Regulating CBF in a normal brain ensures relatively constancy over a wide range of blood pressure changes [38]. Autoregulation is impaired when blood pressure exceeds the compensatory vasoconstrictor or vasodilatory capacity. This event usually occurs when the average blood pressure falls below 50 mm Hg or rises above 150 mm Hg in a normotensive subject (Figure 2).

Furthermore, in the presence of hemodynamic compromise, vasodilation of the intracranial arterial vessels is observed. This underlying vasodilation limits the brain’s ability to further vasodilate in response to other ischemic events, resulting in an increased risk of subsequent accidents. Hypertensive patients show an altered cerebrovascular reserve, which can be corrected by appropriate antihypertensive therapy [38,39].

The exact mechanisms by which hypertension affects autoregulation of CBF are not fully understood but likely include alterations in myogenic tone and remodelling of internal vessels, increasing the wall-to-lumen ratio [7,12,29,38,39].

The cerebral autoregulatory response occurs primarily in the small vessels and has both myogenic and neurogenic components. Myogenic autoregulation occurs in the smallest arterioles and appears to be primarily responsible for cerebral autoregulation at blood pressure above the normal physiological range. On the other hand, neurogenic autoregulation acts at the level of large arterioles and small arteries by the autonomic perivascular sympathetic nerves. Endothelial dysfunction also causes excess nitric oxide (NO) production via upregulation of NO synthase. NO induces vasodilation, particularly in conditions of high intravascular flows, which would tend to counteract the protective effects of vasoconstriction. Some studies have also shown that NO can increase the permeability of cerebral vessels through a mechanism controlled by cyclic guanosine monophosphate (cGMP) [39,40].

The main pathways involved in the main endothelial functions are reported in Table 2, and they include regulation of vascular tone, fibrinolysis and coagulation, inflammation and formation, repair and remodelling of blood vessels.

These endothelium functions, even if they are described separately, are closely interrelated.

The term endothelial dysfunction identifies the transition from a normal to a damaged endothelium that can express itself with a proinflammatory, constricting, proliferative and procoagulative phenotype. Endothelial dysfunction has been studied extensively in peripheral arteries and was found to precede the elevation in blood pressure and, once developed, to correlate with its severity and target organ damage, including brain damage [15,16,44,47,48,49].

The main factors underlying endothelial dysfunction are therefore represented by reduced bioavailability of NO, impaired smooth muscle response to vasodilators, increased production of vasoconstrictive substances or high shear stress [15,16,36,37,44,48]. As previously reported in the introduction section, the endothelium dysfunction is accompanied by increased production of NO supported by alterations in NO metabolism (high NO degradation, NO inactivation or the presence of NO inhibitors) and with a consequent increase in oxidative stress. Oxidative stress is a condition in which ROS production exceeds the capacity of the antioxidant defence system. Excessive ROS generation, reduced antioxidant capacity or a combination of both can lead to oxidative stress. Persistent oxidative stress can deplete the effectiveness of antioxidant molecules, inactivate the enzymes responsible for antioxidant action and, therefore, compromise the cellular defence system [49,50].

This mechanism plays an essential role in the occurrence of cerebrovascular damage induced by hypertension. The alterations in energy production that occur in brain cells represent the basis for several brain alterations. Brain mitochondria produce about 90% of the energy used by brain cells [55], necessary for the performance of their functions, such as intercellular communication and the transmission of stimuli and signals. An adequate energy supply by the mitochondria is essential for neuronal excitability and survival, so they are implicated in the pathogenesis of neurodegenerative diseases and cerebral ischemia.

Recent data have suggested a close link between excessive ROS generation and the development of neuronal death [56]. The brain is particularly prone to oxidative damage, both due to the lack of antioxidant defences [56] and because it can detect targets of free radicals derived from oxygen, such as the high traffic of calcium through neuronal membranes, presence of excitotoxic amino acids and auto-oxidizable neurotransmitters and a high quantity of polyunsaturated fatty acids contained within membrane lipids.

ROS contribute not only to the damage of macromolecules but also to the transduction of apoptotic signals. Overproduction of ROS by the respiratory chain of brain mitochondria can progressively impair mitochondrial energy metabolism in hypertension. This phenomenon may be implicated in the vulnerability to cerebral ischemia, resulting in progressive neuronal cell death [55].

The experimental evidence linking mitochondrial dysfunction with cerebral vascular damage in hypertensive is limited. However, the evidence that an assembly defect in mitochondrial complexes I and V has been demonstrated in the brain mitochondria of SHRs seems interesting [57]. To better understand the consequences of these data, it is necessary to consider the functions and the energetic cellular role of mitochondrial complexes I and V.

Mitochondrial complex I, or nicotinamide adenine dinucleotide phosphate (NADPH) ubiquinone oxidoreductase, is the most important of the mitochondria mediate oxidative phosphorylation (OXPHOS), contributing about 40% to the energy required for the synthesis of adenosine triphosphate (ATP) by ATP synthase [58]. On the other hand, mitochondrial complex V is responsible for the catalytic phosphorylation of adenosine diphosphate (ADP) to form ATP. Therefore, as a consequence of the assembly defects reported, a reduction in the production of ATP is present in the brains of the SHRs, which induces cellular energy deficiency [57].

It is also known that mitochondria contribute to the intracellular homeostasis of the calcium ion [59]. Some studies indicate that, in elderly SHRs, it is possible to document a reduction in the potential of the mitochondrial inner membrane and consequent impairment of the activity of calcium-dependent enzymes, such as mitochondrial isoform of nitric oxide synthase (mtNOS). These alterations can undoubtedly contribute to a progression of mitochondrial dysfunction and neuronal cell death during hypertension [59]. It is interesting to note that mtNOS, located within the mitochondrial membrane, is involved in cell apoptosis by modulating the transmembrane potential, inhibiting chaim respiration and ATP synthesis, and is correlated with the onset of brain damage induced by hypertension [60].

In the course of arterial hypertension, it is possible to document alterations in the cerebral microcirculation consisting of functional alterations that influence the vasomotor capacity, usually capable of transiently modifying the blood flow, which sometimes leads to thrombotic episodes [2,7,10,12,13,32,39,61].

Critical deficits in global or regional brain perfusion induce suppression of brain activity and cognitive dysfunction [62]. Inflammation can also damage the vessel wall through alterations in endothelial function that result in a reduced function of the cerebral microcirculation with irreversible neuronal damage [63]. Cerebral microbleeds associated with hypertension are typically localized in the basal ganglia, thalamus, brainstem and cerebellum, while a lobar distribution linked to cerebral amyloid angiopathy is common [64]. Functional impairment of the endothelium appears to be an early indicator of vascular dysfunction in cSVD and large cerebral vessels [44,48,65]. Unfortunately, direct assessment of cerebral endothelial function in humans is not feasible, limiting the number of studies available. A recent study analysed cerebral and peripheral vessel reactivity measurements in response to CO₂ inhalation using transcranial doppler and duplex ultrasound in lacunar stroke patients and control subjects. Abnormalities in peripheral artery reactivity appeared to be related to vascular risk factors, and the severity of endothelial dysfunction in the cerebral arteries was related to the onset of lacunar stroke in cSVD patients [66]. It has also been shown that endothelial function, assessed by NO metabolism analysis, is impaired in patients with lacunar infarction and associated cSVD [67].

As previously reported, ROS are key mediators of cerebrovascular dysfunction in hypertension, as they contribute to vascular rarefaction and structural remodelling of cerebral blood vessels and, therefore, lead to functional alterations with profound consequences for CBF. In particular, experimental studies, including models of hypertension in rodents, report that targeting the enzyme that produces ROS, NADPH oxidase or its assembly protects against cerebrovascular oxidative stress and, consequently, from alterations endothelium-dependent relaxation and functional hyperaemia [68,69]. However, existing studies showing the direct effect of hypertension on cerebral arterial tone [70] mainly describe endothelial dysfunction in isolated arteries; only a few studies show a direct link between endothelium-dependent vasodilation and CBF. In addition, postmortem histological findings, conducted in patients with severe cSVD, show in these subjects an intact endothelial layer in the small arteries and the frontal white matter, while an apparent loss of myocytes and other wall cells was documented [71].

These conflicting aspects require further studies in order to document the role of endothelial dysfunction of small cerebral arteries exposed to high blood pressure before and during overt cSVD.

Regarding the role of BBB in cSVD, it is necessary to consider the different types of associated brain lesions and that the components mentioned above are closely related to each other. Presumably, endothelial dysfunction is one of the main determinants of cerebral vessels’ structural and functional alterations. As we reported earlier, the imbalance between the production of vasodilator and vasoconstrictor molecules as well as the haemodynamic effects induced by hypertension lead to endothelial activation, a prerequisite for thrombo-inflammation and an early indicator of impaired BBB, which is usually accompanied by an increase in circulating adhesion molecules which condition the rolling, adhesion and migration of leukocytes [74]. Peripheral markers of endothelial activation include soluble vascular cell adhesion molecules-1 (sVCAM-1), soluble intercellular adhesion molecule-1 (sICAM-1), sP-selectin and sE-selectin, which are associated with cSVD markers and a reduction in cognitive performance in hypertensive patients. This fact indicates an essential role of endothelial activation in the pathogenesis of cSVD mediated by hypertension [75].

Levels of sICAM-1 appear to be an important marker for lacunar stroke and early neurological deterioration [67], even if its origin is not exclusively endothelial, which limits its specificity. More specific endothelial activation markers have emerged recently, and among them, sE-selectin seems to correlate with the number of microbleeds in patients with cSVD [76]. Other studies have shown an increase in adhesion molecules, the rolling of leukocytes along the cerebral vessels and the infiltration and accumulation of T cells in the perivascular spaces in hypertensive individuals, suggesting a causal link between endothelial activation and cerebrovascular dysfunction [77,78].

Other data indicate the presence of a thrombo-inflammatory state in elderly hypertensive subjects with cSVD. Thrombo-inflammation favours vascular occlusions, resulting in greater permeability of the BBB and triggering damage to the vessel wall with consequent vascular ruptures [79]. It has also been described that impairment of BBB resulting from alterations of its cellular components occurs during neurodegenerative diseases [80] but has also been associated with cSVD and lacunar infarcts [81]. However, it still remains uncertain whether BBB dysfunction is the main cause of the changes occurring in the cerebral microcirculation, since, to date, studies have only reported an increase in BBB permeability in cSVD associated with arterial hypertension [82,83,84]. Nevertheless, the simultaneous presence in hypertensive subjects of an increased number of activated circulating immunocompetent cells, of endothelial activation and the presence of microglial activation in the central nervous system indicates the existence of a causal relationship between elevated arterial pressure and impaired BBB [2,7,29,35].

The inflammatory response to stroke generally develops as a consequence of two sequential but closely related phenomena:

• The activation of microglia and perivascular and parenchymal resident macrophages;
• The infiltration of peripheral inflammatory cells into the brain.

Once activated, the glial cells can produce a large variety of proinflammatory mediators suitable for damaging the cerebral endothelium, which, in addition to constituting the BBB, is involved in the determinism and the modulation of the cerebral inflammatory response [84,85].

For example, releasing proinflammatory vasoactive mediators by the cerebral endothelium during the acute phase of stroke can induce blood clotting in the microcirculation, extending the infarct area. Furthermore, once the “neuroinflammatory” response induced by ischemic or haemorrhagic stroke is triggered, it progresses for several days after the onset of symptoms, contributing to the pathogenesis of the later stages of brain damage and causing a worsening neurological prognosis. These considerations support new therapeutic strategies for stroke for which a deepening of knowledge of the neuroinflammatory process is needed [86,87]. A variety of mediators involved in all stages of neuroinflammation, especially in ischemic stroke, have recently been reported from its initial stage to resolution. They include tumour necrosis factor-alfa (TNF-α), interleukin (IL) 1α, IL-1β, chemokine (C-X-C motif) ligand 7 (CXCL7), chemokine (C-C motif), ligand 5 (CCL5), chemokine (C-X-C motif) ligand 4 (CXCL4), chemokine (C-X3-C motif) ligand 1 (CX3CL1), adhesion molecules, proteases, prostanoids and leukotrienes, while IL-1, IL-10, TNF-α, IL-6, IL-20, IL-17, NADPH oxidase, chemokine (C-X-C motif) ligand 8 (CXCL8), inducible nitric oxide synthase (iNOS) and cyclooxygenase-2 (COX-2) are involved in the enhancement phase of this process. In addition, other molecules, including transforming growth factor β (TGF-β), IL-17, IL-10 and IL-23, mediate the resolution phase [18,19,88,89,90,91,92,93].

Finally, some molecules such as the neutrophil-to-lymphocyte ratio family pyrin domain-containing 3 (NLRP3) inflammasome, Dickkopf WNT Signalling Pathway Inhibitor 3 (DKK-3), dectin-1, MKEY and microRNAs (miRNAs), as well as being implicated in the course of neuroinflammation [18,19,86,88,89,90,91,92,93], have shown to have interesting characteristics that can attribute a target role in the management of brain damage, and they are summarized as follows:

• The NLRP3 inflammasome is responsible for the activation of IL-1β and the release of IL-18, which phosphorylate insulin receptor substrate 1 (IRS-1), worsening insulin resistance and causing neuronal death. The NLRP3 inflammasome is one of the primary mediators contributing to the neuroinflammation process and consequent brain damage [88].
• DKK-3 concentrations are associated with endothelial dysfunction and atherosclerosis. High or low levels of DKK-3 are able of inducing a worsening of outcomes after ischemic stroke [89].
• The interaction between Dectin-1 and damage-associated molecular patterns (DAMPs) determines the phosphorylation of immunoreceptor tyrosine-based activation motifs (ITAMs) and, subsequently, of spleen tyrosine kinase (SYK9), a kinase able to mediate the neuroinflammatory cascade through the release of some cytokines. Therefore, the inflammatory pathway mediated by Dectin-1/SYK plays a fundamental role in postictal neuroinflammation [90].
• The heterodimer CXCL4-CCL5 plays a crucial role in developing brain damage [91].
• MKEY, a cyclic peptide synthesized in mice, can avoid the formation of the heterodimer CXCL4-CCL5, thereby limiting ischemic brain injury and improving neurological deficits [92].
• The expression of some miRNAs, such as 126, 124-3p, 30a and 16, is considerably elevated in patients with acute ischemic brain injury, even if they still cannot be successfully used in clinical routine for obvious reasons, above all being the high cost and their execution in highly specialized laboratories [93].

Several examples have emerged to support the activation and contribution of TLRs to the progression of vascular atherothrombotic diseases. The vasculature system expresses all of the known human TLRs, and under vascular pathophysiology, the levels of TLRs are found to be upregulated (Figure 3).