clear differentiation between results derived from acute and chronic shear stress models. Acute models of in vitro shear stress are those in which endothelial cells that have not previously been accustomed to shear stress, are suddenly exposed to this force for a duration of seconds to several hours [35]. Chronic models of in vitro shear stress should be defined as those in which endothelial cells are cultured for several days or weeks under the influence of shear stress, with or without superimposed acute alterations in the level of shear stress. The chronic model more closely approximates conditions in vivo, where endothelial cells are continuously exposed to shear stress with variations in the level of shear stress due to alterations in blood viscosity, blood flow, or lumen diameter. The acute model only applies to the in vivo conditions of first initiation of blood flow through newly formed vessels and restoration of blood flow after vessel occlusion [36].