Elsevier

Progress in Neurobiology

Volume 64, Issue 6, August 2001, Pages 575-611
Progress in Neurobiology

Cerebral microvascular pathology in aging and Alzheimer's disease

https://doi.org/10.1016/S0301-0082(00)00068-XGet rights and content

Abstract

The aging of the central nervous system and the development of incapacitating neurological diseases like Alzheimer's disease (AD) are generally associated with a wide range of histological and pathophysiological changes eventually leading to a compromised cognitive status. Although the diverse triggers of the neurodegenerative processes and their interactions are still the topic of extensive debate, the possible contribution of cerebrovascular deficiencies has been vigorously promoted in recent years. Various forms of cerebrovascular insufficiency such as reduced blood supply to the brain or disrupted microvascular integrity in cortical regions may occupy an initiating or intermediate position in the chain of events ending with cognitive failure. When, for example, vasoconstriction takes over a dominating role in the cerebral vessels, the perfusion rate of the brain can considerably decrease causing directly or through structural vascular damage a drop in cerebral glucose utilization. Consequently, cerebral metabolism can suffer a setback leading to neuronal damage and a concomitant suboptimal cognitive capacity. The present review focuses on the microvascular aspects of neurodegenerative processes in aging and AD with special attention to cerebral blood flow, neural metabolic changes and the abnormalities in microvascular ultrastructure. In this context, a few of the specific triggers leading to the prominent cerebrovascular pathology, as well as the potential neurological outcome of the compromised cerebral microvascular system are also going to be touched upon to a certain extent, without aiming at total comprehensiveness. Finally, a set of animal models are going to be presented that are frequently used to uncover the functional relationship between cerebrovascular factors and the damage to neural networks.

Introduction

Although the theme of the current review is the age- and dementia-related breakdown of cerebral microvessels, it is essential to have a clear view of the arborization and regional distribution of the larger cerebral blood vessels. The microvascular network of the brain operates strongly dependent on the blood flow and resistance of the large arteries and the smaller, terminal arterioles.

Cerebrovascular research makes use of a range of experimental animal models such as vessel occlusions to unravel the contribution of an optimal cerebral circulation to the physiology and metabolism of the brain. When employing laboratory animal models (the rat and gerbil being the most frequent ones) to tackle the pathophysiology of human cerebrovascular diseases, it should be emphasized that although the organization of the cerebrovascular system is in many respects similar among mammals, some differences between species do exist. For instance, the anterior communicating artery is a well-known anatomical unit in humans but not in rats, while the olfactory artery can be found in rats but not in humans (Fig. 1) (Lee, 1995). Even more remarkable is the incomplete circle of Willis in gerbils (Mayevsky and Breuer, 1992). The description below mainly focuses on the human situation with some remarks related to the cerebral circulation of laboratory animal models popular in cerebrovascular research.

The brain receives its arterial blood supply via two major routes, the internal carotid arteries and the vertebral arteries, the latter forming the unpaired basilar artery at the junction of the medulla and the pons. The carotid system is responsible for the anterior circulation of the brain while the basilar artery provides the blood supply to the posterior cerebral circulation. Obviously, the anterior and posterior circuits are not independent of each other: the two are interconnected by communicating arteries that create the circle of Willis at the base of the brain providing potential shortcuts between the lateral as well as the anterio-posterior cerebral circulation (Fig. 1). However, the vertebral and carotid systems supply distinct brain regions as demonstrated by McDonald and Potter (1951) in rabbits. Under physiologically optimal circumstances the blood streaming through the vertebral arteries does not mix with the blood carried by the internal carotid arteries. This phenomenon can be demonstrated by infusing vital dyes in the carotid or vertebral arteries, which will appear chiefly in the corresponding intracranial vessels. Nevertheless, if the pressure gradient in the circle of Willis changes due to an insufficient flow in either the anterior or posterior circuits, blood from different origin can be re-distributed via the collateral intercommunication in the circle. However, the degree of compensation depends on the individual variation of vessel diameters and the symmetry of the circle of Willis (Dickey et al., 1996). The compensatory mechanisms can play a role when the lumen of an intracranial artery is narrowed due to severe atherosclerosis (Hartkamp et al., 1999) or when the common carotid arteries or the middle cerebral arteries of laboratory animals are experimentally occluded to create a model for cerebral ischemia (Weinachter et al., 1990, Coyle and Heistad, 1991, Mhairi Macrae, 1992).

Arteries emanating from the posterior route, that is from the basilar artery, predominantly furnish the brainstem and midbrain with fresh blood whereas the cerebral hemispheres are vascularized from both the anterior (internal carotid origin) and posterior vessels. The two large pairs of vessels originating from the internal carotid arteries are the anterior and the middle cerebral arteries, the latter carrying 80% of the blood that reaches the cerebral hemispheres. Without presenting a complete and comprehensive list of target areas, it is worth following the major routes of the larger arteries. The anterior cerebral arteries send their arborization to the frontal lobe, the preoptic and supraoptic areas, the globus pallidus and the amygdala, while the ramifications of the middle cerebral arteries are responsible for the blood supply to the temporal and parietal cortex, important subcortical nuclei such as the basal nuclei and the choroid plexus in the lateral ventricles. The posterior route reaches the occipital lobe of the hemispheres and the diencephalon containing the sensory thalamus and the vital autonomic hypothalamic nuclei.

The major arteries enter the skull at the base of the brain and their branches consequently advance dorsally and spread on the surface of the cerebrum in the subarachnoid space above the pia mater. They perforate the brain parenchyma perpendicular to the cerebral surface without establishing anastomoses with each other. As a narrowed continuum of the subarachnoid space, the vessels are surrounded by the so-called Virchow–Robin space, which is embraced by leptomeningeal cells. The space gradually disappears as the artery penetrates deeper in the brain tissue, only the leptomeningeal cell layer remains to form the first, very thin layer of the artery, the tunica adventitia. The second and the thickest layer of the vessel wall, the tunica media, consists of one or two layers of smooth muscle cells which are separated from the tunica adventitia by elastin and collagen fibers, the lamina elastica externa. The smooth muscle cells can regulate the flow in the vessel by contracting or relaxing, which specifies the most important function of arteries in controlling blood pressure and flow. Finally, the luminal layer of the artery is practically equivalent to the endothelial cell layer and is often referred to as tunica intima.

The network of fine cerebral vessels and capillary function in the brain inherently differs from that of arteries. The general notion that arteries regulate blood pressure while brain capillaries maintain the blood–brain barrier (BBB) and sustain continuous nutrient, electrolyte and waste product trafficking between neural tissue and blood is apparently reflected in the microvascular anatomy.

Cerebral capillaries represent the finest branches of the vascular tree and, unlike arteries, they form anastomoses and create a three-dimensional vascular network. The density of this mesh perforating the substance of the brain is highly variable. As a general rule, capillary density in the gray matter was found about three times as much as that of the white matter but it may be more appropriate to note that the observed differences in density apparently correlate with the activity and nutrient demand of the particular brain region. Experimental data supporting this conclusion showed a prominent correlation between capillary length per brain volume and local cerebral blood flow (Gjedde and Diemer, 1985) and between the number of capillaries, local blood flow and glucose utilization in a given brain area (Klein et al., 1986). The phenomenon that metabolically active brain regions are more heavily vascularized than less active zones is supported by the observation that capillary density appears to be most pronounced in areas rich in synapses, followed by cell body populations and finally neural fiber bundles. Furthermore, microvascular density also seems to coincide with the main task of the given brain regions: the sensory and association centers are usually more densely vascularized than motor centers. The laminar structure of the cerebral cortex also displays a typical layer-dependent density pattern where lamina IV followed by lamina I receive the densest vascularization. In addition to this density pattern, the orientation of microvessels can also show a laminar arrangement shown by the cortical capillaries, which run parallel to the surface in lamina I but form a multi-oriented network in lamina IV (Hudetz, 1997).

The cerebral capillaries display a typical ultrastructure crucial to execute BBB function (Fig. 2). The three cellular building blocks that participate in the formation of the capillaries are the endothelial cells, the irregularly occurring pericytes and the astrocytic end feet attached to the vessels’ abluminal surface. The capillary endothelial cells form one layer around the capillary lumen and create tight junctions (also called zonulae occludens) where they are apposed to each other. The tight junctions seal the space between the meeting endothelial surfaces and are considered as the morphological basis for the BBB gaining their full functional integrity with the maturation of the animal (Rubin and Staddon, 1999, Kniesel and Wolburg, 2000, Saunders et al., 2000). Other features that take care of the selective isolation of the brain from the blood are the lack of endothelial fenestrations and an insignificant transport via pinocytic vesicles. The capillary endothelial cells are further characterized by a relatively high number of mitochondria, which can provide the energy needed for the working of the specific BBB transport proteins (e.g. glucose- and amino acid transporters).

The endothelial cells are surrounded by a 30- to 40-nm-thick basement membrane (BM) (Fig. 3) which is often a target of investigation due to its frequently observed malformations under pathophysiological conditions (for example Alzheimer's disease) (Perlmutter and Chui, 1990, Claudio, 1996, Kalaria, 1996, Farkas et al., 2000b). The extracellular matrix components of the BM, namely the intrinsic collagen type IV, heparan sulfate proteoglycan (HSPG), laminin and the extrinsic fibronectin are known to be produced by the cell types of the capillaries. These BM constituents are arranged into a trilaminar structure with an endothelial layer (lamina rara interna), an astrocytic layer (lamina rara externa) and a transitory, fused layer in-between the two (lamina densa) (Fig. 3). Collagen type IV, the major structural element of the BM is preferentially located in the lamina densa while the proteins laminin and HSPG are more closely associated with the two lamina rarae, which promote cell adhesion and attachment (Perlmutter and Chui, 1990). Besides the widely cited BM elements, additional proteins that are deposited in the BM have also been identified. Cablin, synthesized by the endothelial and smooth muscle cells, is such a molecule (Charron et al., 1999), suited to cross-link cells and matrix constituents. The BM has been suggested to provide physical support to the microvessels, control cellular migration, filter macromolecules, influence endothelial function, promote cell adhesion and protect the brain against extravasated proteins (Perlmutter and Chui, 1990).

The second, heterogeneous cell type of cerebral capillaries, the pericyte is inserted in the BM and covers the vascular wall by its extended processes. Some investigators differentiate granular and filamentous pericytes and attribute a phagocytotic role to the granular type (Tagami et al., 1990). The size and appearance of pericytic profiles seen with the electron microscope is highly variable depending on the level of slicing. When compared to endothelial cells, the density and composition of the cytoplasm looks very similar but the pericytes also contain dense bodies or lysosomes. The pericytes are often considered as a supporting cell type of capillaries, which can regulate capillary tone (Kelley et al., 1987). They also participate in the immune response as shown by their relationship with macrophages and their ability to transform into microglia. These proposals were further substantiated by the demonstration of the presence of macrophage markers on the pericytic surface, their phagocytotic activity and antigen presentation (Thomas, 1999). Furthermore, the pericytes can contribute to the regulation of vascular development by inhibiting endothelial cell proliferation and differentiation via chemical signaling (Shepro and Morel, 1993, Hirschi and D'Amore, 1996, Balabanov and Dore-Duffy, 1998, Martin et al., 2000, Rucker et al., 2000).

The cerebral microvessels are supported by astrocytic processes, which are intimately apposed to the abluminal vascular surface. These astrocytic end feet are thought to play a dominant role in the ontogenesis and maintenance of the BBB (Janzer, 1993). In vitro studies have demonstrated that the close apposition of astrocytes to endothelial cells is necessary for the development of typical BBB features such as the formation of tight junctions or the expression of BBB specific proteins (Arthur et al., 1987, Minakawa et al., 1991, Rauh et al., 1992, Hurwitz et al., 1993). The induction of an endothelial BBB phenotype marker, the so-called HT7 surface glycoprotein by an astrocyte-conditioned medium is an adequate example for the latter (Janzer et al., 1993). Furthermore, astrocytes were implicated in the intracerebral regulation of vascular tone and cerebral blood flow indicated by the expression of serotonergic and cholinergic receptors on the perivascular end feet (Cohen et al., 1996, Cohen et al., 1999, Luiten et al., 1996, Elhusseiny et al., 1999) and the close apposition of noradrenergic nerve endings to the vascular astrocytic sheath (Cohen et al., 1997). Besides receiving neuronal innervation, the astrocytes stand in constant biochemical interaction with the endothelial cells (Goldstein, 1988, Abbott et al., 1992) shown for example by their substance-P immunoreactivity (Michel et al., 1986), the presence of endothelial NOS in their cytoplasm (Wiencken and Casagrande, 1999) and the detection of astrocytic NO release (Janigro et al., 1996).

These examples demonstrate that although the anatomical organization of the cerebral microvascular domain appears to be relatively simple at first sight, the functional implications are far more complex. The vascular system of the brain is designed to perform fine and ready adjustments of vascular tone, cerebral blood flow, BBB penetration and immunological status depending on the needs of the neural tissue and environmental changes. In the next chapter, the dynamics and physiological aspects of the cerebral blood supply stand in focus.

The physical pattern of cerebral blood flow (CBF) and its pathological changes in brain microvessels have been reviewed with reference to the general rules of fluid dynamics extended to biologically active systems (de la Torre and Mussivand, 1993). As previously summarized (de la Torre and Mussivand, 1993), a number of major parameters can help characterize the dynamics of blood flow in the cerebral vessels, such as flow velocity, microturbulent flow, viscosity of the blood, shear stress created by the vascular wall and vascular resistance. These factors are inseparably and dynamically interrelated.

The blood flow velocity, which can be routinely determined in larger brain arteries with the use of Doppler sonography (Maulik, 1995) and can also be measured in the cerebral capillary bed with the sophisticated intravital microscopy (Hudetz, 1997), is not equal at all points in the vessel lumen throughout its transversal profile. A flow gradient can be characterized with a decreasing flow velocity approaching from the midline of a vessel towards the vascular wall when looking at the cross section of the vessel. Moreover, near the vascular wall, the blood flow is reduced to a near standstill where the blood has a cell-free plasma layer (Fung, 1981, Fung, 1984). The plasma layer next to the vessel wall also serves a significant biological purpose, namely to allow nutrient and mineral transport from the blood to the brain parenchyma from this slow moving layer thus supplying the brain with energy substrates.

Microturbulent flow can disturb the regular passage of blood and can develop when the usual shape of the vascular lumen becomes irregular, e.g. locally thickened (fibrotic arteries, capillaries with local basement membrane thickening), partially obstructed (atherosclerosis) or compressed (Fig. 4). The flow pattern in this case becomes disrupted and random swirls can build up compromising the slow flow of the cell-free layer near the vessel wall (Fung, 1984). When such abnormalities occur in microvessels, the optimal nutrient transport through the BBB is in jeopardy and can lead to a suboptimal cerebral metabolism.

The third rheological factor of importance is the viscosity of the blood. The viscosity stands in an inverse relationship with flow velocity and CBF meaning that a higher whole blood viscosity is associated with lower flow values. Two major factors having influence on viscosity and thus oxygen-carrying capacity of the blood have been identified as the haematocrit value (Harrison, 1989) and the membrane fluidity and aggregation of erythrocytes (Schmid-Schonbein, 1983). Early indications that an increased haematocrit could contribute to a lowered CBF under neuropathological circumstances were found in clinical studies. For example, an increased haematocrit was shown to coincide with the occlusion of the carotid arteries and associated transient ischemic strokes in humans. In this study, the size of cerebral strokes could be correlated with a decreased CBF, which was suggested to be the result of a high haematocrit value (Harrison et al., 1981). However, claiming a direct causal relationship between an increased haematocrit and the development of ischemic strokes based on these data could well be an overinterpretation of the findings. Yet, a causal relationship between CBF and the haematocrit was convincingly demonstrated in patients in another study: when reducing the haematocrit, a consequent improvement in CBF was measured (Thomas et al., 1977). Supportive animal models experimenting with isovolemic hemodilution also showed that reducing the haematocrit without changing the volume of circulating blood decreased blood viscosity and could consequently enhance cerebral capillary perfusion and oxygen delivery (Lin et al., 1995, Hudetz et al., 1999). Hence, we can conclude with certainty that a lower haematocrit caring for reduced blood viscosity improves CBF. Other properties of erythrocytes like the rigidity of their cell membrane or their affinity to form aggregates can also interfere with CBF. As shown in an experimental rat model, the aggregation of red blood cells compromises microvascular perfusion (Mchedlishvili et al., 1999). In addition, the rigidity of the erythrocyte membranes can also affect CBF by limiting the rate of capillary perfusion. The inflexibility of the cell membrane can hinder the passage of erythrocytes through capillaries therefore the membrane fluidity of erythrocytes indirectly interferes with CBF.

The contribution of shear stress (due to the above described velocity gradient of flow) to altered CBF can be accomplished through changing viscosity (Kee and Wood, 1984) and/or having an effect on vascular autoregulation (Rubanyi et al., 1990). The alteration of CBF by shear stress plays the most important role in curved blood vessel segments, where the difference in flow velocity between the middle axis and the wall of the vessel is highest. The increased shear stress can present a physical stimulus to the endothelium and may impose slowly regenerating endothelial damage (de la Torre and Mussivand, 1993). On the other hand, shear stress has also been suggested to stimulate mechanoreceptors presumably present on endothelial cells, which would activate inward rectifier K+-channels. In turn, NO and PGI2 (prostaglandin I2) could be released initiating an increase of vascular diameter (Rubanyi et al., 1990).

Changes in vascular diameter directly lead to alterations in vascular resistance and CBF, two inversely related physiological parameters. Any change in lumen radius will affect the resistance exponentially. The vascular resistance and CBF can be regulated by myogenic, metabolic, neuronal and biochemical means, which processes are overviewed in the following two sections.

The brain receives probably the most constant blood supply of all body organs maintained by a very finely tuned regulation of CBF. Physiological fluctuations in the cerebral perfusion pressure are normally compensated by the cerebrovascular autoregulation to sustain an optimal, uninterrupted CBF. An intact autoregulation is capable of keeping the CBF independent of perfusion pressure provided the perfusion pressure ranges approximately between 60 and 150 mmHg (Wagner and Traystman, 1985, Paulson et al., 1990). Below or above the given values, the autoregulatory mechanisms become uncoupled from perfusion pressure and lose accurate control of CBF. The dynamic maintenance of CBF is achieved by changes in vascular resistance, which can be controlled by local-chemical factors, endothelial factors, autacoids (e.g. histamine, prostaglandins, leukotrienes) and neurotransmitters (Wahl, 1985, Wahl and Schilling, 1993).

The basic feedback mechanisms of the autoregulatory loop in the brain have been classified as myogenic, chemical/hormonal, neurogenic or endothelial dependent. The myogenic component of cerebral autoregulation was defined as the intrinsic capacity of vascular smooth muscle cells to contract in response to mechanical stress such as an increase in transmural pressure (Ursino, 1991). This contractile response can be visualized by manipulating the transmural pressure in arteries that triggers vasoconstriction when increased. With the help of isolated rat or human brain artery preparations, an increased vascular tone and a decreased lumen diameter were detected when the perfusion pressure was gradually increased (Halpern and Osol, 1985, Wallis et al., 1996). Moreover, increased transmural pressure caused little change in CBF unless the perfusion pressure dropped below 60 mmHg, the lower limit of the autoregulatory capacity (Wagner and Traystman, 1985). Based on these results, one can conclude that stretch-dependent vasoconstriction keeps CBF constant when the perfusion pressure stays within the autoregulatory range. As mentioned above, the cellular components of the myogenic autoregulation were located in the vascular smooth muscle, which depolarizes as mechanical pressure increases (Harder, 1985). Such a pressure-activated contraction of smooth muscle cells was described to depend on the extracellular calcium concentration and to be mediated by an arachidonic acid signal transduction pathway (Harder et al., 1997). A metabolite of arachidonic acid (20-hydroxyeicosatetraenoic acid, 20-HETE) in vascular smooth muscle cells serves as a potent vasoconstrictor by inhibiting the opening of calcium activated potassium channels or by activating L-type calcium currents (Harder et al., 1997). However, other endothelial substances such as endothelins released as a response to stimulation of the vascular endothelium, which is the major focus of the next section, can also indirectly elicit vascular contraction.

The neurogenic regulation of the main cerebral arteries differs from that of cerebral microvessels in that the large vessels receive extracranial innervation while the terminal microvascular beds of the brain lack such a neural supply. Similar to the systemic resistance vessels, the large arteries of the brain surface and their parenchymal branches receive sympathetic, parasympathetic and sensory fibers. A fundamental body of information was accumulated by tract-tracing studies which identified the superior cervical ganglion as the major source of sympathetic fibers (Edvinsson et al., 1990) and the sphenopalatine, otic and internal carotid ganglia as the principal origin of parasympathetic fibers (Branston, 1995). The perivascular sympathetic fibers eliciting vasoconstriction were immuno-positive to several compounds including the classical neurotransmitter noradrenaline and neuropeptides like neuropeptide-Y (NPY) (Uddman and Edvinsson, 1989) while smaller pial arteries were also reported to receive serotonergic, vasoconstrictive input from the dorsal raphe nucleus (Lincoln, 1995). On the other hand, the parasympathetic nerves showed the presence of acetylcholine (ACh) and vasoactive intestinal polypeptide (VIP), both potent vasodilators besides nitric oxide (NO), which also emerged as a significant neurogenic relaxing factor (Suzuki and Hardebo, 1993, Branston, 1995). The sensory projection fibers to cerebral arteries were shown to arise from the trigeminal ganglion and to contain additional vasodilatory peptides such as substance-P (SP) and calcitonin gene-related peptide (CGRP) (Uddman and Edvinsson, 1989).

The control of vasoconstriction mediated by autonomic fibers exerts a basic, global and relatively rough modulation of CBF while the finer tuning of regional flow rates involves several additional mechanisms depending on the vascular endothelium. Biochemical signals acting on or released by the endothelial cells can substantially modify cerebrovascular resistance. The receptors and functional involvement of local, chemical factors (adenosine), endothelial factors (thromboxanes, endothelin, endothelium-derived constrictor/relaxing factors and prostacycline), autacoids (histamine, bradykinin, eicosanoids) and hormones (angiotensin, vasopressin) (Wahl and Schilling, 1993) were widely investigated and discussed. Here, we present a selection of the most important findings of this research that are relevant to the physiology and regulation of cerebromicrovascular blood flow.

The vascular endothelium plays a pivotal role in CBF regulation because an important group of vasoactive biochemical compounds are released by and act on the endothelial cells. These factors are traditionally named as endothelium-derived relaxing factors, nitric oxide (NO) being one of them, and endothelium-derived contracting factors, like endothelins. Most of the data concerning the regulatory function of NO and endothelins were collected from arterial endothelial cells, but the release of these factors from microvascular endothelium was also shown (Yoshimoto et al., 1991, Durieu-Trautmann et al., 1993, Lovick and Key, 1995). In microvessels, the potential target of these factors are the perivascular astrocytes as opposed to the smooth muscle layer in macrovessels (Durieu-Trautmann et al., 1993).

Vascular dilation mediated by nitric oxide (NO) is a well-described phenomenon. NO relaxes vascular smooth muscle and increases regional cerebral blood flow in response to shear stress to the endothelium or stimulation by acetylcholine, bradykinin or other biochemical compounds (Arnal et al., 1999). The mechanical and chemical stimuli can increase the cytosolic calcium concentration and the association of the calcium/calmodulin complex to NO synthase in the endothelial cells (Fleming and Busse, 1999), which in turn modulates NO production by increasing the gene expression and/or the activity of the endothelial NO synthase (eNOS) (Arnal et al., 1999). The origin of NO is, however, not restricted to the endothelium: NO released from neuronal terminals in addition to endothelial sources can also regulate vascular relaxation. In order to visualize the effects of endothelial NO separately from that of neuronal origin, several methods have been applied. The selective blockade of the endothelial NO synthase (eNOS), cell culture of endothelial cells (Weih et al., 1998) or the use of eNOS knockout or mutant mice (Huang et al., 1995, Strauss et al., 2000) all delivered valuable data in NO research. With the help of these models, it was shown that eNOS mediated basal vasodilatation (Huang et al., 1995) and that endothelial NO could buffer blood pressure variability (Strauss et al., 2000). Additional pioneering work using gene therapy to enhance vasorelaxation also made use of eNOS by associating its gene to an adenovirus vector and achieving augmented NO-mediated vasorelaxation in isolated arteries after gene transfer (Ooboshi et al., 1998, Tsutsui et al., 2000).

Endothelins, the very potent vasoconstrictor substances isolated from cultured endothelial cells, were widely investigated for their role in subarachnoid hemorrhage (SAH) Zimmermann and Seifert, 1998). The substances have been held responsible for the delayed cerebral vasospasm after SAH causing considerable damage to the vascular wall. Out of the presently known three endothelin isoforms, ET-1 seems to be the most potent, which probably acts primarily on the endothelin-A receptor (ET-A) (Zimmermann and Seifert, 1998). Although most data on the functional implications of endothelins come from pathological changes after SAH, endothelins can be involved in the control of CBF under physiological circumstances. As supporting evidence, it was demonstrated that when cerebral perfusion pressure was increased with norepinephrine, CBF did not noticeably follow the evoked increase, but when an endothelin-B receptor (ET-B) antagonist, bosentan was administered in combination with norepinephrine, a remarkable rise in CBF was recorded (Mascia et al., 1999). These findings may indicate that ET-B stimulation plays a role in the maintenance of a constant CBF at increasing perfusion pressure under physiological conditions. Because the ET-A and ET-B receptors, as well as the intracellular second messenger of endothelin action, the mitogen-activated protein kinase (MAPK) were identified in the vascular smooth muscle cells (Zimmermann and Seifert, 1998, Zubkov et al., 2000), the suggested regulatory mechanism gains significance in cerebral arteries.

Besides the global regulation of cerebral blood supply via changing the diameter of larger brain arteries, CBF is also regulated locally at the level of microvessels, based on the metabolic activity of the particular brain area examined. Since the brain's fundamental energy source is glucose and its metabolism requires oxygen, the coupling of cerebral glucose utilization (CGU) and cerebral metabolic rate for oxygen (CMRO2) with CBF has been widely investigated in physiological conditions, as well as in neurodegenerative diseases. CGU is generally considered as an indicator of neuronal activity, taken that glucose is used to maintain resting membrane potential and the restoration of ion gradients after an action potential (Jueptner and Weiller, 1995). This theory may also explain the results of the study where a local administration of glutamate or NMDA to the rat cerebral cortex caused a significant rise in CMRO2 and CBF (Lu et al., 1997). Besides consuming oxygen and metabolizing glucose, which can regulate CBF, the firing neurons also release K+. When the extracellular K+ concentration is raised, the ion acts as a vasodilator on nearby vessels and enhances CBF. At the re-establishment of neuronal resting potential, adenosine may also come free and cause an increase in CBF by vasodilatation (Kuschinsky, 1991).

Non-invasive measurements of cerebral CMRO2 in healthy human volunteers showed a correlation between CBF and CMRO2 (Leenders et al., 1990, Hoge et al., 1999) but these findings by themselves may not be sufficient evidence to prove a causal, regulatory relationship between CMRO2 and CBF. Although additional animal studies provided supporting data by demonstrating that reducing blood oxygen concentration elevated CBF proportionally (Jones et al., 1981, Sato et al., 1992), the effect may not be the result of a direct regulatory loop (e.g. the carotid chemoreceptor reflex) because the role of chemoreceptors for blood oxygen concentration could not be unanimously verified (Miyabe et al., 1989). Rather, the reduced oxygen concentration of blood can accompany an increase in CBF, both potentially being a result of increased neuronal activity. Therefore the metabolic regulation of CBF is probably mediated by other by-products of glucose metabolism, the most important being the elevated concentration of CO2 and a consequent increase in blood pH.

CO2 effect can be measured by the CO2 reactivity test, which provides information about the functional state of brain vessels. In man, 1 mmHg increase in blood pCO2 causes a 2–4% increase of CBF mediated by a concomitant change in pH which acts directly on cerebral vessels posing the basic mechanism of regional CBF (rCBF) regulation. An increased [H+] triggers vasodilatation while a decreased [H+] leads to vasoconstriction. Other regulatory mechanisms can potentially modify pH reactivity.

Section snippets

Cerebral blood flow in the aging brain

A gradual functional decline and a concomitant disintegrating morphology typically characterize the aging central nervous system. The physiological neural changes are also accompanied by a well-defined decline in cerebrovascular parameters. A decreasing CBF, lower metabolic rates of glucose and oxygen and a compromised structural integrity of the cerebral vasculature with special attention to microvessels are representative degenerative features of the vascular system of the aging brain.

Cerebral blood flow in Alzheimer's disease

The contribution of vascular factors to the etiology of dementia, with particular attention to Alzheimer's disease (AD) has become a rapidly extending research field in the last decade. Epidemiological studies emphasized the role of peripheral vascular abnormalities like atherosclerosis or hypertension as risk factors aggravating the progression of cognitive decline (Skoog et al., 1996, Skoog, 1997, Hofman et al., 1997, Breteler, 2000), and a further link has been suggested between the systemic

Frequently used animal models

Since it has been anticipated that there is a dynamic interaction between hypertension, reduced CBF, cerebral microvascular pathology, cognitive performance and memory capacity, experimental models were established to investigate the causal relationship between these factors. The laboratory animal models offer the possibility to take the correlation analysis between CBF, vascular parameters and cognitive performance accomplished in human studies one step further since correlation analysis

Conclusions

Cerebral microvascular pathology in aging and to a markedly pronounced degree in AD include the physiological changes in cerebral perfusion — particularly the decrease of regional cerebral blood flow in cortical areas, a reduction of cerebral glucose utilization, the loss of vascular innervation with special focus to the cholinergic breakdown typical of AD and the ultrastructural damage to capillaries in the cerebral cortex represented by extensive basement membrane pathology. The coincidence

References (277)

  • Z Cohen et al.

    Serotonin in the regulation of brain microcirculation

    Prog. Neurobiol.

    (1996)
  • G.I De Jong et al.

    Microvascular changes in aged rat forebrain. Effects of chronic nimodipine treatment

    Neurobiol. Aging

    (1990)
  • G.I De Jong et al.

    Formation of cerebrovascular anomalies in the ageing rat is delayed by chronic nimodipine application

    Mech. Ageing Dev.

    (1992)
  • G.I De Jong et al.

    Cerebral hypoperfusion yields capillary damage in hippocampus CA1 that correlated to spatial memory impairment

    Neuroscience

    (1999)
  • J.C de la Torre et al.

    Chronic cerebrovascular insufficiency induces dementia-like deficits in aged rats

    Brain Res.

    (1992)
  • J.C de la Torre et al.

    Reduced cytochrome oxidase and memory dysfunction after chronic brain ischemia in aged rats

    Neurosci. Lett.

    (1997)
  • J.F Disterhoft et al.

    Calcium-dependent afterhyperpolarization and learning in young and aging hippocampus

    Life Sci.

    (1996)
  • J.L Eberling et al.

    Reduced temporal lobe blood flow in Alzheimer's disease

    Neurobiol. Aging

    (1992)
  • E Farkas et al.

    Are Alzheimer's disease, hypertension, and cerebrocapillary damage related?

    Neurobiol. Aging

    (2000)
  • R.P Friedland et al.

    Alzheimer's disease: anterior-posterior and lateral hemispheric alterations in cortical glucose utilization

    Neurosci. Lett.

    (1985)
  • R.P Gaykema et al.

    Cholinergic fiber aberrations in nucleus basalis lesioned rat and Alzheimer's disease

    Neurobiol. Aging

    (1992)
  • M.J Harrison et al.

    Effect of haematocrit on carotid stenosis and cerebral infarction

    Lancet

    (1981)
  • P Hicks et al.

    Age-related changes in rat brain capillaries

    Neurobiol. Aging

    (1983)
  • A Hofman et al.

    Atherosclerosis, apolipoprotein E, and prevalence of dementia and Alzheimer's disease in the Rotterdam Study

    Lancet

    (1997)
  • S Hoyer

    The effect of age on glucose and energy metabolism in brain cortex of rats

    Arch. Gerontol. Geriatr.

    (1985)
  • A.A Hurwitz et al.

    Human fetal astrocytes induce the expression of blood-brain barrier specific proteins by autologous endothelial cells

    Brain Res.

    (1993)
  • S Imaizumi et al.

    The influence of oxygen free radicals on the permeability of the monolayer of cultured brain endothelial cells

    Neurochem. Int.

    (1996)
  • S Inoue et al.

    Basement membranes, microfibrils and beta amyloid fibrillogenesis in Alzheimer's disease: high resolution ultrastructural findings

    Brain Res. Brain Res. Rev.

    (1999)
  • G Jancso et al.

    Beta-amyloid (1-42) peptide impairs blood-brain barrier function after intracarotid infusion in rats

    Neurosci. Lett.

    (1998)
  • N.J Abbott et al.

    Astrocyte-endothelial interaction: physiology and pathology

    Neuropathol. Appl. Neurobiol.

    (1992)
  • R.G Ackerstaff et al.

    Influence of biological factors on changes in mean cerebral blood flow velocity in normal ageing: a transcranial Doppler study

    Neurol. Res.

    (1990)
  • J.F Arnal et al.

    Endothelium-derived nitric oxide and vascular physiology and pathology

    Cell. Mol. Life Sci.

    (1999)
  • R Balabanov et al.

    Role of the CNS microvascular pericyte in the blood-brain barrier

    J. Neurosci. Res.

    (1998)
  • J.S Beckman

    The double-edged role of nitric oxide in brain function and superoxide-mediated injury

    J. Dev. Physiol.

    (1991)
  • S.A Bennett et al.

    Chronic cerebral hypoperfusion elicits neuronal apoptosis and behavioral impairment

    Neuroreport

    (1998)
  • R.J Boado

    Molecular regulation of the blood-brain barrier GLUT1 glucose transporter by brain-derived factors

    J. Neural Transm.

    (1998)
  • N.M Branston

    The physiology of the cerebrovascular parasympathetic innervation

    Br. J. Neurosurg.

    (1995)
  • M.M Breteler

    Vascular involvement in cognitive decline and dementia. Epidemiologic evidence from the Rotterdam Study and the Rotterdam Scan Study

    Ann. N. Y. Acad. Sci.

    (2000)
  • L Buée et al.

    Brain microvascular changes in Alzheimer's disease and other dementias

    Ann. N. Y. Acad. Sci.

    (1997)
  • P Burgermeister et al.

    Mechanisms of cerebrovascular amyloid deposition. Lessons from mouse models

    Ann. N. Y. Acad. Sci.

    (2000)
  • A.J Charron et al.

    Cablin: a novel protein of the capillary basal lamina

    Am. J. Physiol.

    (1999)
  • L Claudio

    Ultrastructural features of the blood-brain barrier in biopsy tissue from Alzheimer's disease patients

    Acta Neuropathol.

    (1996)
  • Z Cohen et al.

    Astroglial and vascular interactions of noradrenaline terminals in the rat cerebral cortex

    J. Cereb. Blood Flow Metab.

    (1997)
  • Z Cohen et al.

    Multiple microvascular and astroglial 5-hydroxytryptamine receptor subtypes in human brain: molecular and pharmacologic characterization

    J. Cereb. Blood Flow Metab.

    (1999)
  • F Coria et al.

    Cerebral amyloid angiopathies

    Neuropathol. Appl. Neurobiol.

    (1996)
  • D.C Costa et al.

    CBF tomograms with [99mTc]-HM-PAO in patients with dementia (Alzheimer type and HIV) and Parkinson's disease — initial results

    J. Cereb. Blood Flow Metab.

    (1988)
  • P Coyle et al.

    Development of collaterals in the cerebral circulation

    Blood Vessels

    (1991)
  • G.I De Jong et al.

    Nimodipine effects on cerebral microvessels and sciatic nerve in aging rats

    Neurobiol. Aging

    (1991)
  • G.I De Jong et al.

    Cerebrovascular hypoperfusion: a risk factor for Alzheimer's disease? Animal model and postmortem human studies

    Ann. N. Y. Acad. Sci.

    (1997)
  • J.C de la Torre

    Critical threshold cerebral hypoperfusion causes Alzheimer's disease?

    Acta Neuropathol. (Berl.)

    (1999)
  • Cited by (0)

    View full text