There is no doubt that the hemodynamic effects of hypertension are an important determinant of artery remodeling. However, there is mounting evidence suggesting that several circulating factors, which are increased in hypertensive subjects, also play an important role in the remodeling process.

The search for a molecular mechanism for artery remodeling has led to an intense scrutiny of the renin-angiotensin-aldosterone system (RAAS). Studies utilizing angiotensin-converting enzyme (ACE) inhibitors [79, 80] or angiotensin receptor blockers [81, 82] have helped define the blood pressure dependency of the cerebral artery remodeling process. The effects of RAAS inhibition have been compared to the effects of β-blockers, which lower blood pressure by a RAAS-independent mechanism. The key finding from these studies is that blood pressure lowering alone does not improve pial [79, 82] or middle cerebral artery structure [83, 84] in SHR or SHRSP. The β-blockers reduced blood pressure but did not alter artery structure; they did however reduce plasma renin activity and angiotensin II levels [85]. Why β-blockers and ACE inhibitors have disparate effects on artery structure remains a mystery. At a simplistic level one would expect them to have similar effects because both reduced blood pressure and angiotensin II levels. Other studies have utilized combinations of low doses of ACE inhibitors (ramipril) and angiotensin receptor blockers (telmisartan). These studies showed that targeting the RAAS with these drugs together improved pial artery structure and function in young SHR. Given alone the same doses of each drug had no effect [86]. However, the combination therapy was not more effective than a higher dose of either drug used alone [81]. In all these studies the drugs that improved the artery structure also reduced the blood pressure. Several of these studies also assessed cerebral blood flow and found that when the passive lumen diameter of the cerebral arteries was increased the cerebral blood flow was also elevated [81, 86, 87]. This suggests that passive artery structure plays a role in the control of cerebral blood flow.

Two recent studies assessed the effects of start term inhibition of the RAAS on pial and middle cerebral artery structure. In these studies 4–5-month-old SHR and WKY rats were treated with candesartan or telmisartan for 10 days. Both drugs reduced the blood pressure in the SHR, but only telmisartan increased the resting lumen diameter of the pial arterioles. The authors propose that the beneficial effects of telmisartan occur through a peroxisome proliferator-activated receptor γ (PPARγ)-dependent mechanism [88]. The effects of direct PPARγ activation on artery remodeling will be described further later. Interestingly telmisartan had no effect on the structure of the middle cerebral arteries [89] in rats undergoing the same treatment regime. This differential effect of angiotensin receptor blockade on the middle cerebral and pial arteries is interesting, but at present unexplained.

The studies described above are supported by studies utilizing angiotensinogen knockout mice; these mice do not produce angiotensin II and therefore are good correlates for studies using ACE inhibitors and angiotensin receptor blockers. This study focused on the small collateral arteries that connect the anterior and middle cerebral arteries. The lumen diameter of these anastomoses was increased in the mice lacking the angiotensinogen gene compared to wild-type mice. There was no difference in the number of anastomoses present between the two strains. The absence of angiotensin II in these mice also caused a reduction in blood pressure, and it is possible that it had significant effects on the artery structure [90].

One of the issues with many of the studies of hypertension-associated artery remodeling is the age of the rats and the timing of the treatments. In many cases the drug treatments were given as the hypertension was developing, in this situation effective drugs are essentially preventing artery remodeling. Clinically, it is much more important to identify ways to improve artery structure and function after the development of the hypertension and artery remodeling. Of particular note is a study showing that ACE inhibitors effectively improved artery structure in adult SHR; these rats were already hypertensive and artery remodeling would be present. Rats were treated with captopril from 12 months of age, and the studies were conducted in 15-month-old rats. The captopril treatment reduced the wall thickness and the wall-to-lumen ratio; captopril also reduced the systemic blood pressure. This study suggests that the hypertension associated with pial artery remodeling is reversible contingent upon a chronic reduction in blood pressure [87].

Aldosterone is the last signaling molecule produced by the RAAS. Aldosterone activates the mineralocorticoid receptor, and this has detrimental effects on cerebral arteries. Mineralocorticoid receptor activation with DOCA or 11β-hydroxysteroid dehydrogenase inhibition causes marked middle cerebral artery remodeling that is associated with the development of mild hypertension (systolic pressure of approximately 160 mmHg). These treatments induced an inward hypertrophic remodeling and the changes in artery structure translated into a large area of cerebral infarct when ischemia was induced experimentally [67, 68]. Conversely, mineralocorticoid receptor antagonists have beneficial effects on the cerebral vasculature in SHSRP. Treatment of young SHRSP from 6 to 12 weeks of age with spironolactone prevented the development of inward artery remodeling, the spironolactone-treated rats had middle cerebral arteries with larger lumens and the wall-to-lumen ratio was reduced [64]. Spironolactone also effectively reversed middle cerebral artery remodeling in older SHRSP with established hypertension. In this study the spironolactone-treated rats had larger middle cerebral artery lumen diameters than the untreated hypertensive rats [91]. The mineralocorticoid receptor antagonists did not lower the blood pressure in either study suggesting that the prevention and reversal of mineralocorticoid receptor-mediated artery remodeling is not dependent on systemic arterial pressure. Mineralocorticoid receptor activation has also been implicated in hypertension-associated endothelial dysfunction in the cerebral vasculature. This effect of aldosterone in the basilar artery appears to be the result of increased oxidative stress [92].

Rats treated with ACE inhibitors and angiotensin receptor blockers continue to show beneficial effects of the drugs even after treatment withdrawal [93–95]. A recent study assessed if the beneficial effects of mineralocorticoid receptor antagonists on cerebral arteries were sustained after their withdrawal. Surprisingly, withdrawal of spironolactone after a prolonged treatment had marked detrimental effects on the middle cerebral arteries from SHRSP. The rats were treated from 6 to 12 weeks of age and studied at 18 weeks of age. The middle cerebral artery lumen diameter was smaller in the rats treated with spironolactone when compared to the untreated SHRSP. The reduction in lumen diameter occurred without a change in the wall thickness or wall-to-lumen ratio. Although the structural effects of spironolactone withdrawal were deemed to be detrimental, a marked improvement in middle cerebral artery endothelium-dependent dilation was observed after spironolactone treatment and withdrawal [96]; this may be an effort by the vasculature to compensate for the inward artery remodeling. Interestingly, unlike ACE inhibitors and angiotensin receptor blockers, the effects of mineralocorticoid receptor activation on the cerebral arteries occurred in a blood pressure-independent manner [64, 91, 96]. In the future it will be important to investigate whether angiotensin II and aldosterone act synergistically to exacerbate the artery remodeling process.

The molecular mechanisms responsible for the changes in artery structure that occur with RAAS activation have not been completely elucidated. But the RAAS activates intracellular signaling cascades that could stimulate artery remodeling. Both angiotensin II and aldosterone increase reactive oxygen species production by increasing NADPHoxidase expression and activity [97, 98]. Superoxide is the reactive oxygen species most commonly produced in this situation, and increased superoxide production has been linked to artery remodeling in peripheral resistance [99] and coronary arteries [100] from hypertensive rats. Superoxide also stimulates remodeling in cerebral arteries. Treatment of SHRSP with tempol, a superoxide dismutase mimetic, prevented the reduction in the middle cerebral artery lumen diameter that is normally observed as hypertension develops in the SHRSP [101].

All forms of artery remodeling require some breakdown and reorganization of the extracellular matrix. In eutrophic inward remodeling the extracellular matrix breaks down to allow for the smooth muscle cells to rearrange themselves around a smaller lumen. In hypertrophic remodeling the extracellular matrix breaks down to accommodate hypertrophied smooth muscle cells or cell proliferation. Therefore, it is reasonable to propose that the matrix metalloproteinase (MMP) family of enzymes is involved in the artery remodeling process [102]. MMPs are zinc-dependent proteases responsible for breakdown and rearrangement of extracellular matrix components. Importantly, MMP expression is modulated by RAAS activation. Mineralocorticoid receptor antagonism reduces MMP-13 mRNA expression in large cerebral arteries from SHRSP [91], and angiotensin II increases MMP-2 and MMP-9 levels in smooth muscle cells [103]. Inhibiting the actions of MMPs effectively reduces the development of hypertension-associated cerebral artery remodeling. Chronic treatment of young SHRSP (6-week-old) with the nonspecific MMP inhibitor, doxycycline, increased the lumen diameter and reduced the wall-to-lumen ratio of the middle cerebral artery. As is the case with spironolactone, doxycycline had no effect on blood pressure suggesting the prevention of artery remodeling occurs through a blood pressure-independent mechanism. The damage produced by cerebral ischemia was also reduced by doxycycline treatment, and pial artery blood perfusion post-stroke was increased [73]. Importantly, doxycycline was withdrawn several days prior to the induction of cerebral ischemia; thus the improvement in the outcome of cerebral ischemia is probably not a response to acute MMP inhibition at the time of the stroke. This is important because clinical studies suggest that tetracycline antibiotics, like doxycycline and minocycline, have acute beneficial effects post-stroke [104, 105], but these effects do not appear to be vascular in nature.

In recent years the search of mediators of artery remodeling has moved beyond the RAAS system. Several of these potential mediators of the remodeling process will be discussed here.

Artery remodeling has several hemodynamic effects; in hypertensive subjects, artery remodeling is initiated in an effort to normalize cerebral blood flow, which increases as blood pressure increases. Artery remodeling normalizes flow by increasing vascular resistance [136, 137]. Therefore artery remodeling may be responsible for the relatively normal cerebral blood flow observed in hypertensive rats and patients. However, this situation changes as hypertensive subjects age; older hypertensive patients have reduced blood flow in specific brain regions including the occipitotemporal and prefrontal cortex, and the hippocampus [138]. Similarly, patients with poorly controlled blood pressure exhibit significant reductions in their total cerebral blood flow as they age. This cerebral hypoperfusion occurs independently of atherosclerosis [139], and could be the result of artery remodeling and impaired control of constriction and dilation. Aging is a major risk factor for cerebrovascular disorders independently of hypertension. But the incidence of hypertension in the elderly is approximately 70 % [140], which makes a clear separation between these two risk factors difficult. Clinical studies from the 1980s and early 1990s showed a positive correlation between white matter changes, a clinical marker of dementia, and thickening of the wall in small parenchymal arterioles [141, 142] suggesting the link between changes in cerebral artery structure and age-associated dementia. In aged Fischer F344 rats, there is inward hypotrophic remodeling and increased stiffness in pial arterioles [143]. In RORO rats, an inbred strain derived from Wistar rats with high longevity, the media thickness of the basilar artery is reduced in 32-month-old females, but not in 18-month-old females, without changes in lumen diameter [144]. In elderly humans, intracranial posterior arteries show inward eutrophic remodeling coupled to loss of compliance and elasticity [145]. A final structural alteration in aged intracranial arteries is the presence of various degrees of atherosclerosis, which can increase the risk of middle cerebral artery strokes [146].

While the passive structure of the cerebral arteries contributes to the control of cerebral blood flow it is important to note that several other active mechanisms are involved in the regulation of cerebral perfusion. Studies of how cerebral arteries constrict and dilate are important. We know a lot about the mechanisms of constriction and dilation in cerebral arteries; we know much less about how hypertension affects these vasoactive mechanisms. Neurovascular coupling is the mechanism that links neuronal activity to regional cerebral perfusion. The cellular control of neurovascular coupling is complex and requires integrated signaling from neurons, interneurons, perivascular nerves, glia, and the cells within the vasculature [147, 148]. During neurovascular coupling vasoactive agents are released from active neurons, interneurons, astrocytes, and arteries themselves to produce localized vasodilation in parenchymal arterioles and upstream pial arteries; this in turn increases blood flow to the active neurons [24]. Human studies suggest that neurovascular coupling is impaired by hypertension [136].

Neurovascular coupling in rodents is commonly assessed by measuring the increase in blood flow or the increase in artery lumen diameter in response to various sensory stimuli including whisker stimulation. Studies using rodent models of hypertension suggest the hypertension-associated impairment in neurovascular coupling is more nuanced than just being a response to increased blood pressure. In fact increasing blood pressure in a mouse with phenylephrine does not affect neurovascular coupling suggesting that elevated blood pressure alone does not impair the increase in perfusion associated with neuronal activation [149]. Most studies suggest that activation of the renin angiotensin pathway is required to observe impaired neurovascular coupling in hypertensive animal models. Acute and chronic angiotensin II administration reduces the increase in cerebral blood flow observed with whisker simulation in mice [149]. However, this angiotensin II-mediated event is not blood pressure dependent because direct application of angiotensin II to the cerebral cortex has similar effects on cerebral blood flow in the absence of an increase in blood pressure [150]. The argument for a direct effect on angiotensin II on the brain is strengthened by studies showing that neurovascular coupling is impaired before the blood pressure becomes elevated mice given a low dose of angiotensin II that increases the blood pressure slowly [150]. The impaired neurovascular coupling observed in response to angiotensin II appears to require the activation of the type 1 angiotensin receptor and an increase in reactive oxygen species generation [151].

The effects of hypertension on neurovascular coupling in genetically hypertensive rats are less clear and less well studied. Studies using SHR show that the response to whisker simulation is impaired in rats with developed hypertension. Both the maximum increase in blood flow and the duration of the increase in blood flow were reduced in the SHR. Surprisingly, treating the rats with losartan to lower the blood pressure did not improve the functional impairment in neurovascular coupling and the impaired blood flow was not associated with reductions in vessel density or artery lumen diameters [152]. This suggests that in rats with sustained polygenic hypertension angiotensin II may not be the sole determinant of the impaired neurovascular coupling. The reason why blood pressure lowering does not improve the blood flow response to simulation is less clear and may be associated with long-term dysfunction of the cerebral arteries.

The cerebral circulation is endowed with the innate ability to keep parenchymal perfusion constant while the perfusion pressure is fluctuating; this occurs through the mechanism of autoregulation, also known as Bayliss effect (Fig. 6.2). The autoregulatory range is the range of blood pressure over which the cerebral blood flow remains constant. In most adults this range is between mean arterial pressures of 50–60 mmHg and 150 mmHg [153]. Blood flow control is lost when the perfusion pressure is outside the autoregulatory range; in this situation the cerebral blood flow is dependent on the mean arterial pressure [154]. Pressures above the autoregulatory range cause increased cerebral perfusion and eventually vasogenic edema; pressures below the autoregulatory range cause cerebral hypoperfusion and ischemic injury [11]. Studies to investigate the factors that control autoregulation have been fruitful. Myogenic reactivity is an important determinant of cerebral autoregulation (this will be discussed further below). Neuronal NO production modulates myogenic reactivity and therefore autoregulation [153, 155–157]. Recent studies in humans suggest that sympathetic vasoconstriction [158] and cholinergic vasodilation [159] play an important role in autoregulation. Autoregulation is also controlled by blood flow itself; Koller and Toth recently reviewed this concept [160], and there is evidence suggesting that increased flow causes constriction or dilation depending on the artery being studied.

Before discussing myogenic reactivity it is necessary that we define some key terms. Myogenic reactivity is the ability of the artery to change its tone (or degree of constriction) to respond to fluctuations in intraluminal pressure while keeping blood flow constant [161]. Myogenic tone is an intrinsic property of arteries and arterioles to maintain a spontaneous active contractile force in the smooth muscle cells. Several factors contribute to myogenic tone regulation; these include the intraluminal pressure, resting potassium conductance, calcium channel activity, and the sensitivity of the smooth muscle cell contractile machinery to calcium. Tone is also regulated by extrinsic factors, such as hypoxia.

As mentioned previously, myogenic reactivity is key determinant of cerebral autoregulation; this is particularly true at higher perfusion pressures when arteries constrict to maintain blood flow at a constant level [154]. Cerebrovascular resistance is regulated by basal vascular tone, and much of this tone is myogenic [162]. Myogenic tone is increased when the endothelium is removed from arteries [163, 164] and the myogenic response and myogenic reactivity are regulated by the endothelium through several mechanisms. Myogenic constriction is reduced by agents produced in the endothelium, including NO [165], prostacyclin [166], and endothelium derived hyperpolarizing factor [167]. Conversely, 20-HETE may increase myogenic tone and its production is increased in arteries from hypertensive rats [168, 169].

Pressure myography is commonly used to study the myogenic response. Using this technique the responses of the artery to pressure and stretch can be studied in the absence of confounding factors such as circulating vasoactive compounds and metabolites. The myogenic behavior of large cerebral arteries occurs in three stages. Tone develops in phase one; this requires smooth muscle cell depolarization and occurs at intraluminal pressures of 40–60 mmHg. Intracellular calcium is increased during this stage, and tone development reduces artery wall tension and tangential stress. Phase 2 is myogenic reactivity and this occurs between the intraluminal pressures of approximately 60 and 140 mmHg. In this phase the tone that developed in phase 1 is maintained even though the wall tension and intraluminal pressure are increased. Phase 3 is force-mediated dilation, which occurs when the intraluminal pressure increases to the point where the tangential wall stress exceeds the smooth muscle cells ability to contract [154].

A few studies have assessed autoregulation and myogenic reactivity in models of hypertension, and they are conflicting. Pressure myograph studies show that posterior cerebral arteries from normotensive WKY rats exhibit myogenic reactivity over a range of intraluminal pressures from 40 to 150 mmHg with force-mediated dilation occurring at higher pressures. In SHR the range of pressures where myogenic reactivity occurs is higher (65 to 190 mmHg) [170], and the autoregulatory curve is shifted to the right. However, other studies show that myogenic tone generation is increased in posterior cerebral arteries from SHR compared to WKY rats; i.e., at each intraluminal pressure studied the arteries from the SHR were relatively more constricted than those from WKY rats. This increased tone generation occurred over a wide range of intraluminal pressures, but arteries from both strains experienced force-mediated dilation at the same pressure [171]. Middle cerebral arteries from male SHRSP show similar myogenic responses to WKY rats, but the arteries from female rats behave differently. Middle cerebral arteries from female SHRSPs exhibited more myogenic constriction than female WKY rats [172]. Recent studies from Chan et al. show a small but insignificant increase in tone generation in the penetrating arterioles from female SHR compared to WKY rats [74]. These studies were conducted in 18-week-old rats, and it is possible that statistical significance would be reached in older rats.

Feeding SHRSP a high-salt diet causes the middle cerebral arteries to lose their ability to generate myogenic tone [173]. In this situation the downstream arterioles and capillaries are at risk of rupture because of the elevated blood pressure. SHRSP fed a high-salt diet from weaning begin developing hemorrhagic strokes at about 13 weeks of age—the ability of the cerebral arteries to autoregulate is also completely lost at this time, and a pressure-dependent increase in cerebral perfusion is observed [174]. The mechanisms responsible for the loss of myogenic capacity in salt-loaded SHRSP have not been elucidated. It is possible that this is a physical response to the exacerbation of hypertension that occurs with the administration of a high-salt diet to SHRSP [175, 176].

Aging alters the effects of hypertension on cerebral autoregulation. Young (3-month-old) or aged (24-month-old) mice were treated with angiotensin II to develop hypertension, and the myogenic reactivity of the middle cerebral artery was studied. The middle cerebral artery from young angiotensin II hypertensive mice showed the classical increase in tone associated with elevated intraluminal pressures. Myogenic reactivity was lost in the middle cerebral arteries from hypertensive, aged mice. The loss of myogenic reactivity was attributed to impaired 20-HETE production and reduced activity of the TRP channel canonical 6 (TRPC6), which are responsible for maintenance of myogenic responses in cerebral arteries [177]. 20-HETE and TRPC6 contribute to an increase in intracellular Ca²⁺ in smooth muscle cells, leading to constriction in middle cerebral arteries from young, but not from aged, angiotensin II hypertensive mice [178].

The ability of an artery to respond to increases in intraluminal pressures requires that the arteries express some kind of mechanosensor that will link pressure and stretch to constriction. The precise identity of the mechanosensors in the cerebral vasculature is the subject of much debate, but a few candidates have emerged in the recent years. TRPC6 mediate myogenic constriction of cerebral arteries [179]. Similarly, TRP melastatin (TRPM) 4 regulates myogenic tone and reactivity in large intracranial arteries [180]. The currently accepted model of mechanotransduction proposes that membrane stretch causes angiotensin II receptor type 1 activation and Src-tyrosine kinase-dependent activation of PLCγ1. PLCγ1 activation generates IP₃, which activates IP₃ receptors in the sarcoplasmic reticulum, leading to release of Ca²⁺ that activates TRPM4 in the membrane, resulting in Na⁺ influx, depolarization, and smooth muscle cell constriction. At the same time, membrane stretch activates TRPC6, resulting in Ca²⁺ influx that positively modulates IP₃ receptor activity [181]. TRPM4 are also important regulators of myogenic activity in parenchymal arterioles, although it seems in these arterioles that the mechanosensors are P2Y receptors [182]. The effect of hypertension on these mechanosensor pathways has not been studied. It is possible that the activity of these pathways is increased in arteries from hypertensive subjects, and this could account for the increase in myogenic tone and reactivity.

While the upper end of the autoregulatory curve has received a lot of attention we know much less about how hypertension affects the lower end of the curve and we may be missing important physiological effects. The lower limit of autoregulation is the pressure below which blood flow becomes dependent on blood pressure. When the intraluminal pressure falls below this level active dilation no longer occurs and the arteries collapse because the intraluminal pressure is low; this causes blood flow to fall. In animals with hypertension the lower limit of autoregulation is increased [183] and lowering blood pressure with an ACE inhibitor reduces the lower limit of cerebral blood flow regulation [87, 184, 185]. Reducing the lower limit of blood flow regulation may be important in situations like cerebral ischemia. When a major artery is occluded the intraluminal pressure in arteries downstream from the blockage drops—this would normally cause vasodilation, but if the lower limit of autoregulation is breached, the diameter of the artery, and therefore blood flow, will be dependent on its passive diameter. In this situation the inward remodeling observed in cerebral arteries from hypertensive rats becomes particularly detrimental because it limits perfusion.

NO generation is the mechanism that has been best described for the regulation of cerebral artery tone (Fig. 6.3). Endothelial NO synthase (eNOS) is an important source of NO in cerebral arteries; cerebral microvessels from SHR have lower eNOS expression than those from WKY rats [83], and the impaired NO production associated with this reduces endothelium-dependent dilation, which could lead to increased hypoxia and neuronal death after ischemia. Under normal conditions the arteries around an ischemic occlusion dilate to improve blood flow to the region and reduce the ischemic injury. In hypertensive rats this protective mechanism is impaired [186] and this may be the result of impaired NO production. This hypothesis was tested in studies using eNOS knockout mice. These mice have larger cerebral infarcts after the induction of ischemia than control mice and this appears to be the result of impaired cerebral perfusion [187].

Some drugs have the potential to increase eNOS activity and these drugs also reduce the damage caused by ischemia. Cilostazol is a phosphodiesterase-3 inhibitor used to treat peripheral vascular disease and intermittent claudication. Cilostazol increases eNOS phosphorylation, a marker of enzyme activation, in brains from SHR; this increase in eNOS activation was associated with increased cerebral perfusion after the induction of ischemia [188]. In SHRSP fed a high-fat diet, cilostazol increased cerebral blood flow compared to untreated SHRSP and SHRSP treated with aspirin or clopidogrel. The absence of an effect of aspirin or clopidogrel suggests that the beneficial effects of cilostazol on cerebral blood flow are not merely a function of its antiplatelet activity [189], and that the enhanced perfusion may be the result of improved endothelial function. Clinically this finding could be important. The Cilostazol Stroke Prevention Study showed that cilostazol treatment prevents secondary ischemic strokes; interestingly, this effect was found to be independent of cilostazol’s effects on platelet aggregation [190] and may therefore be related to improved cerebrovascular health.

Increased oxidative stress causes impaired NO-dependent dilation in arteries from hypertensive subjects [191]. Under normal physiological conditions, superoxide dismutase converts superoxide to hydrogen peroxide. Excessive superoxide production saturates the cell’s defenses, which increases the superoxide available to react with NO and reduces its bioavailability [192]. This pathway may be particularly important in cerebral arteries because they have an enhanced capacity to produce superoxide when compared to peripheral arteries [193]. This may be because reactive oxygen species play a physiological role in the cerebral circulation. For example, hydrogen peroxide causes pial artery dilation by activating large conductance Ca²⁺-activated potassium channels (BK_Ca) [194]. Conversely, in the middle cerebral artery, hydrogen peroxide causes constriction by opening L-type voltage-gated Ca²⁺ channels [195]. Interestingly, there is little superoxide production in basilar arteries from SHR. This may be the result of enhanced superoxide dismutase expression since superoxide was detected after inhibition of superoxide dismutase [196]. The mechanism responsible for the increased superoxide dismutase expression is unknown, but it may be a response to enhanced superoxide production, which may function as an alternative vasodilator mechanism during chronic NO depletion (Fig. 6.3).

Endothelial-dependent hyperpolarizing factor (EDHF) (Fig. 6.3) is responsible for NO and prostacyclin-independent vasodilation [197]. The identity of EDHF is the subject of intense debate, but calcium-activated K⁺ channels, the small- and intermediate-conductance Ca²⁺-activated K⁺ channels (SK_Ca and IK_Ca, respectively) in particular, are involved in EDHF-mediated dilation [198]. IK_Ca knockout mice develop hypertension and acetylcholine-mediated dilation is impaired in carotid arteries from these mice [199]. In middle cerebral arteries EDHF-dependent dilation requires IK_Ca activity [200]. It is unclear if IK_Ca expression is reduced in cerebral arteries from hypertensive subjects. Interestingly, IK_Ca expression is elevated in mesenteric resistance arteries from SHRSP, but EDHF-mediated dilation is impaired [201]. This study suggests that the IK_Ca channel is not solely responsible for EDHF-mediated dilation.

TRP channels are increasingly being recognized as important regulators of cerebral artery endothelial function (Fig. 6.3). TRP channels are nonselective for cations, but some channel subtypes are partially selective for monovalent or divalent cations (for a review see [202, 203]). The TRPC3 is expressed in cerebral arteries and TRPC3 activation causes vasoconstriction [204]. TRPC3 expression is increased in carotid arteries from SHR and this is linked to augmented contractility [205]. Similarly, increased TRPC3 expression and activity was shown in mesenteric arteries of SHR and linked to higher contractility to endothelin-I. Although TRPC3 expression in cerebral arteries from hypertensive rats has not been studied, it is possible that there is an increase in TRPC3 activity to enhance myogenic constriction.

TRPV4 channels are activated by epoxyeicosatrienoic acids (EETs) and this causes endothelium-dependent dilation in cerebral arteries [206, 207]. The soluble epoxide hydrolase inactivates EETs to dihydroxyeicosatrienoic acids and may be increased in hypertension [208, 209]. TRPV4 activation causes smooth muscle and endothelial cell hyperpolarization, but the mechanism responsible for this in the two cell types differs. In vascular smooth muscle cells, TRPV4 channels form a complex with ryanodine receptors and large conductance Ca²⁺-activated K⁺ (BK_Ca) channels [206]. TRPV4 channel activation allows for a small Ca²⁺ influx; this leads to ryanodine receptor activation in the sarcoplasmic reticulum and the generation of calcium sparks, which are transient and localized increases in Ca²⁺. The calcium sparks cause an increase in BK_Ca opening and this leads to K⁺ efflux and hyperpolarization [206]. In mesenteric resistance artery endothelial cells, muscarinic receptor activation increases TRPV4 channel opening; this generates calcium sparklets, which are transient increases in intracellular calcium in close proximity to a single channel [210]. Calcium sparklets activate IK_Ca and SK_Ca channels and this causes endothelial cell hyperpolarization that is transmitted through gap junctions to the smooth muscle [211]. Calcium sparklets also occur in smooth muscle cells but they are not linked to TRPV4 activity. Reactive oxygen species activate Ca²⁺ sparks in cerebral arteries, coupling this signal to ryanodine receptors and BK_Ca [212]; this may be the result of TRPV4 activation. TRPV4 channels play an important role in regulating endothelial function in cerebral arteries; however, the expression and function of the TRPV4 channel has not been studied in cerebral arteries from hypertensive subjects. TRPV4 channel knockout mice do not develop spontaneous hypertension, but when the mice are treated with l-NAME the resultant hypertension is exacerbated. This enhanced blood pressure response has been linked to impaired endothelium-dependent dilation in peripheral arteries [213]. Recent studies have implicated reactive oxygen species generation by NADPH oxidase in the impaired basilar artery endothelial function in DOCA-salt hypertensive mice [92].